{
  "best_metric": 0.80437648,
  "best_model_checkpoint": "/qlgy0912/llm_sft_output/qwen2_5-7b/v0-20240927-140411/checkpoint-76000",
  "epoch": 1.994779743070036,
  "eval_steps": 500,
  "global_step": 85500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "acc": 0.69601125,
      "epoch": 2.333075722888931e-05,
      "grad_norm": 65.0,
      "learning_rate": 3.888024883359254e-09,
      "loss": 1.2215333,
      "memory(GiB)": 101.91,
      "step": 1,
      "train_speed(iter/s)": 0.081993
    },
    {
      "acc": 0.6643041,
      "epoch": 0.0002333075722888931,
      "grad_norm": 52.25,
      "learning_rate": 3.888024883359254e-08,
      "loss": 1.42277463,
      "memory(GiB)": 103.58,
      "step": 10,
      "train_speed(iter/s)": 0.306137
    },
    {
      "acc": 0.64442787,
      "epoch": 0.0004666151445777862,
      "grad_norm": 63.25,
      "learning_rate": 7.776049766718508e-08,
      "loss": 1.4146143,
      "memory(GiB)": 106.04,
      "step": 20,
      "train_speed(iter/s)": 0.340883
    },
    {
      "acc": 0.6630681,
      "epoch": 0.0006999227168666793,
      "grad_norm": 58.25,
      "learning_rate": 1.1664074650077761e-07,
      "loss": 1.34018517,
      "memory(GiB)": 106.04,
      "step": 30,
      "train_speed(iter/s)": 0.358391
    },
    {
      "acc": 0.65009575,
      "epoch": 0.0009332302891555724,
      "grad_norm": 54.5,
      "learning_rate": 1.5552099533437016e-07,
      "loss": 1.45347633,
      "memory(GiB)": 106.17,
      "step": 40,
      "train_speed(iter/s)": 0.375122
    },
    {
      "acc": 0.6623395,
      "epoch": 0.0011665378614444655,
      "grad_norm": 28.25,
      "learning_rate": 1.944012441679627e-07,
      "loss": 1.3650135,
      "memory(GiB)": 107.6,
      "step": 50,
      "train_speed(iter/s)": 0.374611
    },
    {
      "acc": 0.65479803,
      "epoch": 0.0013998454337333585,
      "grad_norm": 192.0,
      "learning_rate": 2.3328149300155523e-07,
      "loss": 1.39007854,
      "memory(GiB)": 107.6,
      "step": 60,
      "train_speed(iter/s)": 0.384975
    },
    {
      "acc": 0.67412581,
      "epoch": 0.0016331530060222517,
      "grad_norm": 42.75,
      "learning_rate": 2.721617418351478e-07,
      "loss": 1.3340167,
      "memory(GiB)": 107.6,
      "step": 70,
      "train_speed(iter/s)": 0.392443
    },
    {
      "acc": 0.64180751,
      "epoch": 0.0018664605783111448,
      "grad_norm": 46.0,
      "learning_rate": 3.110419906687403e-07,
      "loss": 1.44007759,
      "memory(GiB)": 107.6,
      "step": 80,
      "train_speed(iter/s)": 0.400513
    },
    {
      "acc": 0.65562725,
      "epoch": 0.0020997681506000378,
      "grad_norm": 169.0,
      "learning_rate": 3.4992223950233286e-07,
      "loss": 1.377672,
      "memory(GiB)": 107.6,
      "step": 90,
      "train_speed(iter/s)": 0.404057
    },
    {
      "acc": 0.65258265,
      "epoch": 0.002333075722888931,
      "grad_norm": 201.0,
      "learning_rate": 3.888024883359254e-07,
      "loss": 1.40249634,
      "memory(GiB)": 107.73,
      "step": 100,
      "train_speed(iter/s)": 0.404893
    },
    {
      "acc": 0.66373754,
      "epoch": 0.0025663832951778242,
      "grad_norm": 44.5,
      "learning_rate": 4.2768273716951787e-07,
      "loss": 1.3577816,
      "memory(GiB)": 107.73,
      "step": 110,
      "train_speed(iter/s)": 0.406581
    },
    {
      "acc": 0.66515074,
      "epoch": 0.002799690867466717,
      "grad_norm": 69.5,
      "learning_rate": 4.6656298600311046e-07,
      "loss": 1.33736687,
      "memory(GiB)": 107.73,
      "step": 120,
      "train_speed(iter/s)": 0.406629
    },
    {
      "acc": 0.6434556,
      "epoch": 0.0030329984397556103,
      "grad_norm": 39.75,
      "learning_rate": 5.054432348367029e-07,
      "loss": 1.45937138,
      "memory(GiB)": 107.73,
      "step": 130,
      "train_speed(iter/s)": 0.409521
    },
    {
      "acc": 0.63647475,
      "epoch": 0.0032663060120445035,
      "grad_norm": 43.5,
      "learning_rate": 5.443234836702956e-07,
      "loss": 1.45315819,
      "memory(GiB)": 107.73,
      "step": 140,
      "train_speed(iter/s)": 0.409846
    },
    {
      "acc": 0.66497622,
      "epoch": 0.0034996135843333967,
      "grad_norm": 18.875,
      "learning_rate": 5.832037325038881e-07,
      "loss": 1.33231993,
      "memory(GiB)": 107.73,
      "step": 150,
      "train_speed(iter/s)": 0.41013
    },
    {
      "acc": 0.67984552,
      "epoch": 0.0037329211566222895,
      "grad_norm": 64.0,
      "learning_rate": 6.220839813374806e-07,
      "loss": 1.25875092,
      "memory(GiB)": 109.26,
      "step": 160,
      "train_speed(iter/s)": 0.409049
    },
    {
      "acc": 0.66304803,
      "epoch": 0.003966228728911183,
      "grad_norm": 41.25,
      "learning_rate": 6.609642301710731e-07,
      "loss": 1.35497437,
      "memory(GiB)": 109.26,
      "step": 170,
      "train_speed(iter/s)": 0.408629
    },
    {
      "acc": 0.6695015,
      "epoch": 0.0041995363012000755,
      "grad_norm": 24.125,
      "learning_rate": 6.998444790046657e-07,
      "loss": 1.32009106,
      "memory(GiB)": 109.26,
      "step": 180,
      "train_speed(iter/s)": 0.408194
    },
    {
      "acc": 0.64377413,
      "epoch": 0.004432843873488969,
      "grad_norm": 189.0,
      "learning_rate": 7.387247278382582e-07,
      "loss": 1.45081196,
      "memory(GiB)": 109.26,
      "step": 190,
      "train_speed(iter/s)": 0.40967
    },
    {
      "acc": 0.66846848,
      "epoch": 0.004666151445777862,
      "grad_norm": 32.25,
      "learning_rate": 7.776049766718508e-07,
      "loss": 1.34669209,
      "memory(GiB)": 109.26,
      "step": 200,
      "train_speed(iter/s)": 0.409516
    },
    {
      "acc": 0.66021504,
      "epoch": 0.004899459018066755,
      "grad_norm": 22.25,
      "learning_rate": 8.164852255054432e-07,
      "loss": 1.3446023,
      "memory(GiB)": 109.26,
      "step": 210,
      "train_speed(iter/s)": 0.40949
    },
    {
      "acc": 0.67646599,
      "epoch": 0.0051327665903556485,
      "grad_norm": 22.625,
      "learning_rate": 8.553654743390357e-07,
      "loss": 1.24624052,
      "memory(GiB)": 109.26,
      "step": 220,
      "train_speed(iter/s)": 0.408686
    },
    {
      "acc": 0.67372646,
      "epoch": 0.005366074162644541,
      "grad_norm": 187.0,
      "learning_rate": 8.942457231726284e-07,
      "loss": 1.31202698,
      "memory(GiB)": 109.26,
      "step": 230,
      "train_speed(iter/s)": 0.408586
    },
    {
      "acc": 0.67337198,
      "epoch": 0.005599381734933434,
      "grad_norm": 87.0,
      "learning_rate": 9.331259720062209e-07,
      "loss": 1.32102585,
      "memory(GiB)": 109.26,
      "step": 240,
      "train_speed(iter/s)": 0.409543
    },
    {
      "acc": 0.68346677,
      "epoch": 0.005832689307222328,
      "grad_norm": 19.125,
      "learning_rate": 9.720062208398133e-07,
      "loss": 1.25821629,
      "memory(GiB)": 109.26,
      "step": 250,
      "train_speed(iter/s)": 0.409961
    },
    {
      "acc": 0.6582798,
      "epoch": 0.0060659968795112205,
      "grad_norm": 27.0,
      "learning_rate": 1.0108864696734059e-06,
      "loss": 1.37746887,
      "memory(GiB)": 109.26,
      "step": 260,
      "train_speed(iter/s)": 0.410468
    },
    {
      "acc": 0.65754652,
      "epoch": 0.006299304451800113,
      "grad_norm": 30.375,
      "learning_rate": 1.0497667185069986e-06,
      "loss": 1.34862213,
      "memory(GiB)": 109.64,
      "step": 270,
      "train_speed(iter/s)": 0.410747
    },
    {
      "acc": 0.67330213,
      "epoch": 0.006532612024089007,
      "grad_norm": 49.0,
      "learning_rate": 1.0886469673405912e-06,
      "loss": 1.3070693,
      "memory(GiB)": 109.64,
      "step": 280,
      "train_speed(iter/s)": 0.411653
    },
    {
      "acc": 0.68029757,
      "epoch": 0.0067659195963779,
      "grad_norm": 34.5,
      "learning_rate": 1.1275272161741837e-06,
      "loss": 1.24388609,
      "memory(GiB)": 109.64,
      "step": 290,
      "train_speed(iter/s)": 0.411959
    },
    {
      "acc": 0.69043927,
      "epoch": 0.006999227168666793,
      "grad_norm": 22.875,
      "learning_rate": 1.1664074650077762e-06,
      "loss": 1.24611559,
      "memory(GiB)": 109.77,
      "step": 300,
      "train_speed(iter/s)": 0.411594
    },
    {
      "acc": 0.68904781,
      "epoch": 0.007232534740955686,
      "grad_norm": 12.75,
      "learning_rate": 1.2052877138413686e-06,
      "loss": 1.20471973,
      "memory(GiB)": 109.77,
      "step": 310,
      "train_speed(iter/s)": 0.410459
    },
    {
      "acc": 0.6853941,
      "epoch": 0.007465842313244579,
      "grad_norm": 24.875,
      "learning_rate": 1.2441679626749613e-06,
      "loss": 1.23733406,
      "memory(GiB)": 109.77,
      "step": 320,
      "train_speed(iter/s)": 0.411169
    },
    {
      "acc": 0.69481087,
      "epoch": 0.007699149885533473,
      "grad_norm": 85.0,
      "learning_rate": 1.2830482115085538e-06,
      "loss": 1.23310585,
      "memory(GiB)": 109.77,
      "step": 330,
      "train_speed(iter/s)": 0.411728
    },
    {
      "acc": 0.69631901,
      "epoch": 0.007932457457822365,
      "grad_norm": 35.0,
      "learning_rate": 1.3219284603421462e-06,
      "loss": 1.18948269,
      "memory(GiB)": 109.77,
      "step": 340,
      "train_speed(iter/s)": 0.412145
    },
    {
      "acc": 0.66458092,
      "epoch": 0.00816576503011126,
      "grad_norm": 17.625,
      "learning_rate": 1.360808709175739e-06,
      "loss": 1.3200798,
      "memory(GiB)": 109.77,
      "step": 350,
      "train_speed(iter/s)": 0.411958
    },
    {
      "acc": 0.68819141,
      "epoch": 0.008399072602400151,
      "grad_norm": 32.5,
      "learning_rate": 1.3996889580093314e-06,
      "loss": 1.22151852,
      "memory(GiB)": 109.77,
      "step": 360,
      "train_speed(iter/s)": 0.412697
    },
    {
      "acc": 0.68463111,
      "epoch": 0.008632380174689045,
      "grad_norm": 38.0,
      "learning_rate": 1.4385692068429238e-06,
      "loss": 1.26127262,
      "memory(GiB)": 109.77,
      "step": 370,
      "train_speed(iter/s)": 0.413014
    },
    {
      "acc": 0.66406565,
      "epoch": 0.008865687746977938,
      "grad_norm": 21.75,
      "learning_rate": 1.4774494556765165e-06,
      "loss": 1.33156033,
      "memory(GiB)": 109.77,
      "step": 380,
      "train_speed(iter/s)": 0.412519
    },
    {
      "acc": 0.67032399,
      "epoch": 0.00909899531926683,
      "grad_norm": 22.875,
      "learning_rate": 1.5163297045101088e-06,
      "loss": 1.27387772,
      "memory(GiB)": 109.77,
      "step": 390,
      "train_speed(iter/s)": 0.412242
    },
    {
      "acc": 0.68596916,
      "epoch": 0.009332302891555724,
      "grad_norm": 15.0,
      "learning_rate": 1.5552099533437016e-06,
      "loss": 1.23029289,
      "memory(GiB)": 109.77,
      "step": 400,
      "train_speed(iter/s)": 0.41303
    },
    {
      "acc": 0.68200741,
      "epoch": 0.009565610463844618,
      "grad_norm": 19.75,
      "learning_rate": 1.594090202177294e-06,
      "loss": 1.23422031,
      "memory(GiB)": 109.77,
      "step": 410,
      "train_speed(iter/s)": 0.413378
    },
    {
      "acc": 0.68170118,
      "epoch": 0.00979891803613351,
      "grad_norm": 23.875,
      "learning_rate": 1.6329704510108864e-06,
      "loss": 1.24763041,
      "memory(GiB)": 109.77,
      "step": 420,
      "train_speed(iter/s)": 0.413317
    },
    {
      "acc": 0.68019838,
      "epoch": 0.010032225608422403,
      "grad_norm": 12.5625,
      "learning_rate": 1.6718506998444792e-06,
      "loss": 1.25324621,
      "memory(GiB)": 109.77,
      "step": 430,
      "train_speed(iter/s)": 0.413645
    },
    {
      "acc": 0.68601837,
      "epoch": 0.010265533180711297,
      "grad_norm": 22.125,
      "learning_rate": 1.7107309486780715e-06,
      "loss": 1.21224289,
      "memory(GiB)": 109.77,
      "step": 440,
      "train_speed(iter/s)": 0.413557
    },
    {
      "acc": 0.70243044,
      "epoch": 0.010498840753000189,
      "grad_norm": 32.0,
      "learning_rate": 1.7496111975116642e-06,
      "loss": 1.15849266,
      "memory(GiB)": 109.77,
      "step": 450,
      "train_speed(iter/s)": 0.412938
    },
    {
      "acc": 0.68498225,
      "epoch": 0.010732148325289083,
      "grad_norm": 21.0,
      "learning_rate": 1.7884914463452568e-06,
      "loss": 1.22883425,
      "memory(GiB)": 109.77,
      "step": 460,
      "train_speed(iter/s)": 0.413629
    },
    {
      "acc": 0.71473951,
      "epoch": 0.010965455897577976,
      "grad_norm": 27.75,
      "learning_rate": 1.8273716951788493e-06,
      "loss": 1.14569626,
      "memory(GiB)": 109.77,
      "step": 470,
      "train_speed(iter/s)": 0.413969
    },
    {
      "acc": 0.68733125,
      "epoch": 0.011198763469866868,
      "grad_norm": 17.75,
      "learning_rate": 1.8662519440124418e-06,
      "loss": 1.21703568,
      "memory(GiB)": 109.77,
      "step": 480,
      "train_speed(iter/s)": 0.414002
    },
    {
      "acc": 0.68442454,
      "epoch": 0.011432071042155762,
      "grad_norm": 29.875,
      "learning_rate": 1.9051321928460342e-06,
      "loss": 1.21695004,
      "memory(GiB)": 109.77,
      "step": 490,
      "train_speed(iter/s)": 0.414733
    },
    {
      "acc": 0.69048681,
      "epoch": 0.011665378614444655,
      "grad_norm": 39.5,
      "learning_rate": 1.9440124416796267e-06,
      "loss": 1.17031765,
      "memory(GiB)": 109.77,
      "step": 500,
      "train_speed(iter/s)": 0.414989
    },
    {
      "epoch": 0.011665378614444655,
      "eval_acc": 0.6771181596993516,
      "eval_loss": 1.170393943786621,
      "eval_runtime": 1270.1553,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 500
    },
    {
      "acc": 0.68483567,
      "epoch": 0.011898686186733547,
      "grad_norm": 24.375,
      "learning_rate": 1.9828926905132194e-06,
      "loss": 1.21425552,
      "memory(GiB)": 112.37,
      "step": 510,
      "train_speed(iter/s)": 0.20271
    },
    {
      "acc": 0.71478772,
      "epoch": 0.012131993759022441,
      "grad_norm": 40.75,
      "learning_rate": 2.0217729393468118e-06,
      "loss": 1.08218079,
      "memory(GiB)": 112.37,
      "step": 520,
      "train_speed(iter/s)": 0.204723
    },
    {
      "acc": 0.69291697,
      "epoch": 0.012365301331311335,
      "grad_norm": 13.3125,
      "learning_rate": 2.0606531881804045e-06,
      "loss": 1.18659,
      "memory(GiB)": 112.37,
      "step": 530,
      "train_speed(iter/s)": 0.206783
    },
    {
      "acc": 0.70176544,
      "epoch": 0.012598608903600227,
      "grad_norm": 10.8125,
      "learning_rate": 2.0995334370139973e-06,
      "loss": 1.15439968,
      "memory(GiB)": 112.37,
      "step": 540,
      "train_speed(iter/s)": 0.208567
    },
    {
      "acc": 0.70656576,
      "epoch": 0.01283191647588912,
      "grad_norm": 24.75,
      "learning_rate": 2.1384136858475896e-06,
      "loss": 1.13584023,
      "memory(GiB)": 112.37,
      "step": 550,
      "train_speed(iter/s)": 0.210478
    },
    {
      "acc": 0.69827976,
      "epoch": 0.013065224048178014,
      "grad_norm": 18.375,
      "learning_rate": 2.1772939346811823e-06,
      "loss": 1.16824875,
      "memory(GiB)": 112.37,
      "step": 560,
      "train_speed(iter/s)": 0.212251
    },
    {
      "acc": 0.69722204,
      "epoch": 0.013298531620466908,
      "grad_norm": 20.625,
      "learning_rate": 2.2161741835147746e-06,
      "loss": 1.17354527,
      "memory(GiB)": 112.37,
      "step": 570,
      "train_speed(iter/s)": 0.214144
    },
    {
      "acc": 0.69391127,
      "epoch": 0.0135318391927558,
      "grad_norm": 12.25,
      "learning_rate": 2.2550544323483674e-06,
      "loss": 1.18383684,
      "memory(GiB)": 112.37,
      "step": 580,
      "train_speed(iter/s)": 0.215836
    },
    {
      "acc": 0.7008687,
      "epoch": 0.013765146765044693,
      "grad_norm": 21.125,
      "learning_rate": 2.2939346811819597e-06,
      "loss": 1.13180761,
      "memory(GiB)": 112.37,
      "step": 590,
      "train_speed(iter/s)": 0.217548
    },
    {
      "acc": 0.70401969,
      "epoch": 0.013998454337333587,
      "grad_norm": 14.1875,
      "learning_rate": 2.3328149300155525e-06,
      "loss": 1.14089346,
      "memory(GiB)": 112.37,
      "step": 600,
      "train_speed(iter/s)": 0.219209
    },
    {
      "acc": 0.67588034,
      "epoch": 0.014231761909622479,
      "grad_norm": 152.0,
      "learning_rate": 2.3716951788491448e-06,
      "loss": 1.282197,
      "memory(GiB)": 112.37,
      "step": 610,
      "train_speed(iter/s)": 0.220998
    },
    {
      "acc": 0.71631866,
      "epoch": 0.014465069481911372,
      "grad_norm": 10.0,
      "learning_rate": 2.410575427682737e-06,
      "loss": 1.09193068,
      "memory(GiB)": 112.37,
      "step": 620,
      "train_speed(iter/s)": 0.22271
    },
    {
      "acc": 0.70892181,
      "epoch": 0.014698377054200266,
      "grad_norm": 21.625,
      "learning_rate": 2.44945567651633e-06,
      "loss": 1.10817814,
      "memory(GiB)": 112.37,
      "step": 630,
      "train_speed(iter/s)": 0.224373
    },
    {
      "acc": 0.71971693,
      "epoch": 0.014931684626489158,
      "grad_norm": 22.75,
      "learning_rate": 2.4883359253499226e-06,
      "loss": 1.06725435,
      "memory(GiB)": 112.37,
      "step": 640,
      "train_speed(iter/s)": 0.226002
    },
    {
      "acc": 0.6958858,
      "epoch": 0.015164992198778052,
      "grad_norm": 13.5,
      "learning_rate": 2.527216174183515e-06,
      "loss": 1.14539852,
      "memory(GiB)": 112.37,
      "step": 650,
      "train_speed(iter/s)": 0.227501
    },
    {
      "acc": 0.7064106,
      "epoch": 0.015398299771066945,
      "grad_norm": 14.875,
      "learning_rate": 2.5660964230171077e-06,
      "loss": 1.09452753,
      "memory(GiB)": 112.37,
      "step": 660,
      "train_speed(iter/s)": 0.229076
    },
    {
      "acc": 0.68724928,
      "epoch": 0.01563160734335584,
      "grad_norm": 10.125,
      "learning_rate": 2.6049766718507004e-06,
      "loss": 1.17633553,
      "memory(GiB)": 112.37,
      "step": 670,
      "train_speed(iter/s)": 0.230642
    },
    {
      "acc": 0.71898904,
      "epoch": 0.01586491491564473,
      "grad_norm": 20.25,
      "learning_rate": 2.6438569206842923e-06,
      "loss": 1.08943672,
      "memory(GiB)": 112.37,
      "step": 680,
      "train_speed(iter/s)": 0.232146
    },
    {
      "acc": 0.70866594,
      "epoch": 0.016098222487933623,
      "grad_norm": 41.25,
      "learning_rate": 2.682737169517885e-06,
      "loss": 1.1266736,
      "memory(GiB)": 112.37,
      "step": 690,
      "train_speed(iter/s)": 0.233723
    },
    {
      "acc": 0.69511156,
      "epoch": 0.01633153006022252,
      "grad_norm": 16.75,
      "learning_rate": 2.721617418351478e-06,
      "loss": 1.15227604,
      "memory(GiB)": 112.37,
      "step": 700,
      "train_speed(iter/s)": 0.235191
    },
    {
      "acc": 0.69382067,
      "epoch": 0.01656483763251141,
      "grad_norm": 13.875,
      "learning_rate": 2.76049766718507e-06,
      "loss": 1.16943235,
      "memory(GiB)": 112.37,
      "step": 710,
      "train_speed(iter/s)": 0.236655
    },
    {
      "acc": 0.70306535,
      "epoch": 0.016798145204800302,
      "grad_norm": 55.5,
      "learning_rate": 2.799377916018663e-06,
      "loss": 1.11830854,
      "memory(GiB)": 112.5,
      "step": 720,
      "train_speed(iter/s)": 0.238078
    },
    {
      "acc": 0.7088171,
      "epoch": 0.017031452777089198,
      "grad_norm": 13.875,
      "learning_rate": 2.838258164852255e-06,
      "loss": 1.11293411,
      "memory(GiB)": 112.5,
      "step": 730,
      "train_speed(iter/s)": 0.239489
    },
    {
      "acc": 0.73532681,
      "epoch": 0.01726476034937809,
      "grad_norm": 15.375,
      "learning_rate": 2.8771384136858475e-06,
      "loss": 1.01635103,
      "memory(GiB)": 112.5,
      "step": 740,
      "train_speed(iter/s)": 0.240802
    },
    {
      "acc": 0.73056784,
      "epoch": 0.01749806792166698,
      "grad_norm": 8.6875,
      "learning_rate": 2.9160186625194403e-06,
      "loss": 1.03644848,
      "memory(GiB)": 112.5,
      "step": 750,
      "train_speed(iter/s)": 0.24216
    },
    {
      "acc": 0.71434994,
      "epoch": 0.017731375493955877,
      "grad_norm": 14.0625,
      "learning_rate": 2.954898911353033e-06,
      "loss": 1.09710608,
      "memory(GiB)": 112.5,
      "step": 760,
      "train_speed(iter/s)": 0.24334
    },
    {
      "acc": 0.71459484,
      "epoch": 0.01796468306624477,
      "grad_norm": 11.125,
      "learning_rate": 2.9937791601866257e-06,
      "loss": 1.10940857,
      "memory(GiB)": 112.5,
      "step": 770,
      "train_speed(iter/s)": 0.24472
    },
    {
      "acc": 0.72359238,
      "epoch": 0.01819799063853366,
      "grad_norm": 10.6875,
      "learning_rate": 3.0326594090202176e-06,
      "loss": 1.06084309,
      "memory(GiB)": 112.5,
      "step": 780,
      "train_speed(iter/s)": 0.246022
    },
    {
      "acc": 0.71572514,
      "epoch": 0.018431298210822556,
      "grad_norm": 11.8125,
      "learning_rate": 3.0715396578538104e-06,
      "loss": 1.07304935,
      "memory(GiB)": 112.5,
      "step": 790,
      "train_speed(iter/s)": 0.247402
    },
    {
      "acc": 0.73548241,
      "epoch": 0.018664605783111448,
      "grad_norm": 9.625,
      "learning_rate": 3.110419906687403e-06,
      "loss": 1.0010951,
      "memory(GiB)": 112.5,
      "step": 800,
      "train_speed(iter/s)": 0.248693
    },
    {
      "acc": 0.74713941,
      "epoch": 0.01889791335540034,
      "grad_norm": 58.0,
      "learning_rate": 3.1493001555209955e-06,
      "loss": 0.96250553,
      "memory(GiB)": 112.5,
      "step": 810,
      "train_speed(iter/s)": 0.249902
    },
    {
      "acc": 0.72171469,
      "epoch": 0.019131220927689235,
      "grad_norm": 25.0,
      "learning_rate": 3.188180404354588e-06,
      "loss": 1.06542034,
      "memory(GiB)": 112.5,
      "step": 820,
      "train_speed(iter/s)": 0.251022
    },
    {
      "acc": 0.72154565,
      "epoch": 0.019364528499978127,
      "grad_norm": 10.375,
      "learning_rate": 3.2270606531881805e-06,
      "loss": 1.06000462,
      "memory(GiB)": 112.5,
      "step": 830,
      "train_speed(iter/s)": 0.252274
    },
    {
      "acc": 0.72352486,
      "epoch": 0.01959783607226702,
      "grad_norm": 15.6875,
      "learning_rate": 3.265940902021773e-06,
      "loss": 1.07526264,
      "memory(GiB)": 112.5,
      "step": 840,
      "train_speed(iter/s)": 0.253342
    },
    {
      "acc": 0.70478067,
      "epoch": 0.019831143644555915,
      "grad_norm": 30.75,
      "learning_rate": 3.3048211508553656e-06,
      "loss": 1.0926899,
      "memory(GiB)": 112.5,
      "step": 850,
      "train_speed(iter/s)": 0.254406
    },
    {
      "acc": 0.7308609,
      "epoch": 0.020064451216844807,
      "grad_norm": 9.375,
      "learning_rate": 3.3437013996889583e-06,
      "loss": 1.01046867,
      "memory(GiB)": 112.5,
      "step": 860,
      "train_speed(iter/s)": 0.255641
    },
    {
      "acc": 0.69665556,
      "epoch": 0.0202977587891337,
      "grad_norm": 29.125,
      "learning_rate": 3.382581648522551e-06,
      "loss": 1.16045341,
      "memory(GiB)": 112.5,
      "step": 870,
      "train_speed(iter/s)": 0.256795
    },
    {
      "acc": 0.72442212,
      "epoch": 0.020531066361422594,
      "grad_norm": 19.875,
      "learning_rate": 3.421461897356143e-06,
      "loss": 1.05415115,
      "memory(GiB)": 112.5,
      "step": 880,
      "train_speed(iter/s)": 0.2579
    },
    {
      "acc": 0.71560249,
      "epoch": 0.020764373933711486,
      "grad_norm": 17.5,
      "learning_rate": 3.4603421461897357e-06,
      "loss": 1.08642292,
      "memory(GiB)": 112.5,
      "step": 890,
      "train_speed(iter/s)": 0.259042
    },
    {
      "acc": 0.7090786,
      "epoch": 0.020997681506000378,
      "grad_norm": 20.5,
      "learning_rate": 3.4992223950233285e-06,
      "loss": 1.12236042,
      "memory(GiB)": 112.5,
      "step": 900,
      "train_speed(iter/s)": 0.260196
    },
    {
      "acc": 0.72494149,
      "epoch": 0.021230989078289273,
      "grad_norm": 14.125,
      "learning_rate": 3.5381026438569212e-06,
      "loss": 1.06427841,
      "memory(GiB)": 112.5,
      "step": 910,
      "train_speed(iter/s)": 0.261275
    },
    {
      "acc": 0.73181691,
      "epoch": 0.021464296650578165,
      "grad_norm": 10.3125,
      "learning_rate": 3.5769828926905135e-06,
      "loss": 1.00808945,
      "memory(GiB)": 112.5,
      "step": 920,
      "train_speed(iter/s)": 0.26246
    },
    {
      "acc": 0.70418386,
      "epoch": 0.021697604222867057,
      "grad_norm": 9.0625,
      "learning_rate": 3.615863141524106e-06,
      "loss": 1.1433609,
      "memory(GiB)": 112.5,
      "step": 930,
      "train_speed(iter/s)": 0.263556
    },
    {
      "acc": 0.72886324,
      "epoch": 0.021930911795155952,
      "grad_norm": 6.9375,
      "learning_rate": 3.6547433903576986e-06,
      "loss": 1.02939377,
      "memory(GiB)": 112.5,
      "step": 940,
      "train_speed(iter/s)": 0.264494
    },
    {
      "acc": 0.71419945,
      "epoch": 0.022164219367444844,
      "grad_norm": 16.625,
      "learning_rate": 3.693623639191291e-06,
      "loss": 1.06986065,
      "memory(GiB)": 112.5,
      "step": 950,
      "train_speed(iter/s)": 0.265441
    },
    {
      "acc": 0.73321857,
      "epoch": 0.022397526939733736,
      "grad_norm": 14.3125,
      "learning_rate": 3.7325038880248837e-06,
      "loss": 1.01707897,
      "memory(GiB)": 112.5,
      "step": 960,
      "train_speed(iter/s)": 0.266415
    },
    {
      "acc": 0.70803647,
      "epoch": 0.02263083451202263,
      "grad_norm": 22.5,
      "learning_rate": 3.7713841368584764e-06,
      "loss": 1.15186424,
      "memory(GiB)": 112.5,
      "step": 970,
      "train_speed(iter/s)": 0.267359
    },
    {
      "acc": 0.7166934,
      "epoch": 0.022864142084311524,
      "grad_norm": 11.5,
      "learning_rate": 3.8102643856920683e-06,
      "loss": 1.08353691,
      "memory(GiB)": 112.5,
      "step": 980,
      "train_speed(iter/s)": 0.268331
    },
    {
      "acc": 0.75482187,
      "epoch": 0.023097449656600415,
      "grad_norm": 11.125,
      "learning_rate": 3.849144634525661e-06,
      "loss": 0.92038155,
      "memory(GiB)": 112.5,
      "step": 990,
      "train_speed(iter/s)": 0.269254
    },
    {
      "acc": 0.72072606,
      "epoch": 0.02333075722888931,
      "grad_norm": 17.375,
      "learning_rate": 3.888024883359253e-06,
      "loss": 1.032094,
      "memory(GiB)": 112.5,
      "step": 1000,
      "train_speed(iter/s)": 0.270211
    },
    {
      "epoch": 0.02333075722888931,
      "eval_acc": 0.6944079715955378,
      "eval_loss": 1.03887140750885,
      "eval_runtime": 1269.799,
      "eval_samples_per_second": 28.344,
      "eval_steps_per_second": 14.172,
      "step": 1000
    },
    {
      "acc": 0.73672218,
      "epoch": 0.023564064801178203,
      "grad_norm": 11.6875,
      "learning_rate": 3.9269051321928466e-06,
      "loss": 0.97621746,
      "memory(GiB)": 112.5,
      "step": 1010,
      "train_speed(iter/s)": 0.201516
    },
    {
      "acc": 0.72829971,
      "epoch": 0.023797372373467095,
      "grad_norm": 12.625,
      "learning_rate": 3.965785381026439e-06,
      "loss": 1.04721756,
      "memory(GiB)": 112.5,
      "step": 1020,
      "train_speed(iter/s)": 0.202516
    },
    {
      "acc": 0.74315853,
      "epoch": 0.02403067994575599,
      "grad_norm": 12.0625,
      "learning_rate": 4.004665629860031e-06,
      "loss": 0.96440945,
      "memory(GiB)": 112.5,
      "step": 1030,
      "train_speed(iter/s)": 0.203553
    },
    {
      "acc": 0.71699724,
      "epoch": 0.024263987518044882,
      "grad_norm": 10.4375,
      "learning_rate": 4.0435458786936235e-06,
      "loss": 1.05050545,
      "memory(GiB)": 112.5,
      "step": 1040,
      "train_speed(iter/s)": 0.20455
    },
    {
      "acc": 0.73913536,
      "epoch": 0.024497295090333774,
      "grad_norm": 56.75,
      "learning_rate": 4.082426127527217e-06,
      "loss": 0.98256474,
      "memory(GiB)": 112.5,
      "step": 1050,
      "train_speed(iter/s)": 0.205477
    },
    {
      "acc": 0.72385454,
      "epoch": 0.02473060266262267,
      "grad_norm": 10.1875,
      "learning_rate": 4.121306376360809e-06,
      "loss": 1.03298187,
      "memory(GiB)": 112.5,
      "step": 1060,
      "train_speed(iter/s)": 0.206437
    },
    {
      "acc": 0.73873882,
      "epoch": 0.02496391023491156,
      "grad_norm": 13.625,
      "learning_rate": 4.160186625194401e-06,
      "loss": 1.0152791,
      "memory(GiB)": 112.5,
      "step": 1070,
      "train_speed(iter/s)": 0.207325
    },
    {
      "acc": 0.73708954,
      "epoch": 0.025197217807200453,
      "grad_norm": 12.3125,
      "learning_rate": 4.1990668740279945e-06,
      "loss": 0.98759651,
      "memory(GiB)": 112.5,
      "step": 1080,
      "train_speed(iter/s)": 0.208333
    },
    {
      "acc": 0.71818414,
      "epoch": 0.02543052537948935,
      "grad_norm": 65.0,
      "learning_rate": 4.237947122861587e-06,
      "loss": 1.0364994,
      "memory(GiB)": 112.5,
      "step": 1090,
      "train_speed(iter/s)": 0.209286
    },
    {
      "acc": 0.7241076,
      "epoch": 0.02566383295177824,
      "grad_norm": 16.5,
      "learning_rate": 4.276827371695179e-06,
      "loss": 1.03190594,
      "memory(GiB)": 112.5,
      "step": 1100,
      "train_speed(iter/s)": 0.210209
    },
    {
      "acc": 0.72726188,
      "epoch": 0.025897140524067136,
      "grad_norm": 18.25,
      "learning_rate": 4.3157076205287715e-06,
      "loss": 1.02791977,
      "memory(GiB)": 112.5,
      "step": 1110,
      "train_speed(iter/s)": 0.211143
    },
    {
      "acc": 0.73693681,
      "epoch": 0.026130448096356028,
      "grad_norm": 8.125,
      "learning_rate": 4.354587869362365e-06,
      "loss": 1.01167622,
      "memory(GiB)": 112.5,
      "step": 1120,
      "train_speed(iter/s)": 0.212056
    },
    {
      "acc": 0.73370867,
      "epoch": 0.02636375566864492,
      "grad_norm": 8.3125,
      "learning_rate": 4.393468118195957e-06,
      "loss": 1.00605249,
      "memory(GiB)": 112.5,
      "step": 1130,
      "train_speed(iter/s)": 0.212952
    },
    {
      "acc": 0.73210607,
      "epoch": 0.026597063240933815,
      "grad_norm": 15.8125,
      "learning_rate": 4.432348367029549e-06,
      "loss": 0.99587374,
      "memory(GiB)": 112.5,
      "step": 1140,
      "train_speed(iter/s)": 0.213882
    },
    {
      "acc": 0.73727956,
      "epoch": 0.026830370813222707,
      "grad_norm": 15.5,
      "learning_rate": 4.471228615863142e-06,
      "loss": 0.9840374,
      "memory(GiB)": 112.5,
      "step": 1150,
      "train_speed(iter/s)": 0.214805
    },
    {
      "acc": 0.73756843,
      "epoch": 0.0270636783855116,
      "grad_norm": 13.6875,
      "learning_rate": 4.510108864696735e-06,
      "loss": 1.0077363,
      "memory(GiB)": 112.5,
      "step": 1160,
      "train_speed(iter/s)": 0.215706
    },
    {
      "acc": 0.71364336,
      "epoch": 0.027296985957800494,
      "grad_norm": 9.375,
      "learning_rate": 4.548989113530327e-06,
      "loss": 1.07981796,
      "memory(GiB)": 112.5,
      "step": 1170,
      "train_speed(iter/s)": 0.216546
    },
    {
      "acc": 0.7402174,
      "epoch": 0.027530293530089386,
      "grad_norm": 6.34375,
      "learning_rate": 4.587869362363919e-06,
      "loss": 0.96098614,
      "memory(GiB)": 112.5,
      "step": 1180,
      "train_speed(iter/s)": 0.217392
    },
    {
      "acc": 0.72905722,
      "epoch": 0.02776360110237828,
      "grad_norm": 12.6875,
      "learning_rate": 4.626749611197512e-06,
      "loss": 1.03549309,
      "memory(GiB)": 112.5,
      "step": 1190,
      "train_speed(iter/s)": 0.218326
    },
    {
      "acc": 0.71565552,
      "epoch": 0.027996908674667174,
      "grad_norm": 6.21875,
      "learning_rate": 4.665629860031105e-06,
      "loss": 1.08136559,
      "memory(GiB)": 112.5,
      "step": 1200,
      "train_speed(iter/s)": 0.2192
    },
    {
      "acc": 0.75236349,
      "epoch": 0.028230216246956066,
      "grad_norm": 8.9375,
      "learning_rate": 4.704510108864697e-06,
      "loss": 0.95540047,
      "memory(GiB)": 112.5,
      "step": 1210,
      "train_speed(iter/s)": 0.220119
    },
    {
      "acc": 0.72307405,
      "epoch": 0.028463523819244958,
      "grad_norm": 10.875,
      "learning_rate": 4.7433903576982896e-06,
      "loss": 1.01828804,
      "memory(GiB)": 112.5,
      "step": 1220,
      "train_speed(iter/s)": 0.221032
    },
    {
      "acc": 0.71835833,
      "epoch": 0.028696831391533853,
      "grad_norm": 11.75,
      "learning_rate": 4.782270606531883e-06,
      "loss": 1.05052891,
      "memory(GiB)": 112.5,
      "step": 1230,
      "train_speed(iter/s)": 0.221873
    },
    {
      "acc": 0.72516937,
      "epoch": 0.028930138963822745,
      "grad_norm": 8.9375,
      "learning_rate": 4.821150855365474e-06,
      "loss": 1.02222395,
      "memory(GiB)": 112.5,
      "step": 1240,
      "train_speed(iter/s)": 0.222722
    },
    {
      "acc": 0.72957277,
      "epoch": 0.029163446536111637,
      "grad_norm": 10.1875,
      "learning_rate": 4.860031104199067e-06,
      "loss": 1.01317539,
      "memory(GiB)": 112.5,
      "step": 1250,
      "train_speed(iter/s)": 0.223525
    },
    {
      "acc": 0.74022655,
      "epoch": 0.029396754108400532,
      "grad_norm": 8.875,
      "learning_rate": 4.89891135303266e-06,
      "loss": 0.95849752,
      "memory(GiB)": 112.5,
      "step": 1260,
      "train_speed(iter/s)": 0.224384
    },
    {
      "acc": 0.7379168,
      "epoch": 0.029630061680689424,
      "grad_norm": 9.1875,
      "learning_rate": 4.937791601866253e-06,
      "loss": 0.96695614,
      "memory(GiB)": 112.5,
      "step": 1270,
      "train_speed(iter/s)": 0.225256
    },
    {
      "acc": 0.7323555,
      "epoch": 0.029863369252978316,
      "grad_norm": 8.125,
      "learning_rate": 4.976671850699845e-06,
      "loss": 1.0258522,
      "memory(GiB)": 112.5,
      "step": 1280,
      "train_speed(iter/s)": 0.226036
    },
    {
      "acc": 0.74023056,
      "epoch": 0.03009667682526721,
      "grad_norm": 11.1875,
      "learning_rate": 5.0155520995334375e-06,
      "loss": 0.97798119,
      "memory(GiB)": 112.5,
      "step": 1290,
      "train_speed(iter/s)": 0.226849
    },
    {
      "acc": 0.74629755,
      "epoch": 0.030329984397556103,
      "grad_norm": 10.0625,
      "learning_rate": 5.05443234836703e-06,
      "loss": 0.95969715,
      "memory(GiB)": 112.5,
      "step": 1300,
      "train_speed(iter/s)": 0.227679
    },
    {
      "acc": 0.74427795,
      "epoch": 0.030563291969844995,
      "grad_norm": 11.75,
      "learning_rate": 5.093312597200622e-06,
      "loss": 0.94915905,
      "memory(GiB)": 112.5,
      "step": 1310,
      "train_speed(iter/s)": 0.228474
    },
    {
      "acc": 0.73095655,
      "epoch": 0.03079659954213389,
      "grad_norm": 12.375,
      "learning_rate": 5.132192846034215e-06,
      "loss": 1.00967436,
      "memory(GiB)": 112.5,
      "step": 1320,
      "train_speed(iter/s)": 0.229243
    },
    {
      "acc": 0.74197416,
      "epoch": 0.031029907114422783,
      "grad_norm": 10.0625,
      "learning_rate": 5.171073094867808e-06,
      "loss": 0.95410166,
      "memory(GiB)": 112.5,
      "step": 1330,
      "train_speed(iter/s)": 0.230016
    },
    {
      "acc": 0.72437105,
      "epoch": 0.03126321468671168,
      "grad_norm": 9.6875,
      "learning_rate": 5.209953343701401e-06,
      "loss": 1.02081079,
      "memory(GiB)": 112.5,
      "step": 1340,
      "train_speed(iter/s)": 0.230814
    },
    {
      "acc": 0.71890373,
      "epoch": 0.03149652225900057,
      "grad_norm": 7.25,
      "learning_rate": 5.248833592534993e-06,
      "loss": 1.05639267,
      "memory(GiB)": 112.5,
      "step": 1350,
      "train_speed(iter/s)": 0.231593
    },
    {
      "acc": 0.73581686,
      "epoch": 0.03172982983128946,
      "grad_norm": 26.375,
      "learning_rate": 5.287713841368585e-06,
      "loss": 0.99747047,
      "memory(GiB)": 112.5,
      "step": 1360,
      "train_speed(iter/s)": 0.232336
    },
    {
      "acc": 0.73829999,
      "epoch": 0.031963137403578354,
      "grad_norm": 13.9375,
      "learning_rate": 5.326594090202177e-06,
      "loss": 0.97000275,
      "memory(GiB)": 112.5,
      "step": 1370,
      "train_speed(iter/s)": 0.233092
    },
    {
      "acc": 0.73752613,
      "epoch": 0.032196444975867246,
      "grad_norm": 9.25,
      "learning_rate": 5.36547433903577e-06,
      "loss": 0.97096643,
      "memory(GiB)": 112.5,
      "step": 1380,
      "train_speed(iter/s)": 0.23385
    },
    {
      "acc": 0.73803453,
      "epoch": 0.03242975254815614,
      "grad_norm": 8.0,
      "learning_rate": 5.404354587869362e-06,
      "loss": 0.97089777,
      "memory(GiB)": 112.5,
      "step": 1390,
      "train_speed(iter/s)": 0.234563
    },
    {
      "acc": 0.73658538,
      "epoch": 0.03266306012044504,
      "grad_norm": 12.5625,
      "learning_rate": 5.443234836702956e-06,
      "loss": 0.99308605,
      "memory(GiB)": 112.5,
      "step": 1400,
      "train_speed(iter/s)": 0.235338
    },
    {
      "acc": 0.71785216,
      "epoch": 0.03289636769273393,
      "grad_norm": 8.3125,
      "learning_rate": 5.482115085536548e-06,
      "loss": 1.07402363,
      "memory(GiB)": 112.5,
      "step": 1410,
      "train_speed(iter/s)": 0.236091
    },
    {
      "acc": 0.74998498,
      "epoch": 0.03312967526502282,
      "grad_norm": 8.25,
      "learning_rate": 5.52099533437014e-06,
      "loss": 0.93102665,
      "memory(GiB)": 112.5,
      "step": 1420,
      "train_speed(iter/s)": 0.236812
    },
    {
      "acc": 0.7163455,
      "epoch": 0.03336298283731171,
      "grad_norm": 13.5,
      "learning_rate": 5.559875583203733e-06,
      "loss": 1.08192539,
      "memory(GiB)": 112.5,
      "step": 1430,
      "train_speed(iter/s)": 0.237497
    },
    {
      "acc": 0.74807477,
      "epoch": 0.033596290409600604,
      "grad_norm": 17.125,
      "learning_rate": 5.598755832037326e-06,
      "loss": 0.94381409,
      "memory(GiB)": 112.5,
      "step": 1440,
      "train_speed(iter/s)": 0.238196
    },
    {
      "acc": 0.72739019,
      "epoch": 0.033829597981889496,
      "grad_norm": 9.0,
      "learning_rate": 5.637636080870919e-06,
      "loss": 1.02746716,
      "memory(GiB)": 112.5,
      "step": 1450,
      "train_speed(iter/s)": 0.238917
    },
    {
      "acc": 0.72795725,
      "epoch": 0.034062905554178395,
      "grad_norm": 9.1875,
      "learning_rate": 5.67651632970451e-06,
      "loss": 1.00883026,
      "memory(GiB)": 112.5,
      "step": 1460,
      "train_speed(iter/s)": 0.239627
    },
    {
      "acc": 0.74547176,
      "epoch": 0.03429621312646729,
      "grad_norm": 20.125,
      "learning_rate": 5.715396578538103e-06,
      "loss": 0.94491329,
      "memory(GiB)": 112.5,
      "step": 1470,
      "train_speed(iter/s)": 0.240273
    },
    {
      "acc": 0.76093078,
      "epoch": 0.03452952069875618,
      "grad_norm": 9.4375,
      "learning_rate": 5.754276827371695e-06,
      "loss": 0.89414454,
      "memory(GiB)": 112.5,
      "step": 1480,
      "train_speed(iter/s)": 0.240967
    },
    {
      "acc": 0.71505122,
      "epoch": 0.03476282827104507,
      "grad_norm": 11.6875,
      "learning_rate": 5.793157076205288e-06,
      "loss": 1.07852478,
      "memory(GiB)": 112.5,
      "step": 1490,
      "train_speed(iter/s)": 0.241618
    },
    {
      "acc": 0.73277426,
      "epoch": 0.03499613584333396,
      "grad_norm": 11.875,
      "learning_rate": 5.8320373250388805e-06,
      "loss": 0.99803505,
      "memory(GiB)": 112.5,
      "step": 1500,
      "train_speed(iter/s)": 0.242283
    },
    {
      "epoch": 0.03499613584333396,
      "eval_acc": 0.7026027121066317,
      "eval_loss": 0.9811294078826904,
      "eval_runtime": 1269.74,
      "eval_samples_per_second": 28.345,
      "eval_steps_per_second": 14.173,
      "step": 1500
    },
    {
      "acc": 0.73681936,
      "epoch": 0.03522944341562286,
      "grad_norm": 7.21875,
      "learning_rate": 5.870917573872474e-06,
      "loss": 1.00402145,
      "memory(GiB)": 112.5,
      "step": 1510,
      "train_speed(iter/s)": 0.201315
    },
    {
      "acc": 0.75371494,
      "epoch": 0.035462750987911754,
      "grad_norm": 13.0625,
      "learning_rate": 5.909797822706066e-06,
      "loss": 0.91483002,
      "memory(GiB)": 112.5,
      "step": 1520,
      "train_speed(iter/s)": 0.201999
    },
    {
      "acc": 0.73148508,
      "epoch": 0.035696058560200646,
      "grad_norm": 7.3125,
      "learning_rate": 5.948678071539658e-06,
      "loss": 1.00952492,
      "memory(GiB)": 112.5,
      "step": 1530,
      "train_speed(iter/s)": 0.202708
    },
    {
      "acc": 0.74506388,
      "epoch": 0.03592936613248954,
      "grad_norm": 10.3125,
      "learning_rate": 5.9875583203732515e-06,
      "loss": 0.94095707,
      "memory(GiB)": 112.5,
      "step": 1540,
      "train_speed(iter/s)": 0.20341
    },
    {
      "acc": 0.7385951,
      "epoch": 0.03616267370477843,
      "grad_norm": 9.9375,
      "learning_rate": 6.026438569206844e-06,
      "loss": 0.96724091,
      "memory(GiB)": 114.08,
      "step": 1550,
      "train_speed(iter/s)": 0.204074
    },
    {
      "acc": 0.73922119,
      "epoch": 0.03639598127706732,
      "grad_norm": 17.375,
      "learning_rate": 6.065318818040435e-06,
      "loss": 0.96251907,
      "memory(GiB)": 114.08,
      "step": 1560,
      "train_speed(iter/s)": 0.204769
    },
    {
      "acc": 0.72654696,
      "epoch": 0.03662928884935622,
      "grad_norm": 10.5625,
      "learning_rate": 6.1041990668740285e-06,
      "loss": 1.03361921,
      "memory(GiB)": 114.08,
      "step": 1570,
      "train_speed(iter/s)": 0.20549
    },
    {
      "acc": 0.72640285,
      "epoch": 0.03686259642164511,
      "grad_norm": 17.25,
      "learning_rate": 6.143079315707621e-06,
      "loss": 1.04433689,
      "memory(GiB)": 114.08,
      "step": 1580,
      "train_speed(iter/s)": 0.20617
    },
    {
      "acc": 0.75693703,
      "epoch": 0.037095903993934004,
      "grad_norm": 7.75,
      "learning_rate": 6.181959564541213e-06,
      "loss": 0.89104633,
      "memory(GiB)": 114.08,
      "step": 1590,
      "train_speed(iter/s)": 0.206865
    },
    {
      "acc": 0.74660778,
      "epoch": 0.037329211566222896,
      "grad_norm": 13.375,
      "learning_rate": 6.220839813374806e-06,
      "loss": 0.92481155,
      "memory(GiB)": 114.08,
      "step": 1600,
      "train_speed(iter/s)": 0.207546
    },
    {
      "acc": 0.72870965,
      "epoch": 0.03756251913851179,
      "grad_norm": 7.21875,
      "learning_rate": 6.259720062208399e-06,
      "loss": 1.03226433,
      "memory(GiB)": 114.08,
      "step": 1610,
      "train_speed(iter/s)": 0.208219
    },
    {
      "acc": 0.72618661,
      "epoch": 0.03779582671080068,
      "grad_norm": 9.0,
      "learning_rate": 6.298600311041991e-06,
      "loss": 1.02390461,
      "memory(GiB)": 114.08,
      "step": 1620,
      "train_speed(iter/s)": 0.208905
    },
    {
      "acc": 0.72518511,
      "epoch": 0.03802913428308958,
      "grad_norm": 11.0625,
      "learning_rate": 6.337480559875584e-06,
      "loss": 1.02645645,
      "memory(GiB)": 114.08,
      "step": 1630,
      "train_speed(iter/s)": 0.209512
    },
    {
      "acc": 0.74411564,
      "epoch": 0.03826244185537847,
      "grad_norm": 7.28125,
      "learning_rate": 6.376360808709176e-06,
      "loss": 0.95807343,
      "memory(GiB)": 114.08,
      "step": 1640,
      "train_speed(iter/s)": 0.21016
    },
    {
      "acc": 0.73625383,
      "epoch": 0.03849574942766736,
      "grad_norm": 8.625,
      "learning_rate": 6.4152410575427696e-06,
      "loss": 0.96755209,
      "memory(GiB)": 114.08,
      "step": 1650,
      "train_speed(iter/s)": 0.210758
    },
    {
      "acc": 0.72957053,
      "epoch": 0.038729056999956255,
      "grad_norm": 9.625,
      "learning_rate": 6.454121306376361e-06,
      "loss": 0.99150848,
      "memory(GiB)": 114.08,
      "step": 1660,
      "train_speed(iter/s)": 0.211372
    },
    {
      "acc": 0.74594126,
      "epoch": 0.038962364572245146,
      "grad_norm": 9.1875,
      "learning_rate": 6.493001555209953e-06,
      "loss": 0.92405319,
      "memory(GiB)": 114.08,
      "step": 1670,
      "train_speed(iter/s)": 0.212017
    },
    {
      "acc": 0.73622656,
      "epoch": 0.03919567214453404,
      "grad_norm": 10.3125,
      "learning_rate": 6.531881804043546e-06,
      "loss": 0.98413754,
      "memory(GiB)": 114.08,
      "step": 1680,
      "train_speed(iter/s)": 0.212643
    },
    {
      "acc": 0.72955575,
      "epoch": 0.03942897971682294,
      "grad_norm": 11.125,
      "learning_rate": 6.570762052877139e-06,
      "loss": 1.01020985,
      "memory(GiB)": 114.08,
      "step": 1690,
      "train_speed(iter/s)": 0.21329
    },
    {
      "acc": 0.74046488,
      "epoch": 0.03966228728911183,
      "grad_norm": 10.1875,
      "learning_rate": 6.609642301710731e-06,
      "loss": 0.96516933,
      "memory(GiB)": 114.08,
      "step": 1700,
      "train_speed(iter/s)": 0.21388
    },
    {
      "acc": 0.74028692,
      "epoch": 0.03989559486140072,
      "grad_norm": 7.15625,
      "learning_rate": 6.648522550544324e-06,
      "loss": 0.97764168,
      "memory(GiB)": 114.08,
      "step": 1710,
      "train_speed(iter/s)": 0.214504
    },
    {
      "acc": 0.74346085,
      "epoch": 0.04012890243368961,
      "grad_norm": 7.1875,
      "learning_rate": 6.687402799377917e-06,
      "loss": 0.95602665,
      "memory(GiB)": 114.08,
      "step": 1720,
      "train_speed(iter/s)": 0.215093
    },
    {
      "acc": 0.74011574,
      "epoch": 0.040362210005978505,
      "grad_norm": 18.0,
      "learning_rate": 6.726283048211509e-06,
      "loss": 0.96684675,
      "memory(GiB)": 114.08,
      "step": 1730,
      "train_speed(iter/s)": 0.215721
    },
    {
      "acc": 0.73707952,
      "epoch": 0.0405955175782674,
      "grad_norm": 8.0,
      "learning_rate": 6.765163297045102e-06,
      "loss": 0.97277508,
      "memory(GiB)": 114.08,
      "step": 1740,
      "train_speed(iter/s)": 0.216341
    },
    {
      "acc": 0.72721863,
      "epoch": 0.040828825150556296,
      "grad_norm": 7.6875,
      "learning_rate": 6.8040435458786945e-06,
      "loss": 0.99633102,
      "memory(GiB)": 114.08,
      "step": 1750,
      "train_speed(iter/s)": 0.216947
    },
    {
      "acc": 0.76437578,
      "epoch": 0.04106213272284519,
      "grad_norm": 6.78125,
      "learning_rate": 6.842923794712286e-06,
      "loss": 0.86832066,
      "memory(GiB)": 114.08,
      "step": 1760,
      "train_speed(iter/s)": 0.217536
    },
    {
      "acc": 0.73662996,
      "epoch": 0.04129544029513408,
      "grad_norm": 7.1875,
      "learning_rate": 6.881804043545879e-06,
      "loss": 1.00305567,
      "memory(GiB)": 114.08,
      "step": 1770,
      "train_speed(iter/s)": 0.218134
    },
    {
      "acc": 0.73254414,
      "epoch": 0.04152874786742297,
      "grad_norm": 7.71875,
      "learning_rate": 6.9206842923794715e-06,
      "loss": 1.02277117,
      "memory(GiB)": 114.08,
      "step": 1780,
      "train_speed(iter/s)": 0.218699
    },
    {
      "acc": 0.74180479,
      "epoch": 0.04176205543971186,
      "grad_norm": 9.5,
      "learning_rate": 6.959564541213064e-06,
      "loss": 0.97406406,
      "memory(GiB)": 114.08,
      "step": 1790,
      "train_speed(iter/s)": 0.219286
    },
    {
      "acc": 0.75128908,
      "epoch": 0.041995363012000755,
      "grad_norm": 7.4375,
      "learning_rate": 6.998444790046657e-06,
      "loss": 0.92827549,
      "memory(GiB)": 114.08,
      "step": 1800,
      "train_speed(iter/s)": 0.219875
    },
    {
      "acc": 0.75414939,
      "epoch": 0.042228670584289654,
      "grad_norm": 8.6875,
      "learning_rate": 7.037325038880249e-06,
      "loss": 0.88612537,
      "memory(GiB)": 114.08,
      "step": 1810,
      "train_speed(iter/s)": 0.220455
    },
    {
      "acc": 0.74286633,
      "epoch": 0.042461978156578546,
      "grad_norm": 9.4375,
      "learning_rate": 7.0762052877138424e-06,
      "loss": 0.95695734,
      "memory(GiB)": 114.08,
      "step": 1820,
      "train_speed(iter/s)": 0.221039
    },
    {
      "acc": 0.72784061,
      "epoch": 0.04269528572886744,
      "grad_norm": 9.6875,
      "learning_rate": 7.115085536547435e-06,
      "loss": 1.02230368,
      "memory(GiB)": 114.08,
      "step": 1830,
      "train_speed(iter/s)": 0.221625
    },
    {
      "acc": 0.72988882,
      "epoch": 0.04292859330115633,
      "grad_norm": 14.8125,
      "learning_rate": 7.153965785381027e-06,
      "loss": 0.99553108,
      "memory(GiB)": 114.08,
      "step": 1840,
      "train_speed(iter/s)": 0.222188
    },
    {
      "acc": 0.74705415,
      "epoch": 0.04316190087344522,
      "grad_norm": 7.1875,
      "learning_rate": 7.19284603421462e-06,
      "loss": 0.95047102,
      "memory(GiB)": 114.08,
      "step": 1850,
      "train_speed(iter/s)": 0.222809
    },
    {
      "acc": 0.71940756,
      "epoch": 0.043395208445734114,
      "grad_norm": 6.53125,
      "learning_rate": 7.231726283048212e-06,
      "loss": 1.04833097,
      "memory(GiB)": 114.08,
      "step": 1860,
      "train_speed(iter/s)": 0.223337
    },
    {
      "acc": 0.73792143,
      "epoch": 0.04362851601802301,
      "grad_norm": 10.6875,
      "learning_rate": 7.270606531881804e-06,
      "loss": 0.97211323,
      "memory(GiB)": 114.08,
      "step": 1870,
      "train_speed(iter/s)": 0.223914
    },
    {
      "acc": 0.723668,
      "epoch": 0.043861823590311905,
      "grad_norm": 5.8125,
      "learning_rate": 7.309486780715397e-06,
      "loss": 1.0364295,
      "memory(GiB)": 114.08,
      "step": 1880,
      "train_speed(iter/s)": 0.224504
    },
    {
      "acc": 0.72725635,
      "epoch": 0.0440951311626008,
      "grad_norm": 8.0,
      "learning_rate": 7.3483670295489895e-06,
      "loss": 1.04018497,
      "memory(GiB)": 114.08,
      "step": 1890,
      "train_speed(iter/s)": 0.225068
    },
    {
      "acc": 0.75151863,
      "epoch": 0.04432843873488969,
      "grad_norm": 6.3125,
      "learning_rate": 7.387247278382582e-06,
      "loss": 0.9029954,
      "memory(GiB)": 114.08,
      "step": 1900,
      "train_speed(iter/s)": 0.225623
    },
    {
      "acc": 0.77624922,
      "epoch": 0.04456174630717858,
      "grad_norm": 10.8125,
      "learning_rate": 7.426127527216175e-06,
      "loss": 0.82208271,
      "memory(GiB)": 114.08,
      "step": 1910,
      "train_speed(iter/s)": 0.226166
    },
    {
      "acc": 0.7336926,
      "epoch": 0.04479505387946747,
      "grad_norm": 7.1875,
      "learning_rate": 7.465007776049767e-06,
      "loss": 0.98131351,
      "memory(GiB)": 114.08,
      "step": 1920,
      "train_speed(iter/s)": 0.226693
    },
    {
      "acc": 0.73093643,
      "epoch": 0.04502836145175637,
      "grad_norm": 8.0,
      "learning_rate": 7.5038880248833605e-06,
      "loss": 0.9974226,
      "memory(GiB)": 114.08,
      "step": 1930,
      "train_speed(iter/s)": 0.227229
    },
    {
      "acc": 0.73827853,
      "epoch": 0.04526166902404526,
      "grad_norm": 9.0,
      "learning_rate": 7.542768273716953e-06,
      "loss": 0.95264168,
      "memory(GiB)": 114.08,
      "step": 1940,
      "train_speed(iter/s)": 0.227762
    },
    {
      "acc": 0.74395685,
      "epoch": 0.045494976596334155,
      "grad_norm": 11.125,
      "learning_rate": 7.581648522550545e-06,
      "loss": 0.95124683,
      "memory(GiB)": 114.08,
      "step": 1950,
      "train_speed(iter/s)": 0.228247
    },
    {
      "acc": 0.74083662,
      "epoch": 0.04572828416862305,
      "grad_norm": 7.59375,
      "learning_rate": 7.620528771384137e-06,
      "loss": 0.97542019,
      "memory(GiB)": 114.08,
      "step": 1960,
      "train_speed(iter/s)": 0.228772
    },
    {
      "acc": 0.7502737,
      "epoch": 0.04596159174091194,
      "grad_norm": 5.96875,
      "learning_rate": 7.659409020217729e-06,
      "loss": 0.92420015,
      "memory(GiB)": 114.08,
      "step": 1970,
      "train_speed(iter/s)": 0.229254
    },
    {
      "acc": 0.75313225,
      "epoch": 0.04619489931320083,
      "grad_norm": 20.75,
      "learning_rate": 7.698289269051322e-06,
      "loss": 0.90695429,
      "memory(GiB)": 114.08,
      "step": 1980,
      "train_speed(iter/s)": 0.229765
    },
    {
      "acc": 0.74232001,
      "epoch": 0.04642820688548973,
      "grad_norm": 8.375,
      "learning_rate": 7.737169517884915e-06,
      "loss": 0.97582893,
      "memory(GiB)": 114.08,
      "step": 1990,
      "train_speed(iter/s)": 0.230283
    },
    {
      "acc": 0.757305,
      "epoch": 0.04666151445777862,
      "grad_norm": 6.40625,
      "learning_rate": 7.776049766718507e-06,
      "loss": 0.91142721,
      "memory(GiB)": 114.08,
      "step": 2000,
      "train_speed(iter/s)": 0.230798
    },
    {
      "epoch": 0.04666151445777862,
      "eval_acc": 0.709365603036719,
      "eval_loss": 0.9482490420341492,
      "eval_runtime": 1270.2818,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 2000
    },
    {
      "acc": 0.73707428,
      "epoch": 0.046894822030067514,
      "grad_norm": 10.1875,
      "learning_rate": 7.8149300155521e-06,
      "loss": 0.97233877,
      "memory(GiB)": 114.08,
      "step": 2010,
      "train_speed(iter/s)": 0.201446
    },
    {
      "acc": 0.74674578,
      "epoch": 0.047128129602356406,
      "grad_norm": 11.125,
      "learning_rate": 7.853810264385693e-06,
      "loss": 0.93988552,
      "memory(GiB)": 114.08,
      "step": 2020,
      "train_speed(iter/s)": 0.201937
    },
    {
      "acc": 0.74976053,
      "epoch": 0.0473614371746453,
      "grad_norm": 8.0,
      "learning_rate": 7.892690513219286e-06,
      "loss": 0.91554279,
      "memory(GiB)": 114.08,
      "step": 2030,
      "train_speed(iter/s)": 0.202447
    },
    {
      "acc": 0.74252758,
      "epoch": 0.04759474474693419,
      "grad_norm": 11.625,
      "learning_rate": 7.931570762052878e-06,
      "loss": 0.93326302,
      "memory(GiB)": 114.08,
      "step": 2040,
      "train_speed(iter/s)": 0.202943
    },
    {
      "acc": 0.72653966,
      "epoch": 0.04782805231922309,
      "grad_norm": 10.125,
      "learning_rate": 7.970451010886471e-06,
      "loss": 1.01006451,
      "memory(GiB)": 114.08,
      "step": 2050,
      "train_speed(iter/s)": 0.203443
    },
    {
      "acc": 0.75554819,
      "epoch": 0.04806135989151198,
      "grad_norm": 5.875,
      "learning_rate": 8.009331259720062e-06,
      "loss": 0.90173092,
      "memory(GiB)": 114.08,
      "step": 2060,
      "train_speed(iter/s)": 0.203953
    },
    {
      "acc": 0.7355751,
      "epoch": 0.04829466746380087,
      "grad_norm": 6.78125,
      "learning_rate": 8.048211508553656e-06,
      "loss": 0.98420334,
      "memory(GiB)": 114.08,
      "step": 2070,
      "train_speed(iter/s)": 0.204457
    },
    {
      "acc": 0.74058089,
      "epoch": 0.048527975036089764,
      "grad_norm": 8.8125,
      "learning_rate": 8.087091757387247e-06,
      "loss": 0.96377907,
      "memory(GiB)": 114.08,
      "step": 2080,
      "train_speed(iter/s)": 0.204931
    },
    {
      "acc": 0.75542016,
      "epoch": 0.048761282608378656,
      "grad_norm": 7.125,
      "learning_rate": 8.12597200622084e-06,
      "loss": 0.90528049,
      "memory(GiB)": 114.08,
      "step": 2090,
      "train_speed(iter/s)": 0.205464
    },
    {
      "acc": 0.74561777,
      "epoch": 0.04899459018066755,
      "grad_norm": 9.6875,
      "learning_rate": 8.164852255054433e-06,
      "loss": 0.94761057,
      "memory(GiB)": 114.08,
      "step": 2100,
      "train_speed(iter/s)": 0.205929
    },
    {
      "acc": 0.75310192,
      "epoch": 0.04922789775295645,
      "grad_norm": 7.0625,
      "learning_rate": 8.203732503888025e-06,
      "loss": 0.90652037,
      "memory(GiB)": 114.08,
      "step": 2110,
      "train_speed(iter/s)": 0.206417
    },
    {
      "acc": 0.74553957,
      "epoch": 0.04946120532524534,
      "grad_norm": 8.8125,
      "learning_rate": 8.242612752721618e-06,
      "loss": 0.94698086,
      "memory(GiB)": 114.08,
      "step": 2120,
      "train_speed(iter/s)": 0.206878
    },
    {
      "acc": 0.74139624,
      "epoch": 0.04969451289753423,
      "grad_norm": 5.78125,
      "learning_rate": 8.281493001555211e-06,
      "loss": 0.9554759,
      "memory(GiB)": 114.08,
      "step": 2130,
      "train_speed(iter/s)": 0.207363
    },
    {
      "acc": 0.75223551,
      "epoch": 0.04992782046982312,
      "grad_norm": 7.1875,
      "learning_rate": 8.320373250388803e-06,
      "loss": 0.91746941,
      "memory(GiB)": 114.08,
      "step": 2140,
      "train_speed(iter/s)": 0.20786
    },
    {
      "acc": 0.75747862,
      "epoch": 0.050161128042112015,
      "grad_norm": 8.1875,
      "learning_rate": 8.359253499222396e-06,
      "loss": 0.90032187,
      "memory(GiB)": 114.08,
      "step": 2150,
      "train_speed(iter/s)": 0.208349
    },
    {
      "acc": 0.74738655,
      "epoch": 0.050394435614400906,
      "grad_norm": 8.0,
      "learning_rate": 8.398133748055989e-06,
      "loss": 0.95068922,
      "memory(GiB)": 114.08,
      "step": 2160,
      "train_speed(iter/s)": 0.208799
    },
    {
      "acc": 0.74726391,
      "epoch": 0.050627743186689805,
      "grad_norm": 10.875,
      "learning_rate": 8.43701399688958e-06,
      "loss": 0.92888641,
      "memory(GiB)": 114.08,
      "step": 2170,
      "train_speed(iter/s)": 0.209331
    },
    {
      "acc": 0.7601017,
      "epoch": 0.0508610507589787,
      "grad_norm": 6.59375,
      "learning_rate": 8.475894245723174e-06,
      "loss": 0.88537483,
      "memory(GiB)": 114.08,
      "step": 2180,
      "train_speed(iter/s)": 0.209829
    },
    {
      "acc": 0.74622045,
      "epoch": 0.05109435833126759,
      "grad_norm": 5.90625,
      "learning_rate": 8.514774494556765e-06,
      "loss": 0.9415596,
      "memory(GiB)": 114.08,
      "step": 2190,
      "train_speed(iter/s)": 0.210313
    },
    {
      "acc": 0.74102707,
      "epoch": 0.05132766590355648,
      "grad_norm": 13.25,
      "learning_rate": 8.553654743390358e-06,
      "loss": 0.970403,
      "memory(GiB)": 114.08,
      "step": 2200,
      "train_speed(iter/s)": 0.210819
    },
    {
      "acc": 0.74637518,
      "epoch": 0.05156097347584537,
      "grad_norm": 6.59375,
      "learning_rate": 8.592534992223951e-06,
      "loss": 0.93611584,
      "memory(GiB)": 114.08,
      "step": 2210,
      "train_speed(iter/s)": 0.211296
    },
    {
      "acc": 0.72973704,
      "epoch": 0.05179428104813427,
      "grad_norm": 13.0625,
      "learning_rate": 8.631415241057543e-06,
      "loss": 0.99803753,
      "memory(GiB)": 114.08,
      "step": 2220,
      "train_speed(iter/s)": 0.211789
    },
    {
      "acc": 0.70917277,
      "epoch": 0.052027588620423164,
      "grad_norm": 8.25,
      "learning_rate": 8.670295489891136e-06,
      "loss": 1.06907005,
      "memory(GiB)": 114.08,
      "step": 2230,
      "train_speed(iter/s)": 0.212285
    },
    {
      "acc": 0.74743881,
      "epoch": 0.052260896192712056,
      "grad_norm": 21.0,
      "learning_rate": 8.70917573872473e-06,
      "loss": 0.9445673,
      "memory(GiB)": 114.08,
      "step": 2240,
      "train_speed(iter/s)": 0.212758
    },
    {
      "acc": 0.74076357,
      "epoch": 0.05249420376500095,
      "grad_norm": 5.21875,
      "learning_rate": 8.74805598755832e-06,
      "loss": 0.97673054,
      "memory(GiB)": 114.08,
      "step": 2250,
      "train_speed(iter/s)": 0.213229
    },
    {
      "acc": 0.74062457,
      "epoch": 0.05272751133728984,
      "grad_norm": 5.96875,
      "learning_rate": 8.786936236391914e-06,
      "loss": 0.96884727,
      "memory(GiB)": 114.08,
      "step": 2260,
      "train_speed(iter/s)": 0.213684
    },
    {
      "acc": 0.72948203,
      "epoch": 0.05296081890957873,
      "grad_norm": 6.0625,
      "learning_rate": 8.825816485225505e-06,
      "loss": 1.01084576,
      "memory(GiB)": 114.08,
      "step": 2270,
      "train_speed(iter/s)": 0.214158
    },
    {
      "acc": 0.75192738,
      "epoch": 0.05319412648186763,
      "grad_norm": 6.15625,
      "learning_rate": 8.864696734059099e-06,
      "loss": 0.89396305,
      "memory(GiB)": 114.08,
      "step": 2280,
      "train_speed(iter/s)": 0.214642
    },
    {
      "acc": 0.76000366,
      "epoch": 0.05342743405415652,
      "grad_norm": 6.90625,
      "learning_rate": 8.903576982892692e-06,
      "loss": 0.90859699,
      "memory(GiB)": 114.08,
      "step": 2290,
      "train_speed(iter/s)": 0.215099
    },
    {
      "acc": 0.7458251,
      "epoch": 0.053660741626445414,
      "grad_norm": 7.3125,
      "learning_rate": 8.942457231726283e-06,
      "loss": 0.97098904,
      "memory(GiB)": 114.08,
      "step": 2300,
      "train_speed(iter/s)": 0.21551
    },
    {
      "acc": 0.73837004,
      "epoch": 0.053894049198734306,
      "grad_norm": 5.8125,
      "learning_rate": 8.981337480559876e-06,
      "loss": 0.95030222,
      "memory(GiB)": 114.08,
      "step": 2310,
      "train_speed(iter/s)": 0.215929
    },
    {
      "acc": 0.75400095,
      "epoch": 0.0541273567710232,
      "grad_norm": 8.1875,
      "learning_rate": 9.02021772939347e-06,
      "loss": 0.90010729,
      "memory(GiB)": 114.08,
      "step": 2320,
      "train_speed(iter/s)": 0.216371
    },
    {
      "acc": 0.75261707,
      "epoch": 0.05436066434331209,
      "grad_norm": 6.4375,
      "learning_rate": 9.059097978227061e-06,
      "loss": 0.94032478,
      "memory(GiB)": 114.08,
      "step": 2330,
      "train_speed(iter/s)": 0.216778
    },
    {
      "acc": 0.7515697,
      "epoch": 0.05459397191560099,
      "grad_norm": 5.34375,
      "learning_rate": 9.097978227060654e-06,
      "loss": 0.94090843,
      "memory(GiB)": 114.08,
      "step": 2340,
      "train_speed(iter/s)": 0.217239
    },
    {
      "acc": 0.7397305,
      "epoch": 0.05482727948788988,
      "grad_norm": 10.0625,
      "learning_rate": 9.136858475894247e-06,
      "loss": 0.97343073,
      "memory(GiB)": 114.08,
      "step": 2350,
      "train_speed(iter/s)": 0.217667
    },
    {
      "acc": 0.74915609,
      "epoch": 0.05506058706017877,
      "grad_norm": 6.65625,
      "learning_rate": 9.175738724727839e-06,
      "loss": 0.921453,
      "memory(GiB)": 114.08,
      "step": 2360,
      "train_speed(iter/s)": 0.218073
    },
    {
      "acc": 0.7468821,
      "epoch": 0.055293894632467665,
      "grad_norm": 9.75,
      "learning_rate": 9.21461897356143e-06,
      "loss": 0.95533428,
      "memory(GiB)": 114.08,
      "step": 2370,
      "train_speed(iter/s)": 0.218486
    },
    {
      "acc": 0.73546376,
      "epoch": 0.05552720220475656,
      "grad_norm": 6.96875,
      "learning_rate": 9.253499222395023e-06,
      "loss": 0.99596453,
      "memory(GiB)": 114.08,
      "step": 2380,
      "train_speed(iter/s)": 0.218898
    },
    {
      "acc": 0.73259478,
      "epoch": 0.05576050977704545,
      "grad_norm": 7.96875,
      "learning_rate": 9.292379471228617e-06,
      "loss": 0.99712763,
      "memory(GiB)": 114.08,
      "step": 2390,
      "train_speed(iter/s)": 0.219283
    },
    {
      "acc": 0.73952942,
      "epoch": 0.05599381734933435,
      "grad_norm": 5.71875,
      "learning_rate": 9.33125972006221e-06,
      "loss": 0.94039307,
      "memory(GiB)": 114.08,
      "step": 2400,
      "train_speed(iter/s)": 0.21973
    },
    {
      "acc": 0.74226012,
      "epoch": 0.05622712492162324,
      "grad_norm": 6.125,
      "learning_rate": 9.370139968895801e-06,
      "loss": 0.93415909,
      "memory(GiB)": 114.08,
      "step": 2410,
      "train_speed(iter/s)": 0.220139
    },
    {
      "acc": 0.75117826,
      "epoch": 0.05646043249391213,
      "grad_norm": 24.625,
      "learning_rate": 9.409020217729394e-06,
      "loss": 0.89564247,
      "memory(GiB)": 114.08,
      "step": 2420,
      "train_speed(iter/s)": 0.220555
    },
    {
      "acc": 0.74033728,
      "epoch": 0.05669374006620102,
      "grad_norm": 10.4375,
      "learning_rate": 9.447900466562988e-06,
      "loss": 0.96549892,
      "memory(GiB)": 114.08,
      "step": 2430,
      "train_speed(iter/s)": 0.220978
    },
    {
      "acc": 0.76307154,
      "epoch": 0.056927047638489915,
      "grad_norm": 17.375,
      "learning_rate": 9.486780715396579e-06,
      "loss": 0.86378517,
      "memory(GiB)": 114.08,
      "step": 2440,
      "train_speed(iter/s)": 0.221411
    },
    {
      "acc": 0.73024807,
      "epoch": 0.05716035521077881,
      "grad_norm": 12.6875,
      "learning_rate": 9.525660964230172e-06,
      "loss": 1.01558571,
      "memory(GiB)": 114.08,
      "step": 2450,
      "train_speed(iter/s)": 0.221845
    },
    {
      "acc": 0.75076675,
      "epoch": 0.057393662783067706,
      "grad_norm": 7.1875,
      "learning_rate": 9.564541213063765e-06,
      "loss": 0.91901255,
      "memory(GiB)": 114.08,
      "step": 2460,
      "train_speed(iter/s)": 0.222309
    },
    {
      "acc": 0.76636009,
      "epoch": 0.0576269703553566,
      "grad_norm": 4.5,
      "learning_rate": 9.603421461897357e-06,
      "loss": 0.85223923,
      "memory(GiB)": 114.08,
      "step": 2470,
      "train_speed(iter/s)": 0.222694
    },
    {
      "acc": 0.75716038,
      "epoch": 0.05786027792764549,
      "grad_norm": 6.5625,
      "learning_rate": 9.642301710730948e-06,
      "loss": 0.90162163,
      "memory(GiB)": 114.08,
      "step": 2480,
      "train_speed(iter/s)": 0.223102
    },
    {
      "acc": 0.74022632,
      "epoch": 0.05809358549993438,
      "grad_norm": 7.46875,
      "learning_rate": 9.681181959564542e-06,
      "loss": 0.96824207,
      "memory(GiB)": 114.08,
      "step": 2490,
      "train_speed(iter/s)": 0.223508
    },
    {
      "acc": 0.75365782,
      "epoch": 0.058326893072223274,
      "grad_norm": 6.75,
      "learning_rate": 9.720062208398135e-06,
      "loss": 0.93858395,
      "memory(GiB)": 114.08,
      "step": 2500,
      "train_speed(iter/s)": 0.223925
    },
    {
      "epoch": 0.058326893072223274,
      "eval_acc": 0.7143739537577124,
      "eval_loss": 0.9248969554901123,
      "eval_runtime": 1269.8474,
      "eval_samples_per_second": 28.343,
      "eval_steps_per_second": 14.172,
      "step": 2500
    },
    {
      "acc": 0.7417676,
      "epoch": 0.058560200644512166,
      "grad_norm": 6.03125,
      "learning_rate": 9.758942457231726e-06,
      "loss": 0.95499773,
      "memory(GiB)": 114.08,
      "step": 2510,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.74519234,
      "epoch": 0.058793508216801064,
      "grad_norm": 8.75,
      "learning_rate": 9.79782270606532e-06,
      "loss": 0.94459839,
      "memory(GiB)": 114.08,
      "step": 2520,
      "train_speed(iter/s)": 0.201609
    },
    {
      "acc": 0.75776386,
      "epoch": 0.059026815789089956,
      "grad_norm": 4.96875,
      "learning_rate": 9.836702954898913e-06,
      "loss": 0.88159065,
      "memory(GiB)": 114.08,
      "step": 2530,
      "train_speed(iter/s)": 0.201997
    },
    {
      "acc": 0.75585461,
      "epoch": 0.05926012336137885,
      "grad_norm": 8.625,
      "learning_rate": 9.875583203732506e-06,
      "loss": 0.89656048,
      "memory(GiB)": 114.08,
      "step": 2540,
      "train_speed(iter/s)": 0.202388
    },
    {
      "acc": 0.74211607,
      "epoch": 0.05949343093366774,
      "grad_norm": 15.3125,
      "learning_rate": 9.914463452566097e-06,
      "loss": 0.94759617,
      "memory(GiB)": 114.08,
      "step": 2550,
      "train_speed(iter/s)": 0.202765
    },
    {
      "acc": 0.74785175,
      "epoch": 0.05972673850595663,
      "grad_norm": 6.9375,
      "learning_rate": 9.95334370139969e-06,
      "loss": 0.92067432,
      "memory(GiB)": 114.08,
      "step": 2560,
      "train_speed(iter/s)": 0.203186
    },
    {
      "acc": 0.73263016,
      "epoch": 0.059960046078245524,
      "grad_norm": 7.5625,
      "learning_rate": 9.992223950233282e-06,
      "loss": 0.99488621,
      "memory(GiB)": 114.08,
      "step": 2570,
      "train_speed(iter/s)": 0.203615
    },
    {
      "acc": 0.74043674,
      "epoch": 0.06019335365053442,
      "grad_norm": 7.8125,
      "learning_rate": 9.999999771600465e-06,
      "loss": 0.96276875,
      "memory(GiB)": 114.08,
      "step": 2580,
      "train_speed(iter/s)": 0.204022
    },
    {
      "acc": 0.78092365,
      "epoch": 0.060426661222823315,
      "grad_norm": 6.6875,
      "learning_rate": 9.999998843727385e-06,
      "loss": 0.79039125,
      "memory(GiB)": 114.08,
      "step": 2590,
      "train_speed(iter/s)": 0.204452
    },
    {
      "acc": 0.76478825,
      "epoch": 0.06065996879511221,
      "grad_norm": 5.5,
      "learning_rate": 9.999997202105923e-06,
      "loss": 0.85824947,
      "memory(GiB)": 114.08,
      "step": 2600,
      "train_speed(iter/s)": 0.204846
    },
    {
      "acc": 0.77252517,
      "epoch": 0.0608932763674011,
      "grad_norm": 7.1875,
      "learning_rate": 9.999994846736312e-06,
      "loss": 0.83852539,
      "memory(GiB)": 114.08,
      "step": 2610,
      "train_speed(iter/s)": 0.20524
    },
    {
      "acc": 0.73587317,
      "epoch": 0.06112658393968999,
      "grad_norm": 16.5,
      "learning_rate": 9.99999177761889e-06,
      "loss": 0.97813015,
      "memory(GiB)": 114.08,
      "step": 2620,
      "train_speed(iter/s)": 0.205637
    },
    {
      "acc": 0.75071974,
      "epoch": 0.06135989151197888,
      "grad_norm": 4.71875,
      "learning_rate": 9.999987994754094e-06,
      "loss": 0.92447548,
      "memory(GiB)": 114.08,
      "step": 2630,
      "train_speed(iter/s)": 0.206024
    },
    {
      "acc": 0.74091039,
      "epoch": 0.06159319908426778,
      "grad_norm": 6.21875,
      "learning_rate": 9.999983498142464e-06,
      "loss": 0.95536308,
      "memory(GiB)": 114.08,
      "step": 2640,
      "train_speed(iter/s)": 0.206403
    },
    {
      "acc": 0.74781256,
      "epoch": 0.06182650665655667,
      "grad_norm": 14.75,
      "learning_rate": 9.999978287784642e-06,
      "loss": 0.9136961,
      "memory(GiB)": 114.08,
      "step": 2650,
      "train_speed(iter/s)": 0.206807
    },
    {
      "acc": 0.74666529,
      "epoch": 0.062059814228845565,
      "grad_norm": 5.8125,
      "learning_rate": 9.999972363681371e-06,
      "loss": 0.90506544,
      "memory(GiB)": 114.08,
      "step": 2660,
      "train_speed(iter/s)": 0.207192
    },
    {
      "acc": 0.76904163,
      "epoch": 0.06229312180113446,
      "grad_norm": 5.84375,
      "learning_rate": 9.9999657258335e-06,
      "loss": 0.85019245,
      "memory(GiB)": 114.08,
      "step": 2670,
      "train_speed(iter/s)": 0.207572
    },
    {
      "acc": 0.7517498,
      "epoch": 0.06252642937342336,
      "grad_norm": 8.125,
      "learning_rate": 9.999958374241974e-06,
      "loss": 0.91046963,
      "memory(GiB)": 114.08,
      "step": 2680,
      "train_speed(iter/s)": 0.207944
    },
    {
      "acc": 0.76082239,
      "epoch": 0.06275973694571224,
      "grad_norm": 6.4375,
      "learning_rate": 9.99995030890784e-06,
      "loss": 0.87537355,
      "memory(GiB)": 114.08,
      "step": 2690,
      "train_speed(iter/s)": 0.208348
    },
    {
      "acc": 0.75649023,
      "epoch": 0.06299304451800114,
      "grad_norm": 6.90625,
      "learning_rate": 9.999941529832254e-06,
      "loss": 0.9041748,
      "memory(GiB)": 114.08,
      "step": 2700,
      "train_speed(iter/s)": 0.208755
    },
    {
      "acc": 0.7486876,
      "epoch": 0.06322635209029003,
      "grad_norm": 8.4375,
      "learning_rate": 9.999932037016466e-06,
      "loss": 0.94257088,
      "memory(GiB)": 114.08,
      "step": 2710,
      "train_speed(iter/s)": 0.209144
    },
    {
      "acc": 0.76688719,
      "epoch": 0.06345965966257892,
      "grad_norm": 15.0,
      "learning_rate": 9.999921830461833e-06,
      "loss": 0.88587494,
      "memory(GiB)": 114.08,
      "step": 2720,
      "train_speed(iter/s)": 0.20952
    },
    {
      "acc": 0.76675873,
      "epoch": 0.06369296723486782,
      "grad_norm": 4.90625,
      "learning_rate": 9.99991091016981e-06,
      "loss": 0.84941301,
      "memory(GiB)": 114.08,
      "step": 2730,
      "train_speed(iter/s)": 0.209885
    },
    {
      "acc": 0.76298923,
      "epoch": 0.06392627480715671,
      "grad_norm": 7.125,
      "learning_rate": 9.99989927614196e-06,
      "loss": 0.87009039,
      "memory(GiB)": 114.08,
      "step": 2740,
      "train_speed(iter/s)": 0.210263
    },
    {
      "acc": 0.75599327,
      "epoch": 0.0641595823794456,
      "grad_norm": 6.71875,
      "learning_rate": 9.999886928379939e-06,
      "loss": 0.88294525,
      "memory(GiB)": 114.08,
      "step": 2750,
      "train_speed(iter/s)": 0.2106
    },
    {
      "acc": 0.76202002,
      "epoch": 0.06439288995173449,
      "grad_norm": 7.5625,
      "learning_rate": 9.99987386688551e-06,
      "loss": 0.87443285,
      "memory(GiB)": 114.08,
      "step": 2760,
      "train_speed(iter/s)": 0.210993
    },
    {
      "acc": 0.74971361,
      "epoch": 0.06462619752402339,
      "grad_norm": 5.375,
      "learning_rate": 9.99986009166054e-06,
      "loss": 0.90866261,
      "memory(GiB)": 114.08,
      "step": 2770,
      "train_speed(iter/s)": 0.211356
    },
    {
      "acc": 0.75157547,
      "epoch": 0.06485950509631228,
      "grad_norm": 6.40625,
      "learning_rate": 9.999845602706995e-06,
      "loss": 0.89404106,
      "memory(GiB)": 114.08,
      "step": 2780,
      "train_speed(iter/s)": 0.211714
    },
    {
      "acc": 0.7517457,
      "epoch": 0.06509281266860117,
      "grad_norm": 7.9375,
      "learning_rate": 9.999830400026941e-06,
      "loss": 0.93447304,
      "memory(GiB)": 114.08,
      "step": 2790,
      "train_speed(iter/s)": 0.212101
    },
    {
      "acc": 0.74939804,
      "epoch": 0.06532612024089007,
      "grad_norm": 5.375,
      "learning_rate": 9.999814483622552e-06,
      "loss": 0.90546379,
      "memory(GiB)": 114.08,
      "step": 2800,
      "train_speed(iter/s)": 0.212463
    },
    {
      "acc": 0.74818916,
      "epoch": 0.06555942781317896,
      "grad_norm": 8.625,
      "learning_rate": 9.999797853496097e-06,
      "loss": 0.91956005,
      "memory(GiB)": 114.08,
      "step": 2810,
      "train_speed(iter/s)": 0.212815
    },
    {
      "acc": 0.74374294,
      "epoch": 0.06579273538546786,
      "grad_norm": 6.40625,
      "learning_rate": 9.999780509649952e-06,
      "loss": 0.95786781,
      "memory(GiB)": 114.08,
      "step": 2820,
      "train_speed(iter/s)": 0.213177
    },
    {
      "acc": 0.74510412,
      "epoch": 0.06602604295775674,
      "grad_norm": 6.125,
      "learning_rate": 9.99976245208659e-06,
      "loss": 0.96016006,
      "memory(GiB)": 114.08,
      "step": 2830,
      "train_speed(iter/s)": 0.213555
    },
    {
      "acc": 0.75915327,
      "epoch": 0.06625935053004564,
      "grad_norm": 7.03125,
      "learning_rate": 9.99974368080859e-06,
      "loss": 0.8857523,
      "memory(GiB)": 114.08,
      "step": 2840,
      "train_speed(iter/s)": 0.213924
    },
    {
      "acc": 0.74505186,
      "epoch": 0.06649265810233454,
      "grad_norm": 7.125,
      "learning_rate": 9.999724195818634e-06,
      "loss": 0.949786,
      "memory(GiB)": 114.08,
      "step": 2850,
      "train_speed(iter/s)": 0.21428
    },
    {
      "acc": 0.76169367,
      "epoch": 0.06672596567462342,
      "grad_norm": 7.5,
      "learning_rate": 9.999703997119501e-06,
      "loss": 0.85612011,
      "memory(GiB)": 114.08,
      "step": 2860,
      "train_speed(iter/s)": 0.214653
    },
    {
      "acc": 0.75327644,
      "epoch": 0.06695927324691232,
      "grad_norm": 7.625,
      "learning_rate": 9.999683084714074e-06,
      "loss": 0.8901247,
      "memory(GiB)": 114.09,
      "step": 2870,
      "train_speed(iter/s)": 0.214995
    },
    {
      "acc": 0.73931332,
      "epoch": 0.06719258081920121,
      "grad_norm": 10.0625,
      "learning_rate": 9.999661458605339e-06,
      "loss": 0.95245857,
      "memory(GiB)": 114.09,
      "step": 2880,
      "train_speed(iter/s)": 0.215355
    },
    {
      "acc": 0.76272769,
      "epoch": 0.06742588839149011,
      "grad_norm": 4.28125,
      "learning_rate": 9.999639118796384e-06,
      "loss": 0.86453457,
      "memory(GiB)": 114.09,
      "step": 2890,
      "train_speed(iter/s)": 0.215716
    },
    {
      "acc": 0.77728963,
      "epoch": 0.06765919596377899,
      "grad_norm": 7.96875,
      "learning_rate": 9.999616065290396e-06,
      "loss": 0.8032423,
      "memory(GiB)": 114.09,
      "step": 2900,
      "train_speed(iter/s)": 0.216072
    },
    {
      "acc": 0.75634222,
      "epoch": 0.06789250353606789,
      "grad_norm": 6.15625,
      "learning_rate": 9.999592298090669e-06,
      "loss": 0.86408386,
      "memory(GiB)": 114.09,
      "step": 2910,
      "train_speed(iter/s)": 0.216406
    },
    {
      "acc": 0.77233763,
      "epoch": 0.06812581110835679,
      "grad_norm": 5.5625,
      "learning_rate": 9.999567817200592e-06,
      "loss": 0.83319054,
      "memory(GiB)": 114.09,
      "step": 2920,
      "train_speed(iter/s)": 0.21677
    },
    {
      "acc": 0.73433094,
      "epoch": 0.06835911868064568,
      "grad_norm": 4.84375,
      "learning_rate": 9.999542622623661e-06,
      "loss": 0.98757401,
      "memory(GiB)": 114.09,
      "step": 2930,
      "train_speed(iter/s)": 0.217116
    },
    {
      "acc": 0.76126733,
      "epoch": 0.06859242625293457,
      "grad_norm": 7.125,
      "learning_rate": 9.999516714363475e-06,
      "loss": 0.89113979,
      "memory(GiB)": 114.09,
      "step": 2940,
      "train_speed(iter/s)": 0.217466
    },
    {
      "acc": 0.74799905,
      "epoch": 0.06882573382522346,
      "grad_norm": 4.71875,
      "learning_rate": 9.99949009242373e-06,
      "loss": 0.92414608,
      "memory(GiB)": 114.09,
      "step": 2950,
      "train_speed(iter/s)": 0.217813
    },
    {
      "acc": 0.75590515,
      "epoch": 0.06905904139751236,
      "grad_norm": 5.625,
      "learning_rate": 9.999462756808227e-06,
      "loss": 0.88417921,
      "memory(GiB)": 114.09,
      "step": 2960,
      "train_speed(iter/s)": 0.218171
    },
    {
      "acc": 0.73305931,
      "epoch": 0.06929234896980126,
      "grad_norm": 10.0625,
      "learning_rate": 9.999434707520867e-06,
      "loss": 0.96791725,
      "memory(GiB)": 114.09,
      "step": 2970,
      "train_speed(iter/s)": 0.218519
    },
    {
      "acc": 0.74652576,
      "epoch": 0.06952565654209014,
      "grad_norm": 5.03125,
      "learning_rate": 9.999405944565654e-06,
      "loss": 0.9400857,
      "memory(GiB)": 114.09,
      "step": 2980,
      "train_speed(iter/s)": 0.218871
    },
    {
      "acc": 0.75939126,
      "epoch": 0.06975896411437904,
      "grad_norm": 5.46875,
      "learning_rate": 9.999376467946695e-06,
      "loss": 0.88669376,
      "memory(GiB)": 114.09,
      "step": 2990,
      "train_speed(iter/s)": 0.219224
    },
    {
      "acc": 0.74682531,
      "epoch": 0.06999227168666793,
      "grad_norm": 6.0625,
      "learning_rate": 9.999346277668198e-06,
      "loss": 0.89512215,
      "memory(GiB)": 114.09,
      "step": 3000,
      "train_speed(iter/s)": 0.21959
    },
    {
      "epoch": 0.06999227168666793,
      "eval_acc": 0.7189282955801313,
      "eval_loss": 0.9065292477607727,
      "eval_runtime": 1270.6255,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "step": 3000
    },
    {
      "acc": 0.75453682,
      "epoch": 0.07022557925895682,
      "grad_norm": 5.0,
      "learning_rate": 9.999315373734472e-06,
      "loss": 0.87960033,
      "memory(GiB)": 117.28,
      "step": 3010,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.76882467,
      "epoch": 0.07045888683124572,
      "grad_norm": 10.75,
      "learning_rate": 9.999283756149932e-06,
      "loss": 0.83173504,
      "memory(GiB)": 117.28,
      "step": 3020,
      "train_speed(iter/s)": 0.201327
    },
    {
      "acc": 0.7758441,
      "epoch": 0.07069219440353461,
      "grad_norm": 5.21875,
      "learning_rate": 9.999251424919083e-06,
      "loss": 0.8254406,
      "memory(GiB)": 117.28,
      "step": 3030,
      "train_speed(iter/s)": 0.201672
    },
    {
      "acc": 0.7445199,
      "epoch": 0.07092550197582351,
      "grad_norm": 10.1875,
      "learning_rate": 9.999218380046548e-06,
      "loss": 0.93491154,
      "memory(GiB)": 117.28,
      "step": 3040,
      "train_speed(iter/s)": 0.202018
    },
    {
      "acc": 0.77481375,
      "epoch": 0.07115880954811239,
      "grad_norm": 6.90625,
      "learning_rate": 9.99918462153704e-06,
      "loss": 0.8122179,
      "memory(GiB)": 117.28,
      "step": 3050,
      "train_speed(iter/s)": 0.202354
    },
    {
      "acc": 0.7394393,
      "epoch": 0.07139211712040129,
      "grad_norm": 6.59375,
      "learning_rate": 9.999150149395383e-06,
      "loss": 0.97233334,
      "memory(GiB)": 117.28,
      "step": 3060,
      "train_speed(iter/s)": 0.202695
    },
    {
      "acc": 0.74153099,
      "epoch": 0.07162542469269018,
      "grad_norm": 5.53125,
      "learning_rate": 9.99911496362649e-06,
      "loss": 0.9292963,
      "memory(GiB)": 117.28,
      "step": 3070,
      "train_speed(iter/s)": 0.203018
    },
    {
      "acc": 0.74457722,
      "epoch": 0.07185873226497907,
      "grad_norm": 6.6875,
      "learning_rate": 9.99907906423539e-06,
      "loss": 0.95251703,
      "memory(GiB)": 117.28,
      "step": 3080,
      "train_speed(iter/s)": 0.203382
    },
    {
      "acc": 0.73962574,
      "epoch": 0.07209203983726797,
      "grad_norm": 6.875,
      "learning_rate": 9.999042451227208e-06,
      "loss": 0.9660202,
      "memory(GiB)": 117.28,
      "step": 3090,
      "train_speed(iter/s)": 0.203734
    },
    {
      "acc": 0.75118494,
      "epoch": 0.07232534740955686,
      "grad_norm": 7.65625,
      "learning_rate": 9.999005124607167e-06,
      "loss": 0.89292297,
      "memory(GiB)": 117.28,
      "step": 3100,
      "train_speed(iter/s)": 0.204063
    },
    {
      "acc": 0.77734885,
      "epoch": 0.07255865498184576,
      "grad_norm": 22.0,
      "learning_rate": 9.998967084380596e-06,
      "loss": 0.81176682,
      "memory(GiB)": 117.28,
      "step": 3110,
      "train_speed(iter/s)": 0.204418
    },
    {
      "acc": 0.75207405,
      "epoch": 0.07279196255413464,
      "grad_norm": 7.125,
      "learning_rate": 9.998928330552925e-06,
      "loss": 0.92476635,
      "memory(GiB)": 117.28,
      "step": 3120,
      "train_speed(iter/s)": 0.204741
    },
    {
      "acc": 0.75438986,
      "epoch": 0.07302527012642354,
      "grad_norm": 9.1875,
      "learning_rate": 9.998888863129688e-06,
      "loss": 0.9007534,
      "memory(GiB)": 117.28,
      "step": 3130,
      "train_speed(iter/s)": 0.205082
    },
    {
      "acc": 0.75198278,
      "epoch": 0.07325857769871244,
      "grad_norm": 9.1875,
      "learning_rate": 9.998848682116518e-06,
      "loss": 0.9158268,
      "memory(GiB)": 117.28,
      "step": 3140,
      "train_speed(iter/s)": 0.205427
    },
    {
      "acc": 0.75476503,
      "epoch": 0.07349188527100133,
      "grad_norm": 7.875,
      "learning_rate": 9.998807787519151e-06,
      "loss": 0.87708139,
      "memory(GiB)": 117.28,
      "step": 3150,
      "train_speed(iter/s)": 0.205765
    },
    {
      "acc": 0.75708885,
      "epoch": 0.07372519284329022,
      "grad_norm": 4.6875,
      "learning_rate": 9.998766179343425e-06,
      "loss": 0.87132273,
      "memory(GiB)": 117.28,
      "step": 3160,
      "train_speed(iter/s)": 0.206091
    },
    {
      "acc": 0.76350727,
      "epoch": 0.07395850041557911,
      "grad_norm": 6.375,
      "learning_rate": 9.998723857595278e-06,
      "loss": 0.8524066,
      "memory(GiB)": 117.28,
      "step": 3170,
      "train_speed(iter/s)": 0.206437
    },
    {
      "acc": 0.75130353,
      "epoch": 0.07419180798786801,
      "grad_norm": 6.46875,
      "learning_rate": 9.998680822280752e-06,
      "loss": 0.94219074,
      "memory(GiB)": 117.28,
      "step": 3180,
      "train_speed(iter/s)": 0.20678
    },
    {
      "acc": 0.74551325,
      "epoch": 0.0744251155601569,
      "grad_norm": 6.25,
      "learning_rate": 9.998637073405992e-06,
      "loss": 0.94350052,
      "memory(GiB)": 117.28,
      "step": 3190,
      "train_speed(iter/s)": 0.207101
    },
    {
      "acc": 0.74958072,
      "epoch": 0.07465842313244579,
      "grad_norm": 5.0,
      "learning_rate": 9.998592610977241e-06,
      "loss": 0.9216177,
      "memory(GiB)": 117.28,
      "step": 3200,
      "train_speed(iter/s)": 0.207424
    },
    {
      "acc": 0.75495872,
      "epoch": 0.07489173070473469,
      "grad_norm": 6.03125,
      "learning_rate": 9.998547435000847e-06,
      "loss": 0.89467354,
      "memory(GiB)": 117.28,
      "step": 3210,
      "train_speed(iter/s)": 0.207777
    },
    {
      "acc": 0.74197507,
      "epoch": 0.07512503827702358,
      "grad_norm": 6.5,
      "learning_rate": 9.998501545483259e-06,
      "loss": 0.94241686,
      "memory(GiB)": 117.28,
      "step": 3220,
      "train_speed(iter/s)": 0.208109
    },
    {
      "acc": 0.75202932,
      "epoch": 0.07535834584931247,
      "grad_norm": 5.375,
      "learning_rate": 9.998454942431029e-06,
      "loss": 0.91658039,
      "memory(GiB)": 117.28,
      "step": 3230,
      "train_speed(iter/s)": 0.208416
    },
    {
      "acc": 0.7436245,
      "epoch": 0.07559165342160136,
      "grad_norm": 9.3125,
      "learning_rate": 9.998407625850806e-06,
      "loss": 0.93169861,
      "memory(GiB)": 117.28,
      "step": 3240,
      "train_speed(iter/s)": 0.208753
    },
    {
      "acc": 0.76363297,
      "epoch": 0.07582496099389026,
      "grad_norm": 6.5625,
      "learning_rate": 9.998359595749346e-06,
      "loss": 0.88251057,
      "memory(GiB)": 117.28,
      "step": 3250,
      "train_speed(iter/s)": 0.209087
    },
    {
      "acc": 0.76651325,
      "epoch": 0.07605826856617916,
      "grad_norm": 6.53125,
      "learning_rate": 9.998310852133506e-06,
      "loss": 0.88197155,
      "memory(GiB)": 117.28,
      "step": 3260,
      "train_speed(iter/s)": 0.209407
    },
    {
      "acc": 0.73732243,
      "epoch": 0.07629157613846804,
      "grad_norm": 6.4375,
      "learning_rate": 9.998261395010246e-06,
      "loss": 0.97510824,
      "memory(GiB)": 117.28,
      "step": 3270,
      "train_speed(iter/s)": 0.209726
    },
    {
      "acc": 0.76042881,
      "epoch": 0.07652488371075694,
      "grad_norm": 5.625,
      "learning_rate": 9.998211224386623e-06,
      "loss": 0.88643074,
      "memory(GiB)": 117.28,
      "step": 3280,
      "train_speed(iter/s)": 0.210037
    },
    {
      "acc": 0.73288879,
      "epoch": 0.07675819128304583,
      "grad_norm": 8.1875,
      "learning_rate": 9.998160340269799e-06,
      "loss": 0.98039322,
      "memory(GiB)": 117.28,
      "step": 3290,
      "train_speed(iter/s)": 0.210365
    },
    {
      "acc": 0.74865913,
      "epoch": 0.07699149885533473,
      "grad_norm": 32.5,
      "learning_rate": 9.998108742667038e-06,
      "loss": 0.93164606,
      "memory(GiB)": 117.28,
      "step": 3300,
      "train_speed(iter/s)": 0.210674
    },
    {
      "acc": 0.73399429,
      "epoch": 0.07722480642762361,
      "grad_norm": 5.3125,
      "learning_rate": 9.998056431585707e-06,
      "loss": 0.97385902,
      "memory(GiB)": 117.28,
      "step": 3310,
      "train_speed(iter/s)": 0.210992
    },
    {
      "acc": 0.76379633,
      "epoch": 0.07745811399991251,
      "grad_norm": 4.96875,
      "learning_rate": 9.998003407033271e-06,
      "loss": 0.84101114,
      "memory(GiB)": 117.28,
      "step": 3320,
      "train_speed(iter/s)": 0.211281
    },
    {
      "acc": 0.76484241,
      "epoch": 0.07769142157220141,
      "grad_norm": 5.71875,
      "learning_rate": 9.997949669017302e-06,
      "loss": 0.8537735,
      "memory(GiB)": 117.28,
      "step": 3330,
      "train_speed(iter/s)": 0.211582
    },
    {
      "acc": 0.74536896,
      "epoch": 0.07792472914449029,
      "grad_norm": 9.0,
      "learning_rate": 9.997895217545468e-06,
      "loss": 0.99053888,
      "memory(GiB)": 117.28,
      "step": 3340,
      "train_speed(iter/s)": 0.211872
    },
    {
      "acc": 0.76393571,
      "epoch": 0.07815803671677919,
      "grad_norm": 6.25,
      "learning_rate": 9.997840052625546e-06,
      "loss": 0.8449995,
      "memory(GiB)": 117.28,
      "step": 3350,
      "train_speed(iter/s)": 0.212198
    },
    {
      "acc": 0.75357141,
      "epoch": 0.07839134428906808,
      "grad_norm": 7.96875,
      "learning_rate": 9.997784174265407e-06,
      "loss": 0.89924717,
      "memory(GiB)": 117.28,
      "step": 3360,
      "train_speed(iter/s)": 0.212515
    },
    {
      "acc": 0.74694405,
      "epoch": 0.07862465186135698,
      "grad_norm": 5.375,
      "learning_rate": 9.99772758247303e-06,
      "loss": 0.916609,
      "memory(GiB)": 117.28,
      "step": 3370,
      "train_speed(iter/s)": 0.212802
    },
    {
      "acc": 0.75625582,
      "epoch": 0.07885795943364587,
      "grad_norm": 9.5,
      "learning_rate": 9.99767027725649e-06,
      "loss": 0.89150448,
      "memory(GiB)": 117.28,
      "step": 3380,
      "train_speed(iter/s)": 0.213124
    },
    {
      "acc": 0.77579999,
      "epoch": 0.07909126700593476,
      "grad_norm": 5.0625,
      "learning_rate": 9.997612258623972e-06,
      "loss": 0.80863886,
      "memory(GiB)": 117.28,
      "step": 3390,
      "train_speed(iter/s)": 0.213441
    },
    {
      "acc": 0.72985349,
      "epoch": 0.07932457457822366,
      "grad_norm": 7.90625,
      "learning_rate": 9.997553526583755e-06,
      "loss": 1.00999374,
      "memory(GiB)": 117.28,
      "step": 3400,
      "train_speed(iter/s)": 0.21375
    },
    {
      "acc": 0.75671997,
      "epoch": 0.07955788215051254,
      "grad_norm": 13.6875,
      "learning_rate": 9.997494081144224e-06,
      "loss": 0.89063911,
      "memory(GiB)": 117.28,
      "step": 3410,
      "train_speed(iter/s)": 0.214062
    },
    {
      "acc": 0.78263044,
      "epoch": 0.07979118972280144,
      "grad_norm": 8.75,
      "learning_rate": 9.997433922313863e-06,
      "loss": 0.78714075,
      "memory(GiB)": 117.28,
      "step": 3420,
      "train_speed(iter/s)": 0.21435
    },
    {
      "acc": 0.76892452,
      "epoch": 0.08002449729509033,
      "grad_norm": 5.21875,
      "learning_rate": 9.997373050101265e-06,
      "loss": 0.84681721,
      "memory(GiB)": 117.28,
      "step": 3430,
      "train_speed(iter/s)": 0.21465
    },
    {
      "acc": 0.74426117,
      "epoch": 0.08025780486737923,
      "grad_norm": 7.25,
      "learning_rate": 9.997311464515113e-06,
      "loss": 0.94508867,
      "memory(GiB)": 117.28,
      "step": 3440,
      "train_speed(iter/s)": 0.214929
    },
    {
      "acc": 0.7614994,
      "epoch": 0.08049111243966812,
      "grad_norm": 5.625,
      "learning_rate": 9.997249165564203e-06,
      "loss": 0.83586025,
      "memory(GiB)": 117.28,
      "step": 3450,
      "train_speed(iter/s)": 0.215237
    },
    {
      "acc": 0.75516629,
      "epoch": 0.08072442001195701,
      "grad_norm": 5.40625,
      "learning_rate": 9.997186153257425e-06,
      "loss": 0.89334755,
      "memory(GiB)": 117.28,
      "step": 3460,
      "train_speed(iter/s)": 0.215538
    },
    {
      "acc": 0.7614068,
      "epoch": 0.08095772758424591,
      "grad_norm": 4.96875,
      "learning_rate": 9.997122427603777e-06,
      "loss": 0.87028713,
      "memory(GiB)": 117.28,
      "step": 3470,
      "train_speed(iter/s)": 0.215836
    },
    {
      "acc": 0.73593702,
      "epoch": 0.0811910351565348,
      "grad_norm": 5.96875,
      "learning_rate": 9.997057988612351e-06,
      "loss": 0.96313515,
      "memory(GiB)": 117.28,
      "step": 3480,
      "train_speed(iter/s)": 0.21614
    },
    {
      "acc": 0.74797134,
      "epoch": 0.08142434272882369,
      "grad_norm": 6.3125,
      "learning_rate": 9.996992836292352e-06,
      "loss": 0.92687082,
      "memory(GiB)": 117.28,
      "step": 3490,
      "train_speed(iter/s)": 0.216442
    },
    {
      "acc": 0.77501402,
      "epoch": 0.08165765030111259,
      "grad_norm": 15.3125,
      "learning_rate": 9.996926970653076e-06,
      "loss": 0.80666208,
      "memory(GiB)": 117.28,
      "step": 3500,
      "train_speed(iter/s)": 0.216742
    },
    {
      "epoch": 0.08165765030111259,
      "eval_acc": 0.7207238327126944,
      "eval_loss": 0.8971706628799438,
      "eval_runtime": 1269.6947,
      "eval_samples_per_second": 28.346,
      "eval_steps_per_second": 14.173,
      "step": 3500
    },
    {
      "acc": 0.75318351,
      "epoch": 0.08189095787340148,
      "grad_norm": 4.59375,
      "learning_rate": 9.996860391703925e-06,
      "loss": 0.9090744,
      "memory(GiB)": 117.28,
      "step": 3510,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.76253424,
      "epoch": 0.08212426544569038,
      "grad_norm": 6.125,
      "learning_rate": 9.996793099454407e-06,
      "loss": 0.87270107,
      "memory(GiB)": 117.28,
      "step": 3520,
      "train_speed(iter/s)": 0.20133
    },
    {
      "acc": 0.76890745,
      "epoch": 0.08235757301797926,
      "grad_norm": 6.375,
      "learning_rate": 9.996725093914125e-06,
      "loss": 0.846912,
      "memory(GiB)": 117.28,
      "step": 3530,
      "train_speed(iter/s)": 0.20162
    },
    {
      "acc": 0.75093694,
      "epoch": 0.08259088059026816,
      "grad_norm": 6.40625,
      "learning_rate": 9.996656375092786e-06,
      "loss": 0.91460123,
      "memory(GiB)": 117.28,
      "step": 3540,
      "train_speed(iter/s)": 0.201922
    },
    {
      "acc": 0.74694242,
      "epoch": 0.08282418816255704,
      "grad_norm": 8.25,
      "learning_rate": 9.996586943000203e-06,
      "loss": 0.94491138,
      "memory(GiB)": 117.28,
      "step": 3550,
      "train_speed(iter/s)": 0.202225
    },
    {
      "acc": 0.7550127,
      "epoch": 0.08305749573484594,
      "grad_norm": 6.8125,
      "learning_rate": 9.996516797646285e-06,
      "loss": 0.89832973,
      "memory(GiB)": 117.28,
      "step": 3560,
      "train_speed(iter/s)": 0.202507
    },
    {
      "acc": 0.76115112,
      "epoch": 0.08329080330713484,
      "grad_norm": 7.375,
      "learning_rate": 9.996445939041043e-06,
      "loss": 0.8638464,
      "memory(GiB)": 117.28,
      "step": 3570,
      "train_speed(iter/s)": 0.20277
    },
    {
      "acc": 0.7447835,
      "epoch": 0.08352411087942373,
      "grad_norm": 8.6875,
      "learning_rate": 9.996374367194599e-06,
      "loss": 0.94378824,
      "memory(GiB)": 117.28,
      "step": 3580,
      "train_speed(iter/s)": 0.203058
    },
    {
      "acc": 0.72563133,
      "epoch": 0.08375741845171263,
      "grad_norm": 38.5,
      "learning_rate": 9.996302082117162e-06,
      "loss": 1.1059824,
      "memory(GiB)": 117.28,
      "step": 3590,
      "train_speed(iter/s)": 0.203348
    },
    {
      "acc": 0.77572851,
      "epoch": 0.08399072602400151,
      "grad_norm": 19.75,
      "learning_rate": 9.996229083819055e-06,
      "loss": 0.79340849,
      "memory(GiB)": 117.28,
      "step": 3600,
      "train_speed(iter/s)": 0.203658
    },
    {
      "acc": 0.74407644,
      "epoch": 0.08422403359629041,
      "grad_norm": 4.90625,
      "learning_rate": 9.996155372310699e-06,
      "loss": 0.92524662,
      "memory(GiB)": 117.28,
      "step": 3610,
      "train_speed(iter/s)": 0.203955
    },
    {
      "acc": 0.74106493,
      "epoch": 0.08445734116857931,
      "grad_norm": 5.84375,
      "learning_rate": 9.996080947602615e-06,
      "loss": 0.96179218,
      "memory(GiB)": 117.28,
      "step": 3620,
      "train_speed(iter/s)": 0.204248
    },
    {
      "acc": 0.73815827,
      "epoch": 0.0846906487408682,
      "grad_norm": 6.125,
      "learning_rate": 9.996005809705428e-06,
      "loss": 0.95186195,
      "memory(GiB)": 117.28,
      "step": 3630,
      "train_speed(iter/s)": 0.204526
    },
    {
      "acc": 0.75756698,
      "epoch": 0.08492395631315709,
      "grad_norm": 8.125,
      "learning_rate": 9.99592995862986e-06,
      "loss": 0.8868536,
      "memory(GiB)": 117.28,
      "step": 3640,
      "train_speed(iter/s)": 0.204816
    },
    {
      "acc": 0.7773509,
      "epoch": 0.08515726388544598,
      "grad_norm": 5.6875,
      "learning_rate": 9.995853394386743e-06,
      "loss": 0.80111885,
      "memory(GiB)": 117.28,
      "step": 3650,
      "train_speed(iter/s)": 0.205093
    },
    {
      "acc": 0.75153332,
      "epoch": 0.08539057145773488,
      "grad_norm": 7.28125,
      "learning_rate": 9.995776116987006e-06,
      "loss": 0.92716045,
      "memory(GiB)": 117.28,
      "step": 3660,
      "train_speed(iter/s)": 0.205357
    },
    {
      "acc": 0.76465082,
      "epoch": 0.08562387903002378,
      "grad_norm": 5.53125,
      "learning_rate": 9.995698126441678e-06,
      "loss": 0.84658279,
      "memory(GiB)": 117.28,
      "step": 3670,
      "train_speed(iter/s)": 0.205626
    },
    {
      "acc": 0.75522108,
      "epoch": 0.08585718660231266,
      "grad_norm": 7.6875,
      "learning_rate": 9.995619422761896e-06,
      "loss": 0.89340916,
      "memory(GiB)": 117.28,
      "step": 3680,
      "train_speed(iter/s)": 0.205898
    },
    {
      "acc": 0.76193209,
      "epoch": 0.08609049417460156,
      "grad_norm": 5.34375,
      "learning_rate": 9.995540005958891e-06,
      "loss": 0.85231638,
      "memory(GiB)": 117.28,
      "step": 3690,
      "train_speed(iter/s)": 0.206188
    },
    {
      "acc": 0.74403105,
      "epoch": 0.08632380174689044,
      "grad_norm": 5.1875,
      "learning_rate": 9.995459876044e-06,
      "loss": 0.96531124,
      "memory(GiB)": 117.28,
      "step": 3700,
      "train_speed(iter/s)": 0.206473
    },
    {
      "acc": 0.76611438,
      "epoch": 0.08655710931917934,
      "grad_norm": 4.9375,
      "learning_rate": 9.995379033028666e-06,
      "loss": 0.85632906,
      "memory(GiB)": 117.28,
      "step": 3710,
      "train_speed(iter/s)": 0.206772
    },
    {
      "acc": 0.74768972,
      "epoch": 0.08679041689146823,
      "grad_norm": 4.46875,
      "learning_rate": 9.995297476924424e-06,
      "loss": 0.94205217,
      "memory(GiB)": 117.28,
      "step": 3720,
      "train_speed(iter/s)": 0.207058
    },
    {
      "acc": 0.75818162,
      "epoch": 0.08702372446375713,
      "grad_norm": 4.78125,
      "learning_rate": 9.99521520774292e-06,
      "loss": 0.88370638,
      "memory(GiB)": 117.28,
      "step": 3730,
      "train_speed(iter/s)": 0.207322
    },
    {
      "acc": 0.76598167,
      "epoch": 0.08725703203604603,
      "grad_norm": 4.625,
      "learning_rate": 9.995132225495896e-06,
      "loss": 0.8380312,
      "memory(GiB)": 117.28,
      "step": 3740,
      "train_speed(iter/s)": 0.207593
    },
    {
      "acc": 0.74331932,
      "epoch": 0.08749033960833491,
      "grad_norm": 4.71875,
      "learning_rate": 9.995048530195198e-06,
      "loss": 0.94072809,
      "memory(GiB)": 117.28,
      "step": 3750,
      "train_speed(iter/s)": 0.207867
    },
    {
      "acc": 0.74893398,
      "epoch": 0.08772364718062381,
      "grad_norm": 8.9375,
      "learning_rate": 9.99496412185277e-06,
      "loss": 0.93704576,
      "memory(GiB)": 117.28,
      "step": 3760,
      "train_speed(iter/s)": 0.208126
    },
    {
      "acc": 0.74125042,
      "epoch": 0.0879569547529127,
      "grad_norm": 8.25,
      "learning_rate": 9.994879000480668e-06,
      "loss": 0.96825314,
      "memory(GiB)": 117.28,
      "step": 3770,
      "train_speed(iter/s)": 0.208406
    },
    {
      "acc": 0.77773728,
      "epoch": 0.0881902623252016,
      "grad_norm": 6.5625,
      "learning_rate": 9.994793166091039e-06,
      "loss": 0.81420488,
      "memory(GiB)": 117.28,
      "step": 3780,
      "train_speed(iter/s)": 0.208696
    },
    {
      "acc": 0.75972219,
      "epoch": 0.08842356989749049,
      "grad_norm": 6.59375,
      "learning_rate": 9.994706618696137e-06,
      "loss": 0.89639206,
      "memory(GiB)": 117.28,
      "step": 3790,
      "train_speed(iter/s)": 0.208963
    },
    {
      "acc": 0.76652493,
      "epoch": 0.08865687746977938,
      "grad_norm": 7.15625,
      "learning_rate": 9.994619358308316e-06,
      "loss": 0.83598127,
      "memory(GiB)": 117.28,
      "step": 3800,
      "train_speed(iter/s)": 0.209241
    },
    {
      "acc": 0.76296563,
      "epoch": 0.08889018504206828,
      "grad_norm": 10.4375,
      "learning_rate": 9.994531384940032e-06,
      "loss": 0.85312195,
      "memory(GiB)": 117.28,
      "step": 3810,
      "train_speed(iter/s)": 0.209509
    },
    {
      "acc": 0.76371479,
      "epoch": 0.08912349261435716,
      "grad_norm": 7.96875,
      "learning_rate": 9.994442698603844e-06,
      "loss": 0.8648243,
      "memory(GiB)": 117.28,
      "step": 3820,
      "train_speed(iter/s)": 0.209773
    },
    {
      "acc": 0.76453638,
      "epoch": 0.08935680018664606,
      "grad_norm": 7.59375,
      "learning_rate": 9.99435329931241e-06,
      "loss": 0.85945387,
      "memory(GiB)": 117.28,
      "step": 3830,
      "train_speed(iter/s)": 0.210032
    },
    {
      "acc": 0.76211448,
      "epoch": 0.08959010775893494,
      "grad_norm": 7.1875,
      "learning_rate": 9.994263187078496e-06,
      "loss": 0.86692085,
      "memory(GiB)": 117.28,
      "step": 3840,
      "train_speed(iter/s)": 0.210304
    },
    {
      "acc": 0.76937609,
      "epoch": 0.08982341533122384,
      "grad_norm": 4.90625,
      "learning_rate": 9.994172361914962e-06,
      "loss": 0.84302549,
      "memory(GiB)": 117.28,
      "step": 3850,
      "train_speed(iter/s)": 0.210561
    },
    {
      "acc": 0.76893077,
      "epoch": 0.09005672290351274,
      "grad_norm": 5.21875,
      "learning_rate": 9.994080823834775e-06,
      "loss": 0.84133387,
      "memory(GiB)": 117.54,
      "step": 3860,
      "train_speed(iter/s)": 0.21084
    },
    {
      "acc": 0.75491571,
      "epoch": 0.09029003047580163,
      "grad_norm": 5.53125,
      "learning_rate": 9.993988572851e-06,
      "loss": 0.90440521,
      "memory(GiB)": 117.54,
      "step": 3870,
      "train_speed(iter/s)": 0.2111
    },
    {
      "acc": 0.74396949,
      "epoch": 0.09052333804809053,
      "grad_norm": 8.0625,
      "learning_rate": 9.993895608976806e-06,
      "loss": 0.95679359,
      "memory(GiB)": 117.54,
      "step": 3880,
      "train_speed(iter/s)": 0.211355
    },
    {
      "acc": 0.75566082,
      "epoch": 0.09075664562037941,
      "grad_norm": 8.9375,
      "learning_rate": 9.993801932225466e-06,
      "loss": 0.89417982,
      "memory(GiB)": 117.54,
      "step": 3890,
      "train_speed(iter/s)": 0.211633
    },
    {
      "acc": 0.76257367,
      "epoch": 0.09098995319266831,
      "grad_norm": 5.625,
      "learning_rate": 9.993707542610351e-06,
      "loss": 0.87649937,
      "memory(GiB)": 117.54,
      "step": 3900,
      "train_speed(iter/s)": 0.211899
    },
    {
      "acc": 0.73606167,
      "epoch": 0.09122326076495721,
      "grad_norm": 7.34375,
      "learning_rate": 9.993612440144935e-06,
      "loss": 0.9821991,
      "memory(GiB)": 117.54,
      "step": 3910,
      "train_speed(iter/s)": 0.212158
    },
    {
      "acc": 0.75946436,
      "epoch": 0.0914565683372461,
      "grad_norm": 6.03125,
      "learning_rate": 9.993516624842792e-06,
      "loss": 0.8709816,
      "memory(GiB)": 117.54,
      "step": 3920,
      "train_speed(iter/s)": 0.212416
    },
    {
      "acc": 0.74540901,
      "epoch": 0.09168987590953499,
      "grad_norm": 8.5,
      "learning_rate": 9.993420096717603e-06,
      "loss": 0.94000492,
      "memory(GiB)": 117.54,
      "step": 3930,
      "train_speed(iter/s)": 0.212684
    },
    {
      "acc": 0.77204304,
      "epoch": 0.09192318348182388,
      "grad_norm": 8.0,
      "learning_rate": 9.993322855783146e-06,
      "loss": 0.81134806,
      "memory(GiB)": 117.54,
      "step": 3940,
      "train_speed(iter/s)": 0.212916
    },
    {
      "acc": 0.75332932,
      "epoch": 0.09215649105411278,
      "grad_norm": 7.5,
      "learning_rate": 9.993224902053302e-06,
      "loss": 0.88411512,
      "memory(GiB)": 117.54,
      "step": 3950,
      "train_speed(iter/s)": 0.213189
    },
    {
      "acc": 0.73475533,
      "epoch": 0.09238979862640166,
      "grad_norm": 4.6875,
      "learning_rate": 9.993126235542053e-06,
      "loss": 0.98314781,
      "memory(GiB)": 117.54,
      "step": 3960,
      "train_speed(iter/s)": 0.213459
    },
    {
      "acc": 0.76145267,
      "epoch": 0.09262310619869056,
      "grad_norm": 8.375,
      "learning_rate": 9.993026856263486e-06,
      "loss": 0.86928034,
      "memory(GiB)": 117.54,
      "step": 3970,
      "train_speed(iter/s)": 0.213725
    },
    {
      "acc": 0.76495628,
      "epoch": 0.09285641377097946,
      "grad_norm": 5.5625,
      "learning_rate": 9.992926764231784e-06,
      "loss": 0.83155947,
      "memory(GiB)": 117.54,
      "step": 3980,
      "train_speed(iter/s)": 0.213988
    },
    {
      "acc": 0.75666361,
      "epoch": 0.09308972134326834,
      "grad_norm": 6.84375,
      "learning_rate": 9.992825959461237e-06,
      "loss": 0.88326521,
      "memory(GiB)": 117.54,
      "step": 3990,
      "train_speed(iter/s)": 0.214237
    },
    {
      "acc": 0.774965,
      "epoch": 0.09332302891555724,
      "grad_norm": 5.125,
      "learning_rate": 9.992724441966234e-06,
      "loss": 0.8087532,
      "memory(GiB)": 117.54,
      "step": 4000,
      "train_speed(iter/s)": 0.214481
    },
    {
      "epoch": 0.09332302891555724,
      "eval_acc": 0.7228697654255328,
      "eval_loss": 0.8877829313278198,
      "eval_runtime": 1271.4977,
      "eval_samples_per_second": 28.306,
      "eval_steps_per_second": 14.153,
      "step": 4000
    },
    {
      "acc": 0.76409369,
      "epoch": 0.09355633648784613,
      "grad_norm": 4.625,
      "learning_rate": 9.99262221176127e-06,
      "loss": 0.85309639,
      "memory(GiB)": 117.54,
      "step": 4010,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.76466002,
      "epoch": 0.09378964406013503,
      "grad_norm": 5.84375,
      "learning_rate": 9.992519268860934e-06,
      "loss": 0.87510233,
      "memory(GiB)": 117.54,
      "step": 4020,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.74888363,
      "epoch": 0.09402295163242393,
      "grad_norm": 4.75,
      "learning_rate": 9.992415613279922e-06,
      "loss": 0.91498604,
      "memory(GiB)": 117.54,
      "step": 4030,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.7692903,
      "epoch": 0.09425625920471281,
      "grad_norm": 4.34375,
      "learning_rate": 9.992311245033033e-06,
      "loss": 0.83095226,
      "memory(GiB)": 117.54,
      "step": 4040,
      "train_speed(iter/s)": 0.201636
    },
    {
      "acc": 0.7509232,
      "epoch": 0.09448956677700171,
      "grad_norm": 5.0,
      "learning_rate": 9.992206164135163e-06,
      "loss": 0.92139778,
      "memory(GiB)": 117.54,
      "step": 4050,
      "train_speed(iter/s)": 0.201873
    },
    {
      "acc": 0.77602358,
      "epoch": 0.0947228743492906,
      "grad_norm": 8.125,
      "learning_rate": 9.992100370601313e-06,
      "loss": 0.79812717,
      "memory(GiB)": 117.54,
      "step": 4060,
      "train_speed(iter/s)": 0.202133
    },
    {
      "acc": 0.7557579,
      "epoch": 0.0949561819215795,
      "grad_norm": 6.875,
      "learning_rate": 9.991993864446585e-06,
      "loss": 0.89294071,
      "memory(GiB)": 117.54,
      "step": 4070,
      "train_speed(iter/s)": 0.202411
    },
    {
      "acc": 0.73773766,
      "epoch": 0.09518948949386838,
      "grad_norm": 5.46875,
      "learning_rate": 9.991886645686184e-06,
      "loss": 0.95771408,
      "memory(GiB)": 117.54,
      "step": 4080,
      "train_speed(iter/s)": 0.20265
    },
    {
      "acc": 0.75475817,
      "epoch": 0.09542279706615728,
      "grad_norm": 8.3125,
      "learning_rate": 9.991778714335415e-06,
      "loss": 0.90779686,
      "memory(GiB)": 117.54,
      "step": 4090,
      "train_speed(iter/s)": 0.202915
    },
    {
      "acc": 0.73981633,
      "epoch": 0.09565610463844618,
      "grad_norm": 6.90625,
      "learning_rate": 9.991670070409684e-06,
      "loss": 0.96604061,
      "memory(GiB)": 117.54,
      "step": 4100,
      "train_speed(iter/s)": 0.203177
    },
    {
      "acc": 0.76607499,
      "epoch": 0.09588941221073506,
      "grad_norm": 5.78125,
      "learning_rate": 9.991560713924501e-06,
      "loss": 0.85471153,
      "memory(GiB)": 117.54,
      "step": 4110,
      "train_speed(iter/s)": 0.20342
    },
    {
      "acc": 0.77161045,
      "epoch": 0.09612271978302396,
      "grad_norm": 8.125,
      "learning_rate": 9.991450644895476e-06,
      "loss": 0.84620152,
      "memory(GiB)": 117.54,
      "step": 4120,
      "train_speed(iter/s)": 0.203677
    },
    {
      "acc": 0.77144375,
      "epoch": 0.09635602735531285,
      "grad_norm": 4.625,
      "learning_rate": 9.99133986333832e-06,
      "loss": 0.83797445,
      "memory(GiB)": 117.54,
      "step": 4130,
      "train_speed(iter/s)": 0.20393
    },
    {
      "acc": 0.73537869,
      "epoch": 0.09658933492760174,
      "grad_norm": 6.3125,
      "learning_rate": 9.99122836926885e-06,
      "loss": 0.96916094,
      "memory(GiB)": 117.54,
      "step": 4140,
      "train_speed(iter/s)": 0.204181
    },
    {
      "acc": 0.75800476,
      "epoch": 0.09682264249989064,
      "grad_norm": 5.625,
      "learning_rate": 9.991116162702981e-06,
      "loss": 0.86221294,
      "memory(GiB)": 117.54,
      "step": 4150,
      "train_speed(iter/s)": 0.204423
    },
    {
      "acc": 0.75571017,
      "epoch": 0.09705595007217953,
      "grad_norm": 5.5,
      "learning_rate": 9.991003243656728e-06,
      "loss": 0.88189182,
      "memory(GiB)": 117.54,
      "step": 4160,
      "train_speed(iter/s)": 0.204672
    },
    {
      "acc": 0.75988812,
      "epoch": 0.09728925764446843,
      "grad_norm": 8.5,
      "learning_rate": 9.990889612146213e-06,
      "loss": 0.88356085,
      "memory(GiB)": 117.54,
      "step": 4170,
      "train_speed(iter/s)": 0.204882
    },
    {
      "acc": 0.74842806,
      "epoch": 0.09752256521675731,
      "grad_norm": 4.96875,
      "learning_rate": 9.990775268187654e-06,
      "loss": 0.93167324,
      "memory(GiB)": 117.54,
      "step": 4180,
      "train_speed(iter/s)": 0.20513
    },
    {
      "acc": 0.76515474,
      "epoch": 0.09775587278904621,
      "grad_norm": 35.75,
      "learning_rate": 9.990660211797378e-06,
      "loss": 0.92051849,
      "memory(GiB)": 117.54,
      "step": 4190,
      "train_speed(iter/s)": 0.205376
    },
    {
      "acc": 0.7553896,
      "epoch": 0.0979891803613351,
      "grad_norm": 5.375,
      "learning_rate": 9.990544442991805e-06,
      "loss": 0.89758406,
      "memory(GiB)": 117.54,
      "step": 4200,
      "train_speed(iter/s)": 0.20564
    },
    {
      "acc": 0.75030479,
      "epoch": 0.098222487933624,
      "grad_norm": 5.3125,
      "learning_rate": 9.99042796178746e-06,
      "loss": 0.91539536,
      "memory(GiB)": 117.54,
      "step": 4210,
      "train_speed(iter/s)": 0.205895
    },
    {
      "acc": 0.74988375,
      "epoch": 0.0984557955059129,
      "grad_norm": 4.25,
      "learning_rate": 9.990310768200977e-06,
      "loss": 0.9145546,
      "memory(GiB)": 117.54,
      "step": 4220,
      "train_speed(iter/s)": 0.206142
    },
    {
      "acc": 0.73188667,
      "epoch": 0.09868910307820178,
      "grad_norm": 5.65625,
      "learning_rate": 9.99019286224908e-06,
      "loss": 0.98185387,
      "memory(GiB)": 117.54,
      "step": 4230,
      "train_speed(iter/s)": 0.20638
    },
    {
      "acc": 0.75859137,
      "epoch": 0.09892241065049068,
      "grad_norm": 8.125,
      "learning_rate": 9.990074243948602e-06,
      "loss": 0.90473709,
      "memory(GiB)": 117.54,
      "step": 4240,
      "train_speed(iter/s)": 0.206628
    },
    {
      "acc": 0.76608834,
      "epoch": 0.09915571822277956,
      "grad_norm": 6.34375,
      "learning_rate": 9.989954913316476e-06,
      "loss": 0.85702991,
      "memory(GiB)": 117.54,
      "step": 4250,
      "train_speed(iter/s)": 0.20688
    },
    {
      "acc": 0.76376286,
      "epoch": 0.09938902579506846,
      "grad_norm": 9.625,
      "learning_rate": 9.989834870369735e-06,
      "loss": 0.84923592,
      "memory(GiB)": 117.54,
      "step": 4260,
      "train_speed(iter/s)": 0.207135
    },
    {
      "acc": 0.77528548,
      "epoch": 0.09962233336735736,
      "grad_norm": 5.1875,
      "learning_rate": 9.989714115125515e-06,
      "loss": 0.79958353,
      "memory(GiB)": 117.54,
      "step": 4270,
      "train_speed(iter/s)": 0.207364
    },
    {
      "acc": 0.75993328,
      "epoch": 0.09985564093964625,
      "grad_norm": 10.625,
      "learning_rate": 9.989592647601056e-06,
      "loss": 0.86611481,
      "memory(GiB)": 117.54,
      "step": 4280,
      "train_speed(iter/s)": 0.207609
    },
    {
      "acc": 0.75393734,
      "epoch": 0.10008894851193514,
      "grad_norm": 6.5625,
      "learning_rate": 9.989470467813696e-06,
      "loss": 0.91193056,
      "memory(GiB)": 117.54,
      "step": 4290,
      "train_speed(iter/s)": 0.207846
    },
    {
      "acc": 0.75926132,
      "epoch": 0.10032225608422403,
      "grad_norm": 8.0,
      "learning_rate": 9.989347575780874e-06,
      "loss": 0.88821259,
      "memory(GiB)": 117.54,
      "step": 4300,
      "train_speed(iter/s)": 0.208096
    },
    {
      "acc": 0.75386171,
      "epoch": 0.10055556365651293,
      "grad_norm": 7.65625,
      "learning_rate": 9.989223971520136e-06,
      "loss": 0.89129429,
      "memory(GiB)": 117.54,
      "step": 4310,
      "train_speed(iter/s)": 0.208334
    },
    {
      "acc": 0.75681725,
      "epoch": 0.10078887122880181,
      "grad_norm": 6.65625,
      "learning_rate": 9.989099655049128e-06,
      "loss": 0.87771931,
      "memory(GiB)": 117.54,
      "step": 4320,
      "train_speed(iter/s)": 0.20856
    },
    {
      "acc": 0.7391355,
      "epoch": 0.10102217880109071,
      "grad_norm": 10.0,
      "learning_rate": 9.98897462638559e-06,
      "loss": 0.96863251,
      "memory(GiB)": 117.54,
      "step": 4330,
      "train_speed(iter/s)": 0.20879
    },
    {
      "acc": 0.76924124,
      "epoch": 0.10125548637337961,
      "grad_norm": 5.78125,
      "learning_rate": 9.988848885547376e-06,
      "loss": 0.8352129,
      "memory(GiB)": 117.54,
      "step": 4340,
      "train_speed(iter/s)": 0.209021
    },
    {
      "acc": 0.77654471,
      "epoch": 0.1014887939456685,
      "grad_norm": 7.375,
      "learning_rate": 9.988722432552431e-06,
      "loss": 0.79848795,
      "memory(GiB)": 117.54,
      "step": 4350,
      "train_speed(iter/s)": 0.209261
    },
    {
      "acc": 0.74507241,
      "epoch": 0.1017221015179574,
      "grad_norm": 5.625,
      "learning_rate": 9.988595267418809e-06,
      "loss": 0.92341309,
      "memory(GiB)": 117.54,
      "step": 4360,
      "train_speed(iter/s)": 0.209503
    },
    {
      "acc": 0.76773772,
      "epoch": 0.10195540909024628,
      "grad_norm": 5.4375,
      "learning_rate": 9.988467390164662e-06,
      "loss": 0.84407911,
      "memory(GiB)": 117.54,
      "step": 4370,
      "train_speed(iter/s)": 0.209715
    },
    {
      "acc": 0.74648399,
      "epoch": 0.10218871666253518,
      "grad_norm": 5.0625,
      "learning_rate": 9.988338800808245e-06,
      "loss": 0.93966646,
      "memory(GiB)": 117.54,
      "step": 4380,
      "train_speed(iter/s)": 0.209944
    },
    {
      "acc": 0.76602821,
      "epoch": 0.10242202423482408,
      "grad_norm": 6.6875,
      "learning_rate": 9.988209499367911e-06,
      "loss": 0.8338089,
      "memory(GiB)": 117.54,
      "step": 4390,
      "train_speed(iter/s)": 0.210194
    },
    {
      "acc": 0.76203928,
      "epoch": 0.10265533180711296,
      "grad_norm": 4.34375,
      "learning_rate": 9.988079485862121e-06,
      "loss": 0.87073536,
      "memory(GiB)": 117.54,
      "step": 4400,
      "train_speed(iter/s)": 0.210432
    },
    {
      "acc": 0.74238539,
      "epoch": 0.10288863937940186,
      "grad_norm": 12.375,
      "learning_rate": 9.987948760309434e-06,
      "loss": 0.95618248,
      "memory(GiB)": 117.54,
      "step": 4410,
      "train_speed(iter/s)": 0.210669
    },
    {
      "acc": 0.7657311,
      "epoch": 0.10312194695169075,
      "grad_norm": 5.4375,
      "learning_rate": 9.987817322728509e-06,
      "loss": 0.85881863,
      "memory(GiB)": 117.54,
      "step": 4420,
      "train_speed(iter/s)": 0.210901
    },
    {
      "acc": 0.76233692,
      "epoch": 0.10335525452397964,
      "grad_norm": 4.4375,
      "learning_rate": 9.98768517313811e-06,
      "loss": 0.8707386,
      "memory(GiB)": 117.54,
      "step": 4430,
      "train_speed(iter/s)": 0.211135
    },
    {
      "acc": 0.75146971,
      "epoch": 0.10358856209626854,
      "grad_norm": 6.9375,
      "learning_rate": 9.987552311557103e-06,
      "loss": 0.90720158,
      "memory(GiB)": 117.54,
      "step": 4440,
      "train_speed(iter/s)": 0.211368
    },
    {
      "acc": 0.77758303,
      "epoch": 0.10382186966855743,
      "grad_norm": 18.25,
      "learning_rate": 9.987418738004453e-06,
      "loss": 0.81924438,
      "memory(GiB)": 117.54,
      "step": 4450,
      "train_speed(iter/s)": 0.211579
    },
    {
      "acc": 0.73662348,
      "epoch": 0.10405517724084633,
      "grad_norm": 5.59375,
      "learning_rate": 9.987284452499227e-06,
      "loss": 0.9785923,
      "memory(GiB)": 117.54,
      "step": 4460,
      "train_speed(iter/s)": 0.211822
    },
    {
      "acc": 0.74656315,
      "epoch": 0.10428848481313521,
      "grad_norm": 17.875,
      "learning_rate": 9.987149455060592e-06,
      "loss": 0.9363677,
      "memory(GiB)": 117.54,
      "step": 4470,
      "train_speed(iter/s)": 0.212055
    },
    {
      "acc": 0.75212345,
      "epoch": 0.10452179238542411,
      "grad_norm": 4.71875,
      "learning_rate": 9.987013745707824e-06,
      "loss": 0.93190804,
      "memory(GiB)": 117.54,
      "step": 4480,
      "train_speed(iter/s)": 0.212285
    },
    {
      "acc": 0.7772028,
      "epoch": 0.104755099957713,
      "grad_norm": 4.46875,
      "learning_rate": 9.986877324460288e-06,
      "loss": 0.79044576,
      "memory(GiB)": 117.54,
      "step": 4490,
      "train_speed(iter/s)": 0.212502
    },
    {
      "acc": 0.76467466,
      "epoch": 0.1049884075300019,
      "grad_norm": 4.65625,
      "learning_rate": 9.986740191337467e-06,
      "loss": 0.86383896,
      "memory(GiB)": 117.54,
      "step": 4500,
      "train_speed(iter/s)": 0.212726
    },
    {
      "epoch": 0.1049884075300019,
      "eval_acc": 0.7243065794777408,
      "eval_loss": 0.8816430568695068,
      "eval_runtime": 1268.0964,
      "eval_samples_per_second": 28.382,
      "eval_steps_per_second": 14.191,
      "step": 4500
    },
    {
      "acc": 0.74907675,
      "epoch": 0.1052217151022908,
      "grad_norm": 4.90625,
      "learning_rate": 9.986602346358932e-06,
      "loss": 0.92465839,
      "memory(GiB)": 117.54,
      "step": 4510,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.77655468,
      "epoch": 0.10545502267457968,
      "grad_norm": 5.6875,
      "learning_rate": 9.986463789544359e-06,
      "loss": 0.79614592,
      "memory(GiB)": 117.54,
      "step": 4520,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.74242306,
      "epoch": 0.10568833024686858,
      "grad_norm": 5.625,
      "learning_rate": 9.986324520913528e-06,
      "loss": 0.9382905,
      "memory(GiB)": 117.54,
      "step": 4530,
      "train_speed(iter/s)": 0.201278
    },
    {
      "acc": 0.7789465,
      "epoch": 0.10592163781915746,
      "grad_norm": 5.1875,
      "learning_rate": 9.986184540486322e-06,
      "loss": 0.80827017,
      "memory(GiB)": 117.54,
      "step": 4540,
      "train_speed(iter/s)": 0.201497
    },
    {
      "acc": 0.78276606,
      "epoch": 0.10615494539144636,
      "grad_norm": 5.9375,
      "learning_rate": 9.98604384828272e-06,
      "loss": 0.78076072,
      "memory(GiB)": 117.54,
      "step": 4550,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.75945024,
      "epoch": 0.10638825296373526,
      "grad_norm": 5.375,
      "learning_rate": 9.985902444322809e-06,
      "loss": 0.86157227,
      "memory(GiB)": 117.54,
      "step": 4560,
      "train_speed(iter/s)": 0.201954
    },
    {
      "acc": 0.75581503,
      "epoch": 0.10662156053602415,
      "grad_norm": 5.90625,
      "learning_rate": 9.98576032862677e-06,
      "loss": 0.89451599,
      "memory(GiB)": 117.54,
      "step": 4570,
      "train_speed(iter/s)": 0.202173
    },
    {
      "acc": 0.7623549,
      "epoch": 0.10685486810831304,
      "grad_norm": 5.3125,
      "learning_rate": 9.985617501214895e-06,
      "loss": 0.86804514,
      "memory(GiB)": 117.54,
      "step": 4580,
      "train_speed(iter/s)": 0.202413
    },
    {
      "acc": 0.75258703,
      "epoch": 0.10708817568060193,
      "grad_norm": 4.78125,
      "learning_rate": 9.985473962107568e-06,
      "loss": 0.916399,
      "memory(GiB)": 117.54,
      "step": 4590,
      "train_speed(iter/s)": 0.202637
    },
    {
      "acc": 0.75255208,
      "epoch": 0.10732148325289083,
      "grad_norm": 6.71875,
      "learning_rate": 9.985329711325282e-06,
      "loss": 0.90113773,
      "memory(GiB)": 117.54,
      "step": 4600,
      "train_speed(iter/s)": 0.20288
    },
    {
      "acc": 0.75763159,
      "epoch": 0.10755479082517971,
      "grad_norm": 6.875,
      "learning_rate": 9.985184748888627e-06,
      "loss": 0.91873503,
      "memory(GiB)": 117.54,
      "step": 4610,
      "train_speed(iter/s)": 0.203115
    },
    {
      "acc": 0.77346997,
      "epoch": 0.10778809839746861,
      "grad_norm": 5.4375,
      "learning_rate": 9.985039074818298e-06,
      "loss": 0.8059083,
      "memory(GiB)": 117.54,
      "step": 4620,
      "train_speed(iter/s)": 0.203336
    },
    {
      "acc": 0.75904865,
      "epoch": 0.10802140596975751,
      "grad_norm": 5.59375,
      "learning_rate": 9.98489268913509e-06,
      "loss": 0.87097445,
      "memory(GiB)": 117.54,
      "step": 4630,
      "train_speed(iter/s)": 0.203564
    },
    {
      "acc": 0.75039043,
      "epoch": 0.1082547135420464,
      "grad_norm": 6.65625,
      "learning_rate": 9.984745591859899e-06,
      "loss": 0.90553379,
      "memory(GiB)": 117.54,
      "step": 4640,
      "train_speed(iter/s)": 0.203807
    },
    {
      "acc": 0.76141634,
      "epoch": 0.1084880211143353,
      "grad_norm": 4.21875,
      "learning_rate": 9.98459778301372e-06,
      "loss": 0.8886508,
      "memory(GiB)": 117.54,
      "step": 4650,
      "train_speed(iter/s)": 0.20402
    },
    {
      "acc": 0.76583786,
      "epoch": 0.10872132868662418,
      "grad_norm": 9.625,
      "learning_rate": 9.984449262617659e-06,
      "loss": 0.86086102,
      "memory(GiB)": 117.54,
      "step": 4660,
      "train_speed(iter/s)": 0.20427
    },
    {
      "acc": 0.76089182,
      "epoch": 0.10895463625891308,
      "grad_norm": 6.09375,
      "learning_rate": 9.984300030692913e-06,
      "loss": 0.87257662,
      "memory(GiB)": 117.54,
      "step": 4670,
      "train_speed(iter/s)": 0.204504
    },
    {
      "acc": 0.75297747,
      "epoch": 0.10918794383120198,
      "grad_norm": 9.5,
      "learning_rate": 9.984150087260784e-06,
      "loss": 0.90016375,
      "memory(GiB)": 117.54,
      "step": 4680,
      "train_speed(iter/s)": 0.204719
    },
    {
      "acc": 0.74139366,
      "epoch": 0.10942125140349086,
      "grad_norm": 6.46875,
      "learning_rate": 9.983999432342679e-06,
      "loss": 0.9342598,
      "memory(GiB)": 117.54,
      "step": 4690,
      "train_speed(iter/s)": 0.204946
    },
    {
      "acc": 0.75930262,
      "epoch": 0.10965455897577976,
      "grad_norm": 5.75,
      "learning_rate": 9.983848065960103e-06,
      "loss": 0.8737071,
      "memory(GiB)": 117.54,
      "step": 4700,
      "train_speed(iter/s)": 0.205175
    },
    {
      "acc": 0.77567587,
      "epoch": 0.10988786654806865,
      "grad_norm": 8.125,
      "learning_rate": 9.983695988134662e-06,
      "loss": 0.79876671,
      "memory(GiB)": 117.54,
      "step": 4710,
      "train_speed(iter/s)": 0.205389
    },
    {
      "acc": 0.75503325,
      "epoch": 0.11012117412035755,
      "grad_norm": 6.03125,
      "learning_rate": 9.983543198888069e-06,
      "loss": 0.90450735,
      "memory(GiB)": 117.54,
      "step": 4720,
      "train_speed(iter/s)": 0.205611
    },
    {
      "acc": 0.75266876,
      "epoch": 0.11035448169264643,
      "grad_norm": 5.0625,
      "learning_rate": 9.98338969824213e-06,
      "loss": 0.9201149,
      "memory(GiB)": 117.54,
      "step": 4730,
      "train_speed(iter/s)": 0.205829
    },
    {
      "acc": 0.76564531,
      "epoch": 0.11058778926493533,
      "grad_norm": 6.625,
      "learning_rate": 9.98323548621876e-06,
      "loss": 0.85361395,
      "memory(GiB)": 117.54,
      "step": 4740,
      "train_speed(iter/s)": 0.206042
    },
    {
      "acc": 0.74129553,
      "epoch": 0.11082109683722423,
      "grad_norm": 9.4375,
      "learning_rate": 9.983080562839971e-06,
      "loss": 0.96250782,
      "memory(GiB)": 117.54,
      "step": 4750,
      "train_speed(iter/s)": 0.206248
    },
    {
      "acc": 0.76101308,
      "epoch": 0.11105440440951311,
      "grad_norm": 5.71875,
      "learning_rate": 9.982924928127881e-06,
      "loss": 0.85979137,
      "memory(GiB)": 117.54,
      "step": 4760,
      "train_speed(iter/s)": 0.206465
    },
    {
      "acc": 0.78050613,
      "epoch": 0.11128771198180201,
      "grad_norm": 5.28125,
      "learning_rate": 9.982768582104705e-06,
      "loss": 0.80721207,
      "memory(GiB)": 117.54,
      "step": 4770,
      "train_speed(iter/s)": 0.206675
    },
    {
      "acc": 0.74485474,
      "epoch": 0.1115210195540909,
      "grad_norm": 4.4375,
      "learning_rate": 9.98261152479276e-06,
      "loss": 0.95470104,
      "memory(GiB)": 117.54,
      "step": 4780,
      "train_speed(iter/s)": 0.20689
    },
    {
      "acc": 0.76254826,
      "epoch": 0.1117543271263798,
      "grad_norm": 5.9375,
      "learning_rate": 9.982453756214467e-06,
      "loss": 0.84586992,
      "memory(GiB)": 117.54,
      "step": 4790,
      "train_speed(iter/s)": 0.207109
    },
    {
      "acc": 0.76125956,
      "epoch": 0.1119876346986687,
      "grad_norm": 7.71875,
      "learning_rate": 9.982295276392349e-06,
      "loss": 0.88607883,
      "memory(GiB)": 117.54,
      "step": 4800,
      "train_speed(iter/s)": 0.207333
    },
    {
      "acc": 0.73547726,
      "epoch": 0.11222094227095758,
      "grad_norm": 5.8125,
      "learning_rate": 9.982136085349028e-06,
      "loss": 0.96996174,
      "memory(GiB)": 117.54,
      "step": 4810,
      "train_speed(iter/s)": 0.207542
    },
    {
      "acc": 0.77212734,
      "epoch": 0.11245424984324648,
      "grad_norm": 7.40625,
      "learning_rate": 9.981976183107227e-06,
      "loss": 0.83480778,
      "memory(GiB)": 117.54,
      "step": 4820,
      "train_speed(iter/s)": 0.20776
    },
    {
      "acc": 0.76515455,
      "epoch": 0.11268755741553536,
      "grad_norm": 7.1875,
      "learning_rate": 9.981815569689774e-06,
      "loss": 0.88458767,
      "memory(GiB)": 117.54,
      "step": 4830,
      "train_speed(iter/s)": 0.207986
    },
    {
      "acc": 0.73565607,
      "epoch": 0.11292086498782426,
      "grad_norm": 6.09375,
      "learning_rate": 9.981654245119594e-06,
      "loss": 0.98005428,
      "memory(GiB)": 117.54,
      "step": 4840,
      "train_speed(iter/s)": 0.208198
    },
    {
      "acc": 0.74796729,
      "epoch": 0.11315417256011315,
      "grad_norm": 7.5625,
      "learning_rate": 9.98149220941972e-06,
      "loss": 0.90888004,
      "memory(GiB)": 117.54,
      "step": 4850,
      "train_speed(iter/s)": 0.208403
    },
    {
      "acc": 0.76575413,
      "epoch": 0.11338748013240205,
      "grad_norm": 6.4375,
      "learning_rate": 9.981329462613278e-06,
      "loss": 0.8409193,
      "memory(GiB)": 117.54,
      "step": 4860,
      "train_speed(iter/s)": 0.208616
    },
    {
      "acc": 0.76135168,
      "epoch": 0.11362078770469095,
      "grad_norm": 4.96875,
      "learning_rate": 9.981166004723504e-06,
      "loss": 0.8730361,
      "memory(GiB)": 117.54,
      "step": 4870,
      "train_speed(iter/s)": 0.208832
    },
    {
      "acc": 0.75576491,
      "epoch": 0.11385409527697983,
      "grad_norm": 5.96875,
      "learning_rate": 9.981001835773729e-06,
      "loss": 0.86638641,
      "memory(GiB)": 117.54,
      "step": 4880,
      "train_speed(iter/s)": 0.20906
    },
    {
      "acc": 0.78793383,
      "epoch": 0.11408740284926873,
      "grad_norm": 6.34375,
      "learning_rate": 9.98083695578739e-06,
      "loss": 0.78035274,
      "memory(GiB)": 117.54,
      "step": 4890,
      "train_speed(iter/s)": 0.209275
    },
    {
      "acc": 0.76127825,
      "epoch": 0.11432071042155761,
      "grad_norm": 4.59375,
      "learning_rate": 9.980671364788022e-06,
      "loss": 0.91963701,
      "memory(GiB)": 117.54,
      "step": 4900,
      "train_speed(iter/s)": 0.20949
    },
    {
      "acc": 0.73467045,
      "epoch": 0.11455401799384651,
      "grad_norm": 5.53125,
      "learning_rate": 9.980505062799262e-06,
      "loss": 0.9670414,
      "memory(GiB)": 117.54,
      "step": 4910,
      "train_speed(iter/s)": 0.209715
    },
    {
      "acc": 0.78355579,
      "epoch": 0.11478732556613541,
      "grad_norm": 6.0,
      "learning_rate": 9.980338049844854e-06,
      "loss": 0.78485413,
      "memory(GiB)": 117.54,
      "step": 4920,
      "train_speed(iter/s)": 0.209923
    },
    {
      "acc": 0.74745569,
      "epoch": 0.1150206331384243,
      "grad_norm": 5.40625,
      "learning_rate": 9.980170325948633e-06,
      "loss": 0.92230692,
      "memory(GiB)": 117.54,
      "step": 4930,
      "train_speed(iter/s)": 0.210131
    },
    {
      "acc": 0.75931578,
      "epoch": 0.1152539407107132,
      "grad_norm": 6.53125,
      "learning_rate": 9.980001891134548e-06,
      "loss": 0.87786045,
      "memory(GiB)": 117.54,
      "step": 4940,
      "train_speed(iter/s)": 0.210361
    },
    {
      "acc": 0.7422245,
      "epoch": 0.11548724828300208,
      "grad_norm": 4.875,
      "learning_rate": 9.979832745426637e-06,
      "loss": 0.94051199,
      "memory(GiB)": 117.54,
      "step": 4950,
      "train_speed(iter/s)": 0.21057
    },
    {
      "acc": 0.75386519,
      "epoch": 0.11572055585529098,
      "grad_norm": 8.1875,
      "learning_rate": 9.97966288884905e-06,
      "loss": 0.89630547,
      "memory(GiB)": 117.54,
      "step": 4960,
      "train_speed(iter/s)": 0.210783
    },
    {
      "acc": 0.74389353,
      "epoch": 0.11595386342757986,
      "grad_norm": 6.375,
      "learning_rate": 9.979492321426032e-06,
      "loss": 0.93810577,
      "memory(GiB)": 117.54,
      "step": 4970,
      "train_speed(iter/s)": 0.210987
    },
    {
      "acc": 0.76075001,
      "epoch": 0.11618717099986876,
      "grad_norm": 5.4375,
      "learning_rate": 9.97932104318193e-06,
      "loss": 0.84013653,
      "memory(GiB)": 117.54,
      "step": 4980,
      "train_speed(iter/s)": 0.211194
    },
    {
      "acc": 0.77435451,
      "epoch": 0.11642047857215766,
      "grad_norm": 4.84375,
      "learning_rate": 9.979149054141197e-06,
      "loss": 0.80839243,
      "memory(GiB)": 117.54,
      "step": 4990,
      "train_speed(iter/s)": 0.211393
    },
    {
      "acc": 0.7572638,
      "epoch": 0.11665378614444655,
      "grad_norm": 7.3125,
      "learning_rate": 9.978976354328383e-06,
      "loss": 0.94224863,
      "memory(GiB)": 117.54,
      "step": 5000,
      "train_speed(iter/s)": 0.21161
    },
    {
      "epoch": 0.11665378614444655,
      "eval_acc": 0.7257033574718734,
      "eval_loss": 0.876083493232727,
      "eval_runtime": 1268.8673,
      "eval_samples_per_second": 28.365,
      "eval_steps_per_second": 14.183,
      "step": 5000
    },
    {
      "acc": 0.74820719,
      "epoch": 0.11688709371673545,
      "grad_norm": 4.8125,
      "learning_rate": 9.97880294376814e-06,
      "loss": 0.91807137,
      "memory(GiB)": 117.54,
      "step": 5010,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.75378785,
      "epoch": 0.11712040128902433,
      "grad_norm": 5.59375,
      "learning_rate": 9.978628822485224e-06,
      "loss": 0.92413845,
      "memory(GiB)": 117.54,
      "step": 5020,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.73901176,
      "epoch": 0.11735370886131323,
      "grad_norm": 7.34375,
      "learning_rate": 9.978453990504488e-06,
      "loss": 0.95172367,
      "memory(GiB)": 117.54,
      "step": 5030,
      "train_speed(iter/s)": 0.201322
    },
    {
      "acc": 0.75668182,
      "epoch": 0.11758701643360213,
      "grad_norm": 5.25,
      "learning_rate": 9.978278447850894e-06,
      "loss": 0.89718676,
      "memory(GiB)": 117.54,
      "step": 5040,
      "train_speed(iter/s)": 0.201526
    },
    {
      "acc": 0.75949087,
      "epoch": 0.11782032400589101,
      "grad_norm": 5.59375,
      "learning_rate": 9.978102194549498e-06,
      "loss": 0.89434929,
      "memory(GiB)": 117.54,
      "step": 5050,
      "train_speed(iter/s)": 0.201738
    },
    {
      "acc": 0.75701513,
      "epoch": 0.11805363157817991,
      "grad_norm": 6.4375,
      "learning_rate": 9.977925230625455e-06,
      "loss": 0.86358986,
      "memory(GiB)": 117.54,
      "step": 5060,
      "train_speed(iter/s)": 0.201951
    },
    {
      "acc": 0.76386681,
      "epoch": 0.1182869391504688,
      "grad_norm": 7.3125,
      "learning_rate": 9.977747556104036e-06,
      "loss": 0.87934084,
      "memory(GiB)": 117.54,
      "step": 5070,
      "train_speed(iter/s)": 0.202151
    },
    {
      "acc": 0.75649219,
      "epoch": 0.1185202467227577,
      "grad_norm": 5.15625,
      "learning_rate": 9.9775691710106e-06,
      "loss": 0.88812695,
      "memory(GiB)": 117.54,
      "step": 5080,
      "train_speed(iter/s)": 0.202352
    },
    {
      "acc": 0.75279508,
      "epoch": 0.11875355429504658,
      "grad_norm": 4.875,
      "learning_rate": 9.977390075370607e-06,
      "loss": 0.90212612,
      "memory(GiB)": 117.54,
      "step": 5090,
      "train_speed(iter/s)": 0.202571
    },
    {
      "acc": 0.75354643,
      "epoch": 0.11898686186733548,
      "grad_norm": 6.875,
      "learning_rate": 9.97721026920963e-06,
      "loss": 0.89249134,
      "memory(GiB)": 117.54,
      "step": 5100,
      "train_speed(iter/s)": 0.202776
    },
    {
      "acc": 0.74565663,
      "epoch": 0.11922016943962438,
      "grad_norm": 5.53125,
      "learning_rate": 9.977029752553331e-06,
      "loss": 0.93843594,
      "memory(GiB)": 117.54,
      "step": 5110,
      "train_speed(iter/s)": 0.202986
    },
    {
      "acc": 0.75341768,
      "epoch": 0.11945347701191326,
      "grad_norm": 4.5,
      "learning_rate": 9.97684852542748e-06,
      "loss": 0.88506985,
      "memory(GiB)": 117.54,
      "step": 5120,
      "train_speed(iter/s)": 0.203191
    },
    {
      "acc": 0.74062805,
      "epoch": 0.11968678458420216,
      "grad_norm": 5.65625,
      "learning_rate": 9.976666587857951e-06,
      "loss": 0.94102879,
      "memory(GiB)": 117.54,
      "step": 5130,
      "train_speed(iter/s)": 0.203403
    },
    {
      "acc": 0.76677217,
      "epoch": 0.11992009215649105,
      "grad_norm": 4.96875,
      "learning_rate": 9.97648393987071e-06,
      "loss": 0.85841618,
      "memory(GiB)": 117.54,
      "step": 5140,
      "train_speed(iter/s)": 0.203589
    },
    {
      "acc": 0.7633976,
      "epoch": 0.12015339972877995,
      "grad_norm": 8.0625,
      "learning_rate": 9.976300581491833e-06,
      "loss": 0.83148098,
      "memory(GiB)": 117.54,
      "step": 5150,
      "train_speed(iter/s)": 0.203784
    },
    {
      "acc": 0.7448679,
      "epoch": 0.12038670730106885,
      "grad_norm": 5.65625,
      "learning_rate": 9.976116512747493e-06,
      "loss": 0.94005127,
      "memory(GiB)": 117.54,
      "step": 5160,
      "train_speed(iter/s)": 0.203989
    },
    {
      "acc": 0.76915073,
      "epoch": 0.12062001487335773,
      "grad_norm": 6.8125,
      "learning_rate": 9.975931733663966e-06,
      "loss": 0.83365669,
      "memory(GiB)": 117.54,
      "step": 5170,
      "train_speed(iter/s)": 0.204202
    },
    {
      "acc": 0.7760251,
      "epoch": 0.12085332244564663,
      "grad_norm": 4.40625,
      "learning_rate": 9.97574624426763e-06,
      "loss": 0.810886,
      "memory(GiB)": 117.54,
      "step": 5180,
      "train_speed(iter/s)": 0.204402
    },
    {
      "acc": 0.74410958,
      "epoch": 0.12108663001793551,
      "grad_norm": 5.9375,
      "learning_rate": 9.975560044584964e-06,
      "loss": 0.92947273,
      "memory(GiB)": 117.54,
      "step": 5190,
      "train_speed(iter/s)": 0.204599
    },
    {
      "acc": 0.74937978,
      "epoch": 0.12131993759022441,
      "grad_norm": 8.0,
      "learning_rate": 9.975373134642545e-06,
      "loss": 0.92720146,
      "memory(GiB)": 117.54,
      "step": 5200,
      "train_speed(iter/s)": 0.2048
    },
    {
      "acc": 0.77070284,
      "epoch": 0.12155324516251331,
      "grad_norm": 5.90625,
      "learning_rate": 9.975185514467058e-06,
      "loss": 0.82222948,
      "memory(GiB)": 117.54,
      "step": 5210,
      "train_speed(iter/s)": 0.204998
    },
    {
      "acc": 0.7634182,
      "epoch": 0.1217865527348022,
      "grad_norm": 6.3125,
      "learning_rate": 9.974997184085285e-06,
      "loss": 0.86025238,
      "memory(GiB)": 117.54,
      "step": 5220,
      "train_speed(iter/s)": 0.205205
    },
    {
      "acc": 0.75734425,
      "epoch": 0.1220198603070911,
      "grad_norm": 5.59375,
      "learning_rate": 9.974808143524107e-06,
      "loss": 0.87401457,
      "memory(GiB)": 117.54,
      "step": 5230,
      "train_speed(iter/s)": 0.20541
    },
    {
      "acc": 0.75092173,
      "epoch": 0.12225316787937998,
      "grad_norm": 4.1875,
      "learning_rate": 9.974618392810513e-06,
      "loss": 0.93851089,
      "memory(GiB)": 117.54,
      "step": 5240,
      "train_speed(iter/s)": 0.205619
    },
    {
      "acc": 0.74711943,
      "epoch": 0.12248647545166888,
      "grad_norm": 6.625,
      "learning_rate": 9.974427931971588e-06,
      "loss": 0.96647253,
      "memory(GiB)": 117.54,
      "step": 5250,
      "train_speed(iter/s)": 0.205826
    },
    {
      "acc": 0.74908724,
      "epoch": 0.12271978302395777,
      "grad_norm": 6.34375,
      "learning_rate": 9.97423676103452e-06,
      "loss": 0.93568954,
      "memory(GiB)": 117.54,
      "step": 5260,
      "train_speed(iter/s)": 0.206042
    },
    {
      "acc": 0.74937153,
      "epoch": 0.12295309059624666,
      "grad_norm": 8.125,
      "learning_rate": 9.974044880026602e-06,
      "loss": 0.94437447,
      "memory(GiB)": 117.54,
      "step": 5270,
      "train_speed(iter/s)": 0.206236
    },
    {
      "acc": 0.76040206,
      "epoch": 0.12318639816853556,
      "grad_norm": 7.59375,
      "learning_rate": 9.97385228897522e-06,
      "loss": 0.88464508,
      "memory(GiB)": 117.54,
      "step": 5280,
      "train_speed(iter/s)": 0.206441
    },
    {
      "acc": 0.77492819,
      "epoch": 0.12341970574082445,
      "grad_norm": 12.25,
      "learning_rate": 9.97365898790787e-06,
      "loss": 0.80230036,
      "memory(GiB)": 117.54,
      "step": 5290,
      "train_speed(iter/s)": 0.206638
    },
    {
      "acc": 0.75563636,
      "epoch": 0.12365301331311335,
      "grad_norm": 4.5625,
      "learning_rate": 9.973464976852144e-06,
      "loss": 0.8981823,
      "memory(GiB)": 117.54,
      "step": 5300,
      "train_speed(iter/s)": 0.206834
    },
    {
      "acc": 0.75627747,
      "epoch": 0.12388632088540223,
      "grad_norm": 6.90625,
      "learning_rate": 9.973270255835737e-06,
      "loss": 0.8913538,
      "memory(GiB)": 117.54,
      "step": 5310,
      "train_speed(iter/s)": 0.207026
    },
    {
      "acc": 0.75131683,
      "epoch": 0.12411962845769113,
      "grad_norm": 5.96875,
      "learning_rate": 9.973074824886446e-06,
      "loss": 0.93807335,
      "memory(GiB)": 117.54,
      "step": 5320,
      "train_speed(iter/s)": 0.207209
    },
    {
      "acc": 0.74059019,
      "epoch": 0.12435293602998003,
      "grad_norm": 5.53125,
      "learning_rate": 9.972878684032169e-06,
      "loss": 0.96218138,
      "memory(GiB)": 117.54,
      "step": 5330,
      "train_speed(iter/s)": 0.207402
    },
    {
      "acc": 0.76596565,
      "epoch": 0.12458624360226891,
      "grad_norm": 6.875,
      "learning_rate": 9.972681833300903e-06,
      "loss": 0.86264296,
      "memory(GiB)": 117.54,
      "step": 5340,
      "train_speed(iter/s)": 0.207611
    },
    {
      "acc": 0.74638081,
      "epoch": 0.12481955117455781,
      "grad_norm": 4.78125,
      "learning_rate": 9.972484272720751e-06,
      "loss": 0.91274433,
      "memory(GiB)": 117.54,
      "step": 5350,
      "train_speed(iter/s)": 0.207813
    },
    {
      "acc": 0.75317116,
      "epoch": 0.1250528587468467,
      "grad_norm": 4.78125,
      "learning_rate": 9.972286002319913e-06,
      "loss": 0.91048222,
      "memory(GiB)": 117.54,
      "step": 5360,
      "train_speed(iter/s)": 0.208016
    },
    {
      "acc": 0.77595243,
      "epoch": 0.12528616631913558,
      "grad_norm": 6.90625,
      "learning_rate": 9.972087022126693e-06,
      "loss": 0.81915474,
      "memory(GiB)": 117.54,
      "step": 5370,
      "train_speed(iter/s)": 0.208199
    },
    {
      "acc": 0.76468878,
      "epoch": 0.12551947389142448,
      "grad_norm": 6.75,
      "learning_rate": 9.971887332169494e-06,
      "loss": 0.84793015,
      "memory(GiB)": 117.54,
      "step": 5380,
      "train_speed(iter/s)": 0.208388
    },
    {
      "acc": 0.75001726,
      "epoch": 0.12575278146371338,
      "grad_norm": 7.375,
      "learning_rate": 9.971686932476825e-06,
      "loss": 0.90510178,
      "memory(GiB)": 117.54,
      "step": 5390,
      "train_speed(iter/s)": 0.208579
    },
    {
      "acc": 0.7452601,
      "epoch": 0.12598608903600228,
      "grad_norm": 4.71875,
      "learning_rate": 9.971485823077288e-06,
      "loss": 0.92312012,
      "memory(GiB)": 117.54,
      "step": 5400,
      "train_speed(iter/s)": 0.20877
    },
    {
      "acc": 0.75537529,
      "epoch": 0.12621939660829118,
      "grad_norm": 5.46875,
      "learning_rate": 9.971284003999595e-06,
      "loss": 0.89949427,
      "memory(GiB)": 117.54,
      "step": 5410,
      "train_speed(iter/s)": 0.208957
    },
    {
      "acc": 0.74212222,
      "epoch": 0.12645270418058005,
      "grad_norm": 6.15625,
      "learning_rate": 9.971081475272555e-06,
      "loss": 0.93930664,
      "memory(GiB)": 117.54,
      "step": 5420,
      "train_speed(iter/s)": 0.209154
    },
    {
      "acc": 0.76207142,
      "epoch": 0.12668601175286895,
      "grad_norm": 5.4375,
      "learning_rate": 9.97087823692508e-06,
      "loss": 0.86257935,
      "memory(GiB)": 117.54,
      "step": 5430,
      "train_speed(iter/s)": 0.20934
    },
    {
      "acc": 0.76142607,
      "epoch": 0.12691931932515785,
      "grad_norm": 5.78125,
      "learning_rate": 9.970674288986178e-06,
      "loss": 0.86087112,
      "memory(GiB)": 117.54,
      "step": 5440,
      "train_speed(iter/s)": 0.209527
    },
    {
      "acc": 0.77505727,
      "epoch": 0.12715262689744675,
      "grad_norm": 7.0625,
      "learning_rate": 9.970469631484967e-06,
      "loss": 0.81593513,
      "memory(GiB)": 117.54,
      "step": 5450,
      "train_speed(iter/s)": 0.209699
    },
    {
      "acc": 0.75893412,
      "epoch": 0.12738593446973565,
      "grad_norm": 6.03125,
      "learning_rate": 9.970264264450659e-06,
      "loss": 0.85409393,
      "memory(GiB)": 117.54,
      "step": 5460,
      "train_speed(iter/s)": 0.209887
    },
    {
      "acc": 0.76101146,
      "epoch": 0.12761924204202452,
      "grad_norm": 4.5625,
      "learning_rate": 9.970058187912572e-06,
      "loss": 0.89274902,
      "memory(GiB)": 117.54,
      "step": 5470,
      "train_speed(iter/s)": 0.210074
    },
    {
      "acc": 0.76868429,
      "epoch": 0.12785254961431342,
      "grad_norm": 5.6875,
      "learning_rate": 9.969851401900122e-06,
      "loss": 0.83084993,
      "memory(GiB)": 117.54,
      "step": 5480,
      "train_speed(iter/s)": 0.210247
    },
    {
      "acc": 0.75309086,
      "epoch": 0.12808585718660231,
      "grad_norm": 5.4375,
      "learning_rate": 9.969643906442828e-06,
      "loss": 0.90782976,
      "memory(GiB)": 117.54,
      "step": 5490,
      "train_speed(iter/s)": 0.210435
    },
    {
      "acc": 0.76415911,
      "epoch": 0.1283191647588912,
      "grad_norm": 5.46875,
      "learning_rate": 9.96943570157031e-06,
      "loss": 0.85922518,
      "memory(GiB)": 117.54,
      "step": 5500,
      "train_speed(iter/s)": 0.210616
    },
    {
      "epoch": 0.1283191647588912,
      "eval_acc": 0.7270218249364108,
      "eval_loss": 0.8716417551040649,
      "eval_runtime": 1270.1331,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.169,
      "step": 5500
    },
    {
      "acc": 0.73600473,
      "epoch": 0.1285524723311801,
      "grad_norm": 7.03125,
      "learning_rate": 9.969226787312288e-06,
      "loss": 0.96929722,
      "memory(GiB)": 117.54,
      "step": 5510,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.76587524,
      "epoch": 0.12878577990346898,
      "grad_norm": 8.375,
      "learning_rate": 9.969017163698587e-06,
      "loss": 0.848559,
      "memory(GiB)": 117.54,
      "step": 5520,
      "train_speed(iter/s)": 0.201104
    },
    {
      "acc": 0.75857878,
      "epoch": 0.12901908747575788,
      "grad_norm": 5.0625,
      "learning_rate": 9.96880683075913e-06,
      "loss": 0.88542099,
      "memory(GiB)": 117.54,
      "step": 5530,
      "train_speed(iter/s)": 0.201282
    },
    {
      "acc": 0.77082253,
      "epoch": 0.12925239504804678,
      "grad_norm": 6.09375,
      "learning_rate": 9.96859578852394e-06,
      "loss": 0.81673002,
      "memory(GiB)": 117.54,
      "step": 5540,
      "train_speed(iter/s)": 0.201462
    },
    {
      "acc": 0.76157207,
      "epoch": 0.12948570262033568,
      "grad_norm": 5.625,
      "learning_rate": 9.968384037023147e-06,
      "loss": 0.87119312,
      "memory(GiB)": 117.54,
      "step": 5550,
      "train_speed(iter/s)": 0.201654
    },
    {
      "acc": 0.75955338,
      "epoch": 0.12971901019262455,
      "grad_norm": 5.21875,
      "learning_rate": 9.968171576286973e-06,
      "loss": 0.86215534,
      "memory(GiB)": 117.54,
      "step": 5560,
      "train_speed(iter/s)": 0.20183
    },
    {
      "acc": 0.77258162,
      "epoch": 0.12995231776491345,
      "grad_norm": 4.5,
      "learning_rate": 9.96795840634575e-06,
      "loss": 0.81898003,
      "memory(GiB)": 117.54,
      "step": 5570,
      "train_speed(iter/s)": 0.202015
    },
    {
      "acc": 0.75478101,
      "epoch": 0.13018562533720235,
      "grad_norm": 4.40625,
      "learning_rate": 9.96774452722991e-06,
      "loss": 0.91053543,
      "memory(GiB)": 117.54,
      "step": 5580,
      "train_speed(iter/s)": 0.202209
    },
    {
      "acc": 0.77054396,
      "epoch": 0.13041893290949125,
      "grad_norm": 6.78125,
      "learning_rate": 9.967529938969981e-06,
      "loss": 0.81340065,
      "memory(GiB)": 117.54,
      "step": 5590,
      "train_speed(iter/s)": 0.202403
    },
    {
      "acc": 0.75519128,
      "epoch": 0.13065224048178015,
      "grad_norm": 4.28125,
      "learning_rate": 9.967314641596595e-06,
      "loss": 0.87434855,
      "memory(GiB)": 117.54,
      "step": 5600,
      "train_speed(iter/s)": 0.202598
    },
    {
      "acc": 0.74656882,
      "epoch": 0.13088554805406902,
      "grad_norm": 8.875,
      "learning_rate": 9.967098635140489e-06,
      "loss": 0.92529068,
      "memory(GiB)": 117.54,
      "step": 5610,
      "train_speed(iter/s)": 0.202788
    },
    {
      "acc": 0.76542044,
      "epoch": 0.13111885562635792,
      "grad_norm": 7.9375,
      "learning_rate": 9.966881919632494e-06,
      "loss": 0.86160889,
      "memory(GiB)": 117.54,
      "step": 5620,
      "train_speed(iter/s)": 0.202986
    },
    {
      "acc": 0.76484203,
      "epoch": 0.13135216319864682,
      "grad_norm": 5.0625,
      "learning_rate": 9.966664495103548e-06,
      "loss": 0.85322952,
      "memory(GiB)": 117.54,
      "step": 5630,
      "train_speed(iter/s)": 0.203181
    },
    {
      "acc": 0.75730438,
      "epoch": 0.13158547077093571,
      "grad_norm": 5.25,
      "learning_rate": 9.96644636158469e-06,
      "loss": 0.85542212,
      "memory(GiB)": 117.54,
      "step": 5640,
      "train_speed(iter/s)": 0.203364
    },
    {
      "acc": 0.75253458,
      "epoch": 0.1318187783432246,
      "grad_norm": 19.875,
      "learning_rate": 9.966227519107054e-06,
      "loss": 0.96850071,
      "memory(GiB)": 117.54,
      "step": 5650,
      "train_speed(iter/s)": 0.203551
    },
    {
      "acc": 0.76095395,
      "epoch": 0.13205208591551348,
      "grad_norm": 4.25,
      "learning_rate": 9.966007967701884e-06,
      "loss": 0.88187704,
      "memory(GiB)": 117.54,
      "step": 5660,
      "train_speed(iter/s)": 0.203728
    },
    {
      "acc": 0.76344061,
      "epoch": 0.13228539348780238,
      "grad_norm": 7.3125,
      "learning_rate": 9.965787707400521e-06,
      "loss": 0.87797689,
      "memory(GiB)": 117.54,
      "step": 5670,
      "train_speed(iter/s)": 0.203918
    },
    {
      "acc": 0.7655921,
      "epoch": 0.13251870106009128,
      "grad_norm": 4.15625,
      "learning_rate": 9.965566738234403e-06,
      "loss": 0.83435678,
      "memory(GiB)": 117.54,
      "step": 5680,
      "train_speed(iter/s)": 0.204106
    },
    {
      "acc": 0.76264372,
      "epoch": 0.13275200863238018,
      "grad_norm": 7.1875,
      "learning_rate": 9.965345060235075e-06,
      "loss": 0.86193419,
      "memory(GiB)": 117.54,
      "step": 5690,
      "train_speed(iter/s)": 0.204292
    },
    {
      "acc": 0.75327988,
      "epoch": 0.13298531620466908,
      "grad_norm": 6.875,
      "learning_rate": 9.965122673434182e-06,
      "loss": 0.91502628,
      "memory(GiB)": 117.54,
      "step": 5700,
      "train_speed(iter/s)": 0.204459
    },
    {
      "acc": 0.75798521,
      "epoch": 0.13321862377695795,
      "grad_norm": 7.15625,
      "learning_rate": 9.964899577863472e-06,
      "loss": 0.86956244,
      "memory(GiB)": 117.54,
      "step": 5710,
      "train_speed(iter/s)": 0.204649
    },
    {
      "acc": 0.75251207,
      "epoch": 0.13345193134924685,
      "grad_norm": 6.03125,
      "learning_rate": 9.964675773554789e-06,
      "loss": 0.907057,
      "memory(GiB)": 117.54,
      "step": 5720,
      "train_speed(iter/s)": 0.204833
    },
    {
      "acc": 0.75791512,
      "epoch": 0.13368523892153575,
      "grad_norm": 6.03125,
      "learning_rate": 9.96445126054008e-06,
      "loss": 0.89783096,
      "memory(GiB)": 117.54,
      "step": 5730,
      "train_speed(iter/s)": 0.205017
    },
    {
      "acc": 0.76094255,
      "epoch": 0.13391854649382465,
      "grad_norm": 6.4375,
      "learning_rate": 9.964226038851397e-06,
      "loss": 0.88064919,
      "memory(GiB)": 117.54,
      "step": 5740,
      "train_speed(iter/s)": 0.205195
    },
    {
      "acc": 0.76472521,
      "epoch": 0.13415185406611355,
      "grad_norm": 5.375,
      "learning_rate": 9.964000108520889e-06,
      "loss": 0.84428978,
      "memory(GiB)": 117.54,
      "step": 5750,
      "train_speed(iter/s)": 0.205379
    },
    {
      "acc": 0.76074214,
      "epoch": 0.13438516163840242,
      "grad_norm": 5.40625,
      "learning_rate": 9.963773469580806e-06,
      "loss": 0.85086288,
      "memory(GiB)": 117.54,
      "step": 5760,
      "train_speed(iter/s)": 0.205562
    },
    {
      "acc": 0.76351795,
      "epoch": 0.13461846921069132,
      "grad_norm": 4.40625,
      "learning_rate": 9.963546122063504e-06,
      "loss": 0.86668959,
      "memory(GiB)": 117.54,
      "step": 5770,
      "train_speed(iter/s)": 0.205749
    },
    {
      "acc": 0.76371803,
      "epoch": 0.13485177678298021,
      "grad_norm": 8.5625,
      "learning_rate": 9.963318066001433e-06,
      "loss": 0.86080093,
      "memory(GiB)": 117.54,
      "step": 5780,
      "train_speed(iter/s)": 0.205929
    },
    {
      "acc": 0.7618516,
      "epoch": 0.1350850843552691,
      "grad_norm": 7.1875,
      "learning_rate": 9.963089301427152e-06,
      "loss": 0.84876232,
      "memory(GiB)": 117.54,
      "step": 5790,
      "train_speed(iter/s)": 0.206111
    },
    {
      "acc": 0.77087803,
      "epoch": 0.13531839192755798,
      "grad_norm": 5.0,
      "learning_rate": 9.962859828373315e-06,
      "loss": 0.83783426,
      "memory(GiB)": 117.54,
      "step": 5800,
      "train_speed(iter/s)": 0.206293
    },
    {
      "acc": 0.77675738,
      "epoch": 0.13555169949984688,
      "grad_norm": 7.90625,
      "learning_rate": 9.96262964687268e-06,
      "loss": 0.83521681,
      "memory(GiB)": 117.54,
      "step": 5810,
      "train_speed(iter/s)": 0.206466
    },
    {
      "acc": 0.76050811,
      "epoch": 0.13578500707213578,
      "grad_norm": 6.1875,
      "learning_rate": 9.9623987569581e-06,
      "loss": 0.88420362,
      "memory(GiB)": 117.54,
      "step": 5820,
      "train_speed(iter/s)": 0.206646
    },
    {
      "acc": 0.73890877,
      "epoch": 0.13601831464442468,
      "grad_norm": 5.90625,
      "learning_rate": 9.962167158662543e-06,
      "loss": 0.95491695,
      "memory(GiB)": 117.54,
      "step": 5830,
      "train_speed(iter/s)": 0.206826
    },
    {
      "acc": 0.74483318,
      "epoch": 0.13625162221671358,
      "grad_norm": 4.15625,
      "learning_rate": 9.961934852019066e-06,
      "loss": 0.92905416,
      "memory(GiB)": 117.54,
      "step": 5840,
      "train_speed(iter/s)": 0.207007
    },
    {
      "acc": 0.74667826,
      "epoch": 0.13648492978900245,
      "grad_norm": 4.6875,
      "learning_rate": 9.96170183706083e-06,
      "loss": 0.92713289,
      "memory(GiB)": 117.54,
      "step": 5850,
      "train_speed(iter/s)": 0.207186
    },
    {
      "acc": 0.74078379,
      "epoch": 0.13671823736129135,
      "grad_norm": 7.71875,
      "learning_rate": 9.961468113821096e-06,
      "loss": 0.94992075,
      "memory(GiB)": 117.54,
      "step": 5860,
      "train_speed(iter/s)": 0.207359
    },
    {
      "acc": 0.77699747,
      "epoch": 0.13695154493358025,
      "grad_norm": 5.84375,
      "learning_rate": 9.96123368233323e-06,
      "loss": 0.78689365,
      "memory(GiB)": 117.54,
      "step": 5870,
      "train_speed(iter/s)": 0.207542
    },
    {
      "acc": 0.74916096,
      "epoch": 0.13718485250586915,
      "grad_norm": 7.625,
      "learning_rate": 9.9609985426307e-06,
      "loss": 0.93192005,
      "memory(GiB)": 117.54,
      "step": 5880,
      "train_speed(iter/s)": 0.207728
    },
    {
      "acc": 0.7669405,
      "epoch": 0.13741816007815805,
      "grad_norm": 6.25,
      "learning_rate": 9.960762694747068e-06,
      "loss": 0.87499352,
      "memory(GiB)": 117.54,
      "step": 5890,
      "train_speed(iter/s)": 0.207914
    },
    {
      "acc": 0.77017179,
      "epoch": 0.13765146765044692,
      "grad_norm": 6.4375,
      "learning_rate": 9.960526138716e-06,
      "loss": 0.83586044,
      "memory(GiB)": 117.54,
      "step": 5900,
      "train_speed(iter/s)": 0.208098
    },
    {
      "acc": 0.75412836,
      "epoch": 0.13788477522273582,
      "grad_norm": 7.78125,
      "learning_rate": 9.960288874571271e-06,
      "loss": 0.90887356,
      "memory(GiB)": 117.54,
      "step": 5910,
      "train_speed(iter/s)": 0.208286
    },
    {
      "acc": 0.76210413,
      "epoch": 0.13811808279502472,
      "grad_norm": 6.0,
      "learning_rate": 9.960050902346743e-06,
      "loss": 0.85120649,
      "memory(GiB)": 117.54,
      "step": 5920,
      "train_speed(iter/s)": 0.208456
    },
    {
      "acc": 0.75921898,
      "epoch": 0.13835139036731361,
      "grad_norm": 8.875,
      "learning_rate": 9.959812222076391e-06,
      "loss": 0.87873182,
      "memory(GiB)": 117.54,
      "step": 5930,
      "train_speed(iter/s)": 0.208634
    },
    {
      "acc": 0.75706267,
      "epoch": 0.1385846979396025,
      "grad_norm": 7.96875,
      "learning_rate": 9.959572833794283e-06,
      "loss": 0.86111183,
      "memory(GiB)": 117.54,
      "step": 5940,
      "train_speed(iter/s)": 0.208814
    },
    {
      "acc": 0.74689651,
      "epoch": 0.13881800551189138,
      "grad_norm": 6.34375,
      "learning_rate": 9.959332737534597e-06,
      "loss": 0.92088099,
      "memory(GiB)": 117.54,
      "step": 5950,
      "train_speed(iter/s)": 0.208985
    },
    {
      "acc": 0.7608963,
      "epoch": 0.13905131308418028,
      "grad_norm": 17.625,
      "learning_rate": 9.959091933331601e-06,
      "loss": 0.85624313,
      "memory(GiB)": 117.54,
      "step": 5960,
      "train_speed(iter/s)": 0.209159
    },
    {
      "acc": 0.77537665,
      "epoch": 0.13928462065646918,
      "grad_norm": 4.96875,
      "learning_rate": 9.958850421219675e-06,
      "loss": 0.79838071,
      "memory(GiB)": 117.54,
      "step": 5970,
      "train_speed(iter/s)": 0.209339
    },
    {
      "acc": 0.75666475,
      "epoch": 0.13951792822875808,
      "grad_norm": 7.78125,
      "learning_rate": 9.958608201233288e-06,
      "loss": 0.89333801,
      "memory(GiB)": 117.54,
      "step": 5980,
      "train_speed(iter/s)": 0.209526
    },
    {
      "acc": 0.75140362,
      "epoch": 0.13975123580104698,
      "grad_norm": 4.75,
      "learning_rate": 9.958365273407023e-06,
      "loss": 0.9052248,
      "memory(GiB)": 117.54,
      "step": 5990,
      "train_speed(iter/s)": 0.209691
    },
    {
      "acc": 0.73782969,
      "epoch": 0.13998454337333585,
      "grad_norm": 13.3125,
      "learning_rate": 9.958121637775554e-06,
      "loss": 0.96962986,
      "memory(GiB)": 117.54,
      "step": 6000,
      "train_speed(iter/s)": 0.209875
    },
    {
      "epoch": 0.13998454337333585,
      "eval_acc": 0.727715409606508,
      "eval_loss": 0.8690560460090637,
      "eval_runtime": 1270.3769,
      "eval_samples_per_second": 28.331,
      "eval_steps_per_second": 14.166,
      "step": 6000
    },
    {
      "acc": 0.75419903,
      "epoch": 0.14021785094562475,
      "grad_norm": 6.4375,
      "learning_rate": 9.957877294373665e-06,
      "loss": 0.91555004,
      "memory(GiB)": 117.54,
      "step": 6010,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.74702139,
      "epoch": 0.14045115851791365,
      "grad_norm": 7.0,
      "learning_rate": 9.957632243236231e-06,
      "loss": 0.92673454,
      "memory(GiB)": 117.54,
      "step": 6020,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.75800505,
      "epoch": 0.14068446609020255,
      "grad_norm": 6.53125,
      "learning_rate": 9.957386484398233e-06,
      "loss": 0.87921581,
      "memory(GiB)": 117.54,
      "step": 6030,
      "train_speed(iter/s)": 0.201343
    },
    {
      "acc": 0.74582348,
      "epoch": 0.14091777366249145,
      "grad_norm": 5.21875,
      "learning_rate": 9.957140017894754e-06,
      "loss": 0.91472492,
      "memory(GiB)": 117.54,
      "step": 6040,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.78675618,
      "epoch": 0.14115108123478032,
      "grad_norm": 6.03125,
      "learning_rate": 9.956892843760979e-06,
      "loss": 0.75914721,
      "memory(GiB)": 117.54,
      "step": 6050,
      "train_speed(iter/s)": 0.201696
    },
    {
      "acc": 0.74411964,
      "epoch": 0.14138438880706922,
      "grad_norm": 12.0625,
      "learning_rate": 9.956644962032192e-06,
      "loss": 0.93103161,
      "memory(GiB)": 117.54,
      "step": 6060,
      "train_speed(iter/s)": 0.201878
    },
    {
      "acc": 0.7726965,
      "epoch": 0.14161769637935812,
      "grad_norm": 7.375,
      "learning_rate": 9.956396372743775e-06,
      "loss": 0.82945976,
      "memory(GiB)": 117.54,
      "step": 6070,
      "train_speed(iter/s)": 0.202042
    },
    {
      "acc": 0.78586645,
      "epoch": 0.14185100395164701,
      "grad_norm": 6.875,
      "learning_rate": 9.956147075931215e-06,
      "loss": 0.76539755,
      "memory(GiB)": 117.54,
      "step": 6080,
      "train_speed(iter/s)": 0.202205
    },
    {
      "acc": 0.75904989,
      "epoch": 0.14208431152393589,
      "grad_norm": 8.75,
      "learning_rate": 9.955897071630101e-06,
      "loss": 0.87475338,
      "memory(GiB)": 117.54,
      "step": 6090,
      "train_speed(iter/s)": 0.202374
    },
    {
      "acc": 0.76456523,
      "epoch": 0.14231761909622478,
      "grad_norm": 6.03125,
      "learning_rate": 9.955646359876118e-06,
      "loss": 0.86053829,
      "memory(GiB)": 117.54,
      "step": 6100,
      "train_speed(iter/s)": 0.202548
    },
    {
      "acc": 0.76305366,
      "epoch": 0.14255092666851368,
      "grad_norm": 6.84375,
      "learning_rate": 9.955394940705057e-06,
      "loss": 0.85630341,
      "memory(GiB)": 117.54,
      "step": 6110,
      "train_speed(iter/s)": 0.202712
    },
    {
      "acc": 0.7719203,
      "epoch": 0.14278423424080258,
      "grad_norm": 8.1875,
      "learning_rate": 9.95514281415281e-06,
      "loss": 0.82893257,
      "memory(GiB)": 117.54,
      "step": 6120,
      "train_speed(iter/s)": 0.20288
    },
    {
      "acc": 0.76380386,
      "epoch": 0.14301754181309148,
      "grad_norm": 7.1875,
      "learning_rate": 9.954889980255363e-06,
      "loss": 0.84828529,
      "memory(GiB)": 117.54,
      "step": 6130,
      "train_speed(iter/s)": 0.203048
    },
    {
      "acc": 0.77324157,
      "epoch": 0.14325084938538035,
      "grad_norm": 4.34375,
      "learning_rate": 9.954636439048813e-06,
      "loss": 0.82084036,
      "memory(GiB)": 117.54,
      "step": 6140,
      "train_speed(iter/s)": 0.203199
    },
    {
      "acc": 0.76487913,
      "epoch": 0.14348415695766925,
      "grad_norm": 6.03125,
      "learning_rate": 9.95438219056935e-06,
      "loss": 0.85029163,
      "memory(GiB)": 117.54,
      "step": 6150,
      "train_speed(iter/s)": 0.203348
    },
    {
      "acc": 0.75904469,
      "epoch": 0.14371746452995815,
      "grad_norm": 7.875,
      "learning_rate": 9.954127234853267e-06,
      "loss": 0.89040432,
      "memory(GiB)": 117.54,
      "step": 6160,
      "train_speed(iter/s)": 0.203524
    },
    {
      "acc": 0.76309099,
      "epoch": 0.14395077210224705,
      "grad_norm": 4.75,
      "learning_rate": 9.953871571936962e-06,
      "loss": 0.85744724,
      "memory(GiB)": 117.54,
      "step": 6170,
      "train_speed(iter/s)": 0.203691
    },
    {
      "acc": 0.77011318,
      "epoch": 0.14418407967453595,
      "grad_norm": 6.625,
      "learning_rate": 9.953615201856928e-06,
      "loss": 0.8331212,
      "memory(GiB)": 117.54,
      "step": 6180,
      "train_speed(iter/s)": 0.203861
    },
    {
      "acc": 0.77067161,
      "epoch": 0.14441738724682482,
      "grad_norm": 5.53125,
      "learning_rate": 9.953358124649764e-06,
      "loss": 0.82904701,
      "memory(GiB)": 117.54,
      "step": 6190,
      "train_speed(iter/s)": 0.20402
    },
    {
      "acc": 0.75703244,
      "epoch": 0.14465069481911372,
      "grad_norm": 6.59375,
      "learning_rate": 9.953100340352166e-06,
      "loss": 0.87819614,
      "memory(GiB)": 117.54,
      "step": 6200,
      "train_speed(iter/s)": 0.20419
    },
    {
      "acc": 0.76388779,
      "epoch": 0.14488400239140262,
      "grad_norm": 6.78125,
      "learning_rate": 9.952841849000935e-06,
      "loss": 0.84278336,
      "memory(GiB)": 117.54,
      "step": 6210,
      "train_speed(iter/s)": 0.204357
    },
    {
      "acc": 0.78060641,
      "epoch": 0.14511730996369152,
      "grad_norm": 6.90625,
      "learning_rate": 9.952582650632967e-06,
      "loss": 0.80852413,
      "memory(GiB)": 117.54,
      "step": 6220,
      "train_speed(iter/s)": 0.204527
    },
    {
      "acc": 0.76564684,
      "epoch": 0.14535061753598041,
      "grad_norm": 5.875,
      "learning_rate": 9.952322745285266e-06,
      "loss": 0.83691425,
      "memory(GiB)": 117.54,
      "step": 6230,
      "train_speed(iter/s)": 0.20469
    },
    {
      "acc": 0.77099514,
      "epoch": 0.14558392510826929,
      "grad_norm": 9.75,
      "learning_rate": 9.95206213299493e-06,
      "loss": 0.83298168,
      "memory(GiB)": 117.54,
      "step": 6240,
      "train_speed(iter/s)": 0.204862
    },
    {
      "acc": 0.76582918,
      "epoch": 0.14581723268055818,
      "grad_norm": 4.75,
      "learning_rate": 9.951800813799164e-06,
      "loss": 0.85061855,
      "memory(GiB)": 117.54,
      "step": 6250,
      "train_speed(iter/s)": 0.205033
    },
    {
      "acc": 0.77523489,
      "epoch": 0.14605054025284708,
      "grad_norm": 5.90625,
      "learning_rate": 9.95153878773527e-06,
      "loss": 0.8064352,
      "memory(GiB)": 117.54,
      "step": 6260,
      "train_speed(iter/s)": 0.205197
    },
    {
      "acc": 0.77384329,
      "epoch": 0.14628384782513598,
      "grad_norm": 7.375,
      "learning_rate": 9.951276054840654e-06,
      "loss": 0.80671873,
      "memory(GiB)": 117.54,
      "step": 6270,
      "train_speed(iter/s)": 0.205362
    },
    {
      "acc": 0.75420942,
      "epoch": 0.14651715539742488,
      "grad_norm": 6.5625,
      "learning_rate": 9.951012615152816e-06,
      "loss": 0.92444849,
      "memory(GiB)": 117.54,
      "step": 6280,
      "train_speed(iter/s)": 0.205538
    },
    {
      "acc": 0.78133068,
      "epoch": 0.14675046296971375,
      "grad_norm": 9.5,
      "learning_rate": 9.950748468709368e-06,
      "loss": 0.78833847,
      "memory(GiB)": 117.54,
      "step": 6290,
      "train_speed(iter/s)": 0.205701
    },
    {
      "acc": 0.74810176,
      "epoch": 0.14698377054200265,
      "grad_norm": 5.78125,
      "learning_rate": 9.950483615548014e-06,
      "loss": 0.92472363,
      "memory(GiB)": 117.54,
      "step": 6300,
      "train_speed(iter/s)": 0.205871
    },
    {
      "acc": 0.7479497,
      "epoch": 0.14721707811429155,
      "grad_norm": 5.9375,
      "learning_rate": 9.950218055706563e-06,
      "loss": 0.92439194,
      "memory(GiB)": 117.54,
      "step": 6310,
      "train_speed(iter/s)": 0.206038
    },
    {
      "acc": 0.77187309,
      "epoch": 0.14745038568658045,
      "grad_norm": 5.40625,
      "learning_rate": 9.94995178922292e-06,
      "loss": 0.83006401,
      "memory(GiB)": 117.54,
      "step": 6320,
      "train_speed(iter/s)": 0.206198
    },
    {
      "acc": 0.75937672,
      "epoch": 0.14768369325886932,
      "grad_norm": 7.03125,
      "learning_rate": 9.949684816135098e-06,
      "loss": 0.88418226,
      "memory(GiB)": 117.54,
      "step": 6330,
      "train_speed(iter/s)": 0.206369
    },
    {
      "acc": 0.7517971,
      "epoch": 0.14791700083115822,
      "grad_norm": 4.96875,
      "learning_rate": 9.949417136481207e-06,
      "loss": 0.92034903,
      "memory(GiB)": 117.54,
      "step": 6340,
      "train_speed(iter/s)": 0.206546
    },
    {
      "acc": 0.75734596,
      "epoch": 0.14815030840344712,
      "grad_norm": 5.1875,
      "learning_rate": 9.94914875029946e-06,
      "loss": 0.90017557,
      "memory(GiB)": 117.54,
      "step": 6350,
      "train_speed(iter/s)": 0.206707
    },
    {
      "acc": 0.75256109,
      "epoch": 0.14838361597573602,
      "grad_norm": 7.0625,
      "learning_rate": 9.948879657628164e-06,
      "loss": 0.9080761,
      "memory(GiB)": 117.54,
      "step": 6360,
      "train_speed(iter/s)": 0.206879
    },
    {
      "acc": 0.76637239,
      "epoch": 0.14861692354802492,
      "grad_norm": 13.8125,
      "learning_rate": 9.948609858505734e-06,
      "loss": 0.84904175,
      "memory(GiB)": 117.54,
      "step": 6370,
      "train_speed(iter/s)": 0.207045
    },
    {
      "acc": 0.73998928,
      "epoch": 0.1488502311203138,
      "grad_norm": 5.25,
      "learning_rate": 9.948339352970683e-06,
      "loss": 0.97184429,
      "memory(GiB)": 117.54,
      "step": 6380,
      "train_speed(iter/s)": 0.207209
    },
    {
      "acc": 0.75804882,
      "epoch": 0.14908353869260269,
      "grad_norm": 8.0,
      "learning_rate": 9.948068141061631e-06,
      "loss": 0.85682335,
      "memory(GiB)": 117.54,
      "step": 6390,
      "train_speed(iter/s)": 0.207362
    },
    {
      "acc": 0.72696562,
      "epoch": 0.14931684626489158,
      "grad_norm": 6.6875,
      "learning_rate": 9.947796222817286e-06,
      "loss": 1.01356945,
      "memory(GiB)": 117.54,
      "step": 6400,
      "train_speed(iter/s)": 0.207524
    },
    {
      "acc": 0.76806068,
      "epoch": 0.14955015383718048,
      "grad_norm": 5.125,
      "learning_rate": 9.94752359827647e-06,
      "loss": 0.85701227,
      "memory(GiB)": 117.54,
      "step": 6410,
      "train_speed(iter/s)": 0.207681
    },
    {
      "acc": 0.76032314,
      "epoch": 0.14978346140946938,
      "grad_norm": 7.25,
      "learning_rate": 9.947250267478094e-06,
      "loss": 0.87021475,
      "memory(GiB)": 117.54,
      "step": 6420,
      "train_speed(iter/s)": 0.207838
    },
    {
      "acc": 0.75938177,
      "epoch": 0.15001676898175825,
      "grad_norm": 5.03125,
      "learning_rate": 9.946976230461183e-06,
      "loss": 0.88507252,
      "memory(GiB)": 117.54,
      "step": 6430,
      "train_speed(iter/s)": 0.208015
    },
    {
      "acc": 0.74555626,
      "epoch": 0.15025007655404715,
      "grad_norm": 5.40625,
      "learning_rate": 9.946701487264851e-06,
      "loss": 0.9236105,
      "memory(GiB)": 117.54,
      "step": 6440,
      "train_speed(iter/s)": 0.208181
    },
    {
      "acc": 0.761584,
      "epoch": 0.15048338412633605,
      "grad_norm": 4.71875,
      "learning_rate": 9.946426037928319e-06,
      "loss": 0.86102915,
      "memory(GiB)": 117.54,
      "step": 6450,
      "train_speed(iter/s)": 0.208345
    },
    {
      "acc": 0.7360589,
      "epoch": 0.15071669169862495,
      "grad_norm": 8.5,
      "learning_rate": 9.946149882490907e-06,
      "loss": 0.94918003,
      "memory(GiB)": 117.54,
      "step": 6460,
      "train_speed(iter/s)": 0.208502
    },
    {
      "acc": 0.76649723,
      "epoch": 0.15094999927091385,
      "grad_norm": 4.78125,
      "learning_rate": 9.945873020992036e-06,
      "loss": 0.86450176,
      "memory(GiB)": 117.54,
      "step": 6470,
      "train_speed(iter/s)": 0.208672
    },
    {
      "acc": 0.77010989,
      "epoch": 0.15118330684320272,
      "grad_norm": 5.53125,
      "learning_rate": 9.945595453471228e-06,
      "loss": 0.82765493,
      "memory(GiB)": 117.54,
      "step": 6480,
      "train_speed(iter/s)": 0.20882
    },
    {
      "acc": 0.76764402,
      "epoch": 0.15141661441549162,
      "grad_norm": 13.0625,
      "learning_rate": 9.945317179968105e-06,
      "loss": 0.84424448,
      "memory(GiB)": 117.54,
      "step": 6490,
      "train_speed(iter/s)": 0.208984
    },
    {
      "acc": 0.75681572,
      "epoch": 0.15164992198778052,
      "grad_norm": 4.4375,
      "learning_rate": 9.945038200522392e-06,
      "loss": 0.9044385,
      "memory(GiB)": 117.54,
      "step": 6500,
      "train_speed(iter/s)": 0.209132
    },
    {
      "epoch": 0.15164992198778052,
      "eval_acc": 0.7281223361007858,
      "eval_loss": 0.8657384514808655,
      "eval_runtime": 1269.6135,
      "eval_samples_per_second": 28.348,
      "eval_steps_per_second": 14.174,
      "step": 6500
    },
    {
      "acc": 0.77224231,
      "epoch": 0.15188322956006942,
      "grad_norm": 6.875,
      "learning_rate": 9.944758515173912e-06,
      "loss": 0.82228127,
      "memory(GiB)": 117.54,
      "step": 6510,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.76387882,
      "epoch": 0.15211653713235831,
      "grad_norm": 5.21875,
      "learning_rate": 9.944478123962592e-06,
      "loss": 0.85496302,
      "memory(GiB)": 117.54,
      "step": 6520,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.75840979,
      "epoch": 0.15234984470464719,
      "grad_norm": 4.8125,
      "learning_rate": 9.944197026928454e-06,
      "loss": 0.88294678,
      "memory(GiB)": 117.54,
      "step": 6530,
      "train_speed(iter/s)": 0.201303
    },
    {
      "acc": 0.76922741,
      "epoch": 0.15258315227693608,
      "grad_norm": 4.84375,
      "learning_rate": 9.943915224111627e-06,
      "loss": 0.87327852,
      "memory(GiB)": 117.54,
      "step": 6540,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.75248985,
      "epoch": 0.15281645984922498,
      "grad_norm": 8.4375,
      "learning_rate": 9.943632715552338e-06,
      "loss": 0.88687239,
      "memory(GiB)": 117.54,
      "step": 6550,
      "train_speed(iter/s)": 0.201614
    },
    {
      "acc": 0.75970192,
      "epoch": 0.15304976742151388,
      "grad_norm": 14.4375,
      "learning_rate": 9.943349501290916e-06,
      "loss": 0.85225811,
      "memory(GiB)": 117.54,
      "step": 6560,
      "train_speed(iter/s)": 0.201782
    },
    {
      "acc": 0.77734346,
      "epoch": 0.15328307499380278,
      "grad_norm": 5.3125,
      "learning_rate": 9.943065581367788e-06,
      "loss": 0.81777477,
      "memory(GiB)": 117.54,
      "step": 6570,
      "train_speed(iter/s)": 0.201947
    },
    {
      "acc": 0.75825195,
      "epoch": 0.15351638256609165,
      "grad_norm": 7.625,
      "learning_rate": 9.942780955823485e-06,
      "loss": 0.86885967,
      "memory(GiB)": 117.54,
      "step": 6580,
      "train_speed(iter/s)": 0.202112
    },
    {
      "acc": 0.76747169,
      "epoch": 0.15374969013838055,
      "grad_norm": 4.4375,
      "learning_rate": 9.942495624698636e-06,
      "loss": 0.84972601,
      "memory(GiB)": 117.54,
      "step": 6590,
      "train_speed(iter/s)": 0.202268
    },
    {
      "acc": 0.75881453,
      "epoch": 0.15398299771066945,
      "grad_norm": 5.46875,
      "learning_rate": 9.942209588033973e-06,
      "loss": 0.87035789,
      "memory(GiB)": 117.54,
      "step": 6600,
      "train_speed(iter/s)": 0.202422
    },
    {
      "acc": 0.74656525,
      "epoch": 0.15421630528295835,
      "grad_norm": 5.03125,
      "learning_rate": 9.941922845870326e-06,
      "loss": 0.92789898,
      "memory(GiB)": 117.54,
      "step": 6610,
      "train_speed(iter/s)": 0.202575
    },
    {
      "acc": 0.7642231,
      "epoch": 0.15444961285524722,
      "grad_norm": 5.4375,
      "learning_rate": 9.941635398248628e-06,
      "loss": 0.83677711,
      "memory(GiB)": 117.54,
      "step": 6620,
      "train_speed(iter/s)": 0.202738
    },
    {
      "acc": 0.77601938,
      "epoch": 0.15468292042753612,
      "grad_norm": 6.65625,
      "learning_rate": 9.941347245209914e-06,
      "loss": 0.78568192,
      "memory(GiB)": 117.54,
      "step": 6630,
      "train_speed(iter/s)": 0.202885
    },
    {
      "acc": 0.75074358,
      "epoch": 0.15491622799982502,
      "grad_norm": 6.375,
      "learning_rate": 9.941058386795314e-06,
      "loss": 0.90226822,
      "memory(GiB)": 117.54,
      "step": 6640,
      "train_speed(iter/s)": 0.203031
    },
    {
      "acc": 0.77472057,
      "epoch": 0.15514953557211392,
      "grad_norm": 4.53125,
      "learning_rate": 9.940768823046067e-06,
      "loss": 0.83820248,
      "memory(GiB)": 117.54,
      "step": 6650,
      "train_speed(iter/s)": 0.203189
    },
    {
      "acc": 0.77502432,
      "epoch": 0.15538284314440282,
      "grad_norm": 5.78125,
      "learning_rate": 9.940478554003506e-06,
      "loss": 0.79055486,
      "memory(GiB)": 117.54,
      "step": 6660,
      "train_speed(iter/s)": 0.203346
    },
    {
      "acc": 0.76303005,
      "epoch": 0.1556161507166917,
      "grad_norm": 6.53125,
      "learning_rate": 9.940187579709064e-06,
      "loss": 0.86618824,
      "memory(GiB)": 117.54,
      "step": 6670,
      "train_speed(iter/s)": 0.203512
    },
    {
      "acc": 0.74345989,
      "epoch": 0.15584945828898059,
      "grad_norm": 4.96875,
      "learning_rate": 9.939895900204281e-06,
      "loss": 0.91676807,
      "memory(GiB)": 117.54,
      "step": 6680,
      "train_speed(iter/s)": 0.203669
    },
    {
      "acc": 0.76821184,
      "epoch": 0.15608276586126948,
      "grad_norm": 5.46875,
      "learning_rate": 9.939603515530796e-06,
      "loss": 0.8443615,
      "memory(GiB)": 117.54,
      "step": 6690,
      "train_speed(iter/s)": 0.203826
    },
    {
      "acc": 0.75775089,
      "epoch": 0.15631607343355838,
      "grad_norm": 6.21875,
      "learning_rate": 9.939310425730342e-06,
      "loss": 0.83715219,
      "memory(GiB)": 117.54,
      "step": 6700,
      "train_speed(iter/s)": 0.203982
    },
    {
      "acc": 0.75873032,
      "epoch": 0.15654938100584728,
      "grad_norm": 4.875,
      "learning_rate": 9.939016630844758e-06,
      "loss": 0.87146358,
      "memory(GiB)": 117.54,
      "step": 6710,
      "train_speed(iter/s)": 0.204137
    },
    {
      "acc": 0.76880903,
      "epoch": 0.15678268857813615,
      "grad_norm": 5.53125,
      "learning_rate": 9.938722130915988e-06,
      "loss": 0.81287689,
      "memory(GiB)": 117.54,
      "step": 6720,
      "train_speed(iter/s)": 0.20429
    },
    {
      "acc": 0.76790185,
      "epoch": 0.15701599615042505,
      "grad_norm": 4.84375,
      "learning_rate": 9.938426925986066e-06,
      "loss": 0.85982666,
      "memory(GiB)": 117.54,
      "step": 6730,
      "train_speed(iter/s)": 0.204444
    },
    {
      "acc": 0.75649314,
      "epoch": 0.15724930372271395,
      "grad_norm": 5.15625,
      "learning_rate": 9.938131016097137e-06,
      "loss": 0.89570599,
      "memory(GiB)": 117.54,
      "step": 6740,
      "train_speed(iter/s)": 0.204591
    },
    {
      "acc": 0.75742426,
      "epoch": 0.15748261129500285,
      "grad_norm": 6.375,
      "learning_rate": 9.937834401291437e-06,
      "loss": 0.88885574,
      "memory(GiB)": 117.54,
      "step": 6750,
      "train_speed(iter/s)": 0.204744
    },
    {
      "acc": 0.75880747,
      "epoch": 0.15771591886729175,
      "grad_norm": 6.5,
      "learning_rate": 9.937537081611313e-06,
      "loss": 0.89389744,
      "memory(GiB)": 117.54,
      "step": 6760,
      "train_speed(iter/s)": 0.2049
    },
    {
      "acc": 0.76473279,
      "epoch": 0.15794922643958062,
      "grad_norm": 5.03125,
      "learning_rate": 9.937239057099205e-06,
      "loss": 0.84799232,
      "memory(GiB)": 117.54,
      "step": 6770,
      "train_speed(iter/s)": 0.205043
    },
    {
      "acc": 0.76936255,
      "epoch": 0.15818253401186952,
      "grad_norm": 6.625,
      "learning_rate": 9.936940327797655e-06,
      "loss": 0.8422473,
      "memory(GiB)": 117.54,
      "step": 6780,
      "train_speed(iter/s)": 0.205188
    },
    {
      "acc": 0.75960865,
      "epoch": 0.15841584158415842,
      "grad_norm": 6.15625,
      "learning_rate": 9.936640893749308e-06,
      "loss": 0.86526461,
      "memory(GiB)": 117.54,
      "step": 6790,
      "train_speed(iter/s)": 0.205344
    },
    {
      "acc": 0.75233178,
      "epoch": 0.15864914915644732,
      "grad_norm": 5.59375,
      "learning_rate": 9.936340754996906e-06,
      "loss": 0.89683399,
      "memory(GiB)": 117.54,
      "step": 6800,
      "train_speed(iter/s)": 0.205496
    },
    {
      "acc": 0.75799913,
      "epoch": 0.15888245672873622,
      "grad_norm": 6.40625,
      "learning_rate": 9.936039911583298e-06,
      "loss": 0.88384476,
      "memory(GiB)": 117.54,
      "step": 6810,
      "train_speed(iter/s)": 0.205656
    },
    {
      "acc": 0.76789608,
      "epoch": 0.1591157643010251,
      "grad_norm": 6.0,
      "learning_rate": 9.935738363551424e-06,
      "loss": 0.84808712,
      "memory(GiB)": 117.54,
      "step": 6820,
      "train_speed(iter/s)": 0.205793
    },
    {
      "acc": 0.73517952,
      "epoch": 0.15934907187331399,
      "grad_norm": 5.53125,
      "learning_rate": 9.935436110944335e-06,
      "loss": 0.95522289,
      "memory(GiB)": 117.54,
      "step": 6830,
      "train_speed(iter/s)": 0.205947
    },
    {
      "acc": 0.75344524,
      "epoch": 0.15958237944560288,
      "grad_norm": 6.4375,
      "learning_rate": 9.935133153805172e-06,
      "loss": 0.91215868,
      "memory(GiB)": 117.54,
      "step": 6840,
      "train_speed(iter/s)": 0.206099
    },
    {
      "acc": 0.73823147,
      "epoch": 0.15981568701789178,
      "grad_norm": 6.46875,
      "learning_rate": 9.934829492177187e-06,
      "loss": 0.9458313,
      "memory(GiB)": 117.54,
      "step": 6850,
      "train_speed(iter/s)": 0.206255
    },
    {
      "acc": 0.74221478,
      "epoch": 0.16004899459018065,
      "grad_norm": 5.40625,
      "learning_rate": 9.934525126103725e-06,
      "loss": 0.94343185,
      "memory(GiB)": 117.54,
      "step": 6860,
      "train_speed(iter/s)": 0.206398
    },
    {
      "acc": 0.76774483,
      "epoch": 0.16028230216246955,
      "grad_norm": 7.4375,
      "learning_rate": 9.934220055628233e-06,
      "loss": 0.85056896,
      "memory(GiB)": 117.54,
      "step": 6870,
      "train_speed(iter/s)": 0.206556
    },
    {
      "acc": 0.75645094,
      "epoch": 0.16051560973475845,
      "grad_norm": 5.375,
      "learning_rate": 9.933914280794266e-06,
      "loss": 0.87650585,
      "memory(GiB)": 117.54,
      "step": 6880,
      "train_speed(iter/s)": 0.206713
    },
    {
      "acc": 0.76912999,
      "epoch": 0.16074891730704735,
      "grad_norm": 5.75,
      "learning_rate": 9.933607801645464e-06,
      "loss": 0.85444221,
      "memory(GiB)": 117.54,
      "step": 6890,
      "train_speed(iter/s)": 0.206863
    },
    {
      "acc": 0.76916509,
      "epoch": 0.16098222487933625,
      "grad_norm": 5.84375,
      "learning_rate": 9.933300618225584e-06,
      "loss": 0.83676643,
      "memory(GiB)": 117.54,
      "step": 6900,
      "train_speed(iter/s)": 0.207006
    },
    {
      "acc": 0.7689889,
      "epoch": 0.16121553245162512,
      "grad_norm": 12.5625,
      "learning_rate": 9.932992730578473e-06,
      "loss": 0.85195637,
      "memory(GiB)": 117.54,
      "step": 6910,
      "train_speed(iter/s)": 0.207151
    },
    {
      "acc": 0.7639226,
      "epoch": 0.16144884002391402,
      "grad_norm": 4.8125,
      "learning_rate": 9.932684138748083e-06,
      "loss": 0.8682415,
      "memory(GiB)": 117.54,
      "step": 6920,
      "train_speed(iter/s)": 0.207303
    },
    {
      "acc": 0.74718456,
      "epoch": 0.16168214759620292,
      "grad_norm": 5.28125,
      "learning_rate": 9.932374842778466e-06,
      "loss": 0.95030527,
      "memory(GiB)": 117.54,
      "step": 6930,
      "train_speed(iter/s)": 0.207455
    },
    {
      "acc": 0.76108999,
      "epoch": 0.16191545516849182,
      "grad_norm": 8.9375,
      "learning_rate": 9.932064842713773e-06,
      "loss": 0.88821373,
      "memory(GiB)": 117.54,
      "step": 6940,
      "train_speed(iter/s)": 0.207608
    },
    {
      "acc": 0.76204557,
      "epoch": 0.16214876274078072,
      "grad_norm": 5.3125,
      "learning_rate": 9.931754138598256e-06,
      "loss": 0.83484173,
      "memory(GiB)": 117.54,
      "step": 6950,
      "train_speed(iter/s)": 0.207758
    },
    {
      "acc": 0.77546968,
      "epoch": 0.1623820703130696,
      "grad_norm": 6.3125,
      "learning_rate": 9.931442730476266e-06,
      "loss": 0.8154768,
      "memory(GiB)": 117.54,
      "step": 6960,
      "train_speed(iter/s)": 0.207903
    },
    {
      "acc": 0.74842148,
      "epoch": 0.1626153778853585,
      "grad_norm": 7.5625,
      "learning_rate": 9.931130618392262e-06,
      "loss": 0.91668129,
      "memory(GiB)": 117.54,
      "step": 6970,
      "train_speed(iter/s)": 0.208058
    },
    {
      "acc": 0.75227356,
      "epoch": 0.16284868545764739,
      "grad_norm": 5.59375,
      "learning_rate": 9.930817802390794e-06,
      "loss": 0.90838165,
      "memory(GiB)": 117.54,
      "step": 6980,
      "train_speed(iter/s)": 0.208207
    },
    {
      "acc": 0.77258139,
      "epoch": 0.16308199302993628,
      "grad_norm": 5.5625,
      "learning_rate": 9.930504282516517e-06,
      "loss": 0.81022301,
      "memory(GiB)": 117.54,
      "step": 6990,
      "train_speed(iter/s)": 0.208358
    },
    {
      "acc": 0.76104183,
      "epoch": 0.16331530060222518,
      "grad_norm": 4.125,
      "learning_rate": 9.930190058814185e-06,
      "loss": 0.87542782,
      "memory(GiB)": 117.54,
      "step": 7000,
      "train_speed(iter/s)": 0.208517
    },
    {
      "epoch": 0.16331530060222518,
      "eval_acc": 0.7292125980342936,
      "eval_loss": 0.8633277416229248,
      "eval_runtime": 1270.1232,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 7000
    },
    {
      "acc": 0.76679716,
      "epoch": 0.16354860817451405,
      "grad_norm": 4.59375,
      "learning_rate": 9.929875131328655e-06,
      "loss": 0.84782066,
      "memory(GiB)": 117.54,
      "step": 7010,
      "train_speed(iter/s)": 0.200947
    },
    {
      "acc": 0.74878941,
      "epoch": 0.16378191574680295,
      "grad_norm": 4.15625,
      "learning_rate": 9.929559500104883e-06,
      "loss": 0.91403189,
      "memory(GiB)": 117.54,
      "step": 7020,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.76715469,
      "epoch": 0.16401522331909185,
      "grad_norm": 6.4375,
      "learning_rate": 9.929243165187922e-06,
      "loss": 0.84933138,
      "memory(GiB)": 117.54,
      "step": 7030,
      "train_speed(iter/s)": 0.201244
    },
    {
      "acc": 0.76669478,
      "epoch": 0.16424853089138075,
      "grad_norm": 3.796875,
      "learning_rate": 9.928926126622933e-06,
      "loss": 0.84990997,
      "memory(GiB)": 117.54,
      "step": 7040,
      "train_speed(iter/s)": 0.201391
    },
    {
      "acc": 0.76388216,
      "epoch": 0.16448183846366965,
      "grad_norm": 6.25,
      "learning_rate": 9.928608384455172e-06,
      "loss": 0.87723694,
      "memory(GiB)": 117.54,
      "step": 7050,
      "train_speed(iter/s)": 0.201534
    },
    {
      "acc": 0.74291534,
      "epoch": 0.16471514603595852,
      "grad_norm": 32.5,
      "learning_rate": 9.928289938729996e-06,
      "loss": 0.95859528,
      "memory(GiB)": 117.54,
      "step": 7060,
      "train_speed(iter/s)": 0.201681
    },
    {
      "acc": 0.75394406,
      "epoch": 0.16494845360824742,
      "grad_norm": 5.34375,
      "learning_rate": 9.92797078949286e-06,
      "loss": 0.90898361,
      "memory(GiB)": 117.54,
      "step": 7070,
      "train_speed(iter/s)": 0.201821
    },
    {
      "acc": 0.74871893,
      "epoch": 0.16518176118053632,
      "grad_norm": 5.125,
      "learning_rate": 9.927650936789329e-06,
      "loss": 0.90822897,
      "memory(GiB)": 117.54,
      "step": 7080,
      "train_speed(iter/s)": 0.201961
    },
    {
      "acc": 0.77017407,
      "epoch": 0.16541506875282522,
      "grad_norm": 7.90625,
      "learning_rate": 9.927330380665056e-06,
      "loss": 0.84466286,
      "memory(GiB)": 117.54,
      "step": 7090,
      "train_speed(iter/s)": 0.202117
    },
    {
      "acc": 0.75458536,
      "epoch": 0.1656483763251141,
      "grad_norm": 4.9375,
      "learning_rate": 9.927009121165803e-06,
      "loss": 0.90666466,
      "memory(GiB)": 117.54,
      "step": 7100,
      "train_speed(iter/s)": 0.202266
    },
    {
      "acc": 0.75543022,
      "epoch": 0.165881683897403,
      "grad_norm": 5.46875,
      "learning_rate": 9.92668715833743e-06,
      "loss": 0.8722743,
      "memory(GiB)": 117.54,
      "step": 7110,
      "train_speed(iter/s)": 0.202404
    },
    {
      "acc": 0.77939939,
      "epoch": 0.16611499146969189,
      "grad_norm": 10.75,
      "learning_rate": 9.926364492225894e-06,
      "loss": 0.78749428,
      "memory(GiB)": 117.54,
      "step": 7120,
      "train_speed(iter/s)": 0.202551
    },
    {
      "acc": 0.79106226,
      "epoch": 0.16634829904198078,
      "grad_norm": 7.3125,
      "learning_rate": 9.92604112287726e-06,
      "loss": 0.74738359,
      "memory(GiB)": 117.54,
      "step": 7130,
      "train_speed(iter/s)": 0.202705
    },
    {
      "acc": 0.76939073,
      "epoch": 0.16658160661426968,
      "grad_norm": 4.8125,
      "learning_rate": 9.925717050337686e-06,
      "loss": 0.8409626,
      "memory(GiB)": 117.54,
      "step": 7140,
      "train_speed(iter/s)": 0.202843
    },
    {
      "acc": 0.76826558,
      "epoch": 0.16681491418655855,
      "grad_norm": 5.65625,
      "learning_rate": 9.925392274653435e-06,
      "loss": 0.82716637,
      "memory(GiB)": 117.54,
      "step": 7150,
      "train_speed(iter/s)": 0.202993
    },
    {
      "acc": 0.76554303,
      "epoch": 0.16704822175884745,
      "grad_norm": 4.21875,
      "learning_rate": 9.925066795870868e-06,
      "loss": 0.849158,
      "memory(GiB)": 117.54,
      "step": 7160,
      "train_speed(iter/s)": 0.203138
    },
    {
      "acc": 0.78131075,
      "epoch": 0.16728152933113635,
      "grad_norm": 5.65625,
      "learning_rate": 9.924740614036445e-06,
      "loss": 0.77836246,
      "memory(GiB)": 117.54,
      "step": 7170,
      "train_speed(iter/s)": 0.203275
    },
    {
      "acc": 0.77169647,
      "epoch": 0.16751483690342525,
      "grad_norm": 6.46875,
      "learning_rate": 9.92441372919673e-06,
      "loss": 0.83209658,
      "memory(GiB)": 117.54,
      "step": 7180,
      "train_speed(iter/s)": 0.203419
    },
    {
      "acc": 0.75183039,
      "epoch": 0.16774814447571415,
      "grad_norm": 4.875,
      "learning_rate": 9.924086141398385e-06,
      "loss": 0.92607002,
      "memory(GiB)": 117.54,
      "step": 7190,
      "train_speed(iter/s)": 0.203569
    },
    {
      "acc": 0.77899566,
      "epoch": 0.16798145204800302,
      "grad_norm": 9.9375,
      "learning_rate": 9.923757850688176e-06,
      "loss": 0.78890448,
      "memory(GiB)": 117.54,
      "step": 7200,
      "train_speed(iter/s)": 0.203724
    },
    {
      "acc": 0.7672864,
      "epoch": 0.16821475962029192,
      "grad_norm": 5.375,
      "learning_rate": 9.923428857112963e-06,
      "loss": 0.82765884,
      "memory(GiB)": 117.54,
      "step": 7210,
      "train_speed(iter/s)": 0.203869
    },
    {
      "acc": 0.76710939,
      "epoch": 0.16844806719258082,
      "grad_norm": 7.78125,
      "learning_rate": 9.923099160719711e-06,
      "loss": 0.83284779,
      "memory(GiB)": 117.54,
      "step": 7220,
      "train_speed(iter/s)": 0.204016
    },
    {
      "acc": 0.76684504,
      "epoch": 0.16868137476486972,
      "grad_norm": 7.5,
      "learning_rate": 9.922768761555485e-06,
      "loss": 0.84021397,
      "memory(GiB)": 117.54,
      "step": 7230,
      "train_speed(iter/s)": 0.20416
    },
    {
      "acc": 0.75771241,
      "epoch": 0.16891468233715862,
      "grad_norm": 6.625,
      "learning_rate": 9.922437659667448e-06,
      "loss": 0.8986228,
      "memory(GiB)": 117.54,
      "step": 7240,
      "train_speed(iter/s)": 0.204305
    },
    {
      "acc": 0.77050877,
      "epoch": 0.1691479899094475,
      "grad_norm": 5.0,
      "learning_rate": 9.922105855102864e-06,
      "loss": 0.83454065,
      "memory(GiB)": 117.54,
      "step": 7250,
      "train_speed(iter/s)": 0.204448
    },
    {
      "acc": 0.77722654,
      "epoch": 0.1693812974817364,
      "grad_norm": 6.21875,
      "learning_rate": 9.921773347909098e-06,
      "loss": 0.79222279,
      "memory(GiB)": 117.54,
      "step": 7260,
      "train_speed(iter/s)": 0.204598
    },
    {
      "acc": 0.76436782,
      "epoch": 0.16961460505402529,
      "grad_norm": 6.75,
      "learning_rate": 9.921440138133619e-06,
      "loss": 0.843188,
      "memory(GiB)": 117.54,
      "step": 7270,
      "train_speed(iter/s)": 0.20475
    },
    {
      "acc": 0.75543094,
      "epoch": 0.16984791262631418,
      "grad_norm": 6.15625,
      "learning_rate": 9.921106225823988e-06,
      "loss": 0.91073751,
      "memory(GiB)": 117.54,
      "step": 7280,
      "train_speed(iter/s)": 0.204899
    },
    {
      "acc": 0.76456766,
      "epoch": 0.17008122019860308,
      "grad_norm": 5.625,
      "learning_rate": 9.920771611027875e-06,
      "loss": 0.83511429,
      "memory(GiB)": 117.54,
      "step": 7290,
      "train_speed(iter/s)": 0.205052
    },
    {
      "acc": 0.73825827,
      "epoch": 0.17031452777089195,
      "grad_norm": 5.3125,
      "learning_rate": 9.920436293793043e-06,
      "loss": 0.97240715,
      "memory(GiB)": 117.54,
      "step": 7300,
      "train_speed(iter/s)": 0.205195
    },
    {
      "acc": 0.74200888,
      "epoch": 0.17054783534318085,
      "grad_norm": 6.59375,
      "learning_rate": 9.920100274167359e-06,
      "loss": 0.9331625,
      "memory(GiB)": 117.54,
      "step": 7310,
      "train_speed(iter/s)": 0.205347
    },
    {
      "acc": 0.76395473,
      "epoch": 0.17078114291546975,
      "grad_norm": 7.46875,
      "learning_rate": 9.91976355219879e-06,
      "loss": 0.85346889,
      "memory(GiB)": 117.54,
      "step": 7320,
      "train_speed(iter/s)": 0.205459
    },
    {
      "acc": 0.75636325,
      "epoch": 0.17101445048775865,
      "grad_norm": 5.78125,
      "learning_rate": 9.919426127935404e-06,
      "loss": 0.89509583,
      "memory(GiB)": 117.54,
      "step": 7330,
      "train_speed(iter/s)": 0.205592
    },
    {
      "acc": 0.76570711,
      "epoch": 0.17124775806004755,
      "grad_norm": 5.53125,
      "learning_rate": 9.919088001425367e-06,
      "loss": 0.85378265,
      "memory(GiB)": 117.54,
      "step": 7340,
      "train_speed(iter/s)": 0.205738
    },
    {
      "acc": 0.75373197,
      "epoch": 0.17148106563233642,
      "grad_norm": 8.8125,
      "learning_rate": 9.918749172716946e-06,
      "loss": 0.88216324,
      "memory(GiB)": 117.54,
      "step": 7350,
      "train_speed(iter/s)": 0.205869
    },
    {
      "acc": 0.75796227,
      "epoch": 0.17171437320462532,
      "grad_norm": 7.0,
      "learning_rate": 9.91840964185851e-06,
      "loss": 0.8882925,
      "memory(GiB)": 117.54,
      "step": 7360,
      "train_speed(iter/s)": 0.206007
    },
    {
      "acc": 0.76364655,
      "epoch": 0.17194768077691422,
      "grad_norm": 5.53125,
      "learning_rate": 9.918069408898527e-06,
      "loss": 0.85295391,
      "memory(GiB)": 117.54,
      "step": 7370,
      "train_speed(iter/s)": 0.206146
    },
    {
      "acc": 0.77610121,
      "epoch": 0.17218098834920312,
      "grad_norm": 6.40625,
      "learning_rate": 9.917728473885564e-06,
      "loss": 0.78088756,
      "memory(GiB)": 117.54,
      "step": 7380,
      "train_speed(iter/s)": 0.206278
    },
    {
      "acc": 0.7751833,
      "epoch": 0.172414295921492,
      "grad_norm": 8.0,
      "learning_rate": 9.91738683686829e-06,
      "loss": 0.82343903,
      "memory(GiB)": 117.54,
      "step": 7390,
      "train_speed(iter/s)": 0.206425
    },
    {
      "acc": 0.75816336,
      "epoch": 0.1726476034937809,
      "grad_norm": 6.40625,
      "learning_rate": 9.917044497895474e-06,
      "loss": 0.86748638,
      "memory(GiB)": 117.54,
      "step": 7400,
      "train_speed(iter/s)": 0.206522
    },
    {
      "acc": 0.76400776,
      "epoch": 0.1728809110660698,
      "grad_norm": 5.25,
      "learning_rate": 9.916701457015983e-06,
      "loss": 0.86720943,
      "memory(GiB)": 117.54,
      "step": 7410,
      "train_speed(iter/s)": 0.206658
    },
    {
      "acc": 0.75771341,
      "epoch": 0.17311421863835869,
      "grad_norm": 5.15625,
      "learning_rate": 9.91635771427879e-06,
      "loss": 0.88320751,
      "memory(GiB)": 117.54,
      "step": 7420,
      "train_speed(iter/s)": 0.206797
    },
    {
      "acc": 0.7662806,
      "epoch": 0.17334752621064758,
      "grad_norm": 6.0625,
      "learning_rate": 9.91601326973296e-06,
      "loss": 0.835116,
      "memory(GiB)": 117.54,
      "step": 7430,
      "train_speed(iter/s)": 0.206941
    },
    {
      "acc": 0.77026176,
      "epoch": 0.17358083378293646,
      "grad_norm": 7.625,
      "learning_rate": 9.915668123427662e-06,
      "loss": 0.84814157,
      "memory(GiB)": 117.54,
      "step": 7440,
      "train_speed(iter/s)": 0.207086
    },
    {
      "acc": 0.76674738,
      "epoch": 0.17381414135522535,
      "grad_norm": 4.6875,
      "learning_rate": 9.91532227541217e-06,
      "loss": 0.85983877,
      "memory(GiB)": 117.54,
      "step": 7450,
      "train_speed(iter/s)": 0.207226
    },
    {
      "acc": 0.75598326,
      "epoch": 0.17404744892751425,
      "grad_norm": 6.03125,
      "learning_rate": 9.91497572573585e-06,
      "loss": 0.87792206,
      "memory(GiB)": 117.54,
      "step": 7460,
      "train_speed(iter/s)": 0.207368
    },
    {
      "acc": 0.77461433,
      "epoch": 0.17428075649980315,
      "grad_norm": 4.8125,
      "learning_rate": 9.914628474448173e-06,
      "loss": 0.81929722,
      "memory(GiB)": 117.54,
      "step": 7470,
      "train_speed(iter/s)": 0.207504
    },
    {
      "acc": 0.75475292,
      "epoch": 0.17451406407209205,
      "grad_norm": 5.53125,
      "learning_rate": 9.91428052159871e-06,
      "loss": 0.88305016,
      "memory(GiB)": 117.54,
      "step": 7480,
      "train_speed(iter/s)": 0.207644
    },
    {
      "acc": 0.77698159,
      "epoch": 0.17474737164438092,
      "grad_norm": 5.40625,
      "learning_rate": 9.913931867237129e-06,
      "loss": 0.81536446,
      "memory(GiB)": 117.54,
      "step": 7490,
      "train_speed(iter/s)": 0.207778
    },
    {
      "acc": 0.75837955,
      "epoch": 0.17498067921666982,
      "grad_norm": 4.71875,
      "learning_rate": 9.913582511413201e-06,
      "loss": 0.87997665,
      "memory(GiB)": 117.54,
      "step": 7500,
      "train_speed(iter/s)": 0.207923
    },
    {
      "epoch": 0.17498067921666982,
      "eval_acc": 0.7296547562596777,
      "eval_loss": 0.8604273200035095,
      "eval_runtime": 1269.7388,
      "eval_samples_per_second": 28.345,
      "eval_steps_per_second": 14.173,
      "step": 7500
    },
    {
      "acc": 0.75066805,
      "epoch": 0.17521398678895872,
      "grad_norm": 5.96875,
      "learning_rate": 9.913232454176797e-06,
      "loss": 0.88283882,
      "memory(GiB)": 117.54,
      "step": 7510,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.73445344,
      "epoch": 0.17544729436124762,
      "grad_norm": 6.25,
      "learning_rate": 9.912881695577889e-06,
      "loss": 0.97776871,
      "memory(GiB)": 117.54,
      "step": 7520,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.77131672,
      "epoch": 0.17568060193353652,
      "grad_norm": 6.21875,
      "learning_rate": 9.912530235666546e-06,
      "loss": 0.82929153,
      "memory(GiB)": 117.54,
      "step": 7530,
      "train_speed(iter/s)": 0.201167
    },
    {
      "acc": 0.77159753,
      "epoch": 0.1759139095058254,
      "grad_norm": 8.5625,
      "learning_rate": 9.912178074492937e-06,
      "loss": 0.81753139,
      "memory(GiB)": 117.54,
      "step": 7540,
      "train_speed(iter/s)": 0.201308
    },
    {
      "acc": 0.75888309,
      "epoch": 0.1761472170781143,
      "grad_norm": 5.9375,
      "learning_rate": 9.911825212107337e-06,
      "loss": 0.85590801,
      "memory(GiB)": 117.54,
      "step": 7550,
      "train_speed(iter/s)": 0.20145
    },
    {
      "acc": 0.77193937,
      "epoch": 0.1763805246504032,
      "grad_norm": 6.21875,
      "learning_rate": 9.911471648560114e-06,
      "loss": 0.81675425,
      "memory(GiB)": 117.54,
      "step": 7560,
      "train_speed(iter/s)": 0.201582
    },
    {
      "acc": 0.75802794,
      "epoch": 0.17661383222269209,
      "grad_norm": 5.625,
      "learning_rate": 9.91111738390174e-06,
      "loss": 0.91160278,
      "memory(GiB)": 117.54,
      "step": 7570,
      "train_speed(iter/s)": 0.201724
    },
    {
      "acc": 0.77081351,
      "epoch": 0.17684713979498098,
      "grad_norm": 4.84375,
      "learning_rate": 9.910762418182786e-06,
      "loss": 0.81983862,
      "memory(GiB)": 117.54,
      "step": 7580,
      "train_speed(iter/s)": 0.201858
    },
    {
      "acc": 0.76340055,
      "epoch": 0.17708044736726986,
      "grad_norm": 14.375,
      "learning_rate": 9.910406751453923e-06,
      "loss": 0.85830593,
      "memory(GiB)": 117.54,
      "step": 7590,
      "train_speed(iter/s)": 0.202005
    },
    {
      "acc": 0.75996342,
      "epoch": 0.17731375493955875,
      "grad_norm": 5.0625,
      "learning_rate": 9.910050383765924e-06,
      "loss": 0.86299267,
      "memory(GiB)": 117.54,
      "step": 7600,
      "train_speed(iter/s)": 0.202147
    },
    {
      "acc": 0.73427224,
      "epoch": 0.17754706251184765,
      "grad_norm": 4.875,
      "learning_rate": 9.909693315169657e-06,
      "loss": 0.96497478,
      "memory(GiB)": 117.54,
      "step": 7610,
      "train_speed(iter/s)": 0.202278
    },
    {
      "acc": 0.76901188,
      "epoch": 0.17778037008413655,
      "grad_norm": 5.9375,
      "learning_rate": 9.909335545716097e-06,
      "loss": 0.82169056,
      "memory(GiB)": 117.54,
      "step": 7620,
      "train_speed(iter/s)": 0.202416
    },
    {
      "acc": 0.75650215,
      "epoch": 0.17801367765642542,
      "grad_norm": 6.53125,
      "learning_rate": 9.908977075456314e-06,
      "loss": 0.8937892,
      "memory(GiB)": 117.54,
      "step": 7630,
      "train_speed(iter/s)": 0.202547
    },
    {
      "acc": 0.78488522,
      "epoch": 0.17824698522871432,
      "grad_norm": 5.53125,
      "learning_rate": 9.90861790444148e-06,
      "loss": 0.76624699,
      "memory(GiB)": 117.54,
      "step": 7640,
      "train_speed(iter/s)": 0.202691
    },
    {
      "acc": 0.77700996,
      "epoch": 0.17848029280100322,
      "grad_norm": 5.75,
      "learning_rate": 9.908258032722865e-06,
      "loss": 0.79376793,
      "memory(GiB)": 117.54,
      "step": 7650,
      "train_speed(iter/s)": 0.202822
    },
    {
      "acc": 0.78278093,
      "epoch": 0.17871360037329212,
      "grad_norm": 6.09375,
      "learning_rate": 9.907897460351842e-06,
      "loss": 0.80827179,
      "memory(GiB)": 117.54,
      "step": 7660,
      "train_speed(iter/s)": 0.202955
    },
    {
      "acc": 0.7604249,
      "epoch": 0.17894690794558102,
      "grad_norm": 7.09375,
      "learning_rate": 9.907536187379883e-06,
      "loss": 0.88734035,
      "memory(GiB)": 117.54,
      "step": 7670,
      "train_speed(iter/s)": 0.203091
    },
    {
      "acc": 0.76182446,
      "epoch": 0.1791802155178699,
      "grad_norm": 5.90625,
      "learning_rate": 9.907174213858556e-06,
      "loss": 0.88050871,
      "memory(GiB)": 117.54,
      "step": 7680,
      "train_speed(iter/s)": 0.203227
    },
    {
      "acc": 0.76138544,
      "epoch": 0.1794135230901588,
      "grad_norm": 12.4375,
      "learning_rate": 9.906811539839539e-06,
      "loss": 0.86582584,
      "memory(GiB)": 117.54,
      "step": 7690,
      "train_speed(iter/s)": 0.203366
    },
    {
      "acc": 0.77104373,
      "epoch": 0.1796468306624477,
      "grad_norm": 5.1875,
      "learning_rate": 9.9064481653746e-06,
      "loss": 0.83353386,
      "memory(GiB)": 117.54,
      "step": 7700,
      "train_speed(iter/s)": 0.203486
    },
    {
      "acc": 0.77400684,
      "epoch": 0.1798801382347366,
      "grad_norm": 4.46875,
      "learning_rate": 9.906084090515609e-06,
      "loss": 0.83886662,
      "memory(GiB)": 117.54,
      "step": 7710,
      "train_speed(iter/s)": 0.203625
    },
    {
      "acc": 0.75984573,
      "epoch": 0.18011344580702549,
      "grad_norm": 5.96875,
      "learning_rate": 9.90571931531454e-06,
      "loss": 0.88316107,
      "memory(GiB)": 117.54,
      "step": 7720,
      "train_speed(iter/s)": 0.203764
    },
    {
      "acc": 0.76342249,
      "epoch": 0.18034675337931436,
      "grad_norm": 4.46875,
      "learning_rate": 9.905353839823463e-06,
      "loss": 0.85615129,
      "memory(GiB)": 117.54,
      "step": 7730,
      "train_speed(iter/s)": 0.203908
    },
    {
      "acc": 0.78326778,
      "epoch": 0.18058006095160326,
      "grad_norm": 6.09375,
      "learning_rate": 9.904987664094553e-06,
      "loss": 0.76730537,
      "memory(GiB)": 117.54,
      "step": 7740,
      "train_speed(iter/s)": 0.204033
    },
    {
      "acc": 0.75868654,
      "epoch": 0.18081336852389215,
      "grad_norm": 6.40625,
      "learning_rate": 9.904620788180076e-06,
      "loss": 0.89008236,
      "memory(GiB)": 117.54,
      "step": 7750,
      "train_speed(iter/s)": 0.204174
    },
    {
      "acc": 0.76263795,
      "epoch": 0.18104667609618105,
      "grad_norm": 5.5,
      "learning_rate": 9.904253212132406e-06,
      "loss": 0.87411928,
      "memory(GiB)": 117.54,
      "step": 7760,
      "train_speed(iter/s)": 0.204312
    },
    {
      "acc": 0.76147628,
      "epoch": 0.18127998366846995,
      "grad_norm": 6.0,
      "learning_rate": 9.903884936004017e-06,
      "loss": 0.86704102,
      "memory(GiB)": 117.54,
      "step": 7770,
      "train_speed(iter/s)": 0.204445
    },
    {
      "acc": 0.76194277,
      "epoch": 0.18151329124075882,
      "grad_norm": 7.5,
      "learning_rate": 9.903515959847477e-06,
      "loss": 0.83300629,
      "memory(GiB)": 117.54,
      "step": 7780,
      "train_speed(iter/s)": 0.204576
    },
    {
      "acc": 0.7630332,
      "epoch": 0.18174659881304772,
      "grad_norm": 8.5,
      "learning_rate": 9.903146283715459e-06,
      "loss": 0.85758762,
      "memory(GiB)": 117.54,
      "step": 7790,
      "train_speed(iter/s)": 0.204716
    },
    {
      "acc": 0.74647932,
      "epoch": 0.18197990638533662,
      "grad_norm": 5.5625,
      "learning_rate": 9.902775907660733e-06,
      "loss": 0.91409321,
      "memory(GiB)": 117.54,
      "step": 7800,
      "train_speed(iter/s)": 0.204846
    },
    {
      "acc": 0.74644547,
      "epoch": 0.18221321395762552,
      "grad_norm": 5.1875,
      "learning_rate": 9.90240483173617e-06,
      "loss": 0.95661182,
      "memory(GiB)": 117.54,
      "step": 7810,
      "train_speed(iter/s)": 0.204982
    },
    {
      "acc": 0.75153122,
      "epoch": 0.18244652152991442,
      "grad_norm": 6.4375,
      "learning_rate": 9.902033055994739e-06,
      "loss": 0.90269318,
      "memory(GiB)": 117.54,
      "step": 7820,
      "train_speed(iter/s)": 0.205106
    },
    {
      "acc": 0.75813732,
      "epoch": 0.1826798291022033,
      "grad_norm": 83.5,
      "learning_rate": 9.901660580489517e-06,
      "loss": 0.88405313,
      "memory(GiB)": 117.54,
      "step": 7830,
      "train_speed(iter/s)": 0.20524
    },
    {
      "acc": 0.78136988,
      "epoch": 0.1829131366744922,
      "grad_norm": 9.25,
      "learning_rate": 9.90128740527367e-06,
      "loss": 0.79004688,
      "memory(GiB)": 117.54,
      "step": 7840,
      "train_speed(iter/s)": 0.205378
    },
    {
      "acc": 0.77203064,
      "epoch": 0.1831464442467811,
      "grad_norm": 4.6875,
      "learning_rate": 9.900913530400469e-06,
      "loss": 0.80475359,
      "memory(GiB)": 117.54,
      "step": 7850,
      "train_speed(iter/s)": 0.205502
    },
    {
      "acc": 0.76842432,
      "epoch": 0.18337975181906999,
      "grad_norm": 5.03125,
      "learning_rate": 9.900538955923287e-06,
      "loss": 0.84965286,
      "memory(GiB)": 117.54,
      "step": 7860,
      "train_speed(iter/s)": 0.205631
    },
    {
      "acc": 0.7858345,
      "epoch": 0.18361305939135886,
      "grad_norm": 4.625,
      "learning_rate": 9.900163681895591e-06,
      "loss": 0.7589952,
      "memory(GiB)": 117.54,
      "step": 7870,
      "train_speed(iter/s)": 0.205758
    },
    {
      "acc": 0.75342913,
      "epoch": 0.18384636696364776,
      "grad_norm": 6.03125,
      "learning_rate": 9.899787708370954e-06,
      "loss": 0.89493637,
      "memory(GiB)": 117.54,
      "step": 7880,
      "train_speed(iter/s)": 0.20589
    },
    {
      "acc": 0.7876399,
      "epoch": 0.18407967453593665,
      "grad_norm": 7.21875,
      "learning_rate": 9.899411035403044e-06,
      "loss": 0.74116306,
      "memory(GiB)": 117.54,
      "step": 7890,
      "train_speed(iter/s)": 0.206028
    },
    {
      "acc": 0.78092766,
      "epoch": 0.18431298210822555,
      "grad_norm": 4.71875,
      "learning_rate": 9.899033663045632e-06,
      "loss": 0.79192057,
      "memory(GiB)": 117.54,
      "step": 7900,
      "train_speed(iter/s)": 0.206168
    },
    {
      "acc": 0.75420218,
      "epoch": 0.18454628968051445,
      "grad_norm": 9.8125,
      "learning_rate": 9.898655591352589e-06,
      "loss": 0.88830757,
      "memory(GiB)": 117.54,
      "step": 7910,
      "train_speed(iter/s)": 0.206303
    },
    {
      "acc": 0.75896831,
      "epoch": 0.18477959725280332,
      "grad_norm": 6.8125,
      "learning_rate": 9.898276820377882e-06,
      "loss": 0.8895916,
      "memory(GiB)": 117.54,
      "step": 7920,
      "train_speed(iter/s)": 0.206432
    },
    {
      "acc": 0.75719609,
      "epoch": 0.18501290482509222,
      "grad_norm": 6.875,
      "learning_rate": 9.897897350175583e-06,
      "loss": 0.89884977,
      "memory(GiB)": 117.54,
      "step": 7930,
      "train_speed(iter/s)": 0.206564
    },
    {
      "acc": 0.75096016,
      "epoch": 0.18524621239738112,
      "grad_norm": 6.875,
      "learning_rate": 9.897517180799858e-06,
      "loss": 0.90399666,
      "memory(GiB)": 117.54,
      "step": 7940,
      "train_speed(iter/s)": 0.206695
    },
    {
      "acc": 0.76634254,
      "epoch": 0.18547951996967002,
      "grad_norm": 5.5,
      "learning_rate": 9.89713631230498e-06,
      "loss": 0.84754448,
      "memory(GiB)": 117.54,
      "step": 7950,
      "train_speed(iter/s)": 0.206826
    },
    {
      "acc": 0.76263399,
      "epoch": 0.18571282754195892,
      "grad_norm": 5.1875,
      "learning_rate": 9.896754744745315e-06,
      "loss": 0.87414799,
      "memory(GiB)": 117.54,
      "step": 7960,
      "train_speed(iter/s)": 0.206959
    },
    {
      "acc": 0.76916466,
      "epoch": 0.1859461351142478,
      "grad_norm": 6.25,
      "learning_rate": 9.896372478175336e-06,
      "loss": 0.85287285,
      "memory(GiB)": 117.54,
      "step": 7970,
      "train_speed(iter/s)": 0.207091
    },
    {
      "acc": 0.76025381,
      "epoch": 0.1861794426865367,
      "grad_norm": 23.5,
      "learning_rate": 9.895989512649605e-06,
      "loss": 0.85706806,
      "memory(GiB)": 117.54,
      "step": 7980,
      "train_speed(iter/s)": 0.207216
    },
    {
      "acc": 0.78468599,
      "epoch": 0.1864127502588256,
      "grad_norm": 8.4375,
      "learning_rate": 9.895605848222794e-06,
      "loss": 0.78214846,
      "memory(GiB)": 117.54,
      "step": 7990,
      "train_speed(iter/s)": 0.207344
    },
    {
      "acc": 0.76179285,
      "epoch": 0.1866460578311145,
      "grad_norm": 5.4375,
      "learning_rate": 9.89522148494967e-06,
      "loss": 0.87004957,
      "memory(GiB)": 117.54,
      "step": 8000,
      "train_speed(iter/s)": 0.20748
    },
    {
      "epoch": 0.1866460578311145,
      "eval_acc": 0.7304815809310498,
      "eval_loss": 0.8580639958381653,
      "eval_runtime": 1269.3016,
      "eval_samples_per_second": 28.355,
      "eval_steps_per_second": 14.178,
      "step": 8000
    },
    {
      "acc": 0.75695009,
      "epoch": 0.18687936540340339,
      "grad_norm": 8.125,
      "learning_rate": 9.894836422885101e-06,
      "loss": 0.87596626,
      "memory(GiB)": 117.54,
      "step": 8010,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.76842041,
      "epoch": 0.18711267297569226,
      "grad_norm": 5.1875,
      "learning_rate": 9.894450662084055e-06,
      "loss": 0.84125156,
      "memory(GiB)": 117.54,
      "step": 8020,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.75734787,
      "epoch": 0.18734598054798116,
      "grad_norm": 5.4375,
      "learning_rate": 9.8940642026016e-06,
      "loss": 0.88770895,
      "memory(GiB)": 117.54,
      "step": 8030,
      "train_speed(iter/s)": 0.201174
    },
    {
      "acc": 0.74493351,
      "epoch": 0.18757928812027005,
      "grad_norm": 17.875,
      "learning_rate": 9.8936770444929e-06,
      "loss": 0.93823605,
      "memory(GiB)": 117.54,
      "step": 8040,
      "train_speed(iter/s)": 0.201304
    },
    {
      "acc": 0.77957907,
      "epoch": 0.18781259569255895,
      "grad_norm": 4.5625,
      "learning_rate": 9.893289187813224e-06,
      "loss": 0.79470148,
      "memory(GiB)": 117.54,
      "step": 8050,
      "train_speed(iter/s)": 0.201441
    },
    {
      "acc": 0.76955447,
      "epoch": 0.18804590326484785,
      "grad_norm": 6.0,
      "learning_rate": 9.892900632617939e-06,
      "loss": 0.82833557,
      "memory(GiB)": 117.54,
      "step": 8060,
      "train_speed(iter/s)": 0.201567
    },
    {
      "acc": 0.77944613,
      "epoch": 0.18827921083713672,
      "grad_norm": 5.3125,
      "learning_rate": 9.892511378962509e-06,
      "loss": 0.78586206,
      "memory(GiB)": 117.54,
      "step": 8070,
      "train_speed(iter/s)": 0.201694
    },
    {
      "acc": 0.78304958,
      "epoch": 0.18851251840942562,
      "grad_norm": 6.1875,
      "learning_rate": 9.892121426902502e-06,
      "loss": 0.7807333,
      "memory(GiB)": 117.54,
      "step": 8080,
      "train_speed(iter/s)": 0.201821
    },
    {
      "acc": 0.77091231,
      "epoch": 0.18874582598171452,
      "grad_norm": 7.75,
      "learning_rate": 9.891730776493579e-06,
      "loss": 0.82303715,
      "memory(GiB)": 117.54,
      "step": 8090,
      "train_speed(iter/s)": 0.201941
    },
    {
      "acc": 0.76864967,
      "epoch": 0.18897913355400342,
      "grad_norm": 6.03125,
      "learning_rate": 9.891339427791513e-06,
      "loss": 0.82791605,
      "memory(GiB)": 117.54,
      "step": 8100,
      "train_speed(iter/s)": 0.202068
    },
    {
      "acc": 0.78619051,
      "epoch": 0.18921244112629232,
      "grad_norm": 6.125,
      "learning_rate": 9.890947380852163e-06,
      "loss": 0.78398085,
      "memory(GiB)": 117.54,
      "step": 8110,
      "train_speed(iter/s)": 0.202193
    },
    {
      "acc": 0.763907,
      "epoch": 0.1894457486985812,
      "grad_norm": 5.3125,
      "learning_rate": 9.890554635731496e-06,
      "loss": 0.86447706,
      "memory(GiB)": 117.54,
      "step": 8120,
      "train_speed(iter/s)": 0.202319
    },
    {
      "acc": 0.75075836,
      "epoch": 0.1896790562708701,
      "grad_norm": 4.6875,
      "learning_rate": 9.890161192485573e-06,
      "loss": 0.91182461,
      "memory(GiB)": 117.54,
      "step": 8130,
      "train_speed(iter/s)": 0.202452
    },
    {
      "acc": 0.78794985,
      "epoch": 0.189912363843159,
      "grad_norm": 5.28125,
      "learning_rate": 9.889767051170563e-06,
      "loss": 0.75603533,
      "memory(GiB)": 117.54,
      "step": 8140,
      "train_speed(iter/s)": 0.202579
    },
    {
      "acc": 0.77740474,
      "epoch": 0.1901456714154479,
      "grad_norm": 6.6875,
      "learning_rate": 9.889372211842726e-06,
      "loss": 0.80326691,
      "memory(GiB)": 117.54,
      "step": 8150,
      "train_speed(iter/s)": 0.202706
    },
    {
      "acc": 0.77399597,
      "epoch": 0.19037897898773676,
      "grad_norm": 5.03125,
      "learning_rate": 9.888976674558426e-06,
      "loss": 0.8158947,
      "memory(GiB)": 117.54,
      "step": 8160,
      "train_speed(iter/s)": 0.202837
    },
    {
      "acc": 0.74653988,
      "epoch": 0.19061228656002566,
      "grad_norm": 6.375,
      "learning_rate": 9.888580439374126e-06,
      "loss": 0.91792946,
      "memory(GiB)": 117.54,
      "step": 8170,
      "train_speed(iter/s)": 0.20296
    },
    {
      "acc": 0.74516058,
      "epoch": 0.19084559413231456,
      "grad_norm": 5.34375,
      "learning_rate": 9.888183506346389e-06,
      "loss": 0.91901636,
      "memory(GiB)": 117.54,
      "step": 8180,
      "train_speed(iter/s)": 0.203084
    },
    {
      "acc": 0.76203327,
      "epoch": 0.19107890170460345,
      "grad_norm": 6.4375,
      "learning_rate": 9.887785875531875e-06,
      "loss": 0.87705736,
      "memory(GiB)": 117.54,
      "step": 8190,
      "train_speed(iter/s)": 0.203213
    },
    {
      "acc": 0.75964112,
      "epoch": 0.19131220927689235,
      "grad_norm": 4.25,
      "learning_rate": 9.887387546987349e-06,
      "loss": 0.87349434,
      "memory(GiB)": 117.54,
      "step": 8200,
      "train_speed(iter/s)": 0.203336
    },
    {
      "acc": 0.77561655,
      "epoch": 0.19154551684918122,
      "grad_norm": 4.21875,
      "learning_rate": 9.886988520769669e-06,
      "loss": 0.80881062,
      "memory(GiB)": 117.54,
      "step": 8210,
      "train_speed(iter/s)": 0.203451
    },
    {
      "acc": 0.76098909,
      "epoch": 0.19177882442147012,
      "grad_norm": 6.96875,
      "learning_rate": 9.886588796935797e-06,
      "loss": 0.87041025,
      "memory(GiB)": 117.54,
      "step": 8220,
      "train_speed(iter/s)": 0.203578
    },
    {
      "acc": 0.7447237,
      "epoch": 0.19201213199375902,
      "grad_norm": 5.25,
      "learning_rate": 9.886188375542795e-06,
      "loss": 0.93633842,
      "memory(GiB)": 117.54,
      "step": 8230,
      "train_speed(iter/s)": 0.20371
    },
    {
      "acc": 0.76472883,
      "epoch": 0.19224543956604792,
      "grad_norm": 5.15625,
      "learning_rate": 9.885787256647822e-06,
      "loss": 0.85684052,
      "memory(GiB)": 117.54,
      "step": 8240,
      "train_speed(iter/s)": 0.203835
    },
    {
      "acc": 0.78805923,
      "epoch": 0.19247874713833682,
      "grad_norm": 5.84375,
      "learning_rate": 9.885385440308137e-06,
      "loss": 0.73936048,
      "memory(GiB)": 117.54,
      "step": 8250,
      "train_speed(iter/s)": 0.203964
    },
    {
      "acc": 0.75716724,
      "epoch": 0.1927120547106257,
      "grad_norm": 7.59375,
      "learning_rate": 9.8849829265811e-06,
      "loss": 0.89653664,
      "memory(GiB)": 117.54,
      "step": 8260,
      "train_speed(iter/s)": 0.204088
    },
    {
      "acc": 0.78018217,
      "epoch": 0.1929453622829146,
      "grad_norm": 5.75,
      "learning_rate": 9.884579715524168e-06,
      "loss": 0.77996178,
      "memory(GiB)": 117.54,
      "step": 8270,
      "train_speed(iter/s)": 0.204218
    },
    {
      "acc": 0.74880614,
      "epoch": 0.1931786698552035,
      "grad_norm": 6.28125,
      "learning_rate": 9.884175807194902e-06,
      "loss": 0.90730553,
      "memory(GiB)": 117.54,
      "step": 8280,
      "train_speed(iter/s)": 0.204343
    },
    {
      "acc": 0.77563052,
      "epoch": 0.1934119774274924,
      "grad_norm": 6.25,
      "learning_rate": 9.883771201650958e-06,
      "loss": 0.80887585,
      "memory(GiB)": 117.54,
      "step": 8290,
      "train_speed(iter/s)": 0.204473
    },
    {
      "acc": 0.76742563,
      "epoch": 0.1936452849997813,
      "grad_norm": 5.71875,
      "learning_rate": 9.883365898950094e-06,
      "loss": 0.83605604,
      "memory(GiB)": 117.54,
      "step": 8300,
      "train_speed(iter/s)": 0.204603
    },
    {
      "acc": 0.77535963,
      "epoch": 0.19387859257207016,
      "grad_norm": 5.65625,
      "learning_rate": 9.882959899150166e-06,
      "loss": 0.79957132,
      "memory(GiB)": 117.54,
      "step": 8310,
      "train_speed(iter/s)": 0.204735
    },
    {
      "acc": 0.76650438,
      "epoch": 0.19411190014435906,
      "grad_norm": 8.25,
      "learning_rate": 9.882553202309131e-06,
      "loss": 0.83275509,
      "memory(GiB)": 117.54,
      "step": 8320,
      "train_speed(iter/s)": 0.204866
    },
    {
      "acc": 0.76723166,
      "epoch": 0.19434520771664796,
      "grad_norm": 6.84375,
      "learning_rate": 9.882145808485045e-06,
      "loss": 0.83784151,
      "memory(GiB)": 117.54,
      "step": 8330,
      "train_speed(iter/s)": 0.204991
    },
    {
      "acc": 0.76202507,
      "epoch": 0.19457851528893685,
      "grad_norm": 6.6875,
      "learning_rate": 9.881737717736063e-06,
      "loss": 0.8698616,
      "memory(GiB)": 117.54,
      "step": 8340,
      "train_speed(iter/s)": 0.205119
    },
    {
      "acc": 0.76839499,
      "epoch": 0.19481182286122575,
      "grad_norm": 6.71875,
      "learning_rate": 9.88132893012044e-06,
      "loss": 0.85822887,
      "memory(GiB)": 117.54,
      "step": 8350,
      "train_speed(iter/s)": 0.205249
    },
    {
      "acc": 0.75704322,
      "epoch": 0.19504513043351462,
      "grad_norm": 4.8125,
      "learning_rate": 9.88091944569653e-06,
      "loss": 0.88926487,
      "memory(GiB)": 117.54,
      "step": 8360,
      "train_speed(iter/s)": 0.20537
    },
    {
      "acc": 0.7531908,
      "epoch": 0.19527843800580352,
      "grad_norm": 6.21875,
      "learning_rate": 9.880509264522788e-06,
      "loss": 0.89586658,
      "memory(GiB)": 117.54,
      "step": 8370,
      "train_speed(iter/s)": 0.205492
    },
    {
      "acc": 0.76802006,
      "epoch": 0.19551174557809242,
      "grad_norm": 4.59375,
      "learning_rate": 9.880098386657765e-06,
      "loss": 0.83442373,
      "memory(GiB)": 117.54,
      "step": 8380,
      "train_speed(iter/s)": 0.205617
    },
    {
      "acc": 0.75828171,
      "epoch": 0.19574505315038132,
      "grad_norm": 12.1875,
      "learning_rate": 9.879686812160116e-06,
      "loss": 0.86747961,
      "memory(GiB)": 117.54,
      "step": 8390,
      "train_speed(iter/s)": 0.205742
    },
    {
      "acc": 0.77685285,
      "epoch": 0.1959783607226702,
      "grad_norm": 13.125,
      "learning_rate": 9.87927454108859e-06,
      "loss": 0.81312294,
      "memory(GiB)": 117.54,
      "step": 8400,
      "train_speed(iter/s)": 0.205862
    },
    {
      "acc": 0.75890455,
      "epoch": 0.1962116682949591,
      "grad_norm": 4.625,
      "learning_rate": 9.878861573502044e-06,
      "loss": 0.90135555,
      "memory(GiB)": 117.54,
      "step": 8410,
      "train_speed(iter/s)": 0.205988
    },
    {
      "acc": 0.77306185,
      "epoch": 0.196444975867248,
      "grad_norm": 5.4375,
      "learning_rate": 9.878447909459423e-06,
      "loss": 0.84094868,
      "memory(GiB)": 117.54,
      "step": 8420,
      "train_speed(iter/s)": 0.206109
    },
    {
      "acc": 0.77186542,
      "epoch": 0.1966782834395369,
      "grad_norm": 4.53125,
      "learning_rate": 9.878033549019781e-06,
      "loss": 0.81778107,
      "memory(GiB)": 117.54,
      "step": 8430,
      "train_speed(iter/s)": 0.206226
    },
    {
      "acc": 0.7592206,
      "epoch": 0.1969115910118258,
      "grad_norm": 5.75,
      "learning_rate": 9.877618492242267e-06,
      "loss": 0.90185413,
      "memory(GiB)": 117.54,
      "step": 8440,
      "train_speed(iter/s)": 0.206345
    },
    {
      "acc": 0.77778916,
      "epoch": 0.19714489858411466,
      "grad_norm": 5.28125,
      "learning_rate": 9.877202739186132e-06,
      "loss": 0.79591818,
      "memory(GiB)": 117.54,
      "step": 8450,
      "train_speed(iter/s)": 0.20647
    },
    {
      "acc": 0.77902613,
      "epoch": 0.19737820615640356,
      "grad_norm": 9.125,
      "learning_rate": 9.876786289910721e-06,
      "loss": 0.79726973,
      "memory(GiB)": 117.54,
      "step": 8460,
      "train_speed(iter/s)": 0.206596
    },
    {
      "acc": 0.75946159,
      "epoch": 0.19761151372869246,
      "grad_norm": 8.6875,
      "learning_rate": 9.876369144475484e-06,
      "loss": 0.8990036,
      "memory(GiB)": 117.54,
      "step": 8470,
      "train_speed(iter/s)": 0.206715
    },
    {
      "acc": 0.74261923,
      "epoch": 0.19784482130098135,
      "grad_norm": 6.125,
      "learning_rate": 9.875951302939967e-06,
      "loss": 0.91385193,
      "memory(GiB)": 117.54,
      "step": 8480,
      "train_speed(iter/s)": 0.206839
    },
    {
      "acc": 0.78411551,
      "epoch": 0.19807812887327025,
      "grad_norm": 4.53125,
      "learning_rate": 9.87553276536382e-06,
      "loss": 0.79219027,
      "memory(GiB)": 117.54,
      "step": 8490,
      "train_speed(iter/s)": 0.206957
    },
    {
      "acc": 0.7668088,
      "epoch": 0.19831143644555912,
      "grad_norm": 9.125,
      "learning_rate": 9.875113531806785e-06,
      "loss": 0.83160515,
      "memory(GiB)": 117.54,
      "step": 8500,
      "train_speed(iter/s)": 0.207077
    },
    {
      "epoch": 0.19831143644555912,
      "eval_acc": 0.731228333486271,
      "eval_loss": 0.8559815883636475,
      "eval_runtime": 1270.6586,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "step": 8500
    },
    {
      "acc": 0.75493984,
      "epoch": 0.19854474401784802,
      "grad_norm": 6.3125,
      "learning_rate": 9.874693602328711e-06,
      "loss": 0.896556,
      "memory(GiB)": 117.54,
      "step": 8510,
      "train_speed(iter/s)": 0.200895
    },
    {
      "acc": 0.78136353,
      "epoch": 0.19877805159013692,
      "grad_norm": 5.03125,
      "learning_rate": 9.874272976989541e-06,
      "loss": 0.78143473,
      "memory(GiB)": 117.54,
      "step": 8520,
      "train_speed(iter/s)": 0.201016
    },
    {
      "acc": 0.76888914,
      "epoch": 0.19901135916242582,
      "grad_norm": 6.0625,
      "learning_rate": 9.87385165584932e-06,
      "loss": 0.82431383,
      "memory(GiB)": 117.54,
      "step": 8530,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.75388536,
      "epoch": 0.19924466673471472,
      "grad_norm": 5.84375,
      "learning_rate": 9.873429638968191e-06,
      "loss": 0.89886112,
      "memory(GiB)": 117.54,
      "step": 8540,
      "train_speed(iter/s)": 0.20127
    },
    {
      "acc": 0.74465961,
      "epoch": 0.1994779743070036,
      "grad_norm": 5.5,
      "learning_rate": 9.873006926406397e-06,
      "loss": 0.93650312,
      "memory(GiB)": 117.54,
      "step": 8550,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.76810255,
      "epoch": 0.1997112818792925,
      "grad_norm": 11.0625,
      "learning_rate": 9.872583518224279e-06,
      "loss": 0.83152609,
      "memory(GiB)": 117.54,
      "step": 8560,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.75034895,
      "epoch": 0.1999445894515814,
      "grad_norm": 4.375,
      "learning_rate": 9.872159414482279e-06,
      "loss": 0.90507641,
      "memory(GiB)": 117.54,
      "step": 8570,
      "train_speed(iter/s)": 0.201644
    },
    {
      "acc": 0.76051202,
      "epoch": 0.2001778970238703,
      "grad_norm": 5.40625,
      "learning_rate": 9.871734615240938e-06,
      "loss": 0.88187637,
      "memory(GiB)": 117.54,
      "step": 8580,
      "train_speed(iter/s)": 0.201772
    },
    {
      "acc": 0.75299907,
      "epoch": 0.2004112045961592,
      "grad_norm": 6.875,
      "learning_rate": 9.871309120560897e-06,
      "loss": 0.88183479,
      "memory(GiB)": 117.54,
      "step": 8590,
      "train_speed(iter/s)": 0.201899
    },
    {
      "acc": 0.77017865,
      "epoch": 0.20064451216844806,
      "grad_norm": 7.375,
      "learning_rate": 9.870882930502894e-06,
      "loss": 0.84091225,
      "memory(GiB)": 117.54,
      "step": 8600,
      "train_speed(iter/s)": 0.202013
    },
    {
      "acc": 0.76628962,
      "epoch": 0.20087781974073696,
      "grad_norm": 6.84375,
      "learning_rate": 9.870456045127767e-06,
      "loss": 0.87175293,
      "memory(GiB)": 117.54,
      "step": 8610,
      "train_speed(iter/s)": 0.202136
    },
    {
      "acc": 0.77316351,
      "epoch": 0.20111112731302586,
      "grad_norm": 5.59375,
      "learning_rate": 9.870028464496455e-06,
      "loss": 0.83083458,
      "memory(GiB)": 117.54,
      "step": 8620,
      "train_speed(iter/s)": 0.202255
    },
    {
      "acc": 0.74215651,
      "epoch": 0.20134443488531475,
      "grad_norm": 7.46875,
      "learning_rate": 9.869600188669995e-06,
      "loss": 0.96017342,
      "memory(GiB)": 117.54,
      "step": 8630,
      "train_speed(iter/s)": 0.20237
    },
    {
      "acc": 0.76919737,
      "epoch": 0.20157774245760363,
      "grad_norm": 4.625,
      "learning_rate": 9.869171217709522e-06,
      "loss": 0.82330322,
      "memory(GiB)": 117.54,
      "step": 8640,
      "train_speed(iter/s)": 0.202492
    },
    {
      "acc": 0.76708584,
      "epoch": 0.20181105002989252,
      "grad_norm": 9.375,
      "learning_rate": 9.86874155167627e-06,
      "loss": 0.82590866,
      "memory(GiB)": 117.54,
      "step": 8650,
      "train_speed(iter/s)": 0.202616
    },
    {
      "acc": 0.7566463,
      "epoch": 0.20204435760218142,
      "grad_norm": 5.96875,
      "learning_rate": 9.868311190631578e-06,
      "loss": 0.8966464,
      "memory(GiB)": 117.54,
      "step": 8660,
      "train_speed(iter/s)": 0.202744
    },
    {
      "acc": 0.75115132,
      "epoch": 0.20227766517447032,
      "grad_norm": 4.4375,
      "learning_rate": 9.867880134636877e-06,
      "loss": 0.91379795,
      "memory(GiB)": 117.54,
      "step": 8670,
      "train_speed(iter/s)": 0.202873
    },
    {
      "acc": 0.78020339,
      "epoch": 0.20251097274675922,
      "grad_norm": 5.78125,
      "learning_rate": 9.867448383753702e-06,
      "loss": 0.78962541,
      "memory(GiB)": 117.54,
      "step": 8680,
      "train_speed(iter/s)": 0.202989
    },
    {
      "acc": 0.77505159,
      "epoch": 0.2027442803190481,
      "grad_norm": 10.125,
      "learning_rate": 9.867015938043685e-06,
      "loss": 0.82020912,
      "memory(GiB)": 117.54,
      "step": 8690,
      "train_speed(iter/s)": 0.203111
    },
    {
      "acc": 0.79113216,
      "epoch": 0.202977587891337,
      "grad_norm": 6.03125,
      "learning_rate": 9.866582797568556e-06,
      "loss": 0.7530036,
      "memory(GiB)": 117.54,
      "step": 8700,
      "train_speed(iter/s)": 0.203235
    },
    {
      "acc": 0.75312705,
      "epoch": 0.2032108954636259,
      "grad_norm": 7.125,
      "learning_rate": 9.866148962390146e-06,
      "loss": 0.90598717,
      "memory(GiB)": 117.54,
      "step": 8710,
      "train_speed(iter/s)": 0.203359
    },
    {
      "acc": 0.765062,
      "epoch": 0.2034442030359148,
      "grad_norm": 4.28125,
      "learning_rate": 9.865714432570384e-06,
      "loss": 0.86992626,
      "memory(GiB)": 117.54,
      "step": 8720,
      "train_speed(iter/s)": 0.203478
    },
    {
      "acc": 0.76311955,
      "epoch": 0.2036775106082037,
      "grad_norm": 6.25,
      "learning_rate": 9.8652792081713e-06,
      "loss": 0.85822725,
      "memory(GiB)": 117.54,
      "step": 8730,
      "train_speed(iter/s)": 0.203588
    },
    {
      "acc": 0.74687796,
      "epoch": 0.20391081818049256,
      "grad_norm": 6.71875,
      "learning_rate": 9.864843289255026e-06,
      "loss": 0.91143522,
      "memory(GiB)": 117.54,
      "step": 8740,
      "train_speed(iter/s)": 0.203712
    },
    {
      "acc": 0.75389738,
      "epoch": 0.20414412575278146,
      "grad_norm": 6.9375,
      "learning_rate": 9.864406675883784e-06,
      "loss": 0.90044785,
      "memory(GiB)": 117.54,
      "step": 8750,
      "train_speed(iter/s)": 0.203828
    },
    {
      "acc": 0.75677834,
      "epoch": 0.20437743332507036,
      "grad_norm": 7.6875,
      "learning_rate": 9.863969368119902e-06,
      "loss": 0.88022346,
      "memory(GiB)": 117.54,
      "step": 8760,
      "train_speed(iter/s)": 0.203942
    },
    {
      "acc": 0.76615176,
      "epoch": 0.20461074089735926,
      "grad_norm": 8.4375,
      "learning_rate": 9.863531366025804e-06,
      "loss": 0.8678463,
      "memory(GiB)": 117.54,
      "step": 8770,
      "train_speed(iter/s)": 0.204065
    },
    {
      "acc": 0.76859779,
      "epoch": 0.20484404846964815,
      "grad_norm": 6.6875,
      "learning_rate": 9.863092669664018e-06,
      "loss": 0.8375844,
      "memory(GiB)": 117.54,
      "step": 8780,
      "train_speed(iter/s)": 0.204184
    },
    {
      "acc": 0.76002398,
      "epoch": 0.20507735604193703,
      "grad_norm": 6.4375,
      "learning_rate": 9.862653279097166e-06,
      "loss": 0.86004581,
      "memory(GiB)": 117.54,
      "step": 8790,
      "train_speed(iter/s)": 0.204307
    },
    {
      "acc": 0.75623741,
      "epoch": 0.20531066361422592,
      "grad_norm": 10.0625,
      "learning_rate": 9.86221319438797e-06,
      "loss": 0.87964134,
      "memory(GiB)": 117.54,
      "step": 8800,
      "train_speed(iter/s)": 0.204424
    },
    {
      "acc": 0.74737983,
      "epoch": 0.20554397118651482,
      "grad_norm": 5.0625,
      "learning_rate": 9.861772415599256e-06,
      "loss": 0.93885403,
      "memory(GiB)": 117.54,
      "step": 8810,
      "train_speed(iter/s)": 0.204549
    },
    {
      "acc": 0.74907146,
      "epoch": 0.20577727875880372,
      "grad_norm": 6.125,
      "learning_rate": 9.861330942793939e-06,
      "loss": 0.90780516,
      "memory(GiB)": 117.54,
      "step": 8820,
      "train_speed(iter/s)": 0.204666
    },
    {
      "acc": 0.7763402,
      "epoch": 0.20601058633109262,
      "grad_norm": 4.96875,
      "learning_rate": 9.860888776035043e-06,
      "loss": 0.80836182,
      "memory(GiB)": 117.54,
      "step": 8830,
      "train_speed(iter/s)": 0.204762
    },
    {
      "acc": 0.74543486,
      "epoch": 0.2062438939033815,
      "grad_norm": 5.40625,
      "learning_rate": 9.860445915385687e-06,
      "loss": 0.91186695,
      "memory(GiB)": 117.54,
      "step": 8840,
      "train_speed(iter/s)": 0.204882
    },
    {
      "acc": 0.76045542,
      "epoch": 0.2064772014756704,
      "grad_norm": 5.3125,
      "learning_rate": 9.860002360909086e-06,
      "loss": 0.84907341,
      "memory(GiB)": 117.54,
      "step": 8850,
      "train_speed(iter/s)": 0.20499
    },
    {
      "acc": 0.74592304,
      "epoch": 0.2067105090479593,
      "grad_norm": 4.03125,
      "learning_rate": 9.859558112668563e-06,
      "loss": 0.89834366,
      "memory(GiB)": 117.54,
      "step": 8860,
      "train_speed(iter/s)": 0.20511
    },
    {
      "acc": 0.78678131,
      "epoch": 0.2069438166202482,
      "grad_norm": 6.28125,
      "learning_rate": 9.85911317072753e-06,
      "loss": 0.76607027,
      "memory(GiB)": 117.54,
      "step": 8870,
      "train_speed(iter/s)": 0.205225
    },
    {
      "acc": 0.76762028,
      "epoch": 0.2071771241925371,
      "grad_norm": 8.375,
      "learning_rate": 9.858667535149503e-06,
      "loss": 0.83826656,
      "memory(GiB)": 117.54,
      "step": 8880,
      "train_speed(iter/s)": 0.205346
    },
    {
      "acc": 0.78018045,
      "epoch": 0.20741043176482596,
      "grad_norm": 8.5625,
      "learning_rate": 9.858221205998097e-06,
      "loss": 0.79698944,
      "memory(GiB)": 117.54,
      "step": 8890,
      "train_speed(iter/s)": 0.205468
    },
    {
      "acc": 0.76134815,
      "epoch": 0.20764373933711486,
      "grad_norm": 7.9375,
      "learning_rate": 9.857774183337025e-06,
      "loss": 0.86685352,
      "memory(GiB)": 117.54,
      "step": 8900,
      "train_speed(iter/s)": 0.205593
    },
    {
      "acc": 0.77512984,
      "epoch": 0.20787704690940376,
      "grad_norm": 6.84375,
      "learning_rate": 9.8573264672301e-06,
      "loss": 0.8138916,
      "memory(GiB)": 117.54,
      "step": 8910,
      "train_speed(iter/s)": 0.205715
    },
    {
      "acc": 0.74891701,
      "epoch": 0.20811035448169266,
      "grad_norm": 5.4375,
      "learning_rate": 9.856878057741233e-06,
      "loss": 0.89883404,
      "memory(GiB)": 117.54,
      "step": 8920,
      "train_speed(iter/s)": 0.20584
    },
    {
      "acc": 0.76096349,
      "epoch": 0.20834366205398153,
      "grad_norm": 6.0,
      "learning_rate": 9.856428954934434e-06,
      "loss": 0.85659542,
      "memory(GiB)": 117.54,
      "step": 8930,
      "train_speed(iter/s)": 0.205957
    },
    {
      "acc": 0.76727762,
      "epoch": 0.20857696962627043,
      "grad_norm": 6.125,
      "learning_rate": 9.855979158873812e-06,
      "loss": 0.82931509,
      "memory(GiB)": 117.54,
      "step": 8940,
      "train_speed(iter/s)": 0.206077
    },
    {
      "acc": 0.76898785,
      "epoch": 0.20881027719855932,
      "grad_norm": 7.09375,
      "learning_rate": 9.855528669623576e-06,
      "loss": 0.81118364,
      "memory(GiB)": 117.54,
      "step": 8950,
      "train_speed(iter/s)": 0.206187
    },
    {
      "acc": 0.744419,
      "epoch": 0.20904358477084822,
      "grad_norm": 7.5,
      "learning_rate": 9.855077487248034e-06,
      "loss": 0.93352222,
      "memory(GiB)": 117.54,
      "step": 8960,
      "train_speed(iter/s)": 0.206303
    },
    {
      "acc": 0.7688417,
      "epoch": 0.20927689234313712,
      "grad_norm": 4.75,
      "learning_rate": 9.85462561181159e-06,
      "loss": 0.8121109,
      "memory(GiB)": 117.54,
      "step": 8970,
      "train_speed(iter/s)": 0.206412
    },
    {
      "acc": 0.7531055,
      "epoch": 0.209510199915426,
      "grad_norm": 5.96875,
      "learning_rate": 9.85417304337875e-06,
      "loss": 0.89911957,
      "memory(GiB)": 117.54,
      "step": 8980,
      "train_speed(iter/s)": 0.206523
    },
    {
      "acc": 0.74656286,
      "epoch": 0.2097435074877149,
      "grad_norm": 6.625,
      "learning_rate": 9.85371978201412e-06,
      "loss": 0.93082085,
      "memory(GiB)": 117.54,
      "step": 8990,
      "train_speed(iter/s)": 0.206644
    },
    {
      "acc": 0.7624054,
      "epoch": 0.2099768150600038,
      "grad_norm": 5.90625,
      "learning_rate": 9.8532658277824e-06,
      "loss": 0.86307392,
      "memory(GiB)": 117.54,
      "step": 9000,
      "train_speed(iter/s)": 0.206763
    },
    {
      "epoch": 0.2099768150600038,
      "eval_acc": 0.7314818418060042,
      "eval_loss": 0.8535876274108887,
      "eval_runtime": 1269.9932,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.17,
      "step": 9000
    },
    {
      "acc": 0.76916304,
      "epoch": 0.2102101226322927,
      "grad_norm": 5.125,
      "learning_rate": 9.852811180748391e-06,
      "loss": 0.83703365,
      "memory(GiB)": 117.54,
      "step": 9010,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.77324028,
      "epoch": 0.2104434302045816,
      "grad_norm": 8.75,
      "learning_rate": 9.852355840976996e-06,
      "loss": 0.8149888,
      "memory(GiB)": 117.54,
      "step": 9020,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.76231546,
      "epoch": 0.21067673777687046,
      "grad_norm": 7.28125,
      "learning_rate": 9.851899808533218e-06,
      "loss": 0.85619822,
      "memory(GiB)": 117.54,
      "step": 9030,
      "train_speed(iter/s)": 0.201166
    },
    {
      "acc": 0.76985378,
      "epoch": 0.21091004534915936,
      "grad_norm": 7.0,
      "learning_rate": 9.851443083482149e-06,
      "loss": 0.83200512,
      "memory(GiB)": 117.54,
      "step": 9040,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.7554831,
      "epoch": 0.21114335292144826,
      "grad_norm": 7.3125,
      "learning_rate": 9.850985665888988e-06,
      "loss": 0.89478712,
      "memory(GiB)": 117.54,
      "step": 9050,
      "train_speed(iter/s)": 0.201393
    },
    {
      "acc": 0.76672964,
      "epoch": 0.21137666049373716,
      "grad_norm": 5.65625,
      "learning_rate": 9.850527555819036e-06,
      "loss": 0.85093822,
      "memory(GiB)": 117.54,
      "step": 9060,
      "train_speed(iter/s)": 0.201505
    },
    {
      "acc": 0.76372004,
      "epoch": 0.21160996806602606,
      "grad_norm": 5.0625,
      "learning_rate": 9.850068753337683e-06,
      "loss": 0.86411572,
      "memory(GiB)": 117.54,
      "step": 9070,
      "train_speed(iter/s)": 0.201607
    },
    {
      "acc": 0.76573353,
      "epoch": 0.21184327563831493,
      "grad_norm": 5.25,
      "learning_rate": 9.849609258510423e-06,
      "loss": 0.83601494,
      "memory(GiB)": 117.54,
      "step": 9080,
      "train_speed(iter/s)": 0.201721
    },
    {
      "acc": 0.74441233,
      "epoch": 0.21207658321060383,
      "grad_norm": 5.4375,
      "learning_rate": 9.84914907140285e-06,
      "loss": 0.92128773,
      "memory(GiB)": 117.54,
      "step": 9090,
      "train_speed(iter/s)": 0.201835
    },
    {
      "acc": 0.76748171,
      "epoch": 0.21230989078289272,
      "grad_norm": 8.125,
      "learning_rate": 9.848688192080657e-06,
      "loss": 0.83393631,
      "memory(GiB)": 117.54,
      "step": 9100,
      "train_speed(iter/s)": 0.201944
    },
    {
      "acc": 0.7691443,
      "epoch": 0.21254319835518162,
      "grad_norm": 5.625,
      "learning_rate": 9.848226620609634e-06,
      "loss": 0.85348082,
      "memory(GiB)": 117.54,
      "step": 9110,
      "train_speed(iter/s)": 0.202057
    },
    {
      "acc": 0.75306101,
      "epoch": 0.21277650592747052,
      "grad_norm": 4.78125,
      "learning_rate": 9.847764357055669e-06,
      "loss": 0.90074921,
      "memory(GiB)": 117.54,
      "step": 9120,
      "train_speed(iter/s)": 0.202178
    },
    {
      "acc": 0.76210451,
      "epoch": 0.2130098134997594,
      "grad_norm": 5.71875,
      "learning_rate": 9.84730140148475e-06,
      "loss": 0.870508,
      "memory(GiB)": 117.54,
      "step": 9130,
      "train_speed(iter/s)": 0.202294
    },
    {
      "acc": 0.77442284,
      "epoch": 0.2132431210720483,
      "grad_norm": 6.0625,
      "learning_rate": 9.846837753962964e-06,
      "loss": 0.81348133,
      "memory(GiB)": 117.54,
      "step": 9140,
      "train_speed(iter/s)": 0.202414
    },
    {
      "acc": 0.75884104,
      "epoch": 0.2134764286443372,
      "grad_norm": 5.1875,
      "learning_rate": 9.846373414556495e-06,
      "loss": 0.85764046,
      "memory(GiB)": 117.54,
      "step": 9150,
      "train_speed(iter/s)": 0.202533
    },
    {
      "acc": 0.76104126,
      "epoch": 0.2137097362166261,
      "grad_norm": 20.75,
      "learning_rate": 9.84590838333163e-06,
      "loss": 0.86669235,
      "memory(GiB)": 117.54,
      "step": 9160,
      "train_speed(iter/s)": 0.202649
    },
    {
      "acc": 0.78756847,
      "epoch": 0.21394304378891496,
      "grad_norm": 4.8125,
      "learning_rate": 9.845442660354752e-06,
      "loss": 0.76561289,
      "memory(GiB)": 117.54,
      "step": 9170,
      "train_speed(iter/s)": 0.202771
    },
    {
      "acc": 0.76651459,
      "epoch": 0.21417635136120386,
      "grad_norm": 4.375,
      "learning_rate": 9.844976245692341e-06,
      "loss": 0.84633484,
      "memory(GiB)": 117.54,
      "step": 9180,
      "train_speed(iter/s)": 0.202893
    },
    {
      "acc": 0.75503793,
      "epoch": 0.21440965893349276,
      "grad_norm": 7.46875,
      "learning_rate": 9.84450913941098e-06,
      "loss": 0.88526068,
      "memory(GiB)": 117.54,
      "step": 9190,
      "train_speed(iter/s)": 0.202997
    },
    {
      "acc": 0.76714697,
      "epoch": 0.21464296650578166,
      "grad_norm": 6.0625,
      "learning_rate": 9.844041341577344e-06,
      "loss": 0.82248907,
      "memory(GiB)": 117.54,
      "step": 9200,
      "train_speed(iter/s)": 0.203109
    },
    {
      "acc": 0.74727135,
      "epoch": 0.21487627407807056,
      "grad_norm": 4.125,
      "learning_rate": 9.843572852258216e-06,
      "loss": 0.91963558,
      "memory(GiB)": 117.54,
      "step": 9210,
      "train_speed(iter/s)": 0.203216
    },
    {
      "acc": 0.7496871,
      "epoch": 0.21510958165035943,
      "grad_norm": 4.78125,
      "learning_rate": 9.843103671520469e-06,
      "loss": 0.92044659,
      "memory(GiB)": 117.54,
      "step": 9220,
      "train_speed(iter/s)": 0.203325
    },
    {
      "acc": 0.76813116,
      "epoch": 0.21534288922264833,
      "grad_norm": 7.75,
      "learning_rate": 9.842633799431081e-06,
      "loss": 0.83491936,
      "memory(GiB)": 117.54,
      "step": 9230,
      "train_speed(iter/s)": 0.203434
    },
    {
      "acc": 0.7559175,
      "epoch": 0.21557619679493722,
      "grad_norm": 7.75,
      "learning_rate": 9.842163236057123e-06,
      "loss": 0.86974545,
      "memory(GiB)": 117.54,
      "step": 9240,
      "train_speed(iter/s)": 0.203544
    },
    {
      "acc": 0.75992718,
      "epoch": 0.21580950436722612,
      "grad_norm": 4.46875,
      "learning_rate": 9.841691981465771e-06,
      "loss": 0.87016802,
      "memory(GiB)": 117.54,
      "step": 9250,
      "train_speed(iter/s)": 0.203653
    },
    {
      "acc": 0.74953775,
      "epoch": 0.21604281193951502,
      "grad_norm": 7.09375,
      "learning_rate": 9.841220035724295e-06,
      "loss": 0.92368717,
      "memory(GiB)": 117.54,
      "step": 9260,
      "train_speed(iter/s)": 0.203766
    },
    {
      "acc": 0.77228193,
      "epoch": 0.2162761195118039,
      "grad_norm": 4.53125,
      "learning_rate": 9.840747398900066e-06,
      "loss": 0.80582151,
      "memory(GiB)": 117.54,
      "step": 9270,
      "train_speed(iter/s)": 0.203884
    },
    {
      "acc": 0.74684772,
      "epoch": 0.2165094270840928,
      "grad_norm": 4.84375,
      "learning_rate": 9.840274071060552e-06,
      "loss": 0.92780819,
      "memory(GiB)": 117.54,
      "step": 9280,
      "train_speed(iter/s)": 0.203989
    },
    {
      "acc": 0.77230215,
      "epoch": 0.2167427346563817,
      "grad_norm": 7.28125,
      "learning_rate": 9.839800052273319e-06,
      "loss": 0.83060665,
      "memory(GiB)": 117.54,
      "step": 9290,
      "train_speed(iter/s)": 0.204102
    },
    {
      "acc": 0.7695189,
      "epoch": 0.2169760422286706,
      "grad_norm": 5.6875,
      "learning_rate": 9.839325342606034e-06,
      "loss": 0.83614769,
      "memory(GiB)": 117.54,
      "step": 9300,
      "train_speed(iter/s)": 0.204205
    },
    {
      "acc": 0.75516381,
      "epoch": 0.2172093498009595,
      "grad_norm": 4.90625,
      "learning_rate": 9.838849942126465e-06,
      "loss": 0.88942032,
      "memory(GiB)": 117.54,
      "step": 9310,
      "train_speed(iter/s)": 0.204315
    },
    {
      "acc": 0.75599709,
      "epoch": 0.21744265737324836,
      "grad_norm": 6.5625,
      "learning_rate": 9.83837385090247e-06,
      "loss": 0.90213356,
      "memory(GiB)": 117.54,
      "step": 9320,
      "train_speed(iter/s)": 0.204429
    },
    {
      "acc": 0.7904994,
      "epoch": 0.21767596494553726,
      "grad_norm": 5.40625,
      "learning_rate": 9.837897069002014e-06,
      "loss": 0.7411685,
      "memory(GiB)": 117.54,
      "step": 9330,
      "train_speed(iter/s)": 0.204543
    },
    {
      "acc": 0.76367722,
      "epoch": 0.21790927251782616,
      "grad_norm": 5.78125,
      "learning_rate": 9.837419596493158e-06,
      "loss": 0.86719389,
      "memory(GiB)": 117.54,
      "step": 9340,
      "train_speed(iter/s)": 0.204654
    },
    {
      "acc": 0.77815828,
      "epoch": 0.21814258009011506,
      "grad_norm": 4.59375,
      "learning_rate": 9.836941433444058e-06,
      "loss": 0.78728352,
      "memory(GiB)": 117.54,
      "step": 9350,
      "train_speed(iter/s)": 0.204767
    },
    {
      "acc": 0.75084362,
      "epoch": 0.21837588766240396,
      "grad_norm": 5.125,
      "learning_rate": 9.836462579922977e-06,
      "loss": 0.90360699,
      "memory(GiB)": 117.54,
      "step": 9360,
      "train_speed(iter/s)": 0.204881
    },
    {
      "acc": 0.76905317,
      "epoch": 0.21860919523469283,
      "grad_norm": 6.0,
      "learning_rate": 9.835983035998264e-06,
      "loss": 0.85155258,
      "memory(GiB)": 117.54,
      "step": 9370,
      "train_speed(iter/s)": 0.204998
    },
    {
      "acc": 0.79401808,
      "epoch": 0.21884250280698173,
      "grad_norm": 4.15625,
      "learning_rate": 9.835502801738379e-06,
      "loss": 0.7368948,
      "memory(GiB)": 117.54,
      "step": 9380,
      "train_speed(iter/s)": 0.205109
    },
    {
      "acc": 0.76240873,
      "epoch": 0.21907581037927062,
      "grad_norm": 12.5625,
      "learning_rate": 9.835021877211873e-06,
      "loss": 0.86140232,
      "memory(GiB)": 117.54,
      "step": 9390,
      "train_speed(iter/s)": 0.205222
    },
    {
      "acc": 0.77225218,
      "epoch": 0.21930911795155952,
      "grad_norm": 7.59375,
      "learning_rate": 9.834540262487399e-06,
      "loss": 0.79418178,
      "memory(GiB)": 117.54,
      "step": 9400,
      "train_speed(iter/s)": 0.205338
    },
    {
      "acc": 0.76001101,
      "epoch": 0.2195424255238484,
      "grad_norm": 5.53125,
      "learning_rate": 9.834057957633707e-06,
      "loss": 0.85294189,
      "memory(GiB)": 117.54,
      "step": 9410,
      "train_speed(iter/s)": 0.205444
    },
    {
      "acc": 0.75794129,
      "epoch": 0.2197757330961373,
      "grad_norm": 4.625,
      "learning_rate": 9.833574962719646e-06,
      "loss": 0.8874979,
      "memory(GiB)": 117.54,
      "step": 9420,
      "train_speed(iter/s)": 0.205545
    },
    {
      "acc": 0.75950623,
      "epoch": 0.2200090406684262,
      "grad_norm": 5.4375,
      "learning_rate": 9.833091277814163e-06,
      "loss": 0.86289253,
      "memory(GiB)": 117.54,
      "step": 9430,
      "train_speed(iter/s)": 0.205662
    },
    {
      "acc": 0.75737329,
      "epoch": 0.2202423482407151,
      "grad_norm": 5.4375,
      "learning_rate": 9.832606902986305e-06,
      "loss": 0.94861965,
      "memory(GiB)": 117.54,
      "step": 9440,
      "train_speed(iter/s)": 0.20577
    },
    {
      "acc": 0.78068657,
      "epoch": 0.220475655813004,
      "grad_norm": 8.25,
      "learning_rate": 9.832121838305214e-06,
      "loss": 0.7791955,
      "memory(GiB)": 117.54,
      "step": 9450,
      "train_speed(iter/s)": 0.205886
    },
    {
      "acc": 0.76481562,
      "epoch": 0.22070896338529286,
      "grad_norm": 5.5,
      "learning_rate": 9.831636083840135e-06,
      "loss": 0.84020872,
      "memory(GiB)": 117.54,
      "step": 9460,
      "train_speed(iter/s)": 0.205992
    },
    {
      "acc": 0.78759332,
      "epoch": 0.22094227095758176,
      "grad_norm": 5.8125,
      "learning_rate": 9.831149639660409e-06,
      "loss": 0.78061409,
      "memory(GiB)": 117.54,
      "step": 9470,
      "train_speed(iter/s)": 0.206099
    },
    {
      "acc": 0.76587334,
      "epoch": 0.22117557852987066,
      "grad_norm": 10.6875,
      "learning_rate": 9.830662505835476e-06,
      "loss": 0.84685173,
      "memory(GiB)": 117.54,
      "step": 9480,
      "train_speed(iter/s)": 0.206213
    },
    {
      "acc": 0.75680537,
      "epoch": 0.22140888610215956,
      "grad_norm": 6.6875,
      "learning_rate": 9.830174682434872e-06,
      "loss": 0.88343801,
      "memory(GiB)": 117.54,
      "step": 9490,
      "train_speed(iter/s)": 0.206329
    },
    {
      "acc": 0.76228046,
      "epoch": 0.22164219367444846,
      "grad_norm": 3.953125,
      "learning_rate": 9.829686169528237e-06,
      "loss": 0.87286196,
      "memory(GiB)": 117.54,
      "step": 9500,
      "train_speed(iter/s)": 0.206441
    },
    {
      "epoch": 0.22164219367444846,
      "eval_acc": 0.7321477215239133,
      "eval_loss": 0.8513048887252808,
      "eval_runtime": 1271.0526,
      "eval_samples_per_second": 28.316,
      "eval_steps_per_second": 14.158,
      "step": 9500
    },
    {
      "acc": 0.7736063,
      "epoch": 0.22187550124673733,
      "grad_norm": 4.78125,
      "learning_rate": 9.829196967185302e-06,
      "loss": 0.82306576,
      "memory(GiB)": 117.54,
      "step": 9510,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.77956977,
      "epoch": 0.22210880881902623,
      "grad_norm": 4.65625,
      "learning_rate": 9.828707075475905e-06,
      "loss": 0.78804388,
      "memory(GiB)": 117.54,
      "step": 9520,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.76058235,
      "epoch": 0.22234211639131513,
      "grad_norm": 8.6875,
      "learning_rate": 9.828216494469975e-06,
      "loss": 0.87693319,
      "memory(GiB)": 117.54,
      "step": 9530,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.77363548,
      "epoch": 0.22257542396360402,
      "grad_norm": 4.53125,
      "learning_rate": 9.827725224237542e-06,
      "loss": 0.83595715,
      "memory(GiB)": 117.54,
      "step": 9540,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.7730834,
      "epoch": 0.22280873153589292,
      "grad_norm": 8.8125,
      "learning_rate": 9.827233264848737e-06,
      "loss": 0.81986485,
      "memory(GiB)": 117.54,
      "step": 9550,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.74745684,
      "epoch": 0.2230420391081818,
      "grad_norm": 5.5625,
      "learning_rate": 9.826740616373785e-06,
      "loss": 0.89649019,
      "memory(GiB)": 117.54,
      "step": 9560,
      "train_speed(iter/s)": 0.201461
    },
    {
      "acc": 0.74992027,
      "epoch": 0.2232753466804707,
      "grad_norm": 8.9375,
      "learning_rate": 9.826247278883012e-06,
      "loss": 0.92695465,
      "memory(GiB)": 117.54,
      "step": 9570,
      "train_speed(iter/s)": 0.201565
    },
    {
      "acc": 0.78267574,
      "epoch": 0.2235086542527596,
      "grad_norm": 5.875,
      "learning_rate": 9.825753252446843e-06,
      "loss": 0.79742799,
      "memory(GiB)": 117.54,
      "step": 9580,
      "train_speed(iter/s)": 0.201669
    },
    {
      "acc": 0.78160505,
      "epoch": 0.2237419618250485,
      "grad_norm": 6.4375,
      "learning_rate": 9.825258537135798e-06,
      "loss": 0.8101265,
      "memory(GiB)": 117.54,
      "step": 9590,
      "train_speed(iter/s)": 0.201776
    },
    {
      "acc": 0.77349329,
      "epoch": 0.2239752693973374,
      "grad_norm": 4.71875,
      "learning_rate": 9.8247631330205e-06,
      "loss": 0.83796654,
      "memory(GiB)": 117.54,
      "step": 9600,
      "train_speed(iter/s)": 0.201886
    },
    {
      "acc": 0.78152819,
      "epoch": 0.22420857696962626,
      "grad_norm": 6.71875,
      "learning_rate": 9.824267040171666e-06,
      "loss": 0.80050898,
      "memory(GiB)": 117.54,
      "step": 9610,
      "train_speed(iter/s)": 0.201988
    },
    {
      "acc": 0.75929379,
      "epoch": 0.22444188454191516,
      "grad_norm": 4.90625,
      "learning_rate": 9.823770258660113e-06,
      "loss": 0.87012081,
      "memory(GiB)": 117.54,
      "step": 9620,
      "train_speed(iter/s)": 0.202098
    },
    {
      "acc": 0.78390207,
      "epoch": 0.22467519211420406,
      "grad_norm": 4.65625,
      "learning_rate": 9.823272788556757e-06,
      "loss": 0.76243534,
      "memory(GiB)": 117.54,
      "step": 9630,
      "train_speed(iter/s)": 0.202202
    },
    {
      "acc": 0.7557579,
      "epoch": 0.22490849968649296,
      "grad_norm": 7.53125,
      "learning_rate": 9.822774629932612e-06,
      "loss": 0.90164337,
      "memory(GiB)": 117.54,
      "step": 9640,
      "train_speed(iter/s)": 0.202301
    },
    {
      "acc": 0.76308393,
      "epoch": 0.22514180725878186,
      "grad_norm": 5.75,
      "learning_rate": 9.822275782858788e-06,
      "loss": 0.85689535,
      "memory(GiB)": 117.54,
      "step": 9650,
      "train_speed(iter/s)": 0.202409
    },
    {
      "acc": 0.7550004,
      "epoch": 0.22537511483107073,
      "grad_norm": 11.5625,
      "learning_rate": 9.821776247406498e-06,
      "loss": 0.8669714,
      "memory(GiB)": 117.54,
      "step": 9660,
      "train_speed(iter/s)": 0.202514
    },
    {
      "acc": 0.75697646,
      "epoch": 0.22560842240335963,
      "grad_norm": 5.03125,
      "learning_rate": 9.821276023647049e-06,
      "loss": 0.89136448,
      "memory(GiB)": 117.54,
      "step": 9670,
      "train_speed(iter/s)": 0.202619
    },
    {
      "acc": 0.77591777,
      "epoch": 0.22584172997564853,
      "grad_norm": 5.03125,
      "learning_rate": 9.820775111651849e-06,
      "loss": 0.80667782,
      "memory(GiB)": 117.54,
      "step": 9680,
      "train_speed(iter/s)": 0.202725
    },
    {
      "acc": 0.76816502,
      "epoch": 0.22607503754793742,
      "grad_norm": 5.96875,
      "learning_rate": 9.820273511492401e-06,
      "loss": 0.84152288,
      "memory(GiB)": 117.54,
      "step": 9690,
      "train_speed(iter/s)": 0.202834
    },
    {
      "acc": 0.76578364,
      "epoch": 0.2263083451202263,
      "grad_norm": 6.25,
      "learning_rate": 9.819771223240312e-06,
      "loss": 0.84982014,
      "memory(GiB)": 117.54,
      "step": 9700,
      "train_speed(iter/s)": 0.202942
    },
    {
      "acc": 0.76617751,
      "epoch": 0.2265416526925152,
      "grad_norm": 5.8125,
      "learning_rate": 9.819268246967279e-06,
      "loss": 0.82987022,
      "memory(GiB)": 117.54,
      "step": 9710,
      "train_speed(iter/s)": 0.203048
    },
    {
      "acc": 0.7701932,
      "epoch": 0.2267749602648041,
      "grad_norm": 7.21875,
      "learning_rate": 9.818764582745103e-06,
      "loss": 0.84979038,
      "memory(GiB)": 117.54,
      "step": 9720,
      "train_speed(iter/s)": 0.203157
    },
    {
      "acc": 0.77504597,
      "epoch": 0.227008267837093,
      "grad_norm": 5.25,
      "learning_rate": 9.818260230645684e-06,
      "loss": 0.81210556,
      "memory(GiB)": 117.54,
      "step": 9730,
      "train_speed(iter/s)": 0.203266
    },
    {
      "acc": 0.77041559,
      "epoch": 0.2272415754093819,
      "grad_norm": 5.28125,
      "learning_rate": 9.817755190741018e-06,
      "loss": 0.81857834,
      "memory(GiB)": 117.54,
      "step": 9740,
      "train_speed(iter/s)": 0.203372
    },
    {
      "acc": 0.76968975,
      "epoch": 0.22747488298167076,
      "grad_norm": 4.375,
      "learning_rate": 9.817249463103196e-06,
      "loss": 0.82577667,
      "memory(GiB)": 117.54,
      "step": 9750,
      "train_speed(iter/s)": 0.203475
    },
    {
      "acc": 0.79651542,
      "epoch": 0.22770819055395966,
      "grad_norm": 7.4375,
      "learning_rate": 9.816743047804413e-06,
      "loss": 0.73690844,
      "memory(GiB)": 117.54,
      "step": 9760,
      "train_speed(iter/s)": 0.203583
    },
    {
      "acc": 0.76188879,
      "epoch": 0.22794149812624856,
      "grad_norm": 4.5625,
      "learning_rate": 9.816235944916959e-06,
      "loss": 0.87093134,
      "memory(GiB)": 117.54,
      "step": 9770,
      "train_speed(iter/s)": 0.203682
    },
    {
      "acc": 0.76635122,
      "epoch": 0.22817480569853746,
      "grad_norm": 6.65625,
      "learning_rate": 9.815728154513224e-06,
      "loss": 0.84573898,
      "memory(GiB)": 117.54,
      "step": 9780,
      "train_speed(iter/s)": 0.203777
    },
    {
      "acc": 0.7656188,
      "epoch": 0.22840811327082636,
      "grad_norm": 8.5625,
      "learning_rate": 9.815219676665694e-06,
      "loss": 0.82872429,
      "memory(GiB)": 117.54,
      "step": 9790,
      "train_speed(iter/s)": 0.203881
    },
    {
      "acc": 0.75859795,
      "epoch": 0.22864142084311523,
      "grad_norm": 7.03125,
      "learning_rate": 9.814710511446954e-06,
      "loss": 0.88318329,
      "memory(GiB)": 117.54,
      "step": 9800,
      "train_speed(iter/s)": 0.203992
    },
    {
      "acc": 0.78200207,
      "epoch": 0.22887472841540413,
      "grad_norm": 5.375,
      "learning_rate": 9.814200658929686e-06,
      "loss": 0.78139582,
      "memory(GiB)": 117.54,
      "step": 9810,
      "train_speed(iter/s)": 0.204096
    },
    {
      "acc": 0.77523031,
      "epoch": 0.22910803598769303,
      "grad_norm": 6.53125,
      "learning_rate": 9.813690119186673e-06,
      "loss": 0.83075628,
      "memory(GiB)": 117.54,
      "step": 9820,
      "train_speed(iter/s)": 0.204201
    },
    {
      "acc": 0.76592264,
      "epoch": 0.22934134355998193,
      "grad_norm": 5.625,
      "learning_rate": 9.813178892290793e-06,
      "loss": 0.84281292,
      "memory(GiB)": 117.54,
      "step": 9830,
      "train_speed(iter/s)": 0.204306
    },
    {
      "acc": 0.75805759,
      "epoch": 0.22957465113227082,
      "grad_norm": 5.1875,
      "learning_rate": 9.812666978315026e-06,
      "loss": 0.88515167,
      "memory(GiB)": 117.54,
      "step": 9840,
      "train_speed(iter/s)": 0.204417
    },
    {
      "acc": 0.75032368,
      "epoch": 0.2298079587045597,
      "grad_norm": 5.03125,
      "learning_rate": 9.812154377332446e-06,
      "loss": 0.91492596,
      "memory(GiB)": 117.54,
      "step": 9850,
      "train_speed(iter/s)": 0.204526
    },
    {
      "acc": 0.77257566,
      "epoch": 0.2300412662768486,
      "grad_norm": 8.6875,
      "learning_rate": 9.811641089416225e-06,
      "loss": 0.81846771,
      "memory(GiB)": 117.54,
      "step": 9860,
      "train_speed(iter/s)": 0.20463
    },
    {
      "acc": 0.74543791,
      "epoch": 0.2302745738491375,
      "grad_norm": 7.96875,
      "learning_rate": 9.811127114639637e-06,
      "loss": 0.92871838,
      "memory(GiB)": 117.54,
      "step": 9870,
      "train_speed(iter/s)": 0.204741
    },
    {
      "acc": 0.79538689,
      "epoch": 0.2305078814214264,
      "grad_norm": 5.0625,
      "learning_rate": 9.810612453076052e-06,
      "loss": 0.72743883,
      "memory(GiB)": 117.54,
      "step": 9880,
      "train_speed(iter/s)": 0.204846
    },
    {
      "acc": 0.76809082,
      "epoch": 0.2307411889937153,
      "grad_norm": 5.75,
      "learning_rate": 9.810097104798934e-06,
      "loss": 0.84220142,
      "memory(GiB)": 117.54,
      "step": 9890,
      "train_speed(iter/s)": 0.204953
    },
    {
      "acc": 0.77071457,
      "epoch": 0.23097449656600416,
      "grad_norm": 7.28125,
      "learning_rate": 9.809581069881854e-06,
      "loss": 0.81765766,
      "memory(GiB)": 117.54,
      "step": 9900,
      "train_speed(iter/s)": 0.205051
    },
    {
      "acc": 0.76900225,
      "epoch": 0.23120780413829306,
      "grad_norm": 7.96875,
      "learning_rate": 9.809064348398474e-06,
      "loss": 0.84059877,
      "memory(GiB)": 117.54,
      "step": 9910,
      "train_speed(iter/s)": 0.205156
    },
    {
      "acc": 0.7770957,
      "epoch": 0.23144111171058196,
      "grad_norm": 7.5625,
      "learning_rate": 9.808546940422555e-06,
      "loss": 0.79744329,
      "memory(GiB)": 117.54,
      "step": 9920,
      "train_speed(iter/s)": 0.205261
    },
    {
      "acc": 0.78089809,
      "epoch": 0.23167441928287086,
      "grad_norm": 7.28125,
      "learning_rate": 9.808028846027954e-06,
      "loss": 0.79451566,
      "memory(GiB)": 117.54,
      "step": 9930,
      "train_speed(iter/s)": 0.205354
    },
    {
      "acc": 0.78215675,
      "epoch": 0.23190772685515973,
      "grad_norm": 5.09375,
      "learning_rate": 9.807510065288635e-06,
      "loss": 0.7874649,
      "memory(GiB)": 117.54,
      "step": 9940,
      "train_speed(iter/s)": 0.205452
    },
    {
      "acc": 0.7620882,
      "epoch": 0.23214103442744863,
      "grad_norm": 10.6875,
      "learning_rate": 9.806990598278651e-06,
      "loss": 0.8667222,
      "memory(GiB)": 117.54,
      "step": 9950,
      "train_speed(iter/s)": 0.205557
    },
    {
      "acc": 0.76581755,
      "epoch": 0.23237434199973753,
      "grad_norm": 6.125,
      "learning_rate": 9.806470445072156e-06,
      "loss": 0.83361053,
      "memory(GiB)": 123.09,
      "step": 9960,
      "train_speed(iter/s)": 0.205654
    },
    {
      "acc": 0.76487746,
      "epoch": 0.23260764957202643,
      "grad_norm": 5.1875,
      "learning_rate": 9.8059496057434e-06,
      "loss": 0.82635136,
      "memory(GiB)": 123.09,
      "step": 9970,
      "train_speed(iter/s)": 0.205754
    },
    {
      "acc": 0.75906668,
      "epoch": 0.23284095714431532,
      "grad_norm": 5.5,
      "learning_rate": 9.805428080366733e-06,
      "loss": 0.8668766,
      "memory(GiB)": 123.09,
      "step": 9980,
      "train_speed(iter/s)": 0.205859
    },
    {
      "acc": 0.75760293,
      "epoch": 0.2330742647166042,
      "grad_norm": 7.0,
      "learning_rate": 9.804905869016603e-06,
      "loss": 0.8750042,
      "memory(GiB)": 123.09,
      "step": 9990,
      "train_speed(iter/s)": 0.205958
    },
    {
      "acc": 0.75372882,
      "epoch": 0.2333075722888931,
      "grad_norm": 6.5625,
      "learning_rate": 9.804382971767559e-06,
      "loss": 0.92433701,
      "memory(GiB)": 123.09,
      "step": 10000,
      "train_speed(iter/s)": 0.20606
    },
    {
      "epoch": 0.2333075722888931,
      "eval_acc": 0.7324511948441245,
      "eval_loss": 0.8497293591499329,
      "eval_runtime": 1272.2316,
      "eval_samples_per_second": 28.29,
      "eval_steps_per_second": 14.145,
      "step": 10000
    },
    {
      "acc": 0.77208815,
      "epoch": 0.233540879861182,
      "grad_norm": 5.46875,
      "learning_rate": 9.803859388694238e-06,
      "loss": 0.82612524,
      "memory(GiB)": 123.09,
      "step": 10010,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.76006479,
      "epoch": 0.2337741874334709,
      "grad_norm": 7.21875,
      "learning_rate": 9.803335119871388e-06,
      "loss": 0.87721148,
      "memory(GiB)": 123.09,
      "step": 10020,
      "train_speed(iter/s)": 0.200935
    },
    {
      "acc": 0.74797354,
      "epoch": 0.2340074950057598,
      "grad_norm": 5.125,
      "learning_rate": 9.802810165373845e-06,
      "loss": 0.93788738,
      "memory(GiB)": 123.09,
      "step": 10030,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.75885124,
      "epoch": 0.23424080257804866,
      "grad_norm": 9.375,
      "learning_rate": 9.802284525276544e-06,
      "loss": 0.86786547,
      "memory(GiB)": 123.09,
      "step": 10040,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.77307854,
      "epoch": 0.23447411015033756,
      "grad_norm": 4.6875,
      "learning_rate": 9.801758199654522e-06,
      "loss": 0.81534319,
      "memory(GiB)": 123.09,
      "step": 10050,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.75475931,
      "epoch": 0.23470741772262646,
      "grad_norm": 5.71875,
      "learning_rate": 9.801231188582914e-06,
      "loss": 0.88743048,
      "memory(GiB)": 123.09,
      "step": 10060,
      "train_speed(iter/s)": 0.201359
    },
    {
      "acc": 0.7665451,
      "epoch": 0.23494072529491536,
      "grad_norm": 6.53125,
      "learning_rate": 9.800703492136948e-06,
      "loss": 0.84892178,
      "memory(GiB)": 123.09,
      "step": 10070,
      "train_speed(iter/s)": 0.201466
    },
    {
      "acc": 0.77399416,
      "epoch": 0.23517403286720426,
      "grad_norm": 4.53125,
      "learning_rate": 9.800175110391952e-06,
      "loss": 0.82023621,
      "memory(GiB)": 123.09,
      "step": 10080,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.75192294,
      "epoch": 0.23540734043949313,
      "grad_norm": 4.875,
      "learning_rate": 9.799646043423353e-06,
      "loss": 0.91079044,
      "memory(GiB)": 123.09,
      "step": 10090,
      "train_speed(iter/s)": 0.201671
    },
    {
      "acc": 0.76243572,
      "epoch": 0.23564064801178203,
      "grad_norm": 5.71875,
      "learning_rate": 9.799116291306677e-06,
      "loss": 0.85686026,
      "memory(GiB)": 123.09,
      "step": 10100,
      "train_speed(iter/s)": 0.201774
    },
    {
      "acc": 0.76248827,
      "epoch": 0.23587395558407093,
      "grad_norm": 5.3125,
      "learning_rate": 9.798585854117543e-06,
      "loss": 0.856462,
      "memory(GiB)": 123.09,
      "step": 10110,
      "train_speed(iter/s)": 0.201874
    },
    {
      "acc": 0.77080183,
      "epoch": 0.23610726315635983,
      "grad_norm": 4.4375,
      "learning_rate": 9.798054731931674e-06,
      "loss": 0.8349369,
      "memory(GiB)": 123.09,
      "step": 10120,
      "train_speed(iter/s)": 0.201974
    },
    {
      "acc": 0.76452069,
      "epoch": 0.23634057072864872,
      "grad_norm": 7.125,
      "learning_rate": 9.797522924824886e-06,
      "loss": 0.85023165,
      "memory(GiB)": 123.09,
      "step": 10130,
      "train_speed(iter/s)": 0.202084
    },
    {
      "acc": 0.77273831,
      "epoch": 0.2365738783009376,
      "grad_norm": 4.84375,
      "learning_rate": 9.796990432873093e-06,
      "loss": 0.8241353,
      "memory(GiB)": 123.09,
      "step": 10140,
      "train_speed(iter/s)": 0.202192
    },
    {
      "acc": 0.77139888,
      "epoch": 0.2368071858732265,
      "grad_norm": 7.34375,
      "learning_rate": 9.79645725615231e-06,
      "loss": 0.85567112,
      "memory(GiB)": 123.09,
      "step": 10150,
      "train_speed(iter/s)": 0.202304
    },
    {
      "acc": 0.76190987,
      "epoch": 0.2370404934455154,
      "grad_norm": 4.875,
      "learning_rate": 9.795923394738646e-06,
      "loss": 0.8605545,
      "memory(GiB)": 126.99,
      "step": 10160,
      "train_speed(iter/s)": 0.202404
    },
    {
      "acc": 0.76465502,
      "epoch": 0.2372738010178043,
      "grad_norm": 9.1875,
      "learning_rate": 9.795388848708312e-06,
      "loss": 0.86345005,
      "memory(GiB)": 126.99,
      "step": 10170,
      "train_speed(iter/s)": 0.202507
    },
    {
      "acc": 0.76446505,
      "epoch": 0.23750710859009316,
      "grad_norm": 5.5,
      "learning_rate": 9.794853618137612e-06,
      "loss": 0.86249762,
      "memory(GiB)": 126.99,
      "step": 10180,
      "train_speed(iter/s)": 0.202612
    },
    {
      "acc": 0.76790714,
      "epoch": 0.23774041616238206,
      "grad_norm": 7.15625,
      "learning_rate": 9.794317703102951e-06,
      "loss": 0.82229271,
      "memory(GiB)": 126.99,
      "step": 10190,
      "train_speed(iter/s)": 0.202713
    },
    {
      "acc": 0.76957502,
      "epoch": 0.23797372373467096,
      "grad_norm": 11.125,
      "learning_rate": 9.793781103680833e-06,
      "loss": 0.86161594,
      "memory(GiB)": 126.99,
      "step": 10200,
      "train_speed(iter/s)": 0.202818
    },
    {
      "acc": 0.7653698,
      "epoch": 0.23820703130695986,
      "grad_norm": 4.40625,
      "learning_rate": 9.793243819947851e-06,
      "loss": 0.84829473,
      "memory(GiB)": 126.99,
      "step": 10210,
      "train_speed(iter/s)": 0.202925
    },
    {
      "acc": 0.7630024,
      "epoch": 0.23844033887924876,
      "grad_norm": 6.875,
      "learning_rate": 9.79270585198071e-06,
      "loss": 0.84405813,
      "memory(GiB)": 126.99,
      "step": 10220,
      "train_speed(iter/s)": 0.203028
    },
    {
      "acc": 0.76894646,
      "epoch": 0.23867364645153763,
      "grad_norm": 6.6875,
      "learning_rate": 9.792167199856198e-06,
      "loss": 0.83470745,
      "memory(GiB)": 126.99,
      "step": 10230,
      "train_speed(iter/s)": 0.203128
    },
    {
      "acc": 0.77872562,
      "epoch": 0.23890695402382653,
      "grad_norm": 8.5625,
      "learning_rate": 9.791627863651212e-06,
      "loss": 0.78085318,
      "memory(GiB)": 126.99,
      "step": 10240,
      "train_speed(iter/s)": 0.203233
    },
    {
      "acc": 0.76503386,
      "epoch": 0.23914026159611543,
      "grad_norm": 7.28125,
      "learning_rate": 9.791087843442738e-06,
      "loss": 0.87600975,
      "memory(GiB)": 126.99,
      "step": 10250,
      "train_speed(iter/s)": 0.203335
    },
    {
      "acc": 0.7559289,
      "epoch": 0.23937356916840433,
      "grad_norm": 8.8125,
      "learning_rate": 9.790547139307869e-06,
      "loss": 0.90565586,
      "memory(GiB)": 126.99,
      "step": 10260,
      "train_speed(iter/s)": 0.203439
    },
    {
      "acc": 0.77096276,
      "epoch": 0.23960687674069323,
      "grad_norm": 6.53125,
      "learning_rate": 9.790005751323787e-06,
      "loss": 0.83454618,
      "memory(GiB)": 126.99,
      "step": 10270,
      "train_speed(iter/s)": 0.203535
    },
    {
      "acc": 0.77344503,
      "epoch": 0.2398401843129821,
      "grad_norm": 5.0625,
      "learning_rate": 9.789463679567775e-06,
      "loss": 0.82620983,
      "memory(GiB)": 126.99,
      "step": 10280,
      "train_speed(iter/s)": 0.203636
    },
    {
      "acc": 0.75782394,
      "epoch": 0.240073491885271,
      "grad_norm": 5.625,
      "learning_rate": 9.788920924117213e-06,
      "loss": 0.89050446,
      "memory(GiB)": 126.99,
      "step": 10290,
      "train_speed(iter/s)": 0.203738
    },
    {
      "acc": 0.76385994,
      "epoch": 0.2403067994575599,
      "grad_norm": 5.15625,
      "learning_rate": 9.788377485049583e-06,
      "loss": 0.85634632,
      "memory(GiB)": 126.99,
      "step": 10300,
      "train_speed(iter/s)": 0.203836
    },
    {
      "acc": 0.7727993,
      "epoch": 0.2405401070298488,
      "grad_norm": 4.875,
      "learning_rate": 9.787833362442456e-06,
      "loss": 0.81806173,
      "memory(GiB)": 126.99,
      "step": 10310,
      "train_speed(iter/s)": 0.203937
    },
    {
      "acc": 0.75147309,
      "epoch": 0.2407734146021377,
      "grad_norm": 6.15625,
      "learning_rate": 9.78728855637351e-06,
      "loss": 0.8997839,
      "memory(GiB)": 126.99,
      "step": 10320,
      "train_speed(iter/s)": 0.204036
    },
    {
      "acc": 0.77584658,
      "epoch": 0.24100672217442656,
      "grad_norm": 7.5,
      "learning_rate": 9.786743066920509e-06,
      "loss": 0.79813576,
      "memory(GiB)": 126.99,
      "step": 10330,
      "train_speed(iter/s)": 0.204138
    },
    {
      "acc": 0.7439743,
      "epoch": 0.24124002974671546,
      "grad_norm": 4.90625,
      "learning_rate": 9.786196894161329e-06,
      "loss": 0.9293148,
      "memory(GiB)": 126.99,
      "step": 10340,
      "train_speed(iter/s)": 0.20424
    },
    {
      "acc": 0.77829103,
      "epoch": 0.24147333731900436,
      "grad_norm": 6.625,
      "learning_rate": 9.78565003817393e-06,
      "loss": 0.78798056,
      "memory(GiB)": 126.99,
      "step": 10350,
      "train_speed(iter/s)": 0.204339
    },
    {
      "acc": 0.75890827,
      "epoch": 0.24170664489129326,
      "grad_norm": 5.5625,
      "learning_rate": 9.78510249903638e-06,
      "loss": 0.87458324,
      "memory(GiB)": 126.99,
      "step": 10360,
      "train_speed(iter/s)": 0.204444
    },
    {
      "acc": 0.76287951,
      "epoch": 0.24193995246358216,
      "grad_norm": 4.84375,
      "learning_rate": 9.784554276826839e-06,
      "loss": 0.84869385,
      "memory(GiB)": 126.99,
      "step": 10370,
      "train_speed(iter/s)": 0.204546
    },
    {
      "acc": 0.74085793,
      "epoch": 0.24217326003587103,
      "grad_norm": 6.625,
      "learning_rate": 9.784005371623564e-06,
      "loss": 0.93879375,
      "memory(GiB)": 126.99,
      "step": 10380,
      "train_speed(iter/s)": 0.204649
    },
    {
      "acc": 0.77869744,
      "epoch": 0.24240656760815993,
      "grad_norm": 5.15625,
      "learning_rate": 9.783455783504911e-06,
      "loss": 0.78200302,
      "memory(GiB)": 126.99,
      "step": 10390,
      "train_speed(iter/s)": 0.20475
    },
    {
      "acc": 0.7931366,
      "epoch": 0.24263987518044883,
      "grad_norm": 6.40625,
      "learning_rate": 9.782905512549336e-06,
      "loss": 0.73280854,
      "memory(GiB)": 126.99,
      "step": 10400,
      "train_speed(iter/s)": 0.204851
    },
    {
      "acc": 0.75236487,
      "epoch": 0.24287318275273773,
      "grad_norm": 4.375,
      "learning_rate": 9.78235455883539e-06,
      "loss": 0.90658245,
      "memory(GiB)": 126.99,
      "step": 10410,
      "train_speed(iter/s)": 0.204943
    },
    {
      "acc": 0.7735096,
      "epoch": 0.24310649032502663,
      "grad_norm": 8.75,
      "learning_rate": 9.781802922441716e-06,
      "loss": 0.81277504,
      "memory(GiB)": 126.99,
      "step": 10420,
      "train_speed(iter/s)": 0.205038
    },
    {
      "acc": 0.76845932,
      "epoch": 0.2433397978973155,
      "grad_norm": 5.84375,
      "learning_rate": 9.781250603447069e-06,
      "loss": 0.85649538,
      "memory(GiB)": 126.99,
      "step": 10430,
      "train_speed(iter/s)": 0.205138
    },
    {
      "acc": 0.76778679,
      "epoch": 0.2435731054696044,
      "grad_norm": 7.40625,
      "learning_rate": 9.780697601930282e-06,
      "loss": 0.84032097,
      "memory(GiB)": 126.99,
      "step": 10440,
      "train_speed(iter/s)": 0.205233
    },
    {
      "acc": 0.77375274,
      "epoch": 0.2438064130418933,
      "grad_norm": 4.96875,
      "learning_rate": 9.780143917970304e-06,
      "loss": 0.81406355,
      "memory(GiB)": 126.99,
      "step": 10450,
      "train_speed(iter/s)": 0.205333
    },
    {
      "acc": 0.76430569,
      "epoch": 0.2440397206141822,
      "grad_norm": 6.09375,
      "learning_rate": 9.77958955164617e-06,
      "loss": 0.84869604,
      "memory(GiB)": 126.99,
      "step": 10460,
      "train_speed(iter/s)": 0.205429
    },
    {
      "acc": 0.75566778,
      "epoch": 0.24427302818647106,
      "grad_norm": 6.5,
      "learning_rate": 9.779034503037016e-06,
      "loss": 0.88865891,
      "memory(GiB)": 126.99,
      "step": 10470,
      "train_speed(iter/s)": 0.205519
    },
    {
      "acc": 0.77346563,
      "epoch": 0.24450633575875996,
      "grad_norm": 6.09375,
      "learning_rate": 9.778478772222075e-06,
      "loss": 0.82391491,
      "memory(GiB)": 126.99,
      "step": 10480,
      "train_speed(iter/s)": 0.20562
    },
    {
      "acc": 0.7746438,
      "epoch": 0.24473964333104886,
      "grad_norm": 7.84375,
      "learning_rate": 9.777922359280677e-06,
      "loss": 0.80834398,
      "memory(GiB)": 126.99,
      "step": 10490,
      "train_speed(iter/s)": 0.205719
    },
    {
      "acc": 0.75339031,
      "epoch": 0.24497295090333776,
      "grad_norm": 6.875,
      "learning_rate": 9.777365264292252e-06,
      "loss": 0.88665466,
      "memory(GiB)": 126.99,
      "step": 10500,
      "train_speed(iter/s)": 0.205821
    },
    {
      "epoch": 0.24497295090333776,
      "eval_acc": 0.7332850658617177,
      "eval_loss": 0.8476359844207764,
      "eval_runtime": 1270.6663,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "step": 10500
    },
    {
      "acc": 0.77165108,
      "epoch": 0.24520625847562666,
      "grad_norm": 5.03125,
      "learning_rate": 9.77680748733632e-06,
      "loss": 0.81288023,
      "memory(GiB)": 126.99,
      "step": 10510,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.75663023,
      "epoch": 0.24543956604791553,
      "grad_norm": 4.46875,
      "learning_rate": 9.77624902849251e-06,
      "loss": 0.89590054,
      "memory(GiB)": 126.99,
      "step": 10520,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.76394377,
      "epoch": 0.24567287362020443,
      "grad_norm": 5.53125,
      "learning_rate": 9.775689887840537e-06,
      "loss": 0.86777287,
      "memory(GiB)": 126.99,
      "step": 10530,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.76071887,
      "epoch": 0.24590618119249333,
      "grad_norm": 26.5,
      "learning_rate": 9.775130065460222e-06,
      "loss": 0.88518524,
      "memory(GiB)": 126.99,
      "step": 10540,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.77137909,
      "epoch": 0.24613948876478223,
      "grad_norm": 7.40625,
      "learning_rate": 9.774569561431474e-06,
      "loss": 0.83860283,
      "memory(GiB)": 126.99,
      "step": 10550,
      "train_speed(iter/s)": 0.201256
    },
    {
      "acc": 0.7849823,
      "epoch": 0.24637279633707113,
      "grad_norm": 5.53125,
      "learning_rate": 9.77400837583431e-06,
      "loss": 0.7651825,
      "memory(GiB)": 126.99,
      "step": 10560,
      "train_speed(iter/s)": 0.201362
    },
    {
      "acc": 0.76274223,
      "epoch": 0.24660610390936,
      "grad_norm": 7.5,
      "learning_rate": 9.773446508748836e-06,
      "loss": 0.86177063,
      "memory(GiB)": 126.99,
      "step": 10570,
      "train_speed(iter/s)": 0.201458
    },
    {
      "acc": 0.75613794,
      "epoch": 0.2468394114816489,
      "grad_norm": 5.96875,
      "learning_rate": 9.772883960255261e-06,
      "loss": 0.89295292,
      "memory(GiB)": 126.99,
      "step": 10580,
      "train_speed(iter/s)": 0.201555
    },
    {
      "acc": 0.76609902,
      "epoch": 0.2470727190539378,
      "grad_norm": 6.1875,
      "learning_rate": 9.772320730433886e-06,
      "loss": 0.84143562,
      "memory(GiB)": 126.99,
      "step": 10590,
      "train_speed(iter/s)": 0.201655
    },
    {
      "acc": 0.75564928,
      "epoch": 0.2473060266262267,
      "grad_norm": 6.3125,
      "learning_rate": 9.771756819365114e-06,
      "loss": 0.88248844,
      "memory(GiB)": 126.99,
      "step": 10600,
      "train_speed(iter/s)": 0.201757
    },
    {
      "acc": 0.7873702,
      "epoch": 0.2475393341985156,
      "grad_norm": 7.78125,
      "learning_rate": 9.771192227129442e-06,
      "loss": 0.78887529,
      "memory(GiB)": 126.99,
      "step": 10610,
      "train_speed(iter/s)": 0.201848
    },
    {
      "acc": 0.78069005,
      "epoch": 0.24777264177080446,
      "grad_norm": 4.65625,
      "learning_rate": 9.770626953807468e-06,
      "loss": 0.78519044,
      "memory(GiB)": 126.99,
      "step": 10620,
      "train_speed(iter/s)": 0.201946
    },
    {
      "acc": 0.76918497,
      "epoch": 0.24800594934309336,
      "grad_norm": 6.0,
      "learning_rate": 9.770060999479878e-06,
      "loss": 0.83556156,
      "memory(GiB)": 126.99,
      "step": 10630,
      "train_speed(iter/s)": 0.202046
    },
    {
      "acc": 0.76681919,
      "epoch": 0.24823925691538226,
      "grad_norm": 4.875,
      "learning_rate": 9.769494364227468e-06,
      "loss": 0.83850193,
      "memory(GiB)": 126.99,
      "step": 10640,
      "train_speed(iter/s)": 0.202148
    },
    {
      "acc": 0.78110733,
      "epoch": 0.24847256448767116,
      "grad_norm": 4.84375,
      "learning_rate": 9.768927048131122e-06,
      "loss": 0.77184925,
      "memory(GiB)": 126.99,
      "step": 10650,
      "train_speed(iter/s)": 0.202252
    },
    {
      "acc": 0.76969419,
      "epoch": 0.24870587205996006,
      "grad_norm": 5.875,
      "learning_rate": 9.768359051271827e-06,
      "loss": 0.85380192,
      "memory(GiB)": 126.99,
      "step": 10660,
      "train_speed(iter/s)": 0.202351
    },
    {
      "acc": 0.7685781,
      "epoch": 0.24893917963224893,
      "grad_norm": 11.375,
      "learning_rate": 9.767790373730663e-06,
      "loss": 0.84533386,
      "memory(GiB)": 126.99,
      "step": 10670,
      "train_speed(iter/s)": 0.202452
    },
    {
      "acc": 0.77039485,
      "epoch": 0.24917248720453783,
      "grad_norm": 5.6875,
      "learning_rate": 9.767221015588807e-06,
      "loss": 0.84062462,
      "memory(GiB)": 126.99,
      "step": 10680,
      "train_speed(iter/s)": 0.202549
    },
    {
      "acc": 0.77639532,
      "epoch": 0.24940579477682673,
      "grad_norm": 5.84375,
      "learning_rate": 9.766650976927536e-06,
      "loss": 0.81324234,
      "memory(GiB)": 126.99,
      "step": 10690,
      "train_speed(iter/s)": 0.202647
    },
    {
      "acc": 0.77875633,
      "epoch": 0.24963910234911563,
      "grad_norm": 7.96875,
      "learning_rate": 9.766080257828223e-06,
      "loss": 0.80693998,
      "memory(GiB)": 126.99,
      "step": 10700,
      "train_speed(iter/s)": 0.202741
    },
    {
      "acc": 0.7646112,
      "epoch": 0.2498724099214045,
      "grad_norm": 6.28125,
      "learning_rate": 9.765508858372337e-06,
      "loss": 0.85661631,
      "memory(GiB)": 126.99,
      "step": 10710,
      "train_speed(iter/s)": 0.202837
    },
    {
      "acc": 0.76114054,
      "epoch": 0.2501057174936934,
      "grad_norm": 6.65625,
      "learning_rate": 9.764936778641448e-06,
      "loss": 0.88623295,
      "memory(GiB)": 126.99,
      "step": 10720,
      "train_speed(iter/s)": 0.202941
    },
    {
      "acc": 0.77698221,
      "epoch": 0.2503390250659823,
      "grad_norm": 7.75,
      "learning_rate": 9.764364018717215e-06,
      "loss": 0.79923716,
      "memory(GiB)": 126.99,
      "step": 10730,
      "train_speed(iter/s)": 0.203033
    },
    {
      "acc": 0.74807673,
      "epoch": 0.25057233263827117,
      "grad_norm": 6.4375,
      "learning_rate": 9.763790578681404e-06,
      "loss": 0.92849827,
      "memory(GiB)": 126.99,
      "step": 10740,
      "train_speed(iter/s)": 0.20313
    },
    {
      "acc": 0.76602993,
      "epoch": 0.25080564021056007,
      "grad_norm": 5.9375,
      "learning_rate": 9.763216458615871e-06,
      "loss": 0.8343996,
      "memory(GiB)": 126.99,
      "step": 10750,
      "train_speed(iter/s)": 0.203226
    },
    {
      "acc": 0.77881093,
      "epoch": 0.25103894778284896,
      "grad_norm": 5.25,
      "learning_rate": 9.762641658602575e-06,
      "loss": 0.7768116,
      "memory(GiB)": 126.99,
      "step": 10760,
      "train_speed(iter/s)": 0.203323
    },
    {
      "acc": 0.74730372,
      "epoch": 0.25127225535513786,
      "grad_norm": 5.84375,
      "learning_rate": 9.762066178723562e-06,
      "loss": 0.90061388,
      "memory(GiB)": 126.99,
      "step": 10770,
      "train_speed(iter/s)": 0.203419
    },
    {
      "acc": 0.77457342,
      "epoch": 0.25150556292742676,
      "grad_norm": 5.34375,
      "learning_rate": 9.761490019060988e-06,
      "loss": 0.80694494,
      "memory(GiB)": 126.99,
      "step": 10780,
      "train_speed(iter/s)": 0.203511
    },
    {
      "acc": 0.76837416,
      "epoch": 0.25173887049971566,
      "grad_norm": 5.34375,
      "learning_rate": 9.760913179697095e-06,
      "loss": 0.82134075,
      "memory(GiB)": 126.99,
      "step": 10790,
      "train_speed(iter/s)": 0.203603
    },
    {
      "acc": 0.75787249,
      "epoch": 0.25197217807200456,
      "grad_norm": 6.125,
      "learning_rate": 9.76033566071423e-06,
      "loss": 0.88264313,
      "memory(GiB)": 126.99,
      "step": 10800,
      "train_speed(iter/s)": 0.203694
    },
    {
      "acc": 0.75982161,
      "epoch": 0.25220548564429346,
      "grad_norm": 4.65625,
      "learning_rate": 9.759757462194832e-06,
      "loss": 0.8837141,
      "memory(GiB)": 126.99,
      "step": 10810,
      "train_speed(iter/s)": 0.203783
    },
    {
      "acc": 0.78241043,
      "epoch": 0.25243879321658236,
      "grad_norm": 6.09375,
      "learning_rate": 9.759178584221439e-06,
      "loss": 0.79735069,
      "memory(GiB)": 126.99,
      "step": 10820,
      "train_speed(iter/s)": 0.203875
    },
    {
      "acc": 0.76798344,
      "epoch": 0.2526721007888712,
      "grad_norm": 7.125,
      "learning_rate": 9.758599026876685e-06,
      "loss": 0.83385935,
      "memory(GiB)": 126.99,
      "step": 10830,
      "train_speed(iter/s)": 0.203966
    },
    {
      "acc": 0.77368031,
      "epoch": 0.2529054083611601,
      "grad_norm": 7.65625,
      "learning_rate": 9.758018790243304e-06,
      "loss": 0.8143096,
      "memory(GiB)": 126.99,
      "step": 10840,
      "train_speed(iter/s)": 0.204058
    },
    {
      "acc": 0.75938997,
      "epoch": 0.253138715933449,
      "grad_norm": 5.59375,
      "learning_rate": 9.757437874404121e-06,
      "loss": 0.85309505,
      "memory(GiB)": 126.99,
      "step": 10850,
      "train_speed(iter/s)": 0.204149
    },
    {
      "acc": 0.78056459,
      "epoch": 0.2533720235057379,
      "grad_norm": 5.625,
      "learning_rate": 9.756856279442064e-06,
      "loss": 0.77668962,
      "memory(GiB)": 126.99,
      "step": 10860,
      "train_speed(iter/s)": 0.204244
    },
    {
      "acc": 0.74807405,
      "epoch": 0.2536053310780268,
      "grad_norm": 6.0625,
      "learning_rate": 9.756274005440156e-06,
      "loss": 0.92297764,
      "memory(GiB)": 126.99,
      "step": 10870,
      "train_speed(iter/s)": 0.204338
    },
    {
      "acc": 0.76769743,
      "epoch": 0.2538386386503157,
      "grad_norm": 4.625,
      "learning_rate": 9.755691052481515e-06,
      "loss": 0.81955452,
      "memory(GiB)": 126.99,
      "step": 10880,
      "train_speed(iter/s)": 0.204435
    },
    {
      "acc": 0.7821074,
      "epoch": 0.2540719462226046,
      "grad_norm": 5.78125,
      "learning_rate": 9.755107420649357e-06,
      "loss": 0.77177048,
      "memory(GiB)": 126.99,
      "step": 10890,
      "train_speed(iter/s)": 0.204532
    },
    {
      "acc": 0.76513309,
      "epoch": 0.2543052537948935,
      "grad_norm": 5.1875,
      "learning_rate": 9.754523110026997e-06,
      "loss": 0.84079628,
      "memory(GiB)": 126.99,
      "step": 10900,
      "train_speed(iter/s)": 0.204625
    },
    {
      "acc": 0.77776675,
      "epoch": 0.2545385613671824,
      "grad_norm": 7.125,
      "learning_rate": 9.753938120697843e-06,
      "loss": 0.79896116,
      "memory(GiB)": 126.99,
      "step": 10910,
      "train_speed(iter/s)": 0.204718
    },
    {
      "acc": 0.77102752,
      "epoch": 0.2547718689394713,
      "grad_norm": 5.34375,
      "learning_rate": 9.753352452745406e-06,
      "loss": 0.8230175,
      "memory(GiB)": 126.99,
      "step": 10920,
      "train_speed(iter/s)": 0.204808
    },
    {
      "acc": 0.7826086,
      "epoch": 0.25500517651176013,
      "grad_norm": 8.3125,
      "learning_rate": 9.752766106253285e-06,
      "loss": 0.77752132,
      "memory(GiB)": 126.99,
      "step": 10930,
      "train_speed(iter/s)": 0.204896
    },
    {
      "acc": 0.76253233,
      "epoch": 0.25523848408404903,
      "grad_norm": 4.75,
      "learning_rate": 9.752179081305184e-06,
      "loss": 0.85609999,
      "memory(GiB)": 126.99,
      "step": 10940,
      "train_speed(iter/s)": 0.204985
    },
    {
      "acc": 0.78686914,
      "epoch": 0.25547179165633793,
      "grad_norm": 5.65625,
      "learning_rate": 9.751591377984899e-06,
      "loss": 0.78415165,
      "memory(GiB)": 126.99,
      "step": 10950,
      "train_speed(iter/s)": 0.205081
    },
    {
      "acc": 0.76985493,
      "epoch": 0.25570509922862683,
      "grad_norm": 13.25,
      "learning_rate": 9.751002996376324e-06,
      "loss": 0.87296238,
      "memory(GiB)": 126.99,
      "step": 10960,
      "train_speed(iter/s)": 0.205177
    },
    {
      "acc": 0.77404528,
      "epoch": 0.25593840680091573,
      "grad_norm": 4.375,
      "learning_rate": 9.750413936563454e-06,
      "loss": 0.80378752,
      "memory(GiB)": 126.99,
      "step": 10970,
      "train_speed(iter/s)": 0.20527
    },
    {
      "acc": 0.76339779,
      "epoch": 0.25617171437320463,
      "grad_norm": 6.125,
      "learning_rate": 9.749824198630371e-06,
      "loss": 0.87236614,
      "memory(GiB)": 126.99,
      "step": 10980,
      "train_speed(iter/s)": 0.205367
    },
    {
      "acc": 0.76326017,
      "epoch": 0.2564050219454935,
      "grad_norm": 4.78125,
      "learning_rate": 9.749233782661267e-06,
      "loss": 0.84402294,
      "memory(GiB)": 126.99,
      "step": 10990,
      "train_speed(iter/s)": 0.20546
    },
    {
      "acc": 0.78211966,
      "epoch": 0.2566383295177824,
      "grad_norm": 4.4375,
      "learning_rate": 9.74864268874042e-06,
      "loss": 0.79133849,
      "memory(GiB)": 126.99,
      "step": 11000,
      "train_speed(iter/s)": 0.20555
    },
    {
      "epoch": 0.2566383295177824,
      "eval_acc": 0.73350094028686,
      "eval_loss": 0.8474711775779724,
      "eval_runtime": 1269.9892,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.17,
      "step": 11000
    },
    {
      "acc": 0.76290751,
      "epoch": 0.2568716370900713,
      "grad_norm": 4.28125,
      "learning_rate": 9.748050916952206e-06,
      "loss": 0.85341339,
      "memory(GiB)": 126.99,
      "step": 11010,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.76456928,
      "epoch": 0.2571049446623602,
      "grad_norm": 4.84375,
      "learning_rate": 9.747458467381104e-06,
      "loss": 0.83613319,
      "memory(GiB)": 126.99,
      "step": 11020,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.77785635,
      "epoch": 0.25733825223464907,
      "grad_norm": 6.875,
      "learning_rate": 9.746865340111686e-06,
      "loss": 0.78845282,
      "memory(GiB)": 126.99,
      "step": 11030,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.77325573,
      "epoch": 0.25757155980693797,
      "grad_norm": 4.6875,
      "learning_rate": 9.74627153522862e-06,
      "loss": 0.82584515,
      "memory(GiB)": 126.99,
      "step": 11040,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.75915756,
      "epoch": 0.25780486737922687,
      "grad_norm": 7.75,
      "learning_rate": 9.74567705281667e-06,
      "loss": 0.86543064,
      "memory(GiB)": 126.99,
      "step": 11050,
      "train_speed(iter/s)": 0.201208
    },
    {
      "acc": 0.75994439,
      "epoch": 0.25803817495151576,
      "grad_norm": 5.46875,
      "learning_rate": 9.745081892960699e-06,
      "loss": 0.87644157,
      "memory(GiB)": 126.99,
      "step": 11060,
      "train_speed(iter/s)": 0.201305
    },
    {
      "acc": 0.75638137,
      "epoch": 0.25827148252380466,
      "grad_norm": 6.21875,
      "learning_rate": 9.744486055745667e-06,
      "loss": 0.90656376,
      "memory(GiB)": 126.99,
      "step": 11070,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.76791043,
      "epoch": 0.25850479009609356,
      "grad_norm": 4.8125,
      "learning_rate": 9.743889541256628e-06,
      "loss": 0.85354557,
      "memory(GiB)": 126.99,
      "step": 11080,
      "train_speed(iter/s)": 0.201493
    },
    {
      "acc": 0.78517094,
      "epoch": 0.25873809766838246,
      "grad_norm": 10.5,
      "learning_rate": 9.743292349578737e-06,
      "loss": 0.79755116,
      "memory(GiB)": 126.99,
      "step": 11090,
      "train_speed(iter/s)": 0.201588
    },
    {
      "acc": 0.77367783,
      "epoch": 0.25897140524067136,
      "grad_norm": 6.375,
      "learning_rate": 9.742694480797239e-06,
      "loss": 0.8688302,
      "memory(GiB)": 126.99,
      "step": 11100,
      "train_speed(iter/s)": 0.201679
    },
    {
      "acc": 0.76761403,
      "epoch": 0.25920471281296026,
      "grad_norm": 5.71875,
      "learning_rate": 9.742095934997482e-06,
      "loss": 0.84129333,
      "memory(GiB)": 126.99,
      "step": 11110,
      "train_speed(iter/s)": 0.201772
    },
    {
      "acc": 0.7736413,
      "epoch": 0.2594380203852491,
      "grad_norm": 5.28125,
      "learning_rate": 9.741496712264908e-06,
      "loss": 0.82993221,
      "memory(GiB)": 126.99,
      "step": 11120,
      "train_speed(iter/s)": 0.20186
    },
    {
      "acc": 0.80244122,
      "epoch": 0.259671327957538,
      "grad_norm": 9.8125,
      "learning_rate": 9.740896812685057e-06,
      "loss": 0.70631914,
      "memory(GiB)": 126.99,
      "step": 11130,
      "train_speed(iter/s)": 0.201957
    },
    {
      "acc": 0.78163109,
      "epoch": 0.2599046355298269,
      "grad_norm": 5.21875,
      "learning_rate": 9.740296236343561e-06,
      "loss": 0.7652422,
      "memory(GiB)": 126.99,
      "step": 11140,
      "train_speed(iter/s)": 0.202052
    },
    {
      "acc": 0.79088631,
      "epoch": 0.2601379431021158,
      "grad_norm": 6.8125,
      "learning_rate": 9.739694983326155e-06,
      "loss": 0.74674163,
      "memory(GiB)": 126.99,
      "step": 11150,
      "train_speed(iter/s)": 0.20215
    },
    {
      "acc": 0.75785141,
      "epoch": 0.2603712506744047,
      "grad_norm": 10.1875,
      "learning_rate": 9.739093053718669e-06,
      "loss": 0.86176929,
      "memory(GiB)": 126.99,
      "step": 11160,
      "train_speed(iter/s)": 0.20224
    },
    {
      "acc": 0.77965183,
      "epoch": 0.2606045582466936,
      "grad_norm": 5.71875,
      "learning_rate": 9.738490447607025e-06,
      "loss": 0.78320961,
      "memory(GiB)": 126.99,
      "step": 11170,
      "train_speed(iter/s)": 0.202336
    },
    {
      "acc": 0.77159529,
      "epoch": 0.2608378658189825,
      "grad_norm": 10.3125,
      "learning_rate": 9.737887165077246e-06,
      "loss": 0.82931585,
      "memory(GiB)": 126.99,
      "step": 11180,
      "train_speed(iter/s)": 0.202431
    },
    {
      "acc": 0.76718621,
      "epoch": 0.2610711733912714,
      "grad_norm": 5.09375,
      "learning_rate": 9.73728320621545e-06,
      "loss": 0.8547121,
      "memory(GiB)": 126.99,
      "step": 11190,
      "train_speed(iter/s)": 0.202523
    },
    {
      "acc": 0.76389389,
      "epoch": 0.2613044809635603,
      "grad_norm": 5.09375,
      "learning_rate": 9.736678571107854e-06,
      "loss": 0.86680279,
      "memory(GiB)": 126.99,
      "step": 11200,
      "train_speed(iter/s)": 0.202609
    },
    {
      "acc": 0.77421455,
      "epoch": 0.2615377885358492,
      "grad_norm": 5.28125,
      "learning_rate": 9.736073259840766e-06,
      "loss": 0.80106392,
      "memory(GiB)": 126.99,
      "step": 11210,
      "train_speed(iter/s)": 0.202707
    },
    {
      "acc": 0.74991431,
      "epoch": 0.26177109610813803,
      "grad_norm": 6.28125,
      "learning_rate": 9.735467272500597e-06,
      "loss": 0.92689228,
      "memory(GiB)": 126.99,
      "step": 11220,
      "train_speed(iter/s)": 0.202804
    },
    {
      "acc": 0.78649111,
      "epoch": 0.26200440368042693,
      "grad_norm": 10.5,
      "learning_rate": 9.73486060917385e-06,
      "loss": 0.78013344,
      "memory(GiB)": 126.99,
      "step": 11230,
      "train_speed(iter/s)": 0.202899
    },
    {
      "acc": 0.76939349,
      "epoch": 0.26223771125271583,
      "grad_norm": 9.375,
      "learning_rate": 9.734253269947128e-06,
      "loss": 0.84294491,
      "memory(GiB)": 126.99,
      "step": 11240,
      "train_speed(iter/s)": 0.20299
    },
    {
      "acc": 0.74981251,
      "epoch": 0.26247101882500473,
      "grad_norm": 4.0625,
      "learning_rate": 9.733645254907126e-06,
      "loss": 0.91040516,
      "memory(GiB)": 126.99,
      "step": 11250,
      "train_speed(iter/s)": 0.203082
    },
    {
      "acc": 0.77205968,
      "epoch": 0.26270432639729363,
      "grad_norm": 5.96875,
      "learning_rate": 9.73303656414064e-06,
      "loss": 0.80362463,
      "memory(GiB)": 126.99,
      "step": 11260,
      "train_speed(iter/s)": 0.203173
    },
    {
      "acc": 0.78065724,
      "epoch": 0.26293763396958253,
      "grad_norm": 10.25,
      "learning_rate": 9.732427197734557e-06,
      "loss": 0.77598629,
      "memory(GiB)": 126.99,
      "step": 11270,
      "train_speed(iter/s)": 0.203266
    },
    {
      "acc": 0.77076845,
      "epoch": 0.26317094154187143,
      "grad_norm": 5.84375,
      "learning_rate": 9.73181715577587e-06,
      "loss": 0.82357779,
      "memory(GiB)": 126.99,
      "step": 11280,
      "train_speed(iter/s)": 0.203349
    },
    {
      "acc": 0.78687391,
      "epoch": 0.2634042491141603,
      "grad_norm": 5.0625,
      "learning_rate": 9.731206438351655e-06,
      "loss": 0.77638292,
      "memory(GiB)": 126.99,
      "step": 11290,
      "train_speed(iter/s)": 0.203439
    },
    {
      "acc": 0.77242026,
      "epoch": 0.2636375566864492,
      "grad_norm": 6.4375,
      "learning_rate": 9.730595045549096e-06,
      "loss": 0.80217266,
      "memory(GiB)": 126.99,
      "step": 11300,
      "train_speed(iter/s)": 0.203536
    },
    {
      "acc": 0.76687241,
      "epoch": 0.2638708642587381,
      "grad_norm": 4.375,
      "learning_rate": 9.72998297745547e-06,
      "loss": 0.84325809,
      "memory(GiB)": 126.99,
      "step": 11310,
      "train_speed(iter/s)": 0.203631
    },
    {
      "acc": 0.74472752,
      "epoch": 0.26410417183102697,
      "grad_norm": 5.0625,
      "learning_rate": 9.729370234158147e-06,
      "loss": 0.92086382,
      "memory(GiB)": 126.99,
      "step": 11320,
      "train_speed(iter/s)": 0.203723
    },
    {
      "acc": 0.7770112,
      "epoch": 0.26433747940331587,
      "grad_norm": 6.25,
      "learning_rate": 9.728756815744598e-06,
      "loss": 0.86181021,
      "memory(GiB)": 126.99,
      "step": 11330,
      "train_speed(iter/s)": 0.203818
    },
    {
      "acc": 0.75057278,
      "epoch": 0.26457078697560477,
      "grad_norm": 6.1875,
      "learning_rate": 9.728142722302385e-06,
      "loss": 0.91030331,
      "memory(GiB)": 126.99,
      "step": 11340,
      "train_speed(iter/s)": 0.20391
    },
    {
      "acc": 0.76302509,
      "epoch": 0.26480409454789366,
      "grad_norm": 11.0625,
      "learning_rate": 9.727527953919174e-06,
      "loss": 0.87596025,
      "memory(GiB)": 126.99,
      "step": 11350,
      "train_speed(iter/s)": 0.203997
    },
    {
      "acc": 0.75909495,
      "epoch": 0.26503740212018256,
      "grad_norm": 5.53125,
      "learning_rate": 9.72691251068272e-06,
      "loss": 0.86929102,
      "memory(GiB)": 126.99,
      "step": 11360,
      "train_speed(iter/s)": 0.204091
    },
    {
      "acc": 0.76749086,
      "epoch": 0.26527070969247146,
      "grad_norm": 5.53125,
      "learning_rate": 9.726296392680879e-06,
      "loss": 0.83195705,
      "memory(GiB)": 126.99,
      "step": 11370,
      "train_speed(iter/s)": 0.204179
    },
    {
      "acc": 0.76265335,
      "epoch": 0.26550401726476036,
      "grad_norm": 20.5,
      "learning_rate": 9.7256796000016e-06,
      "loss": 0.84113808,
      "memory(GiB)": 126.99,
      "step": 11380,
      "train_speed(iter/s)": 0.204266
    },
    {
      "acc": 0.7723104,
      "epoch": 0.26573732483704926,
      "grad_norm": 6.53125,
      "learning_rate": 9.725062132732931e-06,
      "loss": 0.82607574,
      "memory(GiB)": 126.99,
      "step": 11390,
      "train_speed(iter/s)": 0.204356
    },
    {
      "acc": 0.76926665,
      "epoch": 0.26597063240933816,
      "grad_norm": 9.25,
      "learning_rate": 9.724443990963017e-06,
      "loss": 0.84509716,
      "memory(GiB)": 126.99,
      "step": 11400,
      "train_speed(iter/s)": 0.204446
    },
    {
      "acc": 0.77471218,
      "epoch": 0.266203939981627,
      "grad_norm": 5.96875,
      "learning_rate": 9.723825174780095e-06,
      "loss": 0.79775171,
      "memory(GiB)": 126.99,
      "step": 11410,
      "train_speed(iter/s)": 0.204537
    },
    {
      "acc": 0.76716242,
      "epoch": 0.2664372475539159,
      "grad_norm": 6.125,
      "learning_rate": 9.723205684272501e-06,
      "loss": 0.82582216,
      "memory(GiB)": 126.99,
      "step": 11420,
      "train_speed(iter/s)": 0.20463
    },
    {
      "acc": 0.77979913,
      "epoch": 0.2666705551262048,
      "grad_norm": 4.59375,
      "learning_rate": 9.722585519528666e-06,
      "loss": 0.79641523,
      "memory(GiB)": 126.99,
      "step": 11430,
      "train_speed(iter/s)": 0.204722
    },
    {
      "acc": 0.77594395,
      "epoch": 0.2669038626984937,
      "grad_norm": 7.5625,
      "learning_rate": 9.721964680637124e-06,
      "loss": 0.81631298,
      "memory(GiB)": 126.99,
      "step": 11440,
      "train_speed(iter/s)": 0.204809
    },
    {
      "acc": 0.77567191,
      "epoch": 0.2671371702707826,
      "grad_norm": 6.46875,
      "learning_rate": 9.721343167686491e-06,
      "loss": 0.80270195,
      "memory(GiB)": 126.99,
      "step": 11450,
      "train_speed(iter/s)": 0.2049
    },
    {
      "acc": 0.76392403,
      "epoch": 0.2673704778430715,
      "grad_norm": 9.3125,
      "learning_rate": 9.720720980765495e-06,
      "loss": 0.85869751,
      "memory(GiB)": 126.99,
      "step": 11460,
      "train_speed(iter/s)": 0.204987
    },
    {
      "acc": 0.76475925,
      "epoch": 0.2676037854153604,
      "grad_norm": 4.75,
      "learning_rate": 9.72009811996295e-06,
      "loss": 0.85545101,
      "memory(GiB)": 126.99,
      "step": 11470,
      "train_speed(iter/s)": 0.205075
    },
    {
      "acc": 0.77414856,
      "epoch": 0.2678370929876493,
      "grad_norm": 6.125,
      "learning_rate": 9.719474585367771e-06,
      "loss": 0.80691338,
      "memory(GiB)": 126.99,
      "step": 11480,
      "train_speed(iter/s)": 0.205167
    },
    {
      "acc": 0.7491497,
      "epoch": 0.2680704005599382,
      "grad_norm": 8.875,
      "learning_rate": 9.718850377068964e-06,
      "loss": 0.95009527,
      "memory(GiB)": 126.99,
      "step": 11490,
      "train_speed(iter/s)": 0.205256
    },
    {
      "acc": 0.77273641,
      "epoch": 0.2683037081322271,
      "grad_norm": 6.34375,
      "learning_rate": 9.718225495155638e-06,
      "loss": 0.82936974,
      "memory(GiB)": 126.99,
      "step": 11500,
      "train_speed(iter/s)": 0.205341
    },
    {
      "epoch": 0.2683037081322271,
      "eval_acc": 0.7336386643266392,
      "eval_loss": 0.8452193737030029,
      "eval_runtime": 1270.1304,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.169,
      "step": 11500
    },
    {
      "acc": 0.7597836,
      "epoch": 0.26853701570451594,
      "grad_norm": 6.1875,
      "learning_rate": 9.717599939716992e-06,
      "loss": 0.84666138,
      "memory(GiB)": 126.99,
      "step": 11510,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.75912447,
      "epoch": 0.26877032327680483,
      "grad_norm": 5.90625,
      "learning_rate": 9.716973710842326e-06,
      "loss": 0.86441269,
      "memory(GiB)": 126.99,
      "step": 11520,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.77214475,
      "epoch": 0.26900363084909373,
      "grad_norm": 6.0,
      "learning_rate": 9.716346808621031e-06,
      "loss": 0.83422747,
      "memory(GiB)": 126.99,
      "step": 11530,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.78334231,
      "epoch": 0.26923693842138263,
      "grad_norm": 5.6875,
      "learning_rate": 9.715719233142601e-06,
      "loss": 0.77969465,
      "memory(GiB)": 126.99,
      "step": 11540,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.77558217,
      "epoch": 0.26947024599367153,
      "grad_norm": 5.46875,
      "learning_rate": 9.71509098449662e-06,
      "loss": 0.82058563,
      "memory(GiB)": 126.99,
      "step": 11550,
      "train_speed(iter/s)": 0.201179
    },
    {
      "acc": 0.75732489,
      "epoch": 0.26970355356596043,
      "grad_norm": 4.5625,
      "learning_rate": 9.71446206277277e-06,
      "loss": 0.88411083,
      "memory(GiB)": 126.99,
      "step": 11560,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.75806885,
      "epoch": 0.26993686113824933,
      "grad_norm": 5.09375,
      "learning_rate": 9.713832468060831e-06,
      "loss": 0.86597509,
      "memory(GiB)": 126.99,
      "step": 11570,
      "train_speed(iter/s)": 0.20135
    },
    {
      "acc": 0.74850397,
      "epoch": 0.2701701687105382,
      "grad_norm": 5.5,
      "learning_rate": 9.713202200450678e-06,
      "loss": 0.9307641,
      "memory(GiB)": 126.99,
      "step": 11580,
      "train_speed(iter/s)": 0.201441
    },
    {
      "acc": 0.76493893,
      "epoch": 0.2704034762828271,
      "grad_norm": 5.75,
      "learning_rate": 9.712571260032277e-06,
      "loss": 0.8510767,
      "memory(GiB)": 126.99,
      "step": 11590,
      "train_speed(iter/s)": 0.201524
    },
    {
      "acc": 0.76068068,
      "epoch": 0.27063678385511597,
      "grad_norm": 5.9375,
      "learning_rate": 9.7119396468957e-06,
      "loss": 0.87124729,
      "memory(GiB)": 126.99,
      "step": 11600,
      "train_speed(iter/s)": 0.201618
    },
    {
      "acc": 0.76412511,
      "epoch": 0.27087009142740487,
      "grad_norm": 4.59375,
      "learning_rate": 9.711307361131107e-06,
      "loss": 0.84404926,
      "memory(GiB)": 126.99,
      "step": 11610,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.76919479,
      "epoch": 0.27110339899969377,
      "grad_norm": 5.375,
      "learning_rate": 9.710674402828755e-06,
      "loss": 0.83093939,
      "memory(GiB)": 126.99,
      "step": 11620,
      "train_speed(iter/s)": 0.201785
    },
    {
      "acc": 0.7725348,
      "epoch": 0.27133670657198267,
      "grad_norm": 4.28125,
      "learning_rate": 9.710040772079001e-06,
      "loss": 0.81993046,
      "memory(GiB)": 126.99,
      "step": 11630,
      "train_speed(iter/s)": 0.201875
    },
    {
      "acc": 0.75861168,
      "epoch": 0.27157001414427157,
      "grad_norm": 6.375,
      "learning_rate": 9.709406468972295e-06,
      "loss": 0.87555952,
      "memory(GiB)": 126.99,
      "step": 11640,
      "train_speed(iter/s)": 0.201964
    },
    {
      "acc": 0.76188645,
      "epoch": 0.27180332171656046,
      "grad_norm": 6.53125,
      "learning_rate": 9.708771493599185e-06,
      "loss": 0.83453884,
      "memory(GiB)": 126.99,
      "step": 11650,
      "train_speed(iter/s)": 0.202059
    },
    {
      "acc": 0.76258478,
      "epoch": 0.27203662928884936,
      "grad_norm": 4.78125,
      "learning_rate": 9.708135846050313e-06,
      "loss": 0.85640106,
      "memory(GiB)": 126.99,
      "step": 11660,
      "train_speed(iter/s)": 0.20215
    },
    {
      "acc": 0.78319798,
      "epoch": 0.27226993686113826,
      "grad_norm": 4.40625,
      "learning_rate": 9.707499526416415e-06,
      "loss": 0.78641605,
      "memory(GiB)": 126.99,
      "step": 11670,
      "train_speed(iter/s)": 0.202232
    },
    {
      "acc": 0.75823736,
      "epoch": 0.27250324443342716,
      "grad_norm": 5.9375,
      "learning_rate": 9.706862534788327e-06,
      "loss": 0.86921701,
      "memory(GiB)": 126.99,
      "step": 11680,
      "train_speed(iter/s)": 0.202322
    },
    {
      "acc": 0.77783613,
      "epoch": 0.27273655200571606,
      "grad_norm": 8.8125,
      "learning_rate": 9.70622487125698e-06,
      "loss": 0.78761058,
      "memory(GiB)": 126.99,
      "step": 11690,
      "train_speed(iter/s)": 0.202412
    },
    {
      "acc": 0.77462397,
      "epoch": 0.2729698595780049,
      "grad_norm": 6.625,
      "learning_rate": 9.7055865359134e-06,
      "loss": 0.81614437,
      "memory(GiB)": 126.99,
      "step": 11700,
      "train_speed(iter/s)": 0.202495
    },
    {
      "acc": 0.78386021,
      "epoch": 0.2732031671502938,
      "grad_norm": 7.15625,
      "learning_rate": 9.704947528848706e-06,
      "loss": 0.7758873,
      "memory(GiB)": 126.99,
      "step": 11710,
      "train_speed(iter/s)": 0.202588
    },
    {
      "acc": 0.77271953,
      "epoch": 0.2734364747225827,
      "grad_norm": 4.46875,
      "learning_rate": 9.704307850154125e-06,
      "loss": 0.80577879,
      "memory(GiB)": 126.99,
      "step": 11720,
      "train_speed(iter/s)": 0.202677
    },
    {
      "acc": 0.76300001,
      "epoch": 0.2736697822948716,
      "grad_norm": 5.0,
      "learning_rate": 9.70366749992096e-06,
      "loss": 0.85106087,
      "memory(GiB)": 126.99,
      "step": 11730,
      "train_speed(iter/s)": 0.202765
    },
    {
      "acc": 0.76563978,
      "epoch": 0.2739030898671605,
      "grad_norm": 4.0625,
      "learning_rate": 9.703026478240627e-06,
      "loss": 0.84070683,
      "memory(GiB)": 126.99,
      "step": 11740,
      "train_speed(iter/s)": 0.202854
    },
    {
      "acc": 0.79070711,
      "epoch": 0.2741363974394494,
      "grad_norm": 5.09375,
      "learning_rate": 9.702384785204631e-06,
      "loss": 0.75290918,
      "memory(GiB)": 126.99,
      "step": 11750,
      "train_speed(iter/s)": 0.202945
    },
    {
      "acc": 0.75680985,
      "epoch": 0.2743697050117383,
      "grad_norm": 5.125,
      "learning_rate": 9.701742420904574e-06,
      "loss": 0.89517021,
      "memory(GiB)": 126.99,
      "step": 11760,
      "train_speed(iter/s)": 0.203037
    },
    {
      "acc": 0.74839983,
      "epoch": 0.2746030125840272,
      "grad_norm": 6.59375,
      "learning_rate": 9.701099385432151e-06,
      "loss": 0.92097282,
      "memory(GiB)": 126.99,
      "step": 11770,
      "train_speed(iter/s)": 0.203121
    },
    {
      "acc": 0.75796165,
      "epoch": 0.2748363201563161,
      "grad_norm": 4.0,
      "learning_rate": 9.700455678879157e-06,
      "loss": 0.8720623,
      "memory(GiB)": 126.99,
      "step": 11780,
      "train_speed(iter/s)": 0.203206
    },
    {
      "acc": 0.76428533,
      "epoch": 0.275069627728605,
      "grad_norm": 11.4375,
      "learning_rate": 9.69981130133748e-06,
      "loss": 0.8485014,
      "memory(GiB)": 126.99,
      "step": 11790,
      "train_speed(iter/s)": 0.203289
    },
    {
      "acc": 0.75732021,
      "epoch": 0.27530293530089384,
      "grad_norm": 4.84375,
      "learning_rate": 9.699166252899104e-06,
      "loss": 0.88501377,
      "memory(GiB)": 126.99,
      "step": 11800,
      "train_speed(iter/s)": 0.203367
    },
    {
      "acc": 0.781742,
      "epoch": 0.27553624287318274,
      "grad_norm": 4.34375,
      "learning_rate": 9.698520533656112e-06,
      "loss": 0.78006496,
      "memory(GiB)": 126.99,
      "step": 11810,
      "train_speed(iter/s)": 0.203457
    },
    {
      "acc": 0.76830721,
      "epoch": 0.27576955044547163,
      "grad_norm": 7.125,
      "learning_rate": 9.697874143700679e-06,
      "loss": 0.81617126,
      "memory(GiB)": 126.99,
      "step": 11820,
      "train_speed(iter/s)": 0.203552
    },
    {
      "acc": 0.77733955,
      "epoch": 0.27600285801776053,
      "grad_norm": 7.65625,
      "learning_rate": 9.697227083125076e-06,
      "loss": 0.80572834,
      "memory(GiB)": 126.99,
      "step": 11830,
      "train_speed(iter/s)": 0.203642
    },
    {
      "acc": 0.77268896,
      "epoch": 0.27623616559004943,
      "grad_norm": 7.03125,
      "learning_rate": 9.69657935202167e-06,
      "loss": 0.80644588,
      "memory(GiB)": 126.99,
      "step": 11840,
      "train_speed(iter/s)": 0.203734
    },
    {
      "acc": 0.7876543,
      "epoch": 0.27646947316233833,
      "grad_norm": 6.1875,
      "learning_rate": 9.695930950482928e-06,
      "loss": 0.747785,
      "memory(GiB)": 126.99,
      "step": 11850,
      "train_speed(iter/s)": 0.203818
    },
    {
      "acc": 0.74815092,
      "epoch": 0.27670278073462723,
      "grad_norm": 5.53125,
      "learning_rate": 9.695281878601406e-06,
      "loss": 0.90503931,
      "memory(GiB)": 126.99,
      "step": 11860,
      "train_speed(iter/s)": 0.203908
    },
    {
      "acc": 0.7729579,
      "epoch": 0.27693608830691613,
      "grad_norm": 5.0625,
      "learning_rate": 9.69463213646976e-06,
      "loss": 0.82826509,
      "memory(GiB)": 126.99,
      "step": 11870,
      "train_speed(iter/s)": 0.203998
    },
    {
      "acc": 0.75472612,
      "epoch": 0.277169395879205,
      "grad_norm": 5.5,
      "learning_rate": 9.69398172418074e-06,
      "loss": 0.87955236,
      "memory(GiB)": 126.99,
      "step": 11880,
      "train_speed(iter/s)": 0.204085
    },
    {
      "acc": 0.78086786,
      "epoch": 0.27740270345149387,
      "grad_norm": 3.859375,
      "learning_rate": 9.693330641827194e-06,
      "loss": 0.75951362,
      "memory(GiB)": 126.99,
      "step": 11890,
      "train_speed(iter/s)": 0.204175
    },
    {
      "acc": 0.78073997,
      "epoch": 0.27763601102378277,
      "grad_norm": 4.46875,
      "learning_rate": 9.69267888950206e-06,
      "loss": 0.78068328,
      "memory(GiB)": 126.99,
      "step": 11900,
      "train_speed(iter/s)": 0.204262
    },
    {
      "acc": 0.77878237,
      "epoch": 0.27786931859607167,
      "grad_norm": 4.375,
      "learning_rate": 9.69202646729838e-06,
      "loss": 0.79412766,
      "memory(GiB)": 126.99,
      "step": 11910,
      "train_speed(iter/s)": 0.204346
    },
    {
      "acc": 0.78074179,
      "epoch": 0.27810262616836057,
      "grad_norm": 5.75,
      "learning_rate": 9.691373375309284e-06,
      "loss": 0.84565754,
      "memory(GiB)": 126.99,
      "step": 11920,
      "train_speed(iter/s)": 0.20443
    },
    {
      "acc": 0.75893049,
      "epoch": 0.27833593374064947,
      "grad_norm": 7.375,
      "learning_rate": 9.690719613628001e-06,
      "loss": 0.88701725,
      "memory(GiB)": 126.99,
      "step": 11930,
      "train_speed(iter/s)": 0.204519
    },
    {
      "acc": 0.77881913,
      "epoch": 0.27856924131293836,
      "grad_norm": 5.46875,
      "learning_rate": 9.690065182347857e-06,
      "loss": 0.8020647,
      "memory(GiB)": 126.99,
      "step": 11940,
      "train_speed(iter/s)": 0.204602
    },
    {
      "acc": 0.75828795,
      "epoch": 0.27880254888522726,
      "grad_norm": 6.03125,
      "learning_rate": 9.68941008156227e-06,
      "loss": 0.88880653,
      "memory(GiB)": 126.99,
      "step": 11950,
      "train_speed(iter/s)": 0.204691
    },
    {
      "acc": 0.7708621,
      "epoch": 0.27903585645751616,
      "grad_norm": 6.0,
      "learning_rate": 9.688754311364755e-06,
      "loss": 0.81502056,
      "memory(GiB)": 126.99,
      "step": 11960,
      "train_speed(iter/s)": 0.204778
    },
    {
      "acc": 0.74046493,
      "epoch": 0.27926916402980506,
      "grad_norm": 6.28125,
      "learning_rate": 9.688097871848925e-06,
      "loss": 0.93770523,
      "memory(GiB)": 126.99,
      "step": 11970,
      "train_speed(iter/s)": 0.204869
    },
    {
      "acc": 0.78146906,
      "epoch": 0.27950247160209396,
      "grad_norm": 6.375,
      "learning_rate": 9.687440763108487e-06,
      "loss": 0.79011135,
      "memory(GiB)": 126.99,
      "step": 11980,
      "train_speed(iter/s)": 0.204953
    },
    {
      "acc": 0.76589327,
      "epoch": 0.2797357791743828,
      "grad_norm": 5.875,
      "learning_rate": 9.68678298523724e-06,
      "loss": 0.86380882,
      "memory(GiB)": 126.99,
      "step": 11990,
      "train_speed(iter/s)": 0.205044
    },
    {
      "acc": 0.7602313,
      "epoch": 0.2799690867466717,
      "grad_norm": 4.84375,
      "learning_rate": 9.686124538329083e-06,
      "loss": 0.86302662,
      "memory(GiB)": 126.99,
      "step": 12000,
      "train_speed(iter/s)": 0.20513
    },
    {
      "epoch": 0.2799690867466717,
      "eval_acc": 0.7339767288010274,
      "eval_loss": 0.8442361354827881,
      "eval_runtime": 1270.1423,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 12000
    },
    {
      "acc": 0.7855062,
      "epoch": 0.2802023943189606,
      "grad_norm": 4.5625,
      "learning_rate": 9.685465422478011e-06,
      "loss": 0.78390985,
      "memory(GiB)": 126.99,
      "step": 12010,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.77192798,
      "epoch": 0.2804357018912495,
      "grad_norm": 4.28125,
      "learning_rate": 9.684805637778109e-06,
      "loss": 0.83765612,
      "memory(GiB)": 126.99,
      "step": 12020,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.77343655,
      "epoch": 0.2806690094635384,
      "grad_norm": 6.65625,
      "learning_rate": 9.684145184323565e-06,
      "loss": 0.81042671,
      "memory(GiB)": 126.99,
      "step": 12030,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.76841726,
      "epoch": 0.2809023170358273,
      "grad_norm": 4.3125,
      "learning_rate": 9.683484062208657e-06,
      "loss": 0.84212866,
      "memory(GiB)": 126.99,
      "step": 12040,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.76003461,
      "epoch": 0.2811356246081162,
      "grad_norm": 7.59375,
      "learning_rate": 9.682822271527758e-06,
      "loss": 0.86936035,
      "memory(GiB)": 126.99,
      "step": 12050,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.76045341,
      "epoch": 0.2813689321804051,
      "grad_norm": 7.15625,
      "learning_rate": 9.682159812375342e-06,
      "loss": 0.87109947,
      "memory(GiB)": 126.99,
      "step": 12060,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.78477945,
      "epoch": 0.281602239752694,
      "grad_norm": 5.25,
      "learning_rate": 9.681496684845973e-06,
      "loss": 0.75963497,
      "memory(GiB)": 126.99,
      "step": 12070,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.76544571,
      "epoch": 0.2818355473249829,
      "grad_norm": 5.09375,
      "learning_rate": 9.68083288903431e-06,
      "loss": 0.84756174,
      "memory(GiB)": 126.99,
      "step": 12080,
      "train_speed(iter/s)": 0.201406
    },
    {
      "acc": 0.77797155,
      "epoch": 0.28206885489727174,
      "grad_norm": 6.65625,
      "learning_rate": 9.680168425035114e-06,
      "loss": 0.8038868,
      "memory(GiB)": 126.99,
      "step": 12090,
      "train_speed(iter/s)": 0.201491
    },
    {
      "acc": 0.75691757,
      "epoch": 0.28230216246956064,
      "grad_norm": 5.75,
      "learning_rate": 9.679503292943234e-06,
      "loss": 0.87715702,
      "memory(GiB)": 126.99,
      "step": 12100,
      "train_speed(iter/s)": 0.201577
    },
    {
      "acc": 0.76970291,
      "epoch": 0.28253547004184953,
      "grad_norm": 5.03125,
      "learning_rate": 9.678837492853619e-06,
      "loss": 0.83921423,
      "memory(GiB)": 126.99,
      "step": 12110,
      "train_speed(iter/s)": 0.201649
    },
    {
      "acc": 0.77343578,
      "epoch": 0.28276877761413843,
      "grad_norm": 6.71875,
      "learning_rate": 9.67817102486131e-06,
      "loss": 0.79686985,
      "memory(GiB)": 126.99,
      "step": 12120,
      "train_speed(iter/s)": 0.201735
    },
    {
      "acc": 0.76627212,
      "epoch": 0.28300208518642733,
      "grad_norm": 5.96875,
      "learning_rate": 9.677503889061446e-06,
      "loss": 0.8474947,
      "memory(GiB)": 126.99,
      "step": 12130,
      "train_speed(iter/s)": 0.201818
    },
    {
      "acc": 0.76688299,
      "epoch": 0.28323539275871623,
      "grad_norm": 5.875,
      "learning_rate": 9.676836085549263e-06,
      "loss": 0.8329731,
      "memory(GiB)": 126.99,
      "step": 12140,
      "train_speed(iter/s)": 0.201905
    },
    {
      "acc": 0.74896669,
      "epoch": 0.28346870033100513,
      "grad_norm": 5.25,
      "learning_rate": 9.676167614420085e-06,
      "loss": 0.91642704,
      "memory(GiB)": 126.99,
      "step": 12150,
      "train_speed(iter/s)": 0.201985
    },
    {
      "acc": 0.76068959,
      "epoch": 0.28370200790329403,
      "grad_norm": 7.15625,
      "learning_rate": 9.67549847576934e-06,
      "loss": 0.8687542,
      "memory(GiB)": 126.99,
      "step": 12160,
      "train_speed(iter/s)": 0.202069
    },
    {
      "acc": 0.76523075,
      "epoch": 0.28393531547558293,
      "grad_norm": 20.0,
      "learning_rate": 9.674828669692545e-06,
      "loss": 0.8203887,
      "memory(GiB)": 126.99,
      "step": 12170,
      "train_speed(iter/s)": 0.202155
    },
    {
      "acc": 0.77215323,
      "epoch": 0.28416862304787177,
      "grad_norm": 6.5625,
      "learning_rate": 9.674158196285316e-06,
      "loss": 0.80515928,
      "memory(GiB)": 126.99,
      "step": 12180,
      "train_speed(iter/s)": 0.202235
    },
    {
      "acc": 0.77063136,
      "epoch": 0.28440193062016067,
      "grad_norm": 7.75,
      "learning_rate": 9.673487055643362e-06,
      "loss": 0.82243662,
      "memory(GiB)": 126.99,
      "step": 12190,
      "train_speed(iter/s)": 0.202321
    },
    {
      "acc": 0.76656513,
      "epoch": 0.28463523819244957,
      "grad_norm": 5.25,
      "learning_rate": 9.672815247862489e-06,
      "loss": 0.82771702,
      "memory(GiB)": 126.99,
      "step": 12200,
      "train_speed(iter/s)": 0.202403
    },
    {
      "acc": 0.78110971,
      "epoch": 0.28486854576473847,
      "grad_norm": 6.15625,
      "learning_rate": 9.672142773038595e-06,
      "loss": 0.80094032,
      "memory(GiB)": 126.99,
      "step": 12210,
      "train_speed(iter/s)": 0.20249
    },
    {
      "acc": 0.75440207,
      "epoch": 0.28510185333702737,
      "grad_norm": 5.0,
      "learning_rate": 9.671469631267678e-06,
      "loss": 0.89850616,
      "memory(GiB)": 126.99,
      "step": 12220,
      "train_speed(iter/s)": 0.20258
    },
    {
      "acc": 0.753054,
      "epoch": 0.28533516090931627,
      "grad_norm": 7.15625,
      "learning_rate": 9.67079582264583e-06,
      "loss": 0.90991497,
      "memory(GiB)": 126.99,
      "step": 12230,
      "train_speed(iter/s)": 0.202672
    },
    {
      "acc": 0.77435451,
      "epoch": 0.28556846848160516,
      "grad_norm": 5.90625,
      "learning_rate": 9.670121347269234e-06,
      "loss": 0.82249737,
      "memory(GiB)": 126.99,
      "step": 12240,
      "train_speed(iter/s)": 0.202753
    },
    {
      "acc": 0.75970974,
      "epoch": 0.28580177605389406,
      "grad_norm": 5.40625,
      "learning_rate": 9.669446205234172e-06,
      "loss": 0.85512848,
      "memory(GiB)": 126.99,
      "step": 12250,
      "train_speed(iter/s)": 0.202839
    },
    {
      "acc": 0.78008022,
      "epoch": 0.28603508362618296,
      "grad_norm": 5.15625,
      "learning_rate": 9.668770396637022e-06,
      "loss": 0.79443493,
      "memory(GiB)": 126.99,
      "step": 12260,
      "train_speed(iter/s)": 0.20292
    },
    {
      "acc": 0.76683731,
      "epoch": 0.28626839119847186,
      "grad_norm": 5.15625,
      "learning_rate": 9.668093921574253e-06,
      "loss": 0.85410089,
      "memory(GiB)": 126.99,
      "step": 12270,
      "train_speed(iter/s)": 0.203005
    },
    {
      "acc": 0.75589981,
      "epoch": 0.2865016987707607,
      "grad_norm": 6.71875,
      "learning_rate": 9.667416780142434e-06,
      "loss": 0.88183594,
      "memory(GiB)": 126.99,
      "step": 12280,
      "train_speed(iter/s)": 0.20309
    },
    {
      "acc": 0.7808938,
      "epoch": 0.2867350063430496,
      "grad_norm": 5.34375,
      "learning_rate": 9.666738972438224e-06,
      "loss": 0.78594069,
      "memory(GiB)": 126.99,
      "step": 12290,
      "train_speed(iter/s)": 0.203169
    },
    {
      "acc": 0.77634277,
      "epoch": 0.2869683139153385,
      "grad_norm": 5.15625,
      "learning_rate": 9.666060498558381e-06,
      "loss": 0.80375671,
      "memory(GiB)": 126.99,
      "step": 12300,
      "train_speed(iter/s)": 0.203253
    },
    {
      "acc": 0.76860728,
      "epoch": 0.2872016214876274,
      "grad_norm": 4.9375,
      "learning_rate": 9.665381358599759e-06,
      "loss": 0.86330442,
      "memory(GiB)": 126.99,
      "step": 12310,
      "train_speed(iter/s)": 0.203335
    },
    {
      "acc": 0.7679863,
      "epoch": 0.2874349290599163,
      "grad_norm": 5.375,
      "learning_rate": 9.664701552659303e-06,
      "loss": 0.838801,
      "memory(GiB)": 126.99,
      "step": 12320,
      "train_speed(iter/s)": 0.20342
    },
    {
      "acc": 0.74797144,
      "epoch": 0.2876682366322052,
      "grad_norm": 8.5,
      "learning_rate": 9.664021080834053e-06,
      "loss": 0.93438997,
      "memory(GiB)": 126.99,
      "step": 12330,
      "train_speed(iter/s)": 0.203504
    },
    {
      "acc": 0.75219126,
      "epoch": 0.2879015442044941,
      "grad_norm": 5.59375,
      "learning_rate": 9.663339943221153e-06,
      "loss": 0.906147,
      "memory(GiB)": 126.99,
      "step": 12340,
      "train_speed(iter/s)": 0.203584
    },
    {
      "acc": 0.76432757,
      "epoch": 0.288134851776783,
      "grad_norm": 5.28125,
      "learning_rate": 9.662658139917827e-06,
      "loss": 0.84146051,
      "memory(GiB)": 126.99,
      "step": 12350,
      "train_speed(iter/s)": 0.203666
    },
    {
      "acc": 0.76514382,
      "epoch": 0.2883681593490719,
      "grad_norm": 5.0625,
      "learning_rate": 9.661975671021408e-06,
      "loss": 0.85332146,
      "memory(GiB)": 126.99,
      "step": 12360,
      "train_speed(iter/s)": 0.203749
    },
    {
      "acc": 0.75728321,
      "epoch": 0.2886014669213608,
      "grad_norm": 4.84375,
      "learning_rate": 9.661292536629316e-06,
      "loss": 0.85679588,
      "memory(GiB)": 126.99,
      "step": 12370,
      "train_speed(iter/s)": 0.203833
    },
    {
      "acc": 0.77875571,
      "epoch": 0.28883477449364964,
      "grad_norm": 7.0,
      "learning_rate": 9.660608736839067e-06,
      "loss": 0.80268106,
      "memory(GiB)": 126.99,
      "step": 12380,
      "train_speed(iter/s)": 0.203916
    },
    {
      "acc": 0.75960855,
      "epoch": 0.28906808206593854,
      "grad_norm": 5.125,
      "learning_rate": 9.659924271748277e-06,
      "loss": 0.86685429,
      "memory(GiB)": 126.99,
      "step": 12390,
      "train_speed(iter/s)": 0.204004
    },
    {
      "acc": 0.76572495,
      "epoch": 0.28930138963822744,
      "grad_norm": 4.09375,
      "learning_rate": 9.65923914145465e-06,
      "loss": 0.8074934,
      "memory(GiB)": 126.99,
      "step": 12400,
      "train_speed(iter/s)": 0.204085
    },
    {
      "acc": 0.76597309,
      "epoch": 0.28953469721051633,
      "grad_norm": 5.59375,
      "learning_rate": 9.65855334605599e-06,
      "loss": 0.82842016,
      "memory(GiB)": 126.99,
      "step": 12410,
      "train_speed(iter/s)": 0.204169
    },
    {
      "acc": 0.75547256,
      "epoch": 0.28976800478280523,
      "grad_norm": 6.5,
      "learning_rate": 9.65786688565019e-06,
      "loss": 0.88945179,
      "memory(GiB)": 126.99,
      "step": 12420,
      "train_speed(iter/s)": 0.204248
    },
    {
      "acc": 0.77330575,
      "epoch": 0.29000131235509413,
      "grad_norm": 5.34375,
      "learning_rate": 9.65717976033525e-06,
      "loss": 0.80908098,
      "memory(GiB)": 126.99,
      "step": 12430,
      "train_speed(iter/s)": 0.204329
    },
    {
      "acc": 0.76689053,
      "epoch": 0.29023461992738303,
      "grad_norm": 4.6875,
      "learning_rate": 9.656491970209248e-06,
      "loss": 0.84457645,
      "memory(GiB)": 126.99,
      "step": 12440,
      "train_speed(iter/s)": 0.204414
    },
    {
      "acc": 0.76981139,
      "epoch": 0.29046792749967193,
      "grad_norm": 6.71875,
      "learning_rate": 9.655803515370373e-06,
      "loss": 0.8389142,
      "memory(GiB)": 126.99,
      "step": 12450,
      "train_speed(iter/s)": 0.204496
    },
    {
      "acc": 0.77279081,
      "epoch": 0.29070123507196083,
      "grad_norm": 4.46875,
      "learning_rate": 9.655114395916896e-06,
      "loss": 0.8170186,
      "memory(GiB)": 126.99,
      "step": 12460,
      "train_speed(iter/s)": 0.20458
    },
    {
      "acc": 0.7852149,
      "epoch": 0.29093454264424967,
      "grad_norm": 6.90625,
      "learning_rate": 9.654424611947194e-06,
      "loss": 0.80533628,
      "memory(GiB)": 126.99,
      "step": 12470,
      "train_speed(iter/s)": 0.204664
    },
    {
      "acc": 0.76608725,
      "epoch": 0.29116785021653857,
      "grad_norm": 5.75,
      "learning_rate": 9.65373416355973e-06,
      "loss": 0.84452038,
      "memory(GiB)": 126.99,
      "step": 12480,
      "train_speed(iter/s)": 0.204746
    },
    {
      "acc": 0.76543536,
      "epoch": 0.29140115778882747,
      "grad_norm": 5.59375,
      "learning_rate": 9.653043050853065e-06,
      "loss": 0.85324554,
      "memory(GiB)": 126.99,
      "step": 12490,
      "train_speed(iter/s)": 0.204826
    },
    {
      "acc": 0.76770477,
      "epoch": 0.29163446536111637,
      "grad_norm": 6.5625,
      "learning_rate": 9.652351273925854e-06,
      "loss": 0.84301796,
      "memory(GiB)": 126.99,
      "step": 12500,
      "train_speed(iter/s)": 0.204912
    },
    {
      "epoch": 0.29163446536111637,
      "eval_acc": 0.7345405966429606,
      "eval_loss": 0.8422214984893799,
      "eval_runtime": 1269.9542,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.171,
      "step": 12500
    },
    {
      "acc": 0.7895587,
      "epoch": 0.29186777293340527,
      "grad_norm": 5.8125,
      "learning_rate": 9.651658832876853e-06,
      "loss": 0.75596642,
      "memory(GiB)": 126.99,
      "step": 12510,
      "train_speed(iter/s)": 0.200765
    },
    {
      "acc": 0.75599642,
      "epoch": 0.29210108050569417,
      "grad_norm": 5.40625,
      "learning_rate": 9.650965727804907e-06,
      "loss": 0.88882275,
      "memory(GiB)": 126.99,
      "step": 12520,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.77908826,
      "epoch": 0.29233438807798307,
      "grad_norm": 6.25,
      "learning_rate": 9.65027195880895e-06,
      "loss": 0.79518585,
      "memory(GiB)": 126.99,
      "step": 12530,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.7476614,
      "epoch": 0.29256769565027196,
      "grad_norm": 10.25,
      "learning_rate": 9.649577525988025e-06,
      "loss": 0.93213406,
      "memory(GiB)": 126.99,
      "step": 12540,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.7507143,
      "epoch": 0.29280100322256086,
      "grad_norm": 5.78125,
      "learning_rate": 9.648882429441258e-06,
      "loss": 0.92582474,
      "memory(GiB)": 126.99,
      "step": 12550,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.76029882,
      "epoch": 0.29303431079484976,
      "grad_norm": 10.0625,
      "learning_rate": 9.648186669267874e-06,
      "loss": 0.86718149,
      "memory(GiB)": 126.99,
      "step": 12560,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.75470057,
      "epoch": 0.2932676183671386,
      "grad_norm": 6.6875,
      "learning_rate": 9.647490245567194e-06,
      "loss": 0.90233717,
      "memory(GiB)": 126.99,
      "step": 12570,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.7683279,
      "epoch": 0.2935009259394275,
      "grad_norm": 7.0625,
      "learning_rate": 9.646793158438632e-06,
      "loss": 0.82610855,
      "memory(GiB)": 126.99,
      "step": 12580,
      "train_speed(iter/s)": 0.201344
    },
    {
      "acc": 0.77332907,
      "epoch": 0.2937342335117164,
      "grad_norm": 9.125,
      "learning_rate": 9.646095407981695e-06,
      "loss": 0.80713387,
      "memory(GiB)": 126.99,
      "step": 12590,
      "train_speed(iter/s)": 0.201425
    },
    {
      "acc": 0.74880519,
      "epoch": 0.2939675410840053,
      "grad_norm": 8.75,
      "learning_rate": 9.64539699429599e-06,
      "loss": 0.91812592,
      "memory(GiB)": 126.99,
      "step": 12600,
      "train_speed(iter/s)": 0.201507
    },
    {
      "acc": 0.76272821,
      "epoch": 0.2942008486562942,
      "grad_norm": 7.25,
      "learning_rate": 9.644697917481212e-06,
      "loss": 0.84828262,
      "memory(GiB)": 126.99,
      "step": 12610,
      "train_speed(iter/s)": 0.201593
    },
    {
      "acc": 0.76495266,
      "epoch": 0.2944341562285831,
      "grad_norm": 5.46875,
      "learning_rate": 9.643998177637157e-06,
      "loss": 0.8520998,
      "memory(GiB)": 126.99,
      "step": 12620,
      "train_speed(iter/s)": 0.201679
    },
    {
      "acc": 0.75189352,
      "epoch": 0.294667463800872,
      "grad_norm": 5.40625,
      "learning_rate": 9.643297774863709e-06,
      "loss": 0.88819065,
      "memory(GiB)": 126.99,
      "step": 12630,
      "train_speed(iter/s)": 0.201765
    },
    {
      "acc": 0.74685659,
      "epoch": 0.2949007713731609,
      "grad_norm": 6.0625,
      "learning_rate": 9.642596709260854e-06,
      "loss": 0.93766823,
      "memory(GiB)": 126.99,
      "step": 12640,
      "train_speed(iter/s)": 0.201846
    },
    {
      "acc": 0.77418585,
      "epoch": 0.2951340789454498,
      "grad_norm": 7.71875,
      "learning_rate": 9.641894980928668e-06,
      "loss": 0.82185469,
      "memory(GiB)": 126.99,
      "step": 12650,
      "train_speed(iter/s)": 0.201928
    },
    {
      "acc": 0.77108278,
      "epoch": 0.29536738651773864,
      "grad_norm": 7.53125,
      "learning_rate": 9.641192589967321e-06,
      "loss": 0.84046984,
      "memory(GiB)": 126.99,
      "step": 12660,
      "train_speed(iter/s)": 0.202013
    },
    {
      "acc": 0.7660718,
      "epoch": 0.29560069409002754,
      "grad_norm": 6.28125,
      "learning_rate": 9.64048953647708e-06,
      "loss": 0.83025084,
      "memory(GiB)": 126.99,
      "step": 12670,
      "train_speed(iter/s)": 0.202095
    },
    {
      "acc": 0.76199236,
      "epoch": 0.29583400166231644,
      "grad_norm": 3.8125,
      "learning_rate": 9.639785820558307e-06,
      "loss": 0.88814316,
      "memory(GiB)": 126.99,
      "step": 12680,
      "train_speed(iter/s)": 0.202176
    },
    {
      "acc": 0.75356421,
      "epoch": 0.29606730923460534,
      "grad_norm": 4.96875,
      "learning_rate": 9.639081442311456e-06,
      "loss": 0.88174667,
      "memory(GiB)": 126.99,
      "step": 12690,
      "train_speed(iter/s)": 0.202261
    },
    {
      "acc": 0.76016741,
      "epoch": 0.29630061680689423,
      "grad_norm": 6.59375,
      "learning_rate": 9.638376401837075e-06,
      "loss": 0.87111378,
      "memory(GiB)": 126.99,
      "step": 12700,
      "train_speed(iter/s)": 0.202342
    },
    {
      "acc": 0.78399467,
      "epoch": 0.29653392437918313,
      "grad_norm": 4.625,
      "learning_rate": 9.63767069923581e-06,
      "loss": 0.77465773,
      "memory(GiB)": 126.99,
      "step": 12710,
      "train_speed(iter/s)": 0.202423
    },
    {
      "acc": 0.76579094,
      "epoch": 0.29676723195147203,
      "grad_norm": 5.28125,
      "learning_rate": 9.636964334608402e-06,
      "loss": 0.85976906,
      "memory(GiB)": 126.99,
      "step": 12720,
      "train_speed(iter/s)": 0.2025
    },
    {
      "acc": 0.76346893,
      "epoch": 0.29700053952376093,
      "grad_norm": 5.5625,
      "learning_rate": 9.636257308055682e-06,
      "loss": 0.85128269,
      "memory(GiB)": 126.99,
      "step": 12730,
      "train_speed(iter/s)": 0.202583
    },
    {
      "acc": 0.76815395,
      "epoch": 0.29723384709604983,
      "grad_norm": 4.65625,
      "learning_rate": 9.635549619678578e-06,
      "loss": 0.84149694,
      "memory(GiB)": 126.99,
      "step": 12740,
      "train_speed(iter/s)": 0.202667
    },
    {
      "acc": 0.76429968,
      "epoch": 0.29746715466833873,
      "grad_norm": 5.15625,
      "learning_rate": 9.63484126957811e-06,
      "loss": 0.8531889,
      "memory(GiB)": 126.99,
      "step": 12750,
      "train_speed(iter/s)": 0.20275
    },
    {
      "acc": 0.75266466,
      "epoch": 0.2977004622406276,
      "grad_norm": 7.625,
      "learning_rate": 9.6341322578554e-06,
      "loss": 0.87924614,
      "memory(GiB)": 126.99,
      "step": 12760,
      "train_speed(iter/s)": 0.202829
    },
    {
      "acc": 0.77012124,
      "epoch": 0.29793376981291647,
      "grad_norm": 4.4375,
      "learning_rate": 9.633422584611654e-06,
      "loss": 0.81980047,
      "memory(GiB)": 126.99,
      "step": 12770,
      "train_speed(iter/s)": 0.202914
    },
    {
      "acc": 0.78082376,
      "epoch": 0.29816707738520537,
      "grad_norm": 4.90625,
      "learning_rate": 9.632712249948182e-06,
      "loss": 0.80720272,
      "memory(GiB)": 126.99,
      "step": 12780,
      "train_speed(iter/s)": 0.202998
    },
    {
      "acc": 0.7774765,
      "epoch": 0.29840038495749427,
      "grad_norm": 4.71875,
      "learning_rate": 9.632001253966381e-06,
      "loss": 0.79807181,
      "memory(GiB)": 126.99,
      "step": 12790,
      "train_speed(iter/s)": 0.203084
    },
    {
      "acc": 0.77757535,
      "epoch": 0.29863369252978317,
      "grad_norm": 5.625,
      "learning_rate": 9.631289596767748e-06,
      "loss": 0.78618836,
      "memory(GiB)": 126.99,
      "step": 12800,
      "train_speed(iter/s)": 0.203164
    },
    {
      "acc": 0.77521505,
      "epoch": 0.29886700010207207,
      "grad_norm": 8.1875,
      "learning_rate": 9.63057727845387e-06,
      "loss": 0.80945616,
      "memory(GiB)": 126.99,
      "step": 12810,
      "train_speed(iter/s)": 0.203252
    },
    {
      "acc": 0.77795644,
      "epoch": 0.29910030767436097,
      "grad_norm": 8.875,
      "learning_rate": 9.62986429912643e-06,
      "loss": 0.78125734,
      "memory(GiB)": 126.99,
      "step": 12820,
      "train_speed(iter/s)": 0.203339
    },
    {
      "acc": 0.77475395,
      "epoch": 0.29933361524664986,
      "grad_norm": 5.8125,
      "learning_rate": 9.629150658887206e-06,
      "loss": 0.81328382,
      "memory(GiB)": 126.99,
      "step": 12830,
      "train_speed(iter/s)": 0.203414
    },
    {
      "acc": 0.77334385,
      "epoch": 0.29956692281893876,
      "grad_norm": 5.1875,
      "learning_rate": 9.628436357838072e-06,
      "loss": 0.81213036,
      "memory(GiB)": 126.99,
      "step": 12840,
      "train_speed(iter/s)": 0.203488
    },
    {
      "acc": 0.79729042,
      "epoch": 0.29980023039122766,
      "grad_norm": 4.84375,
      "learning_rate": 9.627721396080992e-06,
      "loss": 0.73419204,
      "memory(GiB)": 126.99,
      "step": 12850,
      "train_speed(iter/s)": 0.203572
    },
    {
      "acc": 0.76485343,
      "epoch": 0.3000335379635165,
      "grad_norm": 6.78125,
      "learning_rate": 9.627005773718026e-06,
      "loss": 0.85712852,
      "memory(GiB)": 126.99,
      "step": 12860,
      "train_speed(iter/s)": 0.203649
    },
    {
      "acc": 0.77371435,
      "epoch": 0.3002668455358054,
      "grad_norm": 6.25,
      "learning_rate": 9.626289490851329e-06,
      "loss": 0.81916208,
      "memory(GiB)": 126.99,
      "step": 12870,
      "train_speed(iter/s)": 0.203723
    },
    {
      "acc": 0.78421712,
      "epoch": 0.3005001531080943,
      "grad_norm": 3.6875,
      "learning_rate": 9.625572547583153e-06,
      "loss": 0.77294006,
      "memory(GiB)": 126.99,
      "step": 12880,
      "train_speed(iter/s)": 0.203807
    },
    {
      "acc": 0.77695723,
      "epoch": 0.3007334606803832,
      "grad_norm": 5.75,
      "learning_rate": 9.624854944015839e-06,
      "loss": 0.80077038,
      "memory(GiB)": 126.99,
      "step": 12890,
      "train_speed(iter/s)": 0.203892
    },
    {
      "acc": 0.77849216,
      "epoch": 0.3009667682526721,
      "grad_norm": 4.53125,
      "learning_rate": 9.624136680251826e-06,
      "loss": 0.78988638,
      "memory(GiB)": 126.99,
      "step": 12900,
      "train_speed(iter/s)": 0.203969
    },
    {
      "acc": 0.78356485,
      "epoch": 0.301200075824961,
      "grad_norm": 6.28125,
      "learning_rate": 9.623417756393644e-06,
      "loss": 0.78568068,
      "memory(GiB)": 126.99,
      "step": 12910,
      "train_speed(iter/s)": 0.204048
    },
    {
      "acc": 0.75422754,
      "epoch": 0.3014333833972499,
      "grad_norm": 43.25,
      "learning_rate": 9.622698172543921e-06,
      "loss": 0.88511696,
      "memory(GiB)": 126.99,
      "step": 12920,
      "train_speed(iter/s)": 0.20413
    },
    {
      "acc": 0.76167445,
      "epoch": 0.3016666909695388,
      "grad_norm": 5.375,
      "learning_rate": 9.621977928805377e-06,
      "loss": 0.85034475,
      "memory(GiB)": 126.99,
      "step": 12930,
      "train_speed(iter/s)": 0.204214
    },
    {
      "acc": 0.77598619,
      "epoch": 0.3018999985418277,
      "grad_norm": 6.9375,
      "learning_rate": 9.621257025280826e-06,
      "loss": 0.79932165,
      "memory(GiB)": 126.99,
      "step": 12940,
      "train_speed(iter/s)": 0.204302
    },
    {
      "acc": 0.76910601,
      "epoch": 0.30213330611411654,
      "grad_norm": 15.6875,
      "learning_rate": 9.620535462073177e-06,
      "loss": 0.85517778,
      "memory(GiB)": 126.99,
      "step": 12950,
      "train_speed(iter/s)": 0.204382
    },
    {
      "acc": 0.74408941,
      "epoch": 0.30236661368640544,
      "grad_norm": 6.09375,
      "learning_rate": 9.619813239285433e-06,
      "loss": 0.95150757,
      "memory(GiB)": 126.99,
      "step": 12960,
      "train_speed(iter/s)": 0.204462
    },
    {
      "acc": 0.77192564,
      "epoch": 0.30259992125869434,
      "grad_norm": 5.65625,
      "learning_rate": 9.619090357020691e-06,
      "loss": 0.81647863,
      "memory(GiB)": 126.99,
      "step": 12970,
      "train_speed(iter/s)": 0.204548
    },
    {
      "acc": 0.78476696,
      "epoch": 0.30283322883098324,
      "grad_norm": 6.21875,
      "learning_rate": 9.618366815382143e-06,
      "loss": 0.78855925,
      "memory(GiB)": 126.99,
      "step": 12980,
      "train_speed(iter/s)": 0.204626
    },
    {
      "acc": 0.77091103,
      "epoch": 0.30306653640327214,
      "grad_norm": 5.6875,
      "learning_rate": 9.617642614473073e-06,
      "loss": 0.88907585,
      "memory(GiB)": 126.99,
      "step": 12990,
      "train_speed(iter/s)": 0.204708
    },
    {
      "acc": 0.767976,
      "epoch": 0.30329984397556103,
      "grad_norm": 9.375,
      "learning_rate": 9.616917754396861e-06,
      "loss": 0.82190113,
      "memory(GiB)": 126.99,
      "step": 13000,
      "train_speed(iter/s)": 0.204791
    },
    {
      "epoch": 0.30329984397556103,
      "eval_acc": 0.7348525576074836,
      "eval_loss": 0.8418717384338379,
      "eval_runtime": 1269.0169,
      "eval_samples_per_second": 28.361,
      "eval_steps_per_second": 14.181,
      "step": 13000
    },
    {
      "acc": 0.77310801,
      "epoch": 0.30353315154784993,
      "grad_norm": 5.375,
      "learning_rate": 9.616192235256983e-06,
      "loss": 0.80410042,
      "memory(GiB)": 126.99,
      "step": 13010,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.75549717,
      "epoch": 0.30376645912013883,
      "grad_norm": 6.28125,
      "learning_rate": 9.615466057157002e-06,
      "loss": 0.87396736,
      "memory(GiB)": 126.99,
      "step": 13020,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.78147354,
      "epoch": 0.30399976669242773,
      "grad_norm": 6.96875,
      "learning_rate": 9.614739220200583e-06,
      "loss": 0.79309797,
      "memory(GiB)": 126.99,
      "step": 13030,
      "train_speed(iter/s)": 0.20096
    },
    {
      "acc": 0.76732526,
      "epoch": 0.30423307426471663,
      "grad_norm": 6.28125,
      "learning_rate": 9.614011724491481e-06,
      "loss": 0.8398242,
      "memory(GiB)": 126.99,
      "step": 13040,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.75758252,
      "epoch": 0.3044663818370055,
      "grad_norm": 8.5,
      "learning_rate": 9.613283570133547e-06,
      "loss": 0.87248812,
      "memory(GiB)": 126.99,
      "step": 13050,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.76003232,
      "epoch": 0.30469968940929437,
      "grad_norm": 6.84375,
      "learning_rate": 9.612554757230722e-06,
      "loss": 0.87683392,
      "memory(GiB)": 126.99,
      "step": 13060,
      "train_speed(iter/s)": 0.201212
    },
    {
      "acc": 0.77400045,
      "epoch": 0.30493299698158327,
      "grad_norm": 5.78125,
      "learning_rate": 9.611825285887045e-06,
      "loss": 0.81213093,
      "memory(GiB)": 126.99,
      "step": 13070,
      "train_speed(iter/s)": 0.201287
    },
    {
      "acc": 0.76753764,
      "epoch": 0.30516630455387217,
      "grad_norm": 6.21875,
      "learning_rate": 9.61109515620665e-06,
      "loss": 0.84945755,
      "memory(GiB)": 126.99,
      "step": 13080,
      "train_speed(iter/s)": 0.201367
    },
    {
      "acc": 0.79222326,
      "epoch": 0.30539961212616107,
      "grad_norm": 5.5,
      "learning_rate": 9.61036436829376e-06,
      "loss": 0.73132648,
      "memory(GiB)": 126.99,
      "step": 13090,
      "train_speed(iter/s)": 0.201446
    },
    {
      "acc": 0.75894957,
      "epoch": 0.30563291969844997,
      "grad_norm": 5.5625,
      "learning_rate": 9.609632922252695e-06,
      "loss": 0.89140854,
      "memory(GiB)": 126.99,
      "step": 13100,
      "train_speed(iter/s)": 0.201525
    },
    {
      "acc": 0.77855654,
      "epoch": 0.30586622727073887,
      "grad_norm": 5.0625,
      "learning_rate": 9.60890081818787e-06,
      "loss": 0.81173906,
      "memory(GiB)": 126.99,
      "step": 13110,
      "train_speed(iter/s)": 0.201604
    },
    {
      "acc": 0.77230225,
      "epoch": 0.30609953484302777,
      "grad_norm": 5.40625,
      "learning_rate": 9.608168056203792e-06,
      "loss": 0.80290689,
      "memory(GiB)": 126.99,
      "step": 13120,
      "train_speed(iter/s)": 0.201684
    },
    {
      "acc": 0.76812458,
      "epoch": 0.30633284241531666,
      "grad_norm": 7.71875,
      "learning_rate": 9.607434636405063e-06,
      "loss": 0.86059694,
      "memory(GiB)": 126.99,
      "step": 13130,
      "train_speed(iter/s)": 0.201761
    },
    {
      "acc": 0.76915483,
      "epoch": 0.30656614998760556,
      "grad_norm": 4.90625,
      "learning_rate": 9.606700558896376e-06,
      "loss": 0.8283617,
      "memory(GiB)": 126.99,
      "step": 13140,
      "train_speed(iter/s)": 0.201843
    },
    {
      "acc": 0.75852375,
      "epoch": 0.3067994575598944,
      "grad_norm": 5.59375,
      "learning_rate": 9.605965823782525e-06,
      "loss": 0.88993883,
      "memory(GiB)": 126.99,
      "step": 13150,
      "train_speed(iter/s)": 0.201916
    },
    {
      "acc": 0.77059002,
      "epoch": 0.3070327651321833,
      "grad_norm": 6.875,
      "learning_rate": 9.605230431168391e-06,
      "loss": 0.83691521,
      "memory(GiB)": 126.99,
      "step": 13160,
      "train_speed(iter/s)": 0.201993
    },
    {
      "acc": 0.77577543,
      "epoch": 0.3072660727044722,
      "grad_norm": 4.3125,
      "learning_rate": 9.604494381158949e-06,
      "loss": 0.81176348,
      "memory(GiB)": 126.99,
      "step": 13170,
      "train_speed(iter/s)": 0.202066
    },
    {
      "acc": 0.78056211,
      "epoch": 0.3074993802767611,
      "grad_norm": 5.53125,
      "learning_rate": 9.603757673859274e-06,
      "loss": 0.80487776,
      "memory(GiB)": 126.99,
      "step": 13180,
      "train_speed(iter/s)": 0.202144
    },
    {
      "acc": 0.76723247,
      "epoch": 0.30773268784905,
      "grad_norm": 11.375,
      "learning_rate": 9.603020309374526e-06,
      "loss": 0.87342901,
      "memory(GiB)": 126.99,
      "step": 13190,
      "train_speed(iter/s)": 0.202221
    },
    {
      "acc": 0.77724471,
      "epoch": 0.3079659954213389,
      "grad_norm": 8.75,
      "learning_rate": 9.602282287809966e-06,
      "loss": 0.8030695,
      "memory(GiB)": 126.99,
      "step": 13200,
      "train_speed(iter/s)": 0.202298
    },
    {
      "acc": 0.77045097,
      "epoch": 0.3081993029936278,
      "grad_norm": 4.125,
      "learning_rate": 9.601543609270947e-06,
      "loss": 0.83347521,
      "memory(GiB)": 126.99,
      "step": 13210,
      "train_speed(iter/s)": 0.202369
    },
    {
      "acc": 0.76114216,
      "epoch": 0.3084326105659167,
      "grad_norm": 7.5625,
      "learning_rate": 9.600804273862917e-06,
      "loss": 0.86944656,
      "memory(GiB)": 126.99,
      "step": 13220,
      "train_speed(iter/s)": 0.202449
    },
    {
      "acc": 0.76104956,
      "epoch": 0.3086659181382056,
      "grad_norm": 10.1875,
      "learning_rate": 9.60006428169141e-06,
      "loss": 0.87441788,
      "memory(GiB)": 126.99,
      "step": 13230,
      "train_speed(iter/s)": 0.202524
    },
    {
      "acc": 0.78825359,
      "epoch": 0.30889922571049444,
      "grad_norm": 4.5625,
      "learning_rate": 9.599323632862063e-06,
      "loss": 0.7530653,
      "memory(GiB)": 126.99,
      "step": 13240,
      "train_speed(iter/s)": 0.202602
    },
    {
      "acc": 0.77793827,
      "epoch": 0.30913253328278334,
      "grad_norm": 8.1875,
      "learning_rate": 9.598582327480605e-06,
      "loss": 0.78715887,
      "memory(GiB)": 126.99,
      "step": 13250,
      "train_speed(iter/s)": 0.202675
    },
    {
      "acc": 0.76915102,
      "epoch": 0.30936584085507224,
      "grad_norm": 5.6875,
      "learning_rate": 9.597840365652857e-06,
      "loss": 0.82546225,
      "memory(GiB)": 126.99,
      "step": 13260,
      "train_speed(iter/s)": 0.202753
    },
    {
      "acc": 0.7758049,
      "epoch": 0.30959914842736114,
      "grad_norm": 5.5625,
      "learning_rate": 9.597097747484731e-06,
      "loss": 0.8197854,
      "memory(GiB)": 126.99,
      "step": 13270,
      "train_speed(iter/s)": 0.202827
    },
    {
      "acc": 0.77688622,
      "epoch": 0.30983245599965004,
      "grad_norm": 5.0625,
      "learning_rate": 9.596354473082237e-06,
      "loss": 0.80246162,
      "memory(GiB)": 126.99,
      "step": 13280,
      "train_speed(iter/s)": 0.202904
    },
    {
      "acc": 0.77225947,
      "epoch": 0.31006576357193893,
      "grad_norm": 6.46875,
      "learning_rate": 9.595610542551476e-06,
      "loss": 0.82849407,
      "memory(GiB)": 126.99,
      "step": 13290,
      "train_speed(iter/s)": 0.202984
    },
    {
      "acc": 0.76478229,
      "epoch": 0.31029907114422783,
      "grad_norm": 6.09375,
      "learning_rate": 9.594865955998648e-06,
      "loss": 0.85376844,
      "memory(GiB)": 126.99,
      "step": 13300,
      "train_speed(iter/s)": 0.203061
    },
    {
      "acc": 0.77614241,
      "epoch": 0.31053237871651673,
      "grad_norm": 4.1875,
      "learning_rate": 9.594120713530038e-06,
      "loss": 0.82821312,
      "memory(GiB)": 126.99,
      "step": 13310,
      "train_speed(iter/s)": 0.203142
    },
    {
      "acc": 0.76965942,
      "epoch": 0.31076568628880563,
      "grad_norm": 6.125,
      "learning_rate": 9.59337481525203e-06,
      "loss": 0.83385906,
      "memory(GiB)": 126.99,
      "step": 13320,
      "train_speed(iter/s)": 0.20322
    },
    {
      "acc": 0.7630805,
      "epoch": 0.31099899386109453,
      "grad_norm": 6.96875,
      "learning_rate": 9.592628261271102e-06,
      "loss": 0.8761137,
      "memory(GiB)": 126.99,
      "step": 13330,
      "train_speed(iter/s)": 0.203301
    },
    {
      "acc": 0.76680398,
      "epoch": 0.3112323014333834,
      "grad_norm": 5.28125,
      "learning_rate": 9.591881051693826e-06,
      "loss": 0.85415878,
      "memory(GiB)": 126.99,
      "step": 13340,
      "train_speed(iter/s)": 0.20338
    },
    {
      "acc": 0.77896576,
      "epoch": 0.3114656090056723,
      "grad_norm": 6.53125,
      "learning_rate": 9.591133186626861e-06,
      "loss": 0.80970163,
      "memory(GiB)": 126.99,
      "step": 13350,
      "train_speed(iter/s)": 0.203457
    },
    {
      "acc": 0.75927386,
      "epoch": 0.31169891657796117,
      "grad_norm": 4.84375,
      "learning_rate": 9.590384666176968e-06,
      "loss": 0.86237316,
      "memory(GiB)": 126.99,
      "step": 13360,
      "train_speed(iter/s)": 0.203535
    },
    {
      "acc": 0.77169557,
      "epoch": 0.31193222415025007,
      "grad_norm": 4.15625,
      "learning_rate": 9.589635490450999e-06,
      "loss": 0.80186424,
      "memory(GiB)": 126.99,
      "step": 13370,
      "train_speed(iter/s)": 0.203611
    },
    {
      "acc": 0.78818069,
      "epoch": 0.31216553172253897,
      "grad_norm": 5.0,
      "learning_rate": 9.588885659555895e-06,
      "loss": 0.76719275,
      "memory(GiB)": 126.99,
      "step": 13380,
      "train_speed(iter/s)": 0.20369
    },
    {
      "acc": 0.74860768,
      "epoch": 0.31239883929482787,
      "grad_norm": 7.09375,
      "learning_rate": 9.588135173598696e-06,
      "loss": 0.91373577,
      "memory(GiB)": 126.99,
      "step": 13390,
      "train_speed(iter/s)": 0.203769
    },
    {
      "acc": 0.77999992,
      "epoch": 0.31263214686711677,
      "grad_norm": 9.8125,
      "learning_rate": 9.587384032686536e-06,
      "loss": 0.78641224,
      "memory(GiB)": 126.99,
      "step": 13400,
      "train_speed(iter/s)": 0.203846
    },
    {
      "acc": 0.78604879,
      "epoch": 0.31286545443940567,
      "grad_norm": 5.6875,
      "learning_rate": 9.586632236926637e-06,
      "loss": 0.76967993,
      "memory(GiB)": 126.99,
      "step": 13410,
      "train_speed(iter/s)": 0.203919
    },
    {
      "acc": 0.76556487,
      "epoch": 0.31309876201169456,
      "grad_norm": 5.1875,
      "learning_rate": 9.585879786426317e-06,
      "loss": 0.83489494,
      "memory(GiB)": 126.99,
      "step": 13420,
      "train_speed(iter/s)": 0.203997
    },
    {
      "acc": 0.77812591,
      "epoch": 0.3133320695839834,
      "grad_norm": 4.3125,
      "learning_rate": 9.585126681292991e-06,
      "loss": 0.7988245,
      "memory(GiB)": 126.99,
      "step": 13430,
      "train_speed(iter/s)": 0.204078
    },
    {
      "acc": 0.77678261,
      "epoch": 0.3135653771562723,
      "grad_norm": 6.71875,
      "learning_rate": 9.584372921634164e-06,
      "loss": 0.82126627,
      "memory(GiB)": 126.99,
      "step": 13440,
      "train_speed(iter/s)": 0.204156
    },
    {
      "acc": 0.77149763,
      "epoch": 0.3137986847285612,
      "grad_norm": 6.34375,
      "learning_rate": 9.583618507557433e-06,
      "loss": 0.82525063,
      "memory(GiB)": 126.99,
      "step": 13450,
      "train_speed(iter/s)": 0.204235
    },
    {
      "acc": 0.76586752,
      "epoch": 0.3140319923008501,
      "grad_norm": 4.6875,
      "learning_rate": 9.582863439170493e-06,
      "loss": 0.83788195,
      "memory(GiB)": 126.99,
      "step": 13460,
      "train_speed(iter/s)": 0.204314
    },
    {
      "acc": 0.77512035,
      "epoch": 0.314265299873139,
      "grad_norm": 4.8125,
      "learning_rate": 9.582107716581125e-06,
      "loss": 0.82719707,
      "memory(GiB)": 126.99,
      "step": 13470,
      "train_speed(iter/s)": 0.204394
    },
    {
      "acc": 0.79411778,
      "epoch": 0.3144986074454279,
      "grad_norm": 12.1875,
      "learning_rate": 9.581351339897215e-06,
      "loss": 0.73287711,
      "memory(GiB)": 126.99,
      "step": 13480,
      "train_speed(iter/s)": 0.204473
    },
    {
      "acc": 0.78208966,
      "epoch": 0.3147319150177168,
      "grad_norm": 5.03125,
      "learning_rate": 9.580594309226731e-06,
      "loss": 0.75941992,
      "memory(GiB)": 126.99,
      "step": 13490,
      "train_speed(iter/s)": 0.204553
    },
    {
      "acc": 0.75446653,
      "epoch": 0.3149652225900057,
      "grad_norm": 6.65625,
      "learning_rate": 9.579836624677742e-06,
      "loss": 0.89500237,
      "memory(GiB)": 126.99,
      "step": 13500,
      "train_speed(iter/s)": 0.204631
    },
    {
      "epoch": 0.3149652225900057,
      "eval_acc": 0.7354596643921382,
      "eval_loss": 0.8395183086395264,
      "eval_runtime": 1270.619,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "step": 13500
    },
    {
      "acc": 0.7628726,
      "epoch": 0.3151985301622946,
      "grad_norm": 4.3125,
      "learning_rate": 9.579078286358403e-06,
      "loss": 0.86440277,
      "memory(GiB)": 126.99,
      "step": 13510,
      "train_speed(iter/s)": 0.200788
    },
    {
      "acc": 0.76125512,
      "epoch": 0.3154318377345835,
      "grad_norm": 5.4375,
      "learning_rate": 9.578319294376968e-06,
      "loss": 0.86319227,
      "memory(GiB)": 126.99,
      "step": 13520,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.76013141,
      "epoch": 0.31566514530687234,
      "grad_norm": 5.59375,
      "learning_rate": 9.577559648841785e-06,
      "loss": 0.85533056,
      "memory(GiB)": 126.99,
      "step": 13530,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.76009645,
      "epoch": 0.31589845287916124,
      "grad_norm": 6.59375,
      "learning_rate": 9.576799349861292e-06,
      "loss": 0.86631908,
      "memory(GiB)": 126.99,
      "step": 13540,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.76818829,
      "epoch": 0.31613176045145014,
      "grad_norm": 5.6875,
      "learning_rate": 9.576038397544021e-06,
      "loss": 0.83353624,
      "memory(GiB)": 126.99,
      "step": 13550,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.76592422,
      "epoch": 0.31636506802373904,
      "grad_norm": 4.8125,
      "learning_rate": 9.5752767919986e-06,
      "loss": 0.8563282,
      "memory(GiB)": 126.99,
      "step": 13560,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.76527843,
      "epoch": 0.31659837559602794,
      "grad_norm": 8.125,
      "learning_rate": 9.574514533333744e-06,
      "loss": 0.86033154,
      "memory(GiB)": 126.99,
      "step": 13570,
      "train_speed(iter/s)": 0.201252
    },
    {
      "acc": 0.77543097,
      "epoch": 0.31683168316831684,
      "grad_norm": 17.25,
      "learning_rate": 9.573751621658267e-06,
      "loss": 0.79135704,
      "memory(GiB)": 126.99,
      "step": 13580,
      "train_speed(iter/s)": 0.201323
    },
    {
      "acc": 0.76395407,
      "epoch": 0.31706499074060573,
      "grad_norm": 6.03125,
      "learning_rate": 9.572988057081076e-06,
      "loss": 0.85698547,
      "memory(GiB)": 126.99,
      "step": 13590,
      "train_speed(iter/s)": 0.2014
    },
    {
      "acc": 0.74935603,
      "epoch": 0.31729829831289463,
      "grad_norm": 5.8125,
      "learning_rate": 9.572223839711168e-06,
      "loss": 0.89535446,
      "memory(GiB)": 126.99,
      "step": 13600,
      "train_speed(iter/s)": 0.20148
    },
    {
      "acc": 0.76039143,
      "epoch": 0.31753160588518353,
      "grad_norm": 4.4375,
      "learning_rate": 9.571458969657634e-06,
      "loss": 0.85859356,
      "memory(GiB)": 126.99,
      "step": 13610,
      "train_speed(iter/s)": 0.201561
    },
    {
      "acc": 0.79370561,
      "epoch": 0.31776491345747243,
      "grad_norm": 6.375,
      "learning_rate": 9.570693447029662e-06,
      "loss": 0.73161249,
      "memory(GiB)": 126.99,
      "step": 13620,
      "train_speed(iter/s)": 0.20163
    },
    {
      "acc": 0.77307386,
      "epoch": 0.3179982210297613,
      "grad_norm": 4.8125,
      "learning_rate": 9.569927271936528e-06,
      "loss": 0.81770248,
      "memory(GiB)": 126.99,
      "step": 13630,
      "train_speed(iter/s)": 0.201711
    },
    {
      "acc": 0.77128267,
      "epoch": 0.3182315286020502,
      "grad_norm": 6.75,
      "learning_rate": 9.569160444487602e-06,
      "loss": 0.82179031,
      "memory(GiB)": 126.99,
      "step": 13640,
      "train_speed(iter/s)": 0.20179
    },
    {
      "acc": 0.78777056,
      "epoch": 0.31846483617433907,
      "grad_norm": 6.375,
      "learning_rate": 9.56839296479235e-06,
      "loss": 0.74274769,
      "memory(GiB)": 126.99,
      "step": 13650,
      "train_speed(iter/s)": 0.201867
    },
    {
      "acc": 0.79498763,
      "epoch": 0.31869814374662797,
      "grad_norm": 6.5,
      "learning_rate": 9.56762483296033e-06,
      "loss": 0.72544613,
      "memory(GiB)": 126.99,
      "step": 13660,
      "train_speed(iter/s)": 0.201949
    },
    {
      "acc": 0.76066217,
      "epoch": 0.31893145131891687,
      "grad_norm": 5.625,
      "learning_rate": 9.566856049101192e-06,
      "loss": 0.86741734,
      "memory(GiB)": 126.99,
      "step": 13670,
      "train_speed(iter/s)": 0.20202
    },
    {
      "acc": 0.77316117,
      "epoch": 0.31916475889120577,
      "grad_norm": 5.09375,
      "learning_rate": 9.56608661332468e-06,
      "loss": 0.80373688,
      "memory(GiB)": 126.99,
      "step": 13680,
      "train_speed(iter/s)": 0.202091
    },
    {
      "acc": 0.76230974,
      "epoch": 0.31939806646349467,
      "grad_norm": 3.96875,
      "learning_rate": 9.56531652574063e-06,
      "loss": 0.8677124,
      "memory(GiB)": 126.99,
      "step": 13690,
      "train_speed(iter/s)": 0.20217
    },
    {
      "acc": 0.77418733,
      "epoch": 0.31963137403578357,
      "grad_norm": 10.375,
      "learning_rate": 9.564545786458971e-06,
      "loss": 0.81346025,
      "memory(GiB)": 126.99,
      "step": 13700,
      "train_speed(iter/s)": 0.202244
    },
    {
      "acc": 0.77347922,
      "epoch": 0.31986468160807247,
      "grad_norm": 3.984375,
      "learning_rate": 9.563774395589728e-06,
      "loss": 0.79495201,
      "memory(GiB)": 126.99,
      "step": 13710,
      "train_speed(iter/s)": 0.202319
    },
    {
      "acc": 0.76973238,
      "epoch": 0.3200979891803613,
      "grad_norm": 5.28125,
      "learning_rate": 9.563002353243019e-06,
      "loss": 0.84068336,
      "memory(GiB)": 126.99,
      "step": 13720,
      "train_speed(iter/s)": 0.202395
    },
    {
      "acc": 0.74852266,
      "epoch": 0.3203312967526502,
      "grad_norm": 9.8125,
      "learning_rate": 9.562229659529046e-06,
      "loss": 0.91825085,
      "memory(GiB)": 126.99,
      "step": 13730,
      "train_speed(iter/s)": 0.202472
    },
    {
      "acc": 0.77733259,
      "epoch": 0.3205646043249391,
      "grad_norm": 4.71875,
      "learning_rate": 9.561456314558116e-06,
      "loss": 0.79782739,
      "memory(GiB)": 126.99,
      "step": 13740,
      "train_speed(iter/s)": 0.202545
    },
    {
      "acc": 0.75323248,
      "epoch": 0.320797911897228,
      "grad_norm": 8.3125,
      "learning_rate": 9.560682318440619e-06,
      "loss": 0.88182631,
      "memory(GiB)": 126.99,
      "step": 13750,
      "train_speed(iter/s)": 0.202621
    },
    {
      "acc": 0.76970348,
      "epoch": 0.3210312194695169,
      "grad_norm": 6.5625,
      "learning_rate": 9.55990767128705e-06,
      "loss": 0.84244089,
      "memory(GiB)": 126.99,
      "step": 13760,
      "train_speed(iter/s)": 0.202694
    },
    {
      "acc": 0.7580492,
      "epoch": 0.3212645270418058,
      "grad_norm": 5.46875,
      "learning_rate": 9.559132373207984e-06,
      "loss": 0.87328644,
      "memory(GiB)": 126.99,
      "step": 13770,
      "train_speed(iter/s)": 0.202774
    },
    {
      "acc": 0.74232302,
      "epoch": 0.3214978346140947,
      "grad_norm": 5.09375,
      "learning_rate": 9.558356424314095e-06,
      "loss": 0.94130011,
      "memory(GiB)": 126.99,
      "step": 13780,
      "train_speed(iter/s)": 0.202853
    },
    {
      "acc": 0.77524061,
      "epoch": 0.3217311421863836,
      "grad_norm": 5.375,
      "learning_rate": 9.557579824716152e-06,
      "loss": 0.81230278,
      "memory(GiB)": 126.99,
      "step": 13790,
      "train_speed(iter/s)": 0.20293
    },
    {
      "acc": 0.76253085,
      "epoch": 0.3219644497586725,
      "grad_norm": 6.25,
      "learning_rate": 9.556802574525013e-06,
      "loss": 0.86086216,
      "memory(GiB)": 126.99,
      "step": 13800,
      "train_speed(iter/s)": 0.203005
    },
    {
      "acc": 0.7655097,
      "epoch": 0.3221977573309614,
      "grad_norm": 5.125,
      "learning_rate": 9.556024673851629e-06,
      "loss": 0.86178303,
      "memory(GiB)": 126.99,
      "step": 13810,
      "train_speed(iter/s)": 0.203083
    },
    {
      "acc": 0.74736996,
      "epoch": 0.32243106490325024,
      "grad_norm": 6.4375,
      "learning_rate": 9.555246122807047e-06,
      "loss": 0.92181644,
      "memory(GiB)": 126.99,
      "step": 13820,
      "train_speed(iter/s)": 0.20316
    },
    {
      "acc": 0.77784414,
      "epoch": 0.32266437247553914,
      "grad_norm": 3.796875,
      "learning_rate": 9.554466921502405e-06,
      "loss": 0.80609112,
      "memory(GiB)": 126.99,
      "step": 13830,
      "train_speed(iter/s)": 0.203234
    },
    {
      "acc": 0.76699972,
      "epoch": 0.32289768004782804,
      "grad_norm": 6.21875,
      "learning_rate": 9.553687070048934e-06,
      "loss": 0.8321928,
      "memory(GiB)": 126.99,
      "step": 13840,
      "train_speed(iter/s)": 0.203305
    },
    {
      "acc": 0.76653528,
      "epoch": 0.32313098762011694,
      "grad_norm": 9.6875,
      "learning_rate": 9.552906568557953e-06,
      "loss": 0.85399723,
      "memory(GiB)": 126.99,
      "step": 13850,
      "train_speed(iter/s)": 0.203377
    },
    {
      "acc": 0.78444996,
      "epoch": 0.32336429519240584,
      "grad_norm": 4.78125,
      "learning_rate": 9.552125417140885e-06,
      "loss": 0.75022516,
      "memory(GiB)": 126.99,
      "step": 13860,
      "train_speed(iter/s)": 0.203452
    },
    {
      "acc": 0.75712585,
      "epoch": 0.32359760276469474,
      "grad_norm": 5.1875,
      "learning_rate": 9.551343615909236e-06,
      "loss": 0.87373152,
      "memory(GiB)": 126.99,
      "step": 13870,
      "train_speed(iter/s)": 0.203521
    },
    {
      "acc": 0.77796164,
      "epoch": 0.32383091033698364,
      "grad_norm": 5.125,
      "learning_rate": 9.550561164974606e-06,
      "loss": 0.79385543,
      "memory(GiB)": 126.99,
      "step": 13880,
      "train_speed(iter/s)": 0.203594
    },
    {
      "acc": 0.77821608,
      "epoch": 0.32406421790927253,
      "grad_norm": 7.8125,
      "learning_rate": 9.549778064448693e-06,
      "loss": 0.79453735,
      "memory(GiB)": 126.99,
      "step": 13890,
      "train_speed(iter/s)": 0.203671
    },
    {
      "acc": 0.77085085,
      "epoch": 0.32429752548156143,
      "grad_norm": 6.875,
      "learning_rate": 9.548994314443284e-06,
      "loss": 0.83549728,
      "memory(GiB)": 126.99,
      "step": 13900,
      "train_speed(iter/s)": 0.203749
    },
    {
      "acc": 0.7559999,
      "epoch": 0.32453083305385033,
      "grad_norm": 5.65625,
      "learning_rate": 9.548209915070256e-06,
      "loss": 0.88760204,
      "memory(GiB)": 126.99,
      "step": 13910,
      "train_speed(iter/s)": 0.203819
    },
    {
      "acc": 0.7560101,
      "epoch": 0.3247641406261392,
      "grad_norm": 5.65625,
      "learning_rate": 9.547424866441586e-06,
      "loss": 0.8699544,
      "memory(GiB)": 126.99,
      "step": 13920,
      "train_speed(iter/s)": 0.203893
    },
    {
      "acc": 0.77840662,
      "epoch": 0.3249974481984281,
      "grad_norm": 6.53125,
      "learning_rate": 9.546639168669336e-06,
      "loss": 0.80326424,
      "memory(GiB)": 126.99,
      "step": 13930,
      "train_speed(iter/s)": 0.203969
    },
    {
      "acc": 0.77628956,
      "epoch": 0.325230755770717,
      "grad_norm": 5.75,
      "learning_rate": 9.545852821865667e-06,
      "loss": 0.81750832,
      "memory(GiB)": 126.99,
      "step": 13940,
      "train_speed(iter/s)": 0.204043
    },
    {
      "acc": 0.76031036,
      "epoch": 0.32546406334300587,
      "grad_norm": 8.125,
      "learning_rate": 9.545065826142825e-06,
      "loss": 0.87485161,
      "memory(GiB)": 126.99,
      "step": 13950,
      "train_speed(iter/s)": 0.204122
    },
    {
      "acc": 0.76914015,
      "epoch": 0.32569737091529477,
      "grad_norm": 4.15625,
      "learning_rate": 9.544278181613158e-06,
      "loss": 0.82654829,
      "memory(GiB)": 126.99,
      "step": 13960,
      "train_speed(iter/s)": 0.204195
    },
    {
      "acc": 0.77492094,
      "epoch": 0.32593067848758367,
      "grad_norm": 5.1875,
      "learning_rate": 9.543489888389103e-06,
      "loss": 0.83645306,
      "memory(GiB)": 126.99,
      "step": 13970,
      "train_speed(iter/s)": 0.20427
    },
    {
      "acc": 0.77900114,
      "epoch": 0.32616398605987257,
      "grad_norm": 5.28125,
      "learning_rate": 9.542700946583184e-06,
      "loss": 0.79197259,
      "memory(GiB)": 126.99,
      "step": 13980,
      "train_speed(iter/s)": 0.204347
    },
    {
      "acc": 0.77285557,
      "epoch": 0.32639729363216147,
      "grad_norm": 5.5,
      "learning_rate": 9.541911356308025e-06,
      "loss": 0.83121672,
      "memory(GiB)": 126.99,
      "step": 13990,
      "train_speed(iter/s)": 0.204424
    },
    {
      "acc": 0.76080709,
      "epoch": 0.32663060120445037,
      "grad_norm": 4.71875,
      "learning_rate": 9.541121117676339e-06,
      "loss": 0.86064758,
      "memory(GiB)": 126.99,
      "step": 14000,
      "train_speed(iter/s)": 0.204499
    },
    {
      "epoch": 0.32663060120445037,
      "eval_acc": 0.7354247529494964,
      "eval_loss": 0.8388969898223877,
      "eval_runtime": 1270.5843,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.164,
      "step": 14000
    },
    {
      "acc": 0.7774581,
      "epoch": 0.3268639087767392,
      "grad_norm": 4.84375,
      "learning_rate": 9.540330230800935e-06,
      "loss": 0.79988861,
      "memory(GiB)": 126.99,
      "step": 14010,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.78027916,
      "epoch": 0.3270972163490281,
      "grad_norm": 5.875,
      "learning_rate": 9.539538695794708e-06,
      "loss": 0.76487494,
      "memory(GiB)": 126.99,
      "step": 14020,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.77316322,
      "epoch": 0.327330523921317,
      "grad_norm": 5.46875,
      "learning_rate": 9.53874651277065e-06,
      "loss": 0.80076408,
      "memory(GiB)": 126.99,
      "step": 14030,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.79859195,
      "epoch": 0.3275638314936059,
      "grad_norm": 4.53125,
      "learning_rate": 9.537953681841847e-06,
      "loss": 0.72243781,
      "memory(GiB)": 126.99,
      "step": 14040,
      "train_speed(iter/s)": 0.201027
    },
    {
      "acc": 0.77445354,
      "epoch": 0.3277971390658948,
      "grad_norm": 5.5,
      "learning_rate": 9.537160203121474e-06,
      "loss": 0.80300465,
      "memory(GiB)": 126.99,
      "step": 14050,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.78483415,
      "epoch": 0.3280304466381837,
      "grad_norm": 4.71875,
      "learning_rate": 9.536366076722799e-06,
      "loss": 0.77799454,
      "memory(GiB)": 126.99,
      "step": 14060,
      "train_speed(iter/s)": 0.201175
    },
    {
      "acc": 0.76576405,
      "epoch": 0.3282637542104726,
      "grad_norm": 6.125,
      "learning_rate": 9.535571302759184e-06,
      "loss": 0.84655552,
      "memory(GiB)": 126.99,
      "step": 14070,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.76197848,
      "epoch": 0.3284970617827615,
      "grad_norm": 8.1875,
      "learning_rate": 9.534775881344086e-06,
      "loss": 0.8686224,
      "memory(GiB)": 126.99,
      "step": 14080,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.7604651,
      "epoch": 0.3287303693550504,
      "grad_norm": 6.40625,
      "learning_rate": 9.533979812591046e-06,
      "loss": 0.83883457,
      "memory(GiB)": 126.99,
      "step": 14090,
      "train_speed(iter/s)": 0.201389
    },
    {
      "acc": 0.77590809,
      "epoch": 0.3289636769273393,
      "grad_norm": 4.96875,
      "learning_rate": 9.533183096613705e-06,
      "loss": 0.80737171,
      "memory(GiB)": 126.99,
      "step": 14100,
      "train_speed(iter/s)": 0.201463
    },
    {
      "acc": 0.7858882,
      "epoch": 0.32919698449962814,
      "grad_norm": 5.0,
      "learning_rate": 9.532385733525793e-06,
      "loss": 0.7538044,
      "memory(GiB)": 126.99,
      "step": 14110,
      "train_speed(iter/s)": 0.201532
    },
    {
      "acc": 0.78518639,
      "epoch": 0.32943029207191704,
      "grad_norm": 5.84375,
      "learning_rate": 9.531587723441136e-06,
      "loss": 0.77603545,
      "memory(GiB)": 126.99,
      "step": 14120,
      "train_speed(iter/s)": 0.201604
    },
    {
      "acc": 0.7789793,
      "epoch": 0.32966359964420594,
      "grad_norm": 4.5625,
      "learning_rate": 9.530789066473648e-06,
      "loss": 0.78307705,
      "memory(GiB)": 126.99,
      "step": 14130,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.75844407,
      "epoch": 0.32989690721649484,
      "grad_norm": 5.75,
      "learning_rate": 9.529989762737336e-06,
      "loss": 0.85525703,
      "memory(GiB)": 126.99,
      "step": 14140,
      "train_speed(iter/s)": 0.201751
    },
    {
      "acc": 0.76479483,
      "epoch": 0.33013021478878374,
      "grad_norm": 6.8125,
      "learning_rate": 9.529189812346303e-06,
      "loss": 0.85921402,
      "memory(GiB)": 126.99,
      "step": 14150,
      "train_speed(iter/s)": 0.201826
    },
    {
      "acc": 0.77646847,
      "epoch": 0.33036352236107264,
      "grad_norm": 6.0,
      "learning_rate": 9.528389215414737e-06,
      "loss": 0.81476002,
      "memory(GiB)": 126.99,
      "step": 14160,
      "train_speed(iter/s)": 0.201901
    },
    {
      "acc": 0.77587495,
      "epoch": 0.33059682993336154,
      "grad_norm": 6.09375,
      "learning_rate": 9.527587972056929e-06,
      "loss": 0.79444175,
      "memory(GiB)": 126.99,
      "step": 14170,
      "train_speed(iter/s)": 0.201978
    },
    {
      "acc": 0.787714,
      "epoch": 0.33083013750565043,
      "grad_norm": 5.28125,
      "learning_rate": 9.526786082387251e-06,
      "loss": 0.73652754,
      "memory(GiB)": 126.99,
      "step": 14180,
      "train_speed(iter/s)": 0.20205
    },
    {
      "acc": 0.77357411,
      "epoch": 0.33106344507793933,
      "grad_norm": 4.4375,
      "learning_rate": 9.525983546520176e-06,
      "loss": 0.83334599,
      "memory(GiB)": 126.99,
      "step": 14190,
      "train_speed(iter/s)": 0.202123
    },
    {
      "acc": 0.78118382,
      "epoch": 0.3312967526502282,
      "grad_norm": 7.0,
      "learning_rate": 9.525180364570265e-06,
      "loss": 0.80493126,
      "memory(GiB)": 126.99,
      "step": 14200,
      "train_speed(iter/s)": 0.202197
    },
    {
      "acc": 0.75183415,
      "epoch": 0.3315300602225171,
      "grad_norm": 4.5625,
      "learning_rate": 9.52437653665217e-06,
      "loss": 0.91054926,
      "memory(GiB)": 126.99,
      "step": 14210,
      "train_speed(iter/s)": 0.202276
    },
    {
      "acc": 0.78139944,
      "epoch": 0.331763367794806,
      "grad_norm": 9.0,
      "learning_rate": 9.52357206288064e-06,
      "loss": 0.79247675,
      "memory(GiB)": 126.99,
      "step": 14220,
      "train_speed(iter/s)": 0.20235
    },
    {
      "acc": 0.75987701,
      "epoch": 0.3319966753670949,
      "grad_norm": 6.0625,
      "learning_rate": 9.522766943370512e-06,
      "loss": 0.87182941,
      "memory(GiB)": 126.99,
      "step": 14230,
      "train_speed(iter/s)": 0.202422
    },
    {
      "acc": 0.77394571,
      "epoch": 0.33222998293938377,
      "grad_norm": 7.53125,
      "learning_rate": 9.521961178236716e-06,
      "loss": 0.81149826,
      "memory(GiB)": 126.99,
      "step": 14240,
      "train_speed(iter/s)": 0.202493
    },
    {
      "acc": 0.77338943,
      "epoch": 0.33246329051167267,
      "grad_norm": 4.3125,
      "learning_rate": 9.521154767594276e-06,
      "loss": 0.81733627,
      "memory(GiB)": 126.99,
      "step": 14250,
      "train_speed(iter/s)": 0.202568
    },
    {
      "acc": 0.75781255,
      "epoch": 0.33269659808396157,
      "grad_norm": 6.9375,
      "learning_rate": 9.520347711558306e-06,
      "loss": 0.88546219,
      "memory(GiB)": 126.99,
      "step": 14260,
      "train_speed(iter/s)": 0.202638
    },
    {
      "acc": 0.76396008,
      "epoch": 0.33292990565625047,
      "grad_norm": 5.53125,
      "learning_rate": 9.519540010244013e-06,
      "loss": 0.86021147,
      "memory(GiB)": 126.99,
      "step": 14270,
      "train_speed(iter/s)": 0.202717
    },
    {
      "acc": 0.76865139,
      "epoch": 0.33316321322853937,
      "grad_norm": 4.78125,
      "learning_rate": 9.518731663766697e-06,
      "loss": 0.8223979,
      "memory(GiB)": 126.99,
      "step": 14280,
      "train_speed(iter/s)": 0.202789
    },
    {
      "acc": 0.78064613,
      "epoch": 0.33339652080082827,
      "grad_norm": 5.71875,
      "learning_rate": 9.517922672241748e-06,
      "loss": 0.78749752,
      "memory(GiB)": 126.99,
      "step": 14290,
      "train_speed(iter/s)": 0.202864
    },
    {
      "acc": 0.79933729,
      "epoch": 0.3336298283731171,
      "grad_norm": 7.34375,
      "learning_rate": 9.517113035784651e-06,
      "loss": 0.7244029,
      "memory(GiB)": 126.99,
      "step": 14300,
      "train_speed(iter/s)": 0.20294
    },
    {
      "acc": 0.78056884,
      "epoch": 0.333863135945406,
      "grad_norm": 7.25,
      "learning_rate": 9.51630275451098e-06,
      "loss": 0.79498949,
      "memory(GiB)": 126.99,
      "step": 14310,
      "train_speed(iter/s)": 0.203011
    },
    {
      "acc": 0.7816524,
      "epoch": 0.3340964435176949,
      "grad_norm": 6.0625,
      "learning_rate": 9.515491828536403e-06,
      "loss": 0.77081575,
      "memory(GiB)": 126.99,
      "step": 14320,
      "train_speed(iter/s)": 0.203084
    },
    {
      "acc": 0.76975269,
      "epoch": 0.3343297510899838,
      "grad_norm": 5.0625,
      "learning_rate": 9.51468025797668e-06,
      "loss": 0.8435153,
      "memory(GiB)": 126.99,
      "step": 14330,
      "train_speed(iter/s)": 0.203157
    },
    {
      "acc": 0.79847164,
      "epoch": 0.3345630586622727,
      "grad_norm": 6.90625,
      "learning_rate": 9.51386804294766e-06,
      "loss": 0.7151453,
      "memory(GiB)": 126.99,
      "step": 14340,
      "train_speed(iter/s)": 0.20323
    },
    {
      "acc": 0.78779755,
      "epoch": 0.3347963662345616,
      "grad_norm": 5.21875,
      "learning_rate": 9.51305518356529e-06,
      "loss": 0.75147953,
      "memory(GiB)": 126.99,
      "step": 14350,
      "train_speed(iter/s)": 0.2033
    },
    {
      "acc": 0.7548315,
      "epoch": 0.3350296738068505,
      "grad_norm": 7.09375,
      "learning_rate": 9.512241679945602e-06,
      "loss": 0.89756851,
      "memory(GiB)": 126.99,
      "step": 14360,
      "train_speed(iter/s)": 0.203372
    },
    {
      "acc": 0.7766715,
      "epoch": 0.3352629813791394,
      "grad_norm": 6.3125,
      "learning_rate": 9.511427532204725e-06,
      "loss": 0.79562111,
      "memory(GiB)": 126.99,
      "step": 14370,
      "train_speed(iter/s)": 0.20344
    },
    {
      "acc": 0.77851067,
      "epoch": 0.3354962889514283,
      "grad_norm": 6.46875,
      "learning_rate": 9.51061274045888e-06,
      "loss": 0.7912365,
      "memory(GiB)": 126.99,
      "step": 14380,
      "train_speed(iter/s)": 0.203516
    },
    {
      "acc": 0.76406937,
      "epoch": 0.3357295965237172,
      "grad_norm": 5.46875,
      "learning_rate": 9.509797304824376e-06,
      "loss": 0.87257404,
      "memory(GiB)": 126.99,
      "step": 14390,
      "train_speed(iter/s)": 0.203587
    },
    {
      "acc": 0.77162414,
      "epoch": 0.33596290409600604,
      "grad_norm": 6.21875,
      "learning_rate": 9.508981225417615e-06,
      "loss": 0.81675949,
      "memory(GiB)": 126.99,
      "step": 14400,
      "train_speed(iter/s)": 0.203654
    },
    {
      "acc": 0.76914759,
      "epoch": 0.33619621166829494,
      "grad_norm": 5.53125,
      "learning_rate": 9.508164502355095e-06,
      "loss": 0.83779526,
      "memory(GiB)": 126.99,
      "step": 14410,
      "train_speed(iter/s)": 0.203724
    },
    {
      "acc": 0.77464485,
      "epoch": 0.33642951924058384,
      "grad_norm": 4.53125,
      "learning_rate": 9.507347135753403e-06,
      "loss": 0.79855938,
      "memory(GiB)": 126.99,
      "step": 14420,
      "train_speed(iter/s)": 0.203795
    },
    {
      "acc": 0.77766848,
      "epoch": 0.33666282681287274,
      "grad_norm": 7.875,
      "learning_rate": 9.506529125729216e-06,
      "loss": 0.80243073,
      "memory(GiB)": 126.99,
      "step": 14430,
      "train_speed(iter/s)": 0.203867
    },
    {
      "acc": 0.77303267,
      "epoch": 0.33689613438516164,
      "grad_norm": 5.90625,
      "learning_rate": 9.505710472399306e-06,
      "loss": 0.79739962,
      "memory(GiB)": 126.99,
      "step": 14440,
      "train_speed(iter/s)": 0.203937
    },
    {
      "acc": 0.76884651,
      "epoch": 0.33712944195745054,
      "grad_norm": 4.78125,
      "learning_rate": 9.504891175880533e-06,
      "loss": 0.82474422,
      "memory(GiB)": 126.99,
      "step": 14450,
      "train_speed(iter/s)": 0.204006
    },
    {
      "acc": 0.76067557,
      "epoch": 0.33736274952973944,
      "grad_norm": 6.1875,
      "learning_rate": 9.504071236289856e-06,
      "loss": 0.87364044,
      "memory(GiB)": 126.99,
      "step": 14460,
      "train_speed(iter/s)": 0.204078
    },
    {
      "acc": 0.77003412,
      "epoch": 0.33759605710202834,
      "grad_norm": 5.3125,
      "learning_rate": 9.503250653744316e-06,
      "loss": 0.81374006,
      "memory(GiB)": 126.99,
      "step": 14470,
      "train_speed(iter/s)": 0.204147
    },
    {
      "acc": 0.77818384,
      "epoch": 0.33782936467431723,
      "grad_norm": 5.4375,
      "learning_rate": 9.502429428361055e-06,
      "loss": 0.81029873,
      "memory(GiB)": 126.99,
      "step": 14480,
      "train_speed(iter/s)": 0.204217
    },
    {
      "acc": 0.74643278,
      "epoch": 0.3380626722466061,
      "grad_norm": 5.4375,
      "learning_rate": 9.5016075602573e-06,
      "loss": 0.92115335,
      "memory(GiB)": 126.99,
      "step": 14490,
      "train_speed(iter/s)": 0.20429
    },
    {
      "acc": 0.76585412,
      "epoch": 0.338295979818895,
      "grad_norm": 9.25,
      "learning_rate": 9.500785049550373e-06,
      "loss": 0.83918991,
      "memory(GiB)": 126.99,
      "step": 14500,
      "train_speed(iter/s)": 0.204369
    },
    {
      "epoch": 0.338295979818895,
      "eval_acc": 0.7356140434320767,
      "eval_loss": 0.8385112881660461,
      "eval_runtime": 1269.8549,
      "eval_samples_per_second": 28.343,
      "eval_steps_per_second": 14.172,
      "step": 14500
    },
    {
      "acc": 0.77105675,
      "epoch": 0.3385292873911839,
      "grad_norm": 4.25,
      "learning_rate": 9.49996189635769e-06,
      "loss": 0.81464977,
      "memory(GiB)": 126.99,
      "step": 14510,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.76173067,
      "epoch": 0.3387625949634728,
      "grad_norm": 4.75,
      "learning_rate": 9.499138100796752e-06,
      "loss": 0.85862141,
      "memory(GiB)": 126.99,
      "step": 14520,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.76991301,
      "epoch": 0.3389959025357617,
      "grad_norm": 4.5,
      "learning_rate": 9.498313662985159e-06,
      "loss": 0.83255396,
      "memory(GiB)": 126.99,
      "step": 14530,
      "train_speed(iter/s)": 0.200942
    },
    {
      "acc": 0.77943153,
      "epoch": 0.33922921010805057,
      "grad_norm": 5.21875,
      "learning_rate": 9.497488583040595e-06,
      "loss": 0.79724607,
      "memory(GiB)": 126.99,
      "step": 14540,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.76977105,
      "epoch": 0.33946251768033947,
      "grad_norm": 6.40625,
      "learning_rate": 9.496662861080842e-06,
      "loss": 0.83135042,
      "memory(GiB)": 126.99,
      "step": 14550,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.78565388,
      "epoch": 0.33969582525262837,
      "grad_norm": 7.28125,
      "learning_rate": 9.495836497223775e-06,
      "loss": 0.74578938,
      "memory(GiB)": 126.99,
      "step": 14560,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.75958099,
      "epoch": 0.33992913282491727,
      "grad_norm": 5.78125,
      "learning_rate": 9.49500949158735e-06,
      "loss": 0.88202744,
      "memory(GiB)": 126.99,
      "step": 14570,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.78799295,
      "epoch": 0.34016244039720617,
      "grad_norm": 5.90625,
      "learning_rate": 9.494181844289629e-06,
      "loss": 0.75782499,
      "memory(GiB)": 126.99,
      "step": 14580,
      "train_speed(iter/s)": 0.201295
    },
    {
      "acc": 0.78360558,
      "epoch": 0.340395747969495,
      "grad_norm": 4.78125,
      "learning_rate": 9.493353555448754e-06,
      "loss": 0.75555897,
      "memory(GiB)": 135.77,
      "step": 14590,
      "train_speed(iter/s)": 0.201362
    },
    {
      "acc": 0.78057513,
      "epoch": 0.3406290555417839,
      "grad_norm": 6.8125,
      "learning_rate": 9.492524625182965e-06,
      "loss": 0.76722956,
      "memory(GiB)": 135.77,
      "step": 14600,
      "train_speed(iter/s)": 0.201432
    },
    {
      "acc": 0.77534466,
      "epoch": 0.3408623631140728,
      "grad_norm": 4.8125,
      "learning_rate": 9.49169505361059e-06,
      "loss": 0.8156085,
      "memory(GiB)": 135.77,
      "step": 14610,
      "train_speed(iter/s)": 0.201504
    },
    {
      "acc": 0.7402894,
      "epoch": 0.3410956706863617,
      "grad_norm": 10.5,
      "learning_rate": 9.490864840850051e-06,
      "loss": 0.96359291,
      "memory(GiB)": 135.77,
      "step": 14620,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.7706079,
      "epoch": 0.3413289782586506,
      "grad_norm": 7.40625,
      "learning_rate": 9.490033987019862e-06,
      "loss": 0.83333149,
      "memory(GiB)": 135.77,
      "step": 14630,
      "train_speed(iter/s)": 0.201651
    },
    {
      "acc": 0.76909857,
      "epoch": 0.3415622858309395,
      "grad_norm": 4.6875,
      "learning_rate": 9.489202492238624e-06,
      "loss": 0.82511196,
      "memory(GiB)": 135.77,
      "step": 14640,
      "train_speed(iter/s)": 0.201726
    },
    {
      "acc": 0.75274253,
      "epoch": 0.3417955934032284,
      "grad_norm": 5.3125,
      "learning_rate": 9.488370356625035e-06,
      "loss": 0.8952776,
      "memory(GiB)": 135.77,
      "step": 14650,
      "train_speed(iter/s)": 0.2018
    },
    {
      "acc": 0.78073397,
      "epoch": 0.3420289009755173,
      "grad_norm": 12.375,
      "learning_rate": 9.487537580297881e-06,
      "loss": 0.7810813,
      "memory(GiB)": 135.77,
      "step": 14660,
      "train_speed(iter/s)": 0.201871
    },
    {
      "acc": 0.75856628,
      "epoch": 0.3422622085478062,
      "grad_norm": 4.5625,
      "learning_rate": 9.486704163376041e-06,
      "loss": 0.89207325,
      "memory(GiB)": 135.77,
      "step": 14670,
      "train_speed(iter/s)": 0.201942
    },
    {
      "acc": 0.77476802,
      "epoch": 0.3424955161200951,
      "grad_norm": 6.28125,
      "learning_rate": 9.485870105978487e-06,
      "loss": 0.78234882,
      "memory(GiB)": 135.77,
      "step": 14680,
      "train_speed(iter/s)": 0.202016
    },
    {
      "acc": 0.76214299,
      "epoch": 0.34272882369238394,
      "grad_norm": 4.78125,
      "learning_rate": 9.485035408224277e-06,
      "loss": 0.84753475,
      "memory(GiB)": 135.77,
      "step": 14690,
      "train_speed(iter/s)": 0.202093
    },
    {
      "acc": 0.77669897,
      "epoch": 0.34296213126467284,
      "grad_norm": 7.3125,
      "learning_rate": 9.484200070232565e-06,
      "loss": 0.82079849,
      "memory(GiB)": 135.77,
      "step": 14700,
      "train_speed(iter/s)": 0.202169
    },
    {
      "acc": 0.76222053,
      "epoch": 0.34319543883696174,
      "grad_norm": 5.125,
      "learning_rate": 9.483364092122595e-06,
      "loss": 0.86403046,
      "memory(GiB)": 135.77,
      "step": 14710,
      "train_speed(iter/s)": 0.20224
    },
    {
      "acc": 0.77041807,
      "epoch": 0.34342874640925064,
      "grad_norm": 6.03125,
      "learning_rate": 9.482527474013705e-06,
      "loss": 0.84138517,
      "memory(GiB)": 135.77,
      "step": 14720,
      "train_speed(iter/s)": 0.202309
    },
    {
      "acc": 0.769876,
      "epoch": 0.34366205398153954,
      "grad_norm": 8.125,
      "learning_rate": 9.481690216025321e-06,
      "loss": 0.83939104,
      "memory(GiB)": 135.77,
      "step": 14730,
      "train_speed(iter/s)": 0.202379
    },
    {
      "acc": 0.76854458,
      "epoch": 0.34389536155382844,
      "grad_norm": 7.75,
      "learning_rate": 9.480852318276958e-06,
      "loss": 0.8580862,
      "memory(GiB)": 135.77,
      "step": 14740,
      "train_speed(iter/s)": 0.202448
    },
    {
      "acc": 0.79126053,
      "epoch": 0.34412866912611734,
      "grad_norm": 7.03125,
      "learning_rate": 9.48001378088823e-06,
      "loss": 0.74607038,
      "memory(GiB)": 135.77,
      "step": 14750,
      "train_speed(iter/s)": 0.202516
    },
    {
      "acc": 0.77793612,
      "epoch": 0.34436197669840624,
      "grad_norm": 5.53125,
      "learning_rate": 9.479174603978836e-06,
      "loss": 0.7889802,
      "memory(GiB)": 135.77,
      "step": 14760,
      "train_speed(iter/s)": 0.202585
    },
    {
      "acc": 0.78102379,
      "epoch": 0.34459528427069513,
      "grad_norm": 5.34375,
      "learning_rate": 9.478334787668569e-06,
      "loss": 0.79738607,
      "memory(GiB)": 135.77,
      "step": 14770,
      "train_speed(iter/s)": 0.202653
    },
    {
      "acc": 0.77394485,
      "epoch": 0.344828591842984,
      "grad_norm": 8.1875,
      "learning_rate": 9.477494332077311e-06,
      "loss": 0.81590576,
      "memory(GiB)": 135.77,
      "step": 14780,
      "train_speed(iter/s)": 0.202721
    },
    {
      "acc": 0.76899199,
      "epoch": 0.3450618994152729,
      "grad_norm": 4.78125,
      "learning_rate": 9.476653237325037e-06,
      "loss": 0.84678192,
      "memory(GiB)": 135.77,
      "step": 14790,
      "train_speed(iter/s)": 0.202792
    },
    {
      "acc": 0.77013807,
      "epoch": 0.3452952069875618,
      "grad_norm": 5.21875,
      "learning_rate": 9.475811503531815e-06,
      "loss": 0.82642403,
      "memory(GiB)": 135.77,
      "step": 14800,
      "train_speed(iter/s)": 0.202859
    },
    {
      "acc": 0.7706193,
      "epoch": 0.3455285145598507,
      "grad_norm": 6.15625,
      "learning_rate": 9.474969130817801e-06,
      "loss": 0.82289953,
      "memory(GiB)": 135.77,
      "step": 14810,
      "train_speed(iter/s)": 0.20293
    },
    {
      "acc": 0.76709123,
      "epoch": 0.3457618221321396,
      "grad_norm": 5.34375,
      "learning_rate": 9.474126119303245e-06,
      "loss": 0.85336533,
      "memory(GiB)": 135.77,
      "step": 14820,
      "train_speed(iter/s)": 0.202989
    },
    {
      "acc": 0.77197552,
      "epoch": 0.3459951297044285,
      "grad_norm": 7.625,
      "learning_rate": 9.473282469108483e-06,
      "loss": 0.81618519,
      "memory(GiB)": 135.77,
      "step": 14830,
      "train_speed(iter/s)": 0.203057
    },
    {
      "acc": 0.76123796,
      "epoch": 0.34622843727671737,
      "grad_norm": 4.625,
      "learning_rate": 9.472438180353948e-06,
      "loss": 0.87118511,
      "memory(GiB)": 135.77,
      "step": 14840,
      "train_speed(iter/s)": 0.203124
    },
    {
      "acc": 0.79290915,
      "epoch": 0.34646174484900627,
      "grad_norm": 5.125,
      "learning_rate": 9.471593253160162e-06,
      "loss": 0.73074284,
      "memory(GiB)": 135.77,
      "step": 14850,
      "train_speed(iter/s)": 0.203196
    },
    {
      "acc": 0.76475649,
      "epoch": 0.34669505242129517,
      "grad_norm": 5.15625,
      "learning_rate": 9.470747687647741e-06,
      "loss": 0.85227308,
      "memory(GiB)": 135.77,
      "step": 14860,
      "train_speed(iter/s)": 0.203268
    },
    {
      "acc": 0.7734736,
      "epoch": 0.34692835999358407,
      "grad_norm": 7.03125,
      "learning_rate": 9.469901483937384e-06,
      "loss": 0.79763713,
      "memory(GiB)": 135.77,
      "step": 14870,
      "train_speed(iter/s)": 0.203343
    },
    {
      "acc": 0.79150887,
      "epoch": 0.3471616675658729,
      "grad_norm": 4.8125,
      "learning_rate": 9.469054642149889e-06,
      "loss": 0.73478098,
      "memory(GiB)": 135.77,
      "step": 14880,
      "train_speed(iter/s)": 0.203411
    },
    {
      "acc": 0.77741909,
      "epoch": 0.3473949751381618,
      "grad_norm": 6.25,
      "learning_rate": 9.468207162406143e-06,
      "loss": 0.81160421,
      "memory(GiB)": 135.77,
      "step": 14890,
      "train_speed(iter/s)": 0.203483
    },
    {
      "acc": 0.76978407,
      "epoch": 0.3476282827104507,
      "grad_norm": 5.0,
      "learning_rate": 9.46735904482712e-06,
      "loss": 0.81876183,
      "memory(GiB)": 135.77,
      "step": 14900,
      "train_speed(iter/s)": 0.203552
    },
    {
      "acc": 0.77738295,
      "epoch": 0.3478615902827396,
      "grad_norm": 6.3125,
      "learning_rate": 9.466510289533894e-06,
      "loss": 0.79115467,
      "memory(GiB)": 135.77,
      "step": 14910,
      "train_speed(iter/s)": 0.203622
    },
    {
      "acc": 0.78486333,
      "epoch": 0.3480948978550285,
      "grad_norm": 5.96875,
      "learning_rate": 9.46566089664762e-06,
      "loss": 0.74208999,
      "memory(GiB)": 135.77,
      "step": 14920,
      "train_speed(iter/s)": 0.203692
    },
    {
      "acc": 0.78259773,
      "epoch": 0.3483282054273174,
      "grad_norm": 8.75,
      "learning_rate": 9.46481086628955e-06,
      "loss": 0.79701777,
      "memory(GiB)": 135.77,
      "step": 14930,
      "train_speed(iter/s)": 0.20376
    },
    {
      "acc": 0.77444029,
      "epoch": 0.3485615129996063,
      "grad_norm": 5.53125,
      "learning_rate": 9.463960198581028e-06,
      "loss": 0.82572346,
      "memory(GiB)": 135.77,
      "step": 14940,
      "train_speed(iter/s)": 0.203828
    },
    {
      "acc": 0.76832108,
      "epoch": 0.3487948205718952,
      "grad_norm": 7.4375,
      "learning_rate": 9.463108893643483e-06,
      "loss": 0.81358547,
      "memory(GiB)": 135.77,
      "step": 14950,
      "train_speed(iter/s)": 0.203898
    },
    {
      "acc": 0.76551371,
      "epoch": 0.3490281281441841,
      "grad_norm": 6.84375,
      "learning_rate": 9.46225695159844e-06,
      "loss": 0.86378431,
      "memory(GiB)": 135.77,
      "step": 14960,
      "train_speed(iter/s)": 0.203957
    },
    {
      "acc": 0.79054885,
      "epoch": 0.34926143571647295,
      "grad_norm": 4.71875,
      "learning_rate": 9.461404372567513e-06,
      "loss": 0.75698428,
      "memory(GiB)": 135.77,
      "step": 14970,
      "train_speed(iter/s)": 0.204027
    },
    {
      "acc": 0.76183825,
      "epoch": 0.34949474328876184,
      "grad_norm": 4.84375,
      "learning_rate": 9.460551156672408e-06,
      "loss": 0.87834749,
      "memory(GiB)": 135.77,
      "step": 14980,
      "train_speed(iter/s)": 0.204097
    },
    {
      "acc": 0.75772481,
      "epoch": 0.34972805086105074,
      "grad_norm": 5.0,
      "learning_rate": 9.459697304034923e-06,
      "loss": 0.87366276,
      "memory(GiB)": 135.77,
      "step": 14990,
      "train_speed(iter/s)": 0.204166
    },
    {
      "acc": 0.74876041,
      "epoch": 0.34996135843333964,
      "grad_norm": 5.53125,
      "learning_rate": 9.458842814776941e-06,
      "loss": 0.91464996,
      "memory(GiB)": 135.77,
      "step": 15000,
      "train_speed(iter/s)": 0.204229
    },
    {
      "epoch": 0.34996135843333964,
      "eval_acc": 0.7360832660327197,
      "eval_loss": 0.8364920616149902,
      "eval_runtime": 1270.2256,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.168,
      "step": 15000
    },
    {
      "acc": 0.78980742,
      "epoch": 0.35019466600562854,
      "grad_norm": 5.09375,
      "learning_rate": 9.457987689020444e-06,
      "loss": 0.74719791,
      "memory(GiB)": 135.77,
      "step": 15010,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.76756434,
      "epoch": 0.35042797357791744,
      "grad_norm": 6.4375,
      "learning_rate": 9.457131926887498e-06,
      "loss": 0.84798584,
      "memory(GiB)": 135.77,
      "step": 15020,
      "train_speed(iter/s)": 0.200849
    },
    {
      "acc": 0.7941637,
      "epoch": 0.35066128115020634,
      "grad_norm": 6.28125,
      "learning_rate": 9.456275528500264e-06,
      "loss": 0.75351415,
      "memory(GiB)": 135.77,
      "step": 15030,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.76890516,
      "epoch": 0.35089458872249524,
      "grad_norm": 5.53125,
      "learning_rate": 9.455418493980996e-06,
      "loss": 0.84603033,
      "memory(GiB)": 135.77,
      "step": 15040,
      "train_speed(iter/s)": 0.200988
    },
    {
      "acc": 0.76987128,
      "epoch": 0.35112789629478414,
      "grad_norm": 8.5625,
      "learning_rate": 9.454560823452031e-06,
      "loss": 0.82527428,
      "memory(GiB)": 135.77,
      "step": 15050,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.76635885,
      "epoch": 0.35136120386707304,
      "grad_norm": 6.21875,
      "learning_rate": 9.4537025170358e-06,
      "loss": 0.84537125,
      "memory(GiB)": 135.77,
      "step": 15060,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.78066635,
      "epoch": 0.3515945114393619,
      "grad_norm": 4.71875,
      "learning_rate": 9.45284357485483e-06,
      "loss": 0.79466629,
      "memory(GiB)": 135.77,
      "step": 15070,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.77449002,
      "epoch": 0.3518278190116508,
      "grad_norm": 4.65625,
      "learning_rate": 9.451983997031736e-06,
      "loss": 0.80630569,
      "memory(GiB)": 135.77,
      "step": 15080,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.77382059,
      "epoch": 0.3520611265839397,
      "grad_norm": 5.1875,
      "learning_rate": 9.451123783689216e-06,
      "loss": 0.81937408,
      "memory(GiB)": 135.77,
      "step": 15090,
      "train_speed(iter/s)": 0.201336
    },
    {
      "acc": 0.75354958,
      "epoch": 0.3522944341562286,
      "grad_norm": 6.28125,
      "learning_rate": 9.450262934950069e-06,
      "loss": 0.88799534,
      "memory(GiB)": 135.77,
      "step": 15100,
      "train_speed(iter/s)": 0.201403
    },
    {
      "acc": 0.76095815,
      "epoch": 0.3525277417285175,
      "grad_norm": 6.25,
      "learning_rate": 9.449401450937184e-06,
      "loss": 0.88658123,
      "memory(GiB)": 135.77,
      "step": 15110,
      "train_speed(iter/s)": 0.201471
    },
    {
      "acc": 0.77167931,
      "epoch": 0.3527610493008064,
      "grad_norm": 5.53125,
      "learning_rate": 9.448539331773532e-06,
      "loss": 0.83307276,
      "memory(GiB)": 135.77,
      "step": 15120,
      "train_speed(iter/s)": 0.201538
    },
    {
      "acc": 0.77130241,
      "epoch": 0.35299435687309527,
      "grad_norm": 6.75,
      "learning_rate": 9.447676577582184e-06,
      "loss": 0.8277441,
      "memory(GiB)": 135.77,
      "step": 15130,
      "train_speed(iter/s)": 0.201605
    },
    {
      "acc": 0.76398301,
      "epoch": 0.35322766444538417,
      "grad_norm": 5.5,
      "learning_rate": 9.446813188486294e-06,
      "loss": 0.85096407,
      "memory(GiB)": 135.77,
      "step": 15140,
      "train_speed(iter/s)": 0.201673
    },
    {
      "acc": 0.76413918,
      "epoch": 0.35346097201767307,
      "grad_norm": 7.375,
      "learning_rate": 9.445949164609116e-06,
      "loss": 0.87346344,
      "memory(GiB)": 135.77,
      "step": 15150,
      "train_speed(iter/s)": 0.201742
    },
    {
      "acc": 0.78450708,
      "epoch": 0.35369427958996197,
      "grad_norm": 5.125,
      "learning_rate": 9.445084506073985e-06,
      "loss": 0.77033629,
      "memory(GiB)": 135.77,
      "step": 15160,
      "train_speed(iter/s)": 0.201812
    },
    {
      "acc": 0.76229258,
      "epoch": 0.3539275871622508,
      "grad_norm": 7.3125,
      "learning_rate": 9.444219213004333e-06,
      "loss": 0.86626339,
      "memory(GiB)": 135.77,
      "step": 15170,
      "train_speed(iter/s)": 0.201878
    },
    {
      "acc": 0.76085272,
      "epoch": 0.3541608947345397,
      "grad_norm": 4.71875,
      "learning_rate": 9.443353285523678e-06,
      "loss": 0.86795692,
      "memory(GiB)": 135.77,
      "step": 15180,
      "train_speed(iter/s)": 0.20195
    },
    {
      "acc": 0.76711388,
      "epoch": 0.3543942023068286,
      "grad_norm": 5.5,
      "learning_rate": 9.442486723755633e-06,
      "loss": 0.82695503,
      "memory(GiB)": 135.77,
      "step": 15190,
      "train_speed(iter/s)": 0.202021
    },
    {
      "acc": 0.76617556,
      "epoch": 0.3546275098791175,
      "grad_norm": 7.375,
      "learning_rate": 9.4416195278239e-06,
      "loss": 0.82697554,
      "memory(GiB)": 135.77,
      "step": 15200,
      "train_speed(iter/s)": 0.202089
    },
    {
      "acc": 0.76866422,
      "epoch": 0.3548608174514064,
      "grad_norm": 4.28125,
      "learning_rate": 9.440751697852268e-06,
      "loss": 0.84350128,
      "memory(GiB)": 135.77,
      "step": 15210,
      "train_speed(iter/s)": 0.202153
    },
    {
      "acc": 0.78142757,
      "epoch": 0.3550941250236953,
      "grad_norm": 9.0625,
      "learning_rate": 9.439883233964621e-06,
      "loss": 0.79565501,
      "memory(GiB)": 135.77,
      "step": 15220,
      "train_speed(iter/s)": 0.202225
    },
    {
      "acc": 0.77310419,
      "epoch": 0.3553274325959842,
      "grad_norm": 4.875,
      "learning_rate": 9.439014136284934e-06,
      "loss": 0.83620567,
      "memory(GiB)": 135.77,
      "step": 15230,
      "train_speed(iter/s)": 0.202294
    },
    {
      "acc": 0.78746805,
      "epoch": 0.3555607401682731,
      "grad_norm": 4.0625,
      "learning_rate": 9.438144404937266e-06,
      "loss": 0.75883884,
      "memory(GiB)": 135.77,
      "step": 15240,
      "train_speed(iter/s)": 0.202366
    },
    {
      "acc": 0.77289262,
      "epoch": 0.355794047740562,
      "grad_norm": 4.53125,
      "learning_rate": 9.437274040045775e-06,
      "loss": 0.85325108,
      "memory(GiB)": 135.77,
      "step": 15250,
      "train_speed(iter/s)": 0.202433
    },
    {
      "acc": 0.79338937,
      "epoch": 0.35602735531285085,
      "grad_norm": 5.53125,
      "learning_rate": 9.436403041734704e-06,
      "loss": 0.74687109,
      "memory(GiB)": 135.77,
      "step": 15260,
      "train_speed(iter/s)": 0.202505
    },
    {
      "acc": 0.77933197,
      "epoch": 0.35626066288513975,
      "grad_norm": 4.5,
      "learning_rate": 9.435531410128387e-06,
      "loss": 0.78443747,
      "memory(GiB)": 135.77,
      "step": 15270,
      "train_speed(iter/s)": 0.202572
    },
    {
      "acc": 0.76276693,
      "epoch": 0.35649397045742864,
      "grad_norm": 4.125,
      "learning_rate": 9.434659145351251e-06,
      "loss": 0.85257626,
      "memory(GiB)": 135.77,
      "step": 15280,
      "train_speed(iter/s)": 0.202643
    },
    {
      "acc": 0.77083325,
      "epoch": 0.35672727802971754,
      "grad_norm": 5.4375,
      "learning_rate": 9.433786247527809e-06,
      "loss": 0.81012554,
      "memory(GiB)": 135.77,
      "step": 15290,
      "train_speed(iter/s)": 0.202711
    },
    {
      "acc": 0.76903772,
      "epoch": 0.35696058560200644,
      "grad_norm": 4.625,
      "learning_rate": 9.432912716782667e-06,
      "loss": 0.82582378,
      "memory(GiB)": 135.77,
      "step": 15300,
      "train_speed(iter/s)": 0.202778
    },
    {
      "acc": 0.77473183,
      "epoch": 0.35719389317429534,
      "grad_norm": 5.28125,
      "learning_rate": 9.432038553240526e-06,
      "loss": 0.82623148,
      "memory(GiB)": 135.77,
      "step": 15310,
      "train_speed(iter/s)": 0.202843
    },
    {
      "acc": 0.77468882,
      "epoch": 0.35742720074658424,
      "grad_norm": 5.625,
      "learning_rate": 9.431163757026167e-06,
      "loss": 0.79663391,
      "memory(GiB)": 135.77,
      "step": 15320,
      "train_speed(iter/s)": 0.20291
    },
    {
      "acc": 0.7720438,
      "epoch": 0.35766050831887314,
      "grad_norm": 5.9375,
      "learning_rate": 9.430288328264467e-06,
      "loss": 0.8156147,
      "memory(GiB)": 135.77,
      "step": 15330,
      "train_speed(iter/s)": 0.202979
    },
    {
      "acc": 0.77816076,
      "epoch": 0.35789381589116204,
      "grad_norm": 5.59375,
      "learning_rate": 9.429412267080397e-06,
      "loss": 0.78154631,
      "memory(GiB)": 135.77,
      "step": 15340,
      "train_speed(iter/s)": 0.203047
    },
    {
      "acc": 0.76980267,
      "epoch": 0.35812712346345094,
      "grad_norm": 5.8125,
      "learning_rate": 9.428535573599013e-06,
      "loss": 0.83480797,
      "memory(GiB)": 135.77,
      "step": 15350,
      "train_speed(iter/s)": 0.203121
    },
    {
      "acc": 0.76492939,
      "epoch": 0.3583604310357398,
      "grad_norm": 4.65625,
      "learning_rate": 9.427658247945463e-06,
      "loss": 0.8619091,
      "memory(GiB)": 135.77,
      "step": 15360,
      "train_speed(iter/s)": 0.203181
    },
    {
      "acc": 0.75917945,
      "epoch": 0.3585937386080287,
      "grad_norm": 9.375,
      "learning_rate": 9.426780290244983e-06,
      "loss": 0.89382544,
      "memory(GiB)": 135.77,
      "step": 15370,
      "train_speed(iter/s)": 0.203252
    },
    {
      "acc": 0.76189117,
      "epoch": 0.3588270461803176,
      "grad_norm": 8.9375,
      "learning_rate": 9.425901700622904e-06,
      "loss": 0.84197168,
      "memory(GiB)": 135.77,
      "step": 15380,
      "train_speed(iter/s)": 0.20332
    },
    {
      "acc": 0.77091656,
      "epoch": 0.3590603537526065,
      "grad_norm": 7.6875,
      "learning_rate": 9.42502247920464e-06,
      "loss": 0.8417284,
      "memory(GiB)": 135.77,
      "step": 15390,
      "train_speed(iter/s)": 0.20339
    },
    {
      "acc": 0.77439508,
      "epoch": 0.3592936613248954,
      "grad_norm": 4.78125,
      "learning_rate": 9.424142626115706e-06,
      "loss": 0.81533623,
      "memory(GiB)": 135.77,
      "step": 15400,
      "train_speed(iter/s)": 0.203459
    },
    {
      "acc": 0.75681648,
      "epoch": 0.3595269688971843,
      "grad_norm": 6.8125,
      "learning_rate": 9.423262141481695e-06,
      "loss": 0.87958059,
      "memory(GiB)": 135.77,
      "step": 15410,
      "train_speed(iter/s)": 0.203521
    },
    {
      "acc": 0.77736855,
      "epoch": 0.3597602764694732,
      "grad_norm": 6.5,
      "learning_rate": 9.4223810254283e-06,
      "loss": 0.80077677,
      "memory(GiB)": 135.77,
      "step": 15420,
      "train_speed(iter/s)": 0.203593
    },
    {
      "acc": 0.7821209,
      "epoch": 0.35999358404176207,
      "grad_norm": 6.6875,
      "learning_rate": 9.421499278081296e-06,
      "loss": 0.80219059,
      "memory(GiB)": 135.77,
      "step": 15430,
      "train_speed(iter/s)": 0.20366
    },
    {
      "acc": 0.77446098,
      "epoch": 0.36022689161405097,
      "grad_norm": 6.5625,
      "learning_rate": 9.420616899566557e-06,
      "loss": 0.79374628,
      "memory(GiB)": 135.77,
      "step": 15440,
      "train_speed(iter/s)": 0.203725
    },
    {
      "acc": 0.76317291,
      "epoch": 0.36046019918633987,
      "grad_norm": 4.84375,
      "learning_rate": 9.41973389001004e-06,
      "loss": 0.87783279,
      "memory(GiB)": 135.77,
      "step": 15450,
      "train_speed(iter/s)": 0.203795
    },
    {
      "acc": 0.77032871,
      "epoch": 0.3606935067586287,
      "grad_norm": 5.0,
      "learning_rate": 9.418850249537792e-06,
      "loss": 0.83090649,
      "memory(GiB)": 135.77,
      "step": 15460,
      "train_speed(iter/s)": 0.203862
    },
    {
      "acc": 0.75985117,
      "epoch": 0.3609268143309176,
      "grad_norm": 7.0625,
      "learning_rate": 9.417965978275955e-06,
      "loss": 0.87291212,
      "memory(GiB)": 135.77,
      "step": 15470,
      "train_speed(iter/s)": 0.203928
    },
    {
      "acc": 0.78009644,
      "epoch": 0.3611601219032065,
      "grad_norm": 5.90625,
      "learning_rate": 9.417081076350758e-06,
      "loss": 0.79100814,
      "memory(GiB)": 135.77,
      "step": 15480,
      "train_speed(iter/s)": 0.203997
    },
    {
      "acc": 0.78492813,
      "epoch": 0.3613934294754954,
      "grad_norm": 5.75,
      "learning_rate": 9.416195543888522e-06,
      "loss": 0.78736544,
      "memory(GiB)": 135.77,
      "step": 15490,
      "train_speed(iter/s)": 0.204062
    },
    {
      "acc": 0.75766697,
      "epoch": 0.3616267370477843,
      "grad_norm": 6.53125,
      "learning_rate": 9.415309381015654e-06,
      "loss": 0.86035919,
      "memory(GiB)": 135.77,
      "step": 15500,
      "train_speed(iter/s)": 0.204131
    },
    {
      "epoch": 0.3616267370477843,
      "eval_acc": 0.7363535894968445,
      "eval_loss": 0.8363694548606873,
      "eval_runtime": 1270.7457,
      "eval_samples_per_second": 28.323,
      "eval_steps_per_second": 14.162,
      "step": 15500
    },
    {
      "acc": 0.76968145,
      "epoch": 0.3618600446200732,
      "grad_norm": 7.1875,
      "learning_rate": 9.414422587858654e-06,
      "loss": 0.82733736,
      "memory(GiB)": 135.77,
      "step": 15510,
      "train_speed(iter/s)": 0.200797
    },
    {
      "acc": 0.78013411,
      "epoch": 0.3620933521923621,
      "grad_norm": 4.21875,
      "learning_rate": 9.413535164544112e-06,
      "loss": 0.77737093,
      "memory(GiB)": 135.77,
      "step": 15520,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.76417913,
      "epoch": 0.362326659764651,
      "grad_norm": 5.8125,
      "learning_rate": 9.412647111198708e-06,
      "loss": 0.86939831,
      "memory(GiB)": 135.77,
      "step": 15530,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.76209974,
      "epoch": 0.3625599673369399,
      "grad_norm": 3.8125,
      "learning_rate": 9.411758427949211e-06,
      "loss": 0.85637188,
      "memory(GiB)": 135.77,
      "step": 15540,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.79845304,
      "epoch": 0.36279327490922875,
      "grad_norm": 4.3125,
      "learning_rate": 9.410869114922478e-06,
      "loss": 0.72790012,
      "memory(GiB)": 135.77,
      "step": 15550,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.77766886,
      "epoch": 0.36302658248151765,
      "grad_norm": 6.5,
      "learning_rate": 9.409979172245463e-06,
      "loss": 0.79989691,
      "memory(GiB)": 135.77,
      "step": 15560,
      "train_speed(iter/s)": 0.201133
    },
    {
      "acc": 0.77453566,
      "epoch": 0.36325989005380654,
      "grad_norm": 5.625,
      "learning_rate": 9.409088600045202e-06,
      "loss": 0.82451859,
      "memory(GiB)": 135.77,
      "step": 15570,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.77265034,
      "epoch": 0.36349319762609544,
      "grad_norm": 5.78125,
      "learning_rate": 9.408197398448822e-06,
      "loss": 0.79861336,
      "memory(GiB)": 135.77,
      "step": 15580,
      "train_speed(iter/s)": 0.201269
    },
    {
      "acc": 0.77113476,
      "epoch": 0.36372650519838434,
      "grad_norm": 4.9375,
      "learning_rate": 9.407305567583547e-06,
      "loss": 0.8350358,
      "memory(GiB)": 135.77,
      "step": 15590,
      "train_speed(iter/s)": 0.201333
    },
    {
      "acc": 0.76244006,
      "epoch": 0.36395981277067324,
      "grad_norm": 5.1875,
      "learning_rate": 9.40641310757668e-06,
      "loss": 0.86877785,
      "memory(GiB)": 135.77,
      "step": 15600,
      "train_speed(iter/s)": 0.2014
    },
    {
      "acc": 0.7791647,
      "epoch": 0.36419312034296214,
      "grad_norm": 5.75,
      "learning_rate": 9.405520018555624e-06,
      "loss": 0.79430704,
      "memory(GiB)": 135.77,
      "step": 15610,
      "train_speed(iter/s)": 0.201469
    },
    {
      "acc": 0.77130833,
      "epoch": 0.36442642791525104,
      "grad_norm": 5.46875,
      "learning_rate": 9.404626300647864e-06,
      "loss": 0.8362133,
      "memory(GiB)": 135.77,
      "step": 15620,
      "train_speed(iter/s)": 0.201536
    },
    {
      "acc": 0.78146958,
      "epoch": 0.36465973548753994,
      "grad_norm": 5.53125,
      "learning_rate": 9.403731953980978e-06,
      "loss": 0.7925806,
      "memory(GiB)": 135.77,
      "step": 15630,
      "train_speed(iter/s)": 0.201599
    },
    {
      "acc": 0.77909827,
      "epoch": 0.36489304305982884,
      "grad_norm": 4.25,
      "learning_rate": 9.402836978682636e-06,
      "loss": 0.79750004,
      "memory(GiB)": 135.77,
      "step": 15640,
      "train_speed(iter/s)": 0.201662
    },
    {
      "acc": 0.77254696,
      "epoch": 0.3651263506321177,
      "grad_norm": 7.96875,
      "learning_rate": 9.401941374880595e-06,
      "loss": 0.80021648,
      "memory(GiB)": 135.77,
      "step": 15650,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.77365751,
      "epoch": 0.3653596582044066,
      "grad_norm": 5.09375,
      "learning_rate": 9.4010451427027e-06,
      "loss": 0.81189899,
      "memory(GiB)": 135.77,
      "step": 15660,
      "train_speed(iter/s)": 0.201787
    },
    {
      "acc": 0.77464514,
      "epoch": 0.3655929657766955,
      "grad_norm": 4.34375,
      "learning_rate": 9.40014828227689e-06,
      "loss": 0.81965914,
      "memory(GiB)": 135.77,
      "step": 15670,
      "train_speed(iter/s)": 0.201854
    },
    {
      "acc": 0.76693192,
      "epoch": 0.3658262733489844,
      "grad_norm": 12.625,
      "learning_rate": 9.399250793731192e-06,
      "loss": 0.85733891,
      "memory(GiB)": 135.77,
      "step": 15680,
      "train_speed(iter/s)": 0.20192
    },
    {
      "acc": 0.76616225,
      "epoch": 0.3660595809212733,
      "grad_norm": 4.5,
      "learning_rate": 9.398352677193719e-06,
      "loss": 0.85392532,
      "memory(GiB)": 135.77,
      "step": 15690,
      "train_speed(iter/s)": 0.201988
    },
    {
      "acc": 0.77419605,
      "epoch": 0.3662928884935622,
      "grad_norm": 5.375,
      "learning_rate": 9.397453932792681e-06,
      "loss": 0.80535965,
      "memory(GiB)": 135.77,
      "step": 15700,
      "train_speed(iter/s)": 0.202053
    },
    {
      "acc": 0.76591721,
      "epoch": 0.3665261960658511,
      "grad_norm": 32.5,
      "learning_rate": 9.396554560656371e-06,
      "loss": 0.90535355,
      "memory(GiB)": 135.77,
      "step": 15710,
      "train_speed(iter/s)": 0.202119
    },
    {
      "acc": 0.75157509,
      "epoch": 0.36675950363813997,
      "grad_norm": 5.125,
      "learning_rate": 9.395654560913174e-06,
      "loss": 0.90469847,
      "memory(GiB)": 135.77,
      "step": 15720,
      "train_speed(iter/s)": 0.202184
    },
    {
      "acc": 0.7669796,
      "epoch": 0.36699281121042887,
      "grad_norm": 4.5625,
      "learning_rate": 9.394753933691567e-06,
      "loss": 0.86323175,
      "memory(GiB)": 135.77,
      "step": 15730,
      "train_speed(iter/s)": 0.202243
    },
    {
      "acc": 0.76320677,
      "epoch": 0.3672261187827177,
      "grad_norm": 4.65625,
      "learning_rate": 9.393852679120113e-06,
      "loss": 0.85583038,
      "memory(GiB)": 135.77,
      "step": 15740,
      "train_speed(iter/s)": 0.202312
    },
    {
      "acc": 0.78019586,
      "epoch": 0.3674594263550066,
      "grad_norm": 4.875,
      "learning_rate": 9.392950797327463e-06,
      "loss": 0.77571545,
      "memory(GiB)": 135.77,
      "step": 15750,
      "train_speed(iter/s)": 0.202374
    },
    {
      "acc": 0.76291428,
      "epoch": 0.3676927339272955,
      "grad_norm": 7.25,
      "learning_rate": 9.392048288442363e-06,
      "loss": 0.8645957,
      "memory(GiB)": 135.77,
      "step": 15760,
      "train_speed(iter/s)": 0.202441
    },
    {
      "acc": 0.75715952,
      "epoch": 0.3679260414995844,
      "grad_norm": 5.65625,
      "learning_rate": 9.391145152593646e-06,
      "loss": 0.89210663,
      "memory(GiB)": 135.77,
      "step": 15770,
      "train_speed(iter/s)": 0.202507
    },
    {
      "acc": 0.77362776,
      "epoch": 0.3681593490718733,
      "grad_norm": 8.9375,
      "learning_rate": 9.390241389910236e-06,
      "loss": 0.84507351,
      "memory(GiB)": 135.77,
      "step": 15780,
      "train_speed(iter/s)": 0.202567
    },
    {
      "acc": 0.77469263,
      "epoch": 0.3683926566441622,
      "grad_norm": 5.65625,
      "learning_rate": 9.389337000521142e-06,
      "loss": 0.82550621,
      "memory(GiB)": 135.77,
      "step": 15790,
      "train_speed(iter/s)": 0.202635
    },
    {
      "acc": 0.76861792,
      "epoch": 0.3686259642164511,
      "grad_norm": 5.125,
      "learning_rate": 9.388431984555466e-06,
      "loss": 0.84576836,
      "memory(GiB)": 135.77,
      "step": 15800,
      "train_speed(iter/s)": 0.202702
    },
    {
      "acc": 0.79161878,
      "epoch": 0.36885927178874,
      "grad_norm": 4.8125,
      "learning_rate": 9.387526342142398e-06,
      "loss": 0.75349326,
      "memory(GiB)": 135.77,
      "step": 15810,
      "train_speed(iter/s)": 0.202771
    },
    {
      "acc": 0.7592205,
      "epoch": 0.3690925793610289,
      "grad_norm": 6.21875,
      "learning_rate": 9.386620073411221e-06,
      "loss": 0.8692564,
      "memory(GiB)": 135.77,
      "step": 15820,
      "train_speed(iter/s)": 0.202835
    },
    {
      "acc": 0.77718821,
      "epoch": 0.3693258869333178,
      "grad_norm": 7.3125,
      "learning_rate": 9.385713178491302e-06,
      "loss": 0.81367092,
      "memory(GiB)": 135.77,
      "step": 15830,
      "train_speed(iter/s)": 0.202905
    },
    {
      "acc": 0.78268156,
      "epoch": 0.36955919450560665,
      "grad_norm": 7.625,
      "learning_rate": 9.384805657512101e-06,
      "loss": 0.79621563,
      "memory(GiB)": 135.77,
      "step": 15840,
      "train_speed(iter/s)": 0.202971
    },
    {
      "acc": 0.76887393,
      "epoch": 0.36979250207789555,
      "grad_norm": 6.75,
      "learning_rate": 9.383897510603167e-06,
      "loss": 0.84250908,
      "memory(GiB)": 135.77,
      "step": 15850,
      "train_speed(iter/s)": 0.203038
    },
    {
      "acc": 0.78669109,
      "epoch": 0.37002580965018445,
      "grad_norm": 5.90625,
      "learning_rate": 9.382988737894136e-06,
      "loss": 0.7452014,
      "memory(GiB)": 135.77,
      "step": 15860,
      "train_speed(iter/s)": 0.203102
    },
    {
      "acc": 0.76802521,
      "epoch": 0.37025911722247334,
      "grad_norm": 7.375,
      "learning_rate": 9.382079339514736e-06,
      "loss": 0.85868645,
      "memory(GiB)": 135.77,
      "step": 15870,
      "train_speed(iter/s)": 0.203168
    },
    {
      "acc": 0.76643047,
      "epoch": 0.37049242479476224,
      "grad_norm": 5.0,
      "learning_rate": 9.381169315594782e-06,
      "loss": 0.83867474,
      "memory(GiB)": 135.77,
      "step": 15880,
      "train_speed(iter/s)": 0.20323
    },
    {
      "acc": 0.77568998,
      "epoch": 0.37072573236705114,
      "grad_norm": 4.90625,
      "learning_rate": 9.380258666264184e-06,
      "loss": 0.83556032,
      "memory(GiB)": 135.77,
      "step": 15890,
      "train_speed(iter/s)": 0.203294
    },
    {
      "acc": 0.77228613,
      "epoch": 0.37095903993934004,
      "grad_norm": 9.0625,
      "learning_rate": 9.379347391652931e-06,
      "loss": 0.81034203,
      "memory(GiB)": 135.77,
      "step": 15900,
      "train_speed(iter/s)": 0.20336
    },
    {
      "acc": 0.76030293,
      "epoch": 0.37119234751162894,
      "grad_norm": 7.4375,
      "learning_rate": 9.378435491891112e-06,
      "loss": 0.88398857,
      "memory(GiB)": 135.77,
      "step": 15910,
      "train_speed(iter/s)": 0.203423
    },
    {
      "acc": 0.7455883,
      "epoch": 0.37142565508391784,
      "grad_norm": 6.21875,
      "learning_rate": 9.377522967108897e-06,
      "loss": 0.94365292,
      "memory(GiB)": 135.77,
      "step": 15920,
      "train_speed(iter/s)": 0.203489
    },
    {
      "acc": 0.76555414,
      "epoch": 0.37165896265620674,
      "grad_norm": 6.5625,
      "learning_rate": 9.376609817436551e-06,
      "loss": 0.86510744,
      "memory(GiB)": 135.77,
      "step": 15930,
      "train_speed(iter/s)": 0.203548
    },
    {
      "acc": 0.75843372,
      "epoch": 0.3718922702284956,
      "grad_norm": 9.6875,
      "learning_rate": 9.375696043004425e-06,
      "loss": 0.90398607,
      "memory(GiB)": 135.77,
      "step": 15940,
      "train_speed(iter/s)": 0.203613
    },
    {
      "acc": 0.75562124,
      "epoch": 0.3721255778007845,
      "grad_norm": 4.71875,
      "learning_rate": 9.374781643942961e-06,
      "loss": 0.88584518,
      "memory(GiB)": 135.77,
      "step": 15950,
      "train_speed(iter/s)": 0.20368
    },
    {
      "acc": 0.76519918,
      "epoch": 0.3723588853730734,
      "grad_norm": 7.25,
      "learning_rate": 9.373866620382686e-06,
      "loss": 0.8289814,
      "memory(GiB)": 135.77,
      "step": 15960,
      "train_speed(iter/s)": 0.203744
    },
    {
      "acc": 0.78620424,
      "epoch": 0.3725921929453623,
      "grad_norm": 13.9375,
      "learning_rate": 9.372950972454222e-06,
      "loss": 0.74287148,
      "memory(GiB)": 135.77,
      "step": 15970,
      "train_speed(iter/s)": 0.203812
    },
    {
      "acc": 0.77351294,
      "epoch": 0.3728255005176512,
      "grad_norm": 4.84375,
      "learning_rate": 9.372034700288278e-06,
      "loss": 0.79458714,
      "memory(GiB)": 135.77,
      "step": 15980,
      "train_speed(iter/s)": 0.20388
    },
    {
      "acc": 0.74867344,
      "epoch": 0.3730588080899401,
      "grad_norm": 6.75,
      "learning_rate": 9.37111780401565e-06,
      "loss": 0.92256651,
      "memory(GiB)": 135.77,
      "step": 15990,
      "train_speed(iter/s)": 0.203948
    },
    {
      "acc": 0.74160004,
      "epoch": 0.373292115662229,
      "grad_norm": 5.8125,
      "learning_rate": 9.370200283767225e-06,
      "loss": 0.94474297,
      "memory(GiB)": 135.77,
      "step": 16000,
      "train_speed(iter/s)": 0.204015
    },
    {
      "epoch": 0.373292115662229,
      "eval_acc": 0.7363508670448953,
      "eval_loss": 0.8348709940910339,
      "eval_runtime": 1270.1462,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 16000
    },
    {
      "acc": 0.78337979,
      "epoch": 0.3735254232345179,
      "grad_norm": 5.5,
      "learning_rate": 9.369282139673979e-06,
      "loss": 0.79470425,
      "memory(GiB)": 135.77,
      "step": 16010,
      "train_speed(iter/s)": 0.200783
    },
    {
      "acc": 0.76890593,
      "epoch": 0.37375873080680677,
      "grad_norm": 4.6875,
      "learning_rate": 9.368363371866978e-06,
      "loss": 0.82831898,
      "memory(GiB)": 135.77,
      "step": 16020,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.75959797,
      "epoch": 0.3739920383790956,
      "grad_norm": 3.90625,
      "learning_rate": 9.367443980477374e-06,
      "loss": 0.86754074,
      "memory(GiB)": 135.77,
      "step": 16030,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.77851305,
      "epoch": 0.3742253459513845,
      "grad_norm": 5.6875,
      "learning_rate": 9.366523965636412e-06,
      "loss": 0.77482352,
      "memory(GiB)": 135.77,
      "step": 16040,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.75851483,
      "epoch": 0.3744586535236734,
      "grad_norm": 6.1875,
      "learning_rate": 9.36560332747542e-06,
      "loss": 0.86383018,
      "memory(GiB)": 135.77,
      "step": 16050,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.77050686,
      "epoch": 0.3746919610959623,
      "grad_norm": 7.25,
      "learning_rate": 9.364682066125822e-06,
      "loss": 0.84514685,
      "memory(GiB)": 135.77,
      "step": 16060,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.7671258,
      "epoch": 0.3749252686682512,
      "grad_norm": 7.90625,
      "learning_rate": 9.363760181719127e-06,
      "loss": 0.86259499,
      "memory(GiB)": 135.77,
      "step": 16070,
      "train_speed(iter/s)": 0.201169
    },
    {
      "acc": 0.76694169,
      "epoch": 0.3751585762405401,
      "grad_norm": 10.0,
      "learning_rate": 9.362837674386934e-06,
      "loss": 0.85537701,
      "memory(GiB)": 135.77,
      "step": 16080,
      "train_speed(iter/s)": 0.201237
    },
    {
      "acc": 0.79346738,
      "epoch": 0.375391883812829,
      "grad_norm": 6.4375,
      "learning_rate": 9.36191454426093e-06,
      "loss": 0.74549022,
      "memory(GiB)": 135.77,
      "step": 16090,
      "train_speed(iter/s)": 0.2013
    },
    {
      "acc": 0.76404152,
      "epoch": 0.3756251913851179,
      "grad_norm": 5.21875,
      "learning_rate": 9.360990791472893e-06,
      "loss": 0.85988121,
      "memory(GiB)": 135.77,
      "step": 16100,
      "train_speed(iter/s)": 0.20136
    },
    {
      "acc": 0.77790794,
      "epoch": 0.3758584989574068,
      "grad_norm": 7.1875,
      "learning_rate": 9.360066416154687e-06,
      "loss": 0.82874393,
      "memory(GiB)": 135.77,
      "step": 16110,
      "train_speed(iter/s)": 0.201424
    },
    {
      "acc": 0.76713996,
      "epoch": 0.3760918065296957,
      "grad_norm": 6.34375,
      "learning_rate": 9.359141418438266e-06,
      "loss": 0.82495213,
      "memory(GiB)": 135.77,
      "step": 16120,
      "train_speed(iter/s)": 0.201487
    },
    {
      "acc": 0.76586213,
      "epoch": 0.37632511410198455,
      "grad_norm": 7.125,
      "learning_rate": 9.358215798455674e-06,
      "loss": 0.83144331,
      "memory(GiB)": 135.77,
      "step": 16130,
      "train_speed(iter/s)": 0.201555
    },
    {
      "acc": 0.7565609,
      "epoch": 0.37655842167427345,
      "grad_norm": 6.0,
      "learning_rate": 9.357289556339044e-06,
      "loss": 0.8857419,
      "memory(GiB)": 135.77,
      "step": 16140,
      "train_speed(iter/s)": 0.201622
    },
    {
      "acc": 0.76512766,
      "epoch": 0.37679172924656235,
      "grad_norm": 10.5,
      "learning_rate": 9.356362692220593e-06,
      "loss": 0.88596659,
      "memory(GiB)": 135.77,
      "step": 16150,
      "train_speed(iter/s)": 0.201686
    },
    {
      "acc": 0.77160578,
      "epoch": 0.37702503681885124,
      "grad_norm": 7.6875,
      "learning_rate": 9.355435206232635e-06,
      "loss": 0.81468468,
      "memory(GiB)": 135.77,
      "step": 16160,
      "train_speed(iter/s)": 0.201747
    },
    {
      "acc": 0.76007071,
      "epoch": 0.37725834439114014,
      "grad_norm": 5.1875,
      "learning_rate": 9.354507098507568e-06,
      "loss": 0.86362362,
      "memory(GiB)": 135.77,
      "step": 16170,
      "train_speed(iter/s)": 0.201815
    },
    {
      "acc": 0.75717325,
      "epoch": 0.37749165196342904,
      "grad_norm": 5.78125,
      "learning_rate": 9.353578369177876e-06,
      "loss": 0.88568306,
      "memory(GiB)": 135.77,
      "step": 16180,
      "train_speed(iter/s)": 0.201881
    },
    {
      "acc": 0.77196169,
      "epoch": 0.37772495953571794,
      "grad_norm": 5.78125,
      "learning_rate": 9.352649018376136e-06,
      "loss": 0.82890453,
      "memory(GiB)": 135.77,
      "step": 16190,
      "train_speed(iter/s)": 0.201948
    },
    {
      "acc": 0.75310216,
      "epoch": 0.37795826710800684,
      "grad_norm": 5.8125,
      "learning_rate": 9.351719046235013e-06,
      "loss": 0.88745995,
      "memory(GiB)": 135.77,
      "step": 16200,
      "train_speed(iter/s)": 0.202008
    },
    {
      "acc": 0.78186665,
      "epoch": 0.37819157468029574,
      "grad_norm": 4.03125,
      "learning_rate": 9.350788452887262e-06,
      "loss": 0.75842514,
      "memory(GiB)": 135.77,
      "step": 16210,
      "train_speed(iter/s)": 0.202072
    },
    {
      "acc": 0.7654644,
      "epoch": 0.37842488225258464,
      "grad_norm": 7.9375,
      "learning_rate": 9.349857238465723e-06,
      "loss": 0.8322834,
      "memory(GiB)": 135.77,
      "step": 16220,
      "train_speed(iter/s)": 0.202134
    },
    {
      "acc": 0.77073183,
      "epoch": 0.3786581898248735,
      "grad_norm": 5.125,
      "learning_rate": 9.348925403103326e-06,
      "loss": 0.84538736,
      "memory(GiB)": 135.77,
      "step": 16230,
      "train_speed(iter/s)": 0.202201
    },
    {
      "acc": 0.76439257,
      "epoch": 0.3788914973971624,
      "grad_norm": 6.25,
      "learning_rate": 9.347992946933091e-06,
      "loss": 0.86807938,
      "memory(GiB)": 135.77,
      "step": 16240,
      "train_speed(iter/s)": 0.202266
    },
    {
      "acc": 0.77966743,
      "epoch": 0.3791248049694513,
      "grad_norm": 5.0625,
      "learning_rate": 9.347059870088127e-06,
      "loss": 0.75784845,
      "memory(GiB)": 135.77,
      "step": 16250,
      "train_speed(iter/s)": 0.202327
    },
    {
      "acc": 0.76927428,
      "epoch": 0.3793581125417402,
      "grad_norm": 5.96875,
      "learning_rate": 9.346126172701629e-06,
      "loss": 0.83154087,
      "memory(GiB)": 135.77,
      "step": 16260,
      "train_speed(iter/s)": 0.202393
    },
    {
      "acc": 0.77350688,
      "epoch": 0.3795914201140291,
      "grad_norm": 3.765625,
      "learning_rate": 9.345191854906881e-06,
      "loss": 0.80910931,
      "memory(GiB)": 135.77,
      "step": 16270,
      "train_speed(iter/s)": 0.202457
    },
    {
      "acc": 0.7722403,
      "epoch": 0.379824727686318,
      "grad_norm": 4.125,
      "learning_rate": 9.344256916837259e-06,
      "loss": 0.83394699,
      "memory(GiB)": 135.77,
      "step": 16280,
      "train_speed(iter/s)": 0.202518
    },
    {
      "acc": 0.77347498,
      "epoch": 0.3800580352586069,
      "grad_norm": 8.625,
      "learning_rate": 9.343321358626225e-06,
      "loss": 0.78846674,
      "memory(GiB)": 135.77,
      "step": 16290,
      "train_speed(iter/s)": 0.202587
    },
    {
      "acc": 0.76977482,
      "epoch": 0.3802913428308958,
      "grad_norm": 6.1875,
      "learning_rate": 9.342385180407328e-06,
      "loss": 0.83136139,
      "memory(GiB)": 135.77,
      "step": 16300,
      "train_speed(iter/s)": 0.202652
    },
    {
      "acc": 0.77511702,
      "epoch": 0.38052465040318467,
      "grad_norm": 4.34375,
      "learning_rate": 9.341448382314207e-06,
      "loss": 0.80090103,
      "memory(GiB)": 135.77,
      "step": 16310,
      "train_speed(iter/s)": 0.202714
    },
    {
      "acc": 0.7784832,
      "epoch": 0.3807579579754735,
      "grad_norm": 4.28125,
      "learning_rate": 9.340510964480591e-06,
      "loss": 0.77812891,
      "memory(GiB)": 135.77,
      "step": 16320,
      "train_speed(iter/s)": 0.202777
    },
    {
      "acc": 0.78746824,
      "epoch": 0.3809912655477624,
      "grad_norm": 5.40625,
      "learning_rate": 9.339572927040298e-06,
      "loss": 0.76021338,
      "memory(GiB)": 135.77,
      "step": 16330,
      "train_speed(iter/s)": 0.202838
    },
    {
      "acc": 0.77678819,
      "epoch": 0.3812245731200513,
      "grad_norm": 5.46875,
      "learning_rate": 9.338634270127227e-06,
      "loss": 0.81745186,
      "memory(GiB)": 135.77,
      "step": 16340,
      "train_speed(iter/s)": 0.202905
    },
    {
      "acc": 0.75448432,
      "epoch": 0.3814578806923402,
      "grad_norm": 4.875,
      "learning_rate": 9.337694993875376e-06,
      "loss": 0.89182949,
      "memory(GiB)": 135.77,
      "step": 16350,
      "train_speed(iter/s)": 0.202971
    },
    {
      "acc": 0.7632596,
      "epoch": 0.3816911882646291,
      "grad_norm": 7.53125,
      "learning_rate": 9.336755098418824e-06,
      "loss": 0.86298809,
      "memory(GiB)": 135.77,
      "step": 16360,
      "train_speed(iter/s)": 0.203036
    },
    {
      "acc": 0.756043,
      "epoch": 0.381924495836918,
      "grad_norm": 5.5,
      "learning_rate": 9.335814583891743e-06,
      "loss": 0.89383011,
      "memory(GiB)": 135.77,
      "step": 16370,
      "train_speed(iter/s)": 0.203095
    },
    {
      "acc": 0.76111898,
      "epoch": 0.3821578034092069,
      "grad_norm": 4.65625,
      "learning_rate": 9.33487345042839e-06,
      "loss": 0.89134369,
      "memory(GiB)": 135.77,
      "step": 16380,
      "train_speed(iter/s)": 0.203158
    },
    {
      "acc": 0.75423412,
      "epoch": 0.3823911109814958,
      "grad_norm": 7.34375,
      "learning_rate": 9.333931698163107e-06,
      "loss": 0.89891577,
      "memory(GiB)": 135.77,
      "step": 16390,
      "train_speed(iter/s)": 0.203226
    },
    {
      "acc": 0.77302418,
      "epoch": 0.3826244185537847,
      "grad_norm": 5.6875,
      "learning_rate": 9.332989327230337e-06,
      "loss": 0.80847969,
      "memory(GiB)": 135.77,
      "step": 16400,
      "train_speed(iter/s)": 0.203289
    },
    {
      "acc": 0.75313158,
      "epoch": 0.3828577261260736,
      "grad_norm": 6.71875,
      "learning_rate": 9.3320463377646e-06,
      "loss": 0.87891369,
      "memory(GiB)": 135.77,
      "step": 16410,
      "train_speed(iter/s)": 0.20336
    },
    {
      "acc": 0.76052837,
      "epoch": 0.38309103369836245,
      "grad_norm": 5.0625,
      "learning_rate": 9.331102729900505e-06,
      "loss": 0.87168789,
      "memory(GiB)": 135.77,
      "step": 16420,
      "train_speed(iter/s)": 0.203426
    },
    {
      "acc": 0.75323372,
      "epoch": 0.38332434127065135,
      "grad_norm": 4.8125,
      "learning_rate": 9.330158503772753e-06,
      "loss": 0.9150569,
      "memory(GiB)": 135.77,
      "step": 16430,
      "train_speed(iter/s)": 0.203491
    },
    {
      "acc": 0.7864645,
      "epoch": 0.38355764884294025,
      "grad_norm": 5.90625,
      "learning_rate": 9.329213659516134e-06,
      "loss": 0.79203343,
      "memory(GiB)": 135.77,
      "step": 16440,
      "train_speed(iter/s)": 0.203553
    },
    {
      "acc": 0.77691231,
      "epoch": 0.38379095641522915,
      "grad_norm": 4.40625,
      "learning_rate": 9.328268197265523e-06,
      "loss": 0.79185615,
      "memory(GiB)": 135.77,
      "step": 16450,
      "train_speed(iter/s)": 0.203619
    },
    {
      "acc": 0.77853565,
      "epoch": 0.38402426398751804,
      "grad_norm": 6.59375,
      "learning_rate": 9.327322117155881e-06,
      "loss": 0.78379364,
      "memory(GiB)": 135.77,
      "step": 16460,
      "train_speed(iter/s)": 0.20368
    },
    {
      "acc": 0.78826647,
      "epoch": 0.38425757155980694,
      "grad_norm": 4.21875,
      "learning_rate": 9.326375419322267e-06,
      "loss": 0.81094532,
      "memory(GiB)": 135.77,
      "step": 16470,
      "train_speed(iter/s)": 0.203742
    },
    {
      "acc": 0.76101255,
      "epoch": 0.38449087913209584,
      "grad_norm": 4.59375,
      "learning_rate": 9.325428103899818e-06,
      "loss": 0.88430595,
      "memory(GiB)": 135.77,
      "step": 16480,
      "train_speed(iter/s)": 0.203798
    },
    {
      "acc": 0.774227,
      "epoch": 0.38472418670438474,
      "grad_norm": 5.75,
      "learning_rate": 9.324480171023764e-06,
      "loss": 0.81697226,
      "memory(GiB)": 135.77,
      "step": 16490,
      "train_speed(iter/s)": 0.203856
    },
    {
      "acc": 0.76122689,
      "epoch": 0.38495749427667364,
      "grad_norm": 4.5,
      "learning_rate": 9.32353162082942e-06,
      "loss": 0.85380974,
      "memory(GiB)": 135.77,
      "step": 16500,
      "train_speed(iter/s)": 0.203918
    },
    {
      "epoch": 0.38495749427667364,
      "eval_acc": 0.736815605607034,
      "eval_loss": 0.8338123559951782,
      "eval_runtime": 1271.5777,
      "eval_samples_per_second": 28.304,
      "eval_steps_per_second": 14.152,
      "step": 16500
    },
    {
      "acc": 0.76048388,
      "epoch": 0.3851908018489625,
      "grad_norm": 7.8125,
      "learning_rate": 9.322582453452195e-06,
      "loss": 0.85223875,
      "memory(GiB)": 135.77,
      "step": 16510,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.78726511,
      "epoch": 0.3854241094212514,
      "grad_norm": 5.5625,
      "learning_rate": 9.32163266902758e-06,
      "loss": 0.76919699,
      "memory(GiB)": 135.77,
      "step": 16520,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.76031055,
      "epoch": 0.3856574169935403,
      "grad_norm": 6.0625,
      "learning_rate": 9.320682267691157e-06,
      "loss": 0.87715168,
      "memory(GiB)": 135.77,
      "step": 16530,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.76955204,
      "epoch": 0.3858907245658292,
      "grad_norm": 8.6875,
      "learning_rate": 9.319731249578595e-06,
      "loss": 0.84494944,
      "memory(GiB)": 135.77,
      "step": 16540,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.77983017,
      "epoch": 0.3861240321381181,
      "grad_norm": 5.4375,
      "learning_rate": 9.318779614825653e-06,
      "loss": 0.78632112,
      "memory(GiB)": 135.77,
      "step": 16550,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.77373309,
      "epoch": 0.386357339710407,
      "grad_norm": 5.15625,
      "learning_rate": 9.317827363568176e-06,
      "loss": 0.8040472,
      "memory(GiB)": 135.77,
      "step": 16560,
      "train_speed(iter/s)": 0.2011
    },
    {
      "acc": 0.77826786,
      "epoch": 0.3865906472826959,
      "grad_norm": 6.34375,
      "learning_rate": 9.316874495942095e-06,
      "loss": 0.7839695,
      "memory(GiB)": 135.77,
      "step": 16570,
      "train_speed(iter/s)": 0.20116
    },
    {
      "acc": 0.77872405,
      "epoch": 0.3868239548549848,
      "grad_norm": 6.0,
      "learning_rate": 9.315921012083436e-06,
      "loss": 0.77586727,
      "memory(GiB)": 135.77,
      "step": 16580,
      "train_speed(iter/s)": 0.201222
    },
    {
      "acc": 0.7748662,
      "epoch": 0.3870572624272737,
      "grad_norm": 11.875,
      "learning_rate": 9.314966912128305e-06,
      "loss": 0.81466694,
      "memory(GiB)": 135.77,
      "step": 16590,
      "train_speed(iter/s)": 0.201285
    },
    {
      "acc": 0.77469873,
      "epoch": 0.3872905699995626,
      "grad_norm": 6.21875,
      "learning_rate": 9.3140121962129e-06,
      "loss": 0.81014376,
      "memory(GiB)": 135.77,
      "step": 16600,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.78045406,
      "epoch": 0.3875238775718514,
      "grad_norm": 6.5625,
      "learning_rate": 9.313056864473508e-06,
      "loss": 0.77907581,
      "memory(GiB)": 135.77,
      "step": 16610,
      "train_speed(iter/s)": 0.20141
    },
    {
      "acc": 0.76823573,
      "epoch": 0.3877571851441403,
      "grad_norm": 6.46875,
      "learning_rate": 9.312100917046502e-06,
      "loss": 0.8444459,
      "memory(GiB)": 135.77,
      "step": 16620,
      "train_speed(iter/s)": 0.201474
    },
    {
      "acc": 0.76610289,
      "epoch": 0.3879904927164292,
      "grad_norm": 6.03125,
      "learning_rate": 9.311144354068342e-06,
      "loss": 0.84226179,
      "memory(GiB)": 135.77,
      "step": 16630,
      "train_speed(iter/s)": 0.20154
    },
    {
      "acc": 0.78322606,
      "epoch": 0.3882238002887181,
      "grad_norm": 5.84375,
      "learning_rate": 9.310187175675579e-06,
      "loss": 0.78812022,
      "memory(GiB)": 135.77,
      "step": 16640,
      "train_speed(iter/s)": 0.201607
    },
    {
      "acc": 0.75634503,
      "epoch": 0.388457107861007,
      "grad_norm": 6.78125,
      "learning_rate": 9.309229382004847e-06,
      "loss": 0.90151043,
      "memory(GiB)": 135.77,
      "step": 16650,
      "train_speed(iter/s)": 0.20167
    },
    {
      "acc": 0.75544233,
      "epoch": 0.3886904154332959,
      "grad_norm": 5.0625,
      "learning_rate": 9.308270973192875e-06,
      "loss": 0.89699974,
      "memory(GiB)": 135.77,
      "step": 16660,
      "train_speed(iter/s)": 0.201732
    },
    {
      "acc": 0.74633512,
      "epoch": 0.3889237230055848,
      "grad_norm": 5.96875,
      "learning_rate": 9.307311949376472e-06,
      "loss": 0.92126493,
      "memory(GiB)": 135.77,
      "step": 16670,
      "train_speed(iter/s)": 0.201796
    },
    {
      "acc": 0.77705793,
      "epoch": 0.3891570305778737,
      "grad_norm": 6.125,
      "learning_rate": 9.306352310692539e-06,
      "loss": 0.80012226,
      "memory(GiB)": 135.77,
      "step": 16680,
      "train_speed(iter/s)": 0.201858
    },
    {
      "acc": 0.77201366,
      "epoch": 0.3893903381501626,
      "grad_norm": 4.84375,
      "learning_rate": 9.305392057278066e-06,
      "loss": 0.8111311,
      "memory(GiB)": 135.77,
      "step": 16690,
      "train_speed(iter/s)": 0.201919
    },
    {
      "acc": 0.76858664,
      "epoch": 0.3896236457224515,
      "grad_norm": 5.28125,
      "learning_rate": 9.304431189270127e-06,
      "loss": 0.8267766,
      "memory(GiB)": 135.77,
      "step": 16700,
      "train_speed(iter/s)": 0.201983
    },
    {
      "acc": 0.75889578,
      "epoch": 0.38985695329474035,
      "grad_norm": 4.59375,
      "learning_rate": 9.303469706805886e-06,
      "loss": 0.84821739,
      "memory(GiB)": 135.77,
      "step": 16710,
      "train_speed(iter/s)": 0.202043
    },
    {
      "acc": 0.76242075,
      "epoch": 0.39009026086702925,
      "grad_norm": 6.125,
      "learning_rate": 9.302507610022593e-06,
      "loss": 0.87555027,
      "memory(GiB)": 135.77,
      "step": 16720,
      "train_speed(iter/s)": 0.202104
    },
    {
      "acc": 0.79397368,
      "epoch": 0.39032356843931815,
      "grad_norm": 8.0625,
      "learning_rate": 9.30154489905759e-06,
      "loss": 0.72974205,
      "memory(GiB)": 135.77,
      "step": 16730,
      "train_speed(iter/s)": 0.202163
    },
    {
      "acc": 0.74320917,
      "epoch": 0.39055687601160705,
      "grad_norm": 5.5,
      "learning_rate": 9.300581574048303e-06,
      "loss": 0.93350315,
      "memory(GiB)": 135.77,
      "step": 16740,
      "train_speed(iter/s)": 0.202225
    },
    {
      "acc": 0.77537889,
      "epoch": 0.39079018358389594,
      "grad_norm": 5.59375,
      "learning_rate": 9.299617635132243e-06,
      "loss": 0.78615303,
      "memory(GiB)": 135.77,
      "step": 16750,
      "train_speed(iter/s)": 0.202286
    },
    {
      "acc": 0.77060065,
      "epoch": 0.39102349115618484,
      "grad_norm": 5.6875,
      "learning_rate": 9.298653082447019e-06,
      "loss": 0.84748516,
      "memory(GiB)": 135.77,
      "step": 16760,
      "train_speed(iter/s)": 0.202351
    },
    {
      "acc": 0.78617654,
      "epoch": 0.39125679872847374,
      "grad_norm": 5.40625,
      "learning_rate": 9.29768791613031e-06,
      "loss": 0.76330194,
      "memory(GiB)": 135.77,
      "step": 16770,
      "train_speed(iter/s)": 0.202413
    },
    {
      "acc": 0.7793808,
      "epoch": 0.39149010630076264,
      "grad_norm": 4.96875,
      "learning_rate": 9.296722136319904e-06,
      "loss": 0.79507647,
      "memory(GiB)": 135.77,
      "step": 16780,
      "train_speed(iter/s)": 0.202476
    },
    {
      "acc": 0.7710515,
      "epoch": 0.39172341387305154,
      "grad_norm": 6.25,
      "learning_rate": 9.29575574315366e-06,
      "loss": 0.8227356,
      "memory(GiB)": 135.77,
      "step": 16790,
      "train_speed(iter/s)": 0.202539
    },
    {
      "acc": 0.77599983,
      "epoch": 0.3919567214453404,
      "grad_norm": 5.53125,
      "learning_rate": 9.294788736769534e-06,
      "loss": 0.80106039,
      "memory(GiB)": 135.77,
      "step": 16800,
      "train_speed(iter/s)": 0.202601
    },
    {
      "acc": 0.77693524,
      "epoch": 0.3921900290176293,
      "grad_norm": 7.5625,
      "learning_rate": 9.293821117305562e-06,
      "loss": 0.79049306,
      "memory(GiB)": 135.77,
      "step": 16810,
      "train_speed(iter/s)": 0.202663
    },
    {
      "acc": 0.7763052,
      "epoch": 0.3924233365899182,
      "grad_norm": 6.28125,
      "learning_rate": 9.29285288489987e-06,
      "loss": 0.81076241,
      "memory(GiB)": 135.77,
      "step": 16820,
      "train_speed(iter/s)": 0.202723
    },
    {
      "acc": 0.77143803,
      "epoch": 0.3926566441622071,
      "grad_norm": 9.8125,
      "learning_rate": 9.29188403969068e-06,
      "loss": 0.81417742,
      "memory(GiB)": 135.77,
      "step": 16830,
      "train_speed(iter/s)": 0.202787
    },
    {
      "acc": 0.79286661,
      "epoch": 0.392889951734496,
      "grad_norm": 8.375,
      "learning_rate": 9.290914581816287e-06,
      "loss": 0.72028284,
      "memory(GiB)": 135.77,
      "step": 16840,
      "train_speed(iter/s)": 0.20285
    },
    {
      "acc": 0.77796297,
      "epoch": 0.3931232593067849,
      "grad_norm": 6.1875,
      "learning_rate": 9.289944511415086e-06,
      "loss": 0.80314026,
      "memory(GiB)": 135.77,
      "step": 16850,
      "train_speed(iter/s)": 0.202906
    },
    {
      "acc": 0.75296588,
      "epoch": 0.3933565668790738,
      "grad_norm": 4.28125,
      "learning_rate": 9.28897382862555e-06,
      "loss": 0.88018427,
      "memory(GiB)": 135.77,
      "step": 16860,
      "train_speed(iter/s)": 0.202965
    },
    {
      "acc": 0.74555116,
      "epoch": 0.3935898744513627,
      "grad_norm": 5.53125,
      "learning_rate": 9.288002533586247e-06,
      "loss": 0.92258759,
      "memory(GiB)": 135.77,
      "step": 16870,
      "train_speed(iter/s)": 0.203027
    },
    {
      "acc": 0.77875471,
      "epoch": 0.3938231820236516,
      "grad_norm": 5.125,
      "learning_rate": 9.287030626435828e-06,
      "loss": 0.81162071,
      "memory(GiB)": 135.77,
      "step": 16880,
      "train_speed(iter/s)": 0.20309
    },
    {
      "acc": 0.77456217,
      "epoch": 0.3940564895959405,
      "grad_norm": 5.1875,
      "learning_rate": 9.286058107313034e-06,
      "loss": 0.79409857,
      "memory(GiB)": 135.77,
      "step": 16890,
      "train_speed(iter/s)": 0.203154
    },
    {
      "acc": 0.76377172,
      "epoch": 0.3942897971682293,
      "grad_norm": 5.8125,
      "learning_rate": 9.285084976356689e-06,
      "loss": 0.84241772,
      "memory(GiB)": 135.77,
      "step": 16900,
      "train_speed(iter/s)": 0.203216
    },
    {
      "acc": 0.76552515,
      "epoch": 0.3945231047405182,
      "grad_norm": 4.9375,
      "learning_rate": 9.284111233705709e-06,
      "loss": 0.86844196,
      "memory(GiB)": 135.77,
      "step": 16910,
      "train_speed(iter/s)": 0.20328
    },
    {
      "acc": 0.7652967,
      "epoch": 0.3947564123128071,
      "grad_norm": 4.28125,
      "learning_rate": 9.283136879499094e-06,
      "loss": 0.85671463,
      "memory(GiB)": 135.77,
      "step": 16920,
      "train_speed(iter/s)": 0.203337
    },
    {
      "acc": 0.77319775,
      "epoch": 0.394989719885096,
      "grad_norm": 4.84375,
      "learning_rate": 9.282161913875933e-06,
      "loss": 0.83186893,
      "memory(GiB)": 135.77,
      "step": 16930,
      "train_speed(iter/s)": 0.203398
    },
    {
      "acc": 0.76866665,
      "epoch": 0.3952230274573849,
      "grad_norm": 4.71875,
      "learning_rate": 9.281186336975406e-06,
      "loss": 0.83866825,
      "memory(GiB)": 135.77,
      "step": 16940,
      "train_speed(iter/s)": 0.203462
    },
    {
      "acc": 0.76994901,
      "epoch": 0.3954563350296738,
      "grad_norm": 5.78125,
      "learning_rate": 9.28021014893677e-06,
      "loss": 0.83855286,
      "memory(GiB)": 135.77,
      "step": 16950,
      "train_speed(iter/s)": 0.203521
    },
    {
      "acc": 0.77082605,
      "epoch": 0.3956896426019627,
      "grad_norm": 6.09375,
      "learning_rate": 9.27923334989938e-06,
      "loss": 0.83198881,
      "memory(GiB)": 135.77,
      "step": 16960,
      "train_speed(iter/s)": 0.203576
    },
    {
      "acc": 0.75358386,
      "epoch": 0.3959229501742516,
      "grad_norm": 5.28125,
      "learning_rate": 9.278255940002671e-06,
      "loss": 0.89232082,
      "memory(GiB)": 135.77,
      "step": 16970,
      "train_speed(iter/s)": 0.20364
    },
    {
      "acc": 0.76478891,
      "epoch": 0.3961562577465405,
      "grad_norm": 9.0625,
      "learning_rate": 9.27727791938617e-06,
      "loss": 0.84635849,
      "memory(GiB)": 135.77,
      "step": 16980,
      "train_speed(iter/s)": 0.203705
    },
    {
      "acc": 0.7693079,
      "epoch": 0.3963895653188294,
      "grad_norm": 7.15625,
      "learning_rate": 9.27629928818949e-06,
      "loss": 0.83523731,
      "memory(GiB)": 135.77,
      "step": 16990,
      "train_speed(iter/s)": 0.203767
    },
    {
      "acc": 0.76514502,
      "epoch": 0.39662287289111825,
      "grad_norm": 6.40625,
      "learning_rate": 9.275320046552328e-06,
      "loss": 0.85309467,
      "memory(GiB)": 135.77,
      "step": 17000,
      "train_speed(iter/s)": 0.203827
    },
    {
      "epoch": 0.39662287289111825,
      "eval_acc": 0.7371956278702851,
      "eval_loss": 0.8326179385185242,
      "eval_runtime": 1270.543,
      "eval_samples_per_second": 28.327,
      "eval_steps_per_second": 14.164,
      "step": 17000
    },
    {
      "acc": 0.76229954,
      "epoch": 0.39685618046340715,
      "grad_norm": 6.625,
      "learning_rate": 9.274340194614471e-06,
      "loss": 0.83733425,
      "memory(GiB)": 135.77,
      "step": 17010,
      "train_speed(iter/s)": 0.200789
    },
    {
      "acc": 0.77126536,
      "epoch": 0.39708948803569605,
      "grad_norm": 8.75,
      "learning_rate": 9.273359732515793e-06,
      "loss": 0.83545189,
      "memory(GiB)": 135.77,
      "step": 17020,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.76169815,
      "epoch": 0.39732279560798495,
      "grad_norm": 5.84375,
      "learning_rate": 9.272378660396255e-06,
      "loss": 0.87710171,
      "memory(GiB)": 135.77,
      "step": 17030,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.77218542,
      "epoch": 0.39755610318027385,
      "grad_norm": 5.0,
      "learning_rate": 9.271396978395904e-06,
      "loss": 0.83802376,
      "memory(GiB)": 135.77,
      "step": 17040,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.76550169,
      "epoch": 0.39778941075256274,
      "grad_norm": 5.25,
      "learning_rate": 9.270414686654875e-06,
      "loss": 0.85740089,
      "memory(GiB)": 135.77,
      "step": 17050,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.7832613,
      "epoch": 0.39802271832485164,
      "grad_norm": 7.125,
      "learning_rate": 9.269431785313391e-06,
      "loss": 0.76584721,
      "memory(GiB)": 135.77,
      "step": 17060,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.77678032,
      "epoch": 0.39825602589714054,
      "grad_norm": 7.3125,
      "learning_rate": 9.268448274511759e-06,
      "loss": 0.81068668,
      "memory(GiB)": 135.77,
      "step": 17070,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.77908564,
      "epoch": 0.39848933346942944,
      "grad_norm": 5.96875,
      "learning_rate": 9.267464154390375e-06,
      "loss": 0.79487085,
      "memory(GiB)": 135.77,
      "step": 17080,
      "train_speed(iter/s)": 0.201218
    },
    {
      "acc": 0.80031281,
      "epoch": 0.3987226410417183,
      "grad_norm": 4.46875,
      "learning_rate": 9.266479425089725e-06,
      "loss": 0.71730728,
      "memory(GiB)": 135.77,
      "step": 17090,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.77682505,
      "epoch": 0.3989559486140072,
      "grad_norm": 10.0625,
      "learning_rate": 9.265494086750375e-06,
      "loss": 0.80053043,
      "memory(GiB)": 135.77,
      "step": 17100,
      "train_speed(iter/s)": 0.201344
    },
    {
      "acc": 0.77470179,
      "epoch": 0.3991892561862961,
      "grad_norm": 5.125,
      "learning_rate": 9.264508139512985e-06,
      "loss": 0.81498718,
      "memory(GiB)": 135.77,
      "step": 17110,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.77046309,
      "epoch": 0.399422563758585,
      "grad_norm": 6.25,
      "learning_rate": 9.263521583518293e-06,
      "loss": 0.82258339,
      "memory(GiB)": 135.77,
      "step": 17120,
      "train_speed(iter/s)": 0.20146
    },
    {
      "acc": 0.79407024,
      "epoch": 0.3996558713308739,
      "grad_norm": 8.25,
      "learning_rate": 9.262534418907137e-06,
      "loss": 0.72604542,
      "memory(GiB)": 135.77,
      "step": 17130,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.7635807,
      "epoch": 0.3998891789031628,
      "grad_norm": 5.6875,
      "learning_rate": 9.26154664582043e-06,
      "loss": 0.8567337,
      "memory(GiB)": 135.77,
      "step": 17140,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.78804669,
      "epoch": 0.4001224864754517,
      "grad_norm": 6.65625,
      "learning_rate": 9.260558264399177e-06,
      "loss": 0.73743448,
      "memory(GiB)": 135.77,
      "step": 17150,
      "train_speed(iter/s)": 0.201638
    },
    {
      "acc": 0.76328459,
      "epoch": 0.4003557940477406,
      "grad_norm": 5.5,
      "learning_rate": 9.25956927478447e-06,
      "loss": 0.85945473,
      "memory(GiB)": 135.77,
      "step": 17160,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.77089682,
      "epoch": 0.4005891016200295,
      "grad_norm": 5.4375,
      "learning_rate": 9.258579677117486e-06,
      "loss": 0.81945114,
      "memory(GiB)": 135.77,
      "step": 17170,
      "train_speed(iter/s)": 0.201758
    },
    {
      "acc": 0.7839241,
      "epoch": 0.4008224091923184,
      "grad_norm": 5.125,
      "learning_rate": 9.25758947153949e-06,
      "loss": 0.77010636,
      "memory(GiB)": 135.77,
      "step": 17180,
      "train_speed(iter/s)": 0.201814
    },
    {
      "acc": 0.76271248,
      "epoch": 0.4010557167646072,
      "grad_norm": 3.828125,
      "learning_rate": 9.256598658191834e-06,
      "loss": 0.85306063,
      "memory(GiB)": 135.77,
      "step": 17190,
      "train_speed(iter/s)": 0.201875
    },
    {
      "acc": 0.77881155,
      "epoch": 0.4012890243368961,
      "grad_norm": 6.875,
      "learning_rate": 9.255607237215957e-06,
      "loss": 0.7829484,
      "memory(GiB)": 135.77,
      "step": 17200,
      "train_speed(iter/s)": 0.201937
    },
    {
      "acc": 0.75986853,
      "epoch": 0.401522331909185,
      "grad_norm": 5.96875,
      "learning_rate": 9.254615208753381e-06,
      "loss": 0.8725729,
      "memory(GiB)": 135.77,
      "step": 17210,
      "train_speed(iter/s)": 0.202
    },
    {
      "acc": 0.77503152,
      "epoch": 0.4017556394814739,
      "grad_norm": 4.0,
      "learning_rate": 9.253622572945722e-06,
      "loss": 0.82128544,
      "memory(GiB)": 135.77,
      "step": 17220,
      "train_speed(iter/s)": 0.202063
    },
    {
      "acc": 0.75122519,
      "epoch": 0.4019889470537628,
      "grad_norm": 5.53125,
      "learning_rate": 9.252629329934676e-06,
      "loss": 0.91885424,
      "memory(GiB)": 135.77,
      "step": 17230,
      "train_speed(iter/s)": 0.202123
    },
    {
      "acc": 0.78238068,
      "epoch": 0.4022222546260517,
      "grad_norm": 4.25,
      "learning_rate": 9.251635479862029e-06,
      "loss": 0.7822403,
      "memory(GiB)": 135.77,
      "step": 17240,
      "train_speed(iter/s)": 0.202187
    },
    {
      "acc": 0.7734271,
      "epoch": 0.4024555621983406,
      "grad_norm": 5.9375,
      "learning_rate": 9.25064102286965e-06,
      "loss": 0.83377533,
      "memory(GiB)": 135.77,
      "step": 17250,
      "train_speed(iter/s)": 0.202249
    },
    {
      "acc": 0.78481631,
      "epoch": 0.4026888697706295,
      "grad_norm": 8.6875,
      "learning_rate": 9.249645959099503e-06,
      "loss": 0.76742496,
      "memory(GiB)": 135.77,
      "step": 17260,
      "train_speed(iter/s)": 0.202311
    },
    {
      "acc": 0.78067999,
      "epoch": 0.4029221773429184,
      "grad_norm": 6.4375,
      "learning_rate": 9.248650288693628e-06,
      "loss": 0.78222075,
      "memory(GiB)": 135.77,
      "step": 17270,
      "train_speed(iter/s)": 0.202371
    },
    {
      "acc": 0.78589168,
      "epoch": 0.40315548491520725,
      "grad_norm": 4.6875,
      "learning_rate": 9.247654011794158e-06,
      "loss": 0.7706872,
      "memory(GiB)": 135.77,
      "step": 17280,
      "train_speed(iter/s)": 0.20243
    },
    {
      "acc": 0.76477909,
      "epoch": 0.40338879248749615,
      "grad_norm": 9.0,
      "learning_rate": 9.246657128543313e-06,
      "loss": 0.85425339,
      "memory(GiB)": 135.77,
      "step": 17290,
      "train_speed(iter/s)": 0.202491
    },
    {
      "acc": 0.77842503,
      "epoch": 0.40362210005978505,
      "grad_norm": 5.4375,
      "learning_rate": 9.245659639083396e-06,
      "loss": 0.79319229,
      "memory(GiB)": 135.77,
      "step": 17300,
      "train_speed(iter/s)": 0.202549
    },
    {
      "acc": 0.78424797,
      "epoch": 0.40385540763207395,
      "grad_norm": 4.375,
      "learning_rate": 9.244661543556799e-06,
      "loss": 0.7634676,
      "memory(GiB)": 135.77,
      "step": 17310,
      "train_speed(iter/s)": 0.202609
    },
    {
      "acc": 0.77571602,
      "epoch": 0.40408871520436285,
      "grad_norm": 8.375,
      "learning_rate": 9.243662842106e-06,
      "loss": 0.80258932,
      "memory(GiB)": 135.77,
      "step": 17320,
      "train_speed(iter/s)": 0.202667
    },
    {
      "acc": 0.76631498,
      "epoch": 0.40432202277665175,
      "grad_norm": 4.25,
      "learning_rate": 9.242663534873562e-06,
      "loss": 0.85109797,
      "memory(GiB)": 135.77,
      "step": 17330,
      "train_speed(iter/s)": 0.202731
    },
    {
      "acc": 0.76990891,
      "epoch": 0.40455533034894064,
      "grad_norm": 5.40625,
      "learning_rate": 9.241663622002137e-06,
      "loss": 0.84066992,
      "memory(GiB)": 135.77,
      "step": 17340,
      "train_speed(iter/s)": 0.202789
    },
    {
      "acc": 0.76096568,
      "epoch": 0.40478863792122954,
      "grad_norm": 5.59375,
      "learning_rate": 9.240663103634464e-06,
      "loss": 0.86097231,
      "memory(GiB)": 135.77,
      "step": 17350,
      "train_speed(iter/s)": 0.202851
    },
    {
      "acc": 0.78643513,
      "epoch": 0.40502194549351844,
      "grad_norm": 7.78125,
      "learning_rate": 9.239661979913364e-06,
      "loss": 0.7529027,
      "memory(GiB)": 135.77,
      "step": 17360,
      "train_speed(iter/s)": 0.20291
    },
    {
      "acc": 0.78234057,
      "epoch": 0.40525525306580734,
      "grad_norm": 5.71875,
      "learning_rate": 9.238660250981748e-06,
      "loss": 0.76194448,
      "memory(GiB)": 135.77,
      "step": 17370,
      "train_speed(iter/s)": 0.202971
    },
    {
      "acc": 0.7768539,
      "epoch": 0.4054885606380962,
      "grad_norm": 5.0,
      "learning_rate": 9.237657916982612e-06,
      "loss": 0.79744596,
      "memory(GiB)": 135.77,
      "step": 17380,
      "train_speed(iter/s)": 0.203032
    },
    {
      "acc": 0.76710577,
      "epoch": 0.4057218682103851,
      "grad_norm": 13.4375,
      "learning_rate": 9.236654978059039e-06,
      "loss": 0.82225361,
      "memory(GiB)": 135.77,
      "step": 17390,
      "train_speed(iter/s)": 0.203092
    },
    {
      "acc": 0.78320732,
      "epoch": 0.405955175782674,
      "grad_norm": 4.53125,
      "learning_rate": 9.2356514343542e-06,
      "loss": 0.79602137,
      "memory(GiB)": 135.77,
      "step": 17400,
      "train_speed(iter/s)": 0.203151
    },
    {
      "acc": 0.7971508,
      "epoch": 0.4061884833549629,
      "grad_norm": 3.9375,
      "learning_rate": 9.234647286011347e-06,
      "loss": 0.73624454,
      "memory(GiB)": 135.77,
      "step": 17410,
      "train_speed(iter/s)": 0.203208
    },
    {
      "acc": 0.79400434,
      "epoch": 0.4064217909272518,
      "grad_norm": 5.75,
      "learning_rate": 9.233642533173827e-06,
      "loss": 0.73229074,
      "memory(GiB)": 135.77,
      "step": 17420,
      "train_speed(iter/s)": 0.203266
    },
    {
      "acc": 0.78245573,
      "epoch": 0.4066550984995407,
      "grad_norm": 7.125,
      "learning_rate": 9.232637175985064e-06,
      "loss": 0.79270024,
      "memory(GiB)": 135.77,
      "step": 17430,
      "train_speed(iter/s)": 0.203326
    },
    {
      "acc": 0.7708353,
      "epoch": 0.4068884060718296,
      "grad_norm": 5.40625,
      "learning_rate": 9.231631214588572e-06,
      "loss": 0.80469551,
      "memory(GiB)": 135.77,
      "step": 17440,
      "train_speed(iter/s)": 0.203388
    },
    {
      "acc": 0.77979951,
      "epoch": 0.4071217136441185,
      "grad_norm": 5.3125,
      "learning_rate": 9.230624649127956e-06,
      "loss": 0.7915205,
      "memory(GiB)": 135.77,
      "step": 17450,
      "train_speed(iter/s)": 0.203451
    },
    {
      "acc": 0.78297606,
      "epoch": 0.4073550212164074,
      "grad_norm": 5.21875,
      "learning_rate": 9.2296174797469e-06,
      "loss": 0.78460536,
      "memory(GiB)": 135.77,
      "step": 17460,
      "train_speed(iter/s)": 0.20351
    },
    {
      "acc": 0.78100643,
      "epoch": 0.4075883287886963,
      "grad_norm": 4.1875,
      "learning_rate": 9.228609706589175e-06,
      "loss": 0.80171537,
      "memory(GiB)": 135.77,
      "step": 17470,
      "train_speed(iter/s)": 0.203567
    },
    {
      "acc": 0.76642833,
      "epoch": 0.4078216363609851,
      "grad_norm": 5.1875,
      "learning_rate": 9.227601329798645e-06,
      "loss": 0.86399326,
      "memory(GiB)": 135.77,
      "step": 17480,
      "train_speed(iter/s)": 0.203629
    },
    {
      "acc": 0.76468382,
      "epoch": 0.408054943933274,
      "grad_norm": 5.15625,
      "learning_rate": 9.226592349519254e-06,
      "loss": 0.84064198,
      "memory(GiB)": 135.77,
      "step": 17490,
      "train_speed(iter/s)": 0.203689
    },
    {
      "acc": 0.757623,
      "epoch": 0.4082882515055629,
      "grad_norm": 7.5,
      "learning_rate": 9.225582765895032e-06,
      "loss": 0.88375015,
      "memory(GiB)": 135.77,
      "step": 17500,
      "train_speed(iter/s)": 0.203747
    },
    {
      "epoch": 0.4082882515055629,
      "eval_acc": 0.7373759502758565,
      "eval_loss": 0.8317380547523499,
      "eval_runtime": 1270.3611,
      "eval_samples_per_second": 28.331,
      "eval_steps_per_second": 14.166,
      "step": 17500
    },
    {
      "acc": 0.76951265,
      "epoch": 0.4085215590778518,
      "grad_norm": 6.9375,
      "learning_rate": 9.224572579070097e-06,
      "loss": 0.83398447,
      "memory(GiB)": 135.77,
      "step": 17510,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.77614574,
      "epoch": 0.4087548666501407,
      "grad_norm": 5.53125,
      "learning_rate": 9.223561789188655e-06,
      "loss": 0.80068007,
      "memory(GiB)": 135.77,
      "step": 17520,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.76483154,
      "epoch": 0.4089881742224296,
      "grad_norm": 6.59375,
      "learning_rate": 9.222550396394994e-06,
      "loss": 0.82644291,
      "memory(GiB)": 135.77,
      "step": 17530,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.76731434,
      "epoch": 0.4092214817947185,
      "grad_norm": 11.25,
      "learning_rate": 9.221538400833489e-06,
      "loss": 0.84105368,
      "memory(GiB)": 135.77,
      "step": 17540,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.79679441,
      "epoch": 0.4094547893670074,
      "grad_norm": 5.625,
      "learning_rate": 9.220525802648605e-06,
      "loss": 0.72740088,
      "memory(GiB)": 135.77,
      "step": 17550,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.77622986,
      "epoch": 0.4096880969392963,
      "grad_norm": 5.5625,
      "learning_rate": 9.219512601984889e-06,
      "loss": 0.81656227,
      "memory(GiB)": 135.77,
      "step": 17560,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.78093081,
      "epoch": 0.40992140451158515,
      "grad_norm": 5.8125,
      "learning_rate": 9.218498798986975e-06,
      "loss": 0.75747585,
      "memory(GiB)": 135.77,
      "step": 17570,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.75355787,
      "epoch": 0.41015471208387405,
      "grad_norm": 4.84375,
      "learning_rate": 9.217484393799582e-06,
      "loss": 0.91436329,
      "memory(GiB)": 135.77,
      "step": 17580,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.76489058,
      "epoch": 0.41038801965616295,
      "grad_norm": 5.125,
      "learning_rate": 9.216469386567517e-06,
      "loss": 0.86888409,
      "memory(GiB)": 135.77,
      "step": 17590,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.79336662,
      "epoch": 0.41062132722845185,
      "grad_norm": 4.875,
      "learning_rate": 9.215453777435672e-06,
      "loss": 0.71763039,
      "memory(GiB)": 135.77,
      "step": 17600,
      "train_speed(iter/s)": 0.201315
    },
    {
      "acc": 0.80592194,
      "epoch": 0.41085463480074075,
      "grad_norm": 7.0,
      "learning_rate": 9.214437566549026e-06,
      "loss": 0.7039835,
      "memory(GiB)": 135.77,
      "step": 17610,
      "train_speed(iter/s)": 0.201373
    },
    {
      "acc": 0.75960789,
      "epoch": 0.41108794237302965,
      "grad_norm": 5.3125,
      "learning_rate": 9.21342075405264e-06,
      "loss": 0.88718882,
      "memory(GiB)": 135.77,
      "step": 17620,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.75959663,
      "epoch": 0.41132124994531855,
      "grad_norm": 5.09375,
      "learning_rate": 9.212403340091667e-06,
      "loss": 0.88778267,
      "memory(GiB)": 135.77,
      "step": 17630,
      "train_speed(iter/s)": 0.201493
    },
    {
      "acc": 0.76623917,
      "epoch": 0.41155455751760744,
      "grad_norm": 5.65625,
      "learning_rate": 9.21138532481134e-06,
      "loss": 0.84202518,
      "memory(GiB)": 135.77,
      "step": 17640,
      "train_speed(iter/s)": 0.201554
    },
    {
      "acc": 0.76951289,
      "epoch": 0.41178786508989634,
      "grad_norm": 5.375,
      "learning_rate": 9.210366708356982e-06,
      "loss": 0.81451988,
      "memory(GiB)": 135.77,
      "step": 17650,
      "train_speed(iter/s)": 0.201614
    },
    {
      "acc": 0.76205425,
      "epoch": 0.41202117266218524,
      "grad_norm": 5.0625,
      "learning_rate": 9.209347490874e-06,
      "loss": 0.86257401,
      "memory(GiB)": 135.77,
      "step": 17660,
      "train_speed(iter/s)": 0.201674
    },
    {
      "acc": 0.76792526,
      "epoch": 0.4122544802344741,
      "grad_norm": 4.46875,
      "learning_rate": 9.208327672507883e-06,
      "loss": 0.82535515,
      "memory(GiB)": 135.77,
      "step": 17670,
      "train_speed(iter/s)": 0.201736
    },
    {
      "acc": 0.77031803,
      "epoch": 0.412487787806763,
      "grad_norm": 6.40625,
      "learning_rate": 9.207307253404216e-06,
      "loss": 0.83331718,
      "memory(GiB)": 135.77,
      "step": 17680,
      "train_speed(iter/s)": 0.201795
    },
    {
      "acc": 0.78183222,
      "epoch": 0.4127210953790519,
      "grad_norm": 5.59375,
      "learning_rate": 9.20628623370866e-06,
      "loss": 0.7861207,
      "memory(GiB)": 135.77,
      "step": 17690,
      "train_speed(iter/s)": 0.201854
    },
    {
      "acc": 0.78258052,
      "epoch": 0.4129544029513408,
      "grad_norm": 4.71875,
      "learning_rate": 9.205264613566968e-06,
      "loss": 0.78774872,
      "memory(GiB)": 135.77,
      "step": 17700,
      "train_speed(iter/s)": 0.201912
    },
    {
      "acc": 0.76604571,
      "epoch": 0.4131877105236297,
      "grad_norm": 5.78125,
      "learning_rate": 9.204242393124973e-06,
      "loss": 0.85110035,
      "memory(GiB)": 135.77,
      "step": 17710,
      "train_speed(iter/s)": 0.201968
    },
    {
      "acc": 0.77456961,
      "epoch": 0.4134210180959186,
      "grad_norm": 4.25,
      "learning_rate": 9.203219572528597e-06,
      "loss": 0.7974988,
      "memory(GiB)": 135.77,
      "step": 17720,
      "train_speed(iter/s)": 0.202026
    },
    {
      "acc": 0.78755922,
      "epoch": 0.4136543256682075,
      "grad_norm": 6.59375,
      "learning_rate": 9.202196151923849e-06,
      "loss": 0.75697212,
      "memory(GiB)": 135.77,
      "step": 17730,
      "train_speed(iter/s)": 0.202085
    },
    {
      "acc": 0.75414586,
      "epoch": 0.4138876332404964,
      "grad_norm": 6.90625,
      "learning_rate": 9.201172131456821e-06,
      "loss": 0.8852272,
      "memory(GiB)": 135.77,
      "step": 17740,
      "train_speed(iter/s)": 0.202146
    },
    {
      "acc": 0.75268192,
      "epoch": 0.4141209408127853,
      "grad_norm": 6.1875,
      "learning_rate": 9.20014751127369e-06,
      "loss": 0.89101114,
      "memory(GiB)": 135.77,
      "step": 17750,
      "train_speed(iter/s)": 0.202204
    },
    {
      "acc": 0.76879907,
      "epoch": 0.4143542483850742,
      "grad_norm": 4.65625,
      "learning_rate": 9.199122291520724e-06,
      "loss": 0.82659073,
      "memory(GiB)": 135.77,
      "step": 17760,
      "train_speed(iter/s)": 0.202266
    },
    {
      "acc": 0.78848228,
      "epoch": 0.414587555957363,
      "grad_norm": 4.625,
      "learning_rate": 9.198096472344269e-06,
      "loss": 0.76872091,
      "memory(GiB)": 135.77,
      "step": 17770,
      "train_speed(iter/s)": 0.202325
    },
    {
      "acc": 0.76771512,
      "epoch": 0.4148208635296519,
      "grad_norm": 4.34375,
      "learning_rate": 9.197070053890764e-06,
      "loss": 0.82841167,
      "memory(GiB)": 135.77,
      "step": 17780,
      "train_speed(iter/s)": 0.202385
    },
    {
      "acc": 0.78823538,
      "epoch": 0.4150541711019408,
      "grad_norm": 3.953125,
      "learning_rate": 9.196043036306726e-06,
      "loss": 0.76270905,
      "memory(GiB)": 135.77,
      "step": 17790,
      "train_speed(iter/s)": 0.202446
    },
    {
      "acc": 0.78680201,
      "epoch": 0.4152874786742297,
      "grad_norm": 4.9375,
      "learning_rate": 9.195015419738765e-06,
      "loss": 0.77416258,
      "memory(GiB)": 135.77,
      "step": 17800,
      "train_speed(iter/s)": 0.202509
    },
    {
      "acc": 0.77593956,
      "epoch": 0.4155207862465186,
      "grad_norm": 6.6875,
      "learning_rate": 9.193987204333573e-06,
      "loss": 0.79424219,
      "memory(GiB)": 135.77,
      "step": 17810,
      "train_speed(iter/s)": 0.202572
    },
    {
      "acc": 0.76831007,
      "epoch": 0.4157540938188075,
      "grad_norm": 7.0,
      "learning_rate": 9.192958390237923e-06,
      "loss": 0.84353199,
      "memory(GiB)": 135.77,
      "step": 17820,
      "train_speed(iter/s)": 0.202631
    },
    {
      "acc": 0.7490344,
      "epoch": 0.4159874013910964,
      "grad_norm": 4.9375,
      "learning_rate": 9.19192897759868e-06,
      "loss": 0.92404823,
      "memory(GiB)": 135.77,
      "step": 17830,
      "train_speed(iter/s)": 0.202692
    },
    {
      "acc": 0.76985192,
      "epoch": 0.4162207089633853,
      "grad_norm": 9.25,
      "learning_rate": 9.190898966562796e-06,
      "loss": 0.83063097,
      "memory(GiB)": 135.77,
      "step": 17840,
      "train_speed(iter/s)": 0.202753
    },
    {
      "acc": 0.76316862,
      "epoch": 0.4164540165356742,
      "grad_norm": 5.8125,
      "learning_rate": 9.1898683572773e-06,
      "loss": 0.8611824,
      "memory(GiB)": 135.77,
      "step": 17850,
      "train_speed(iter/s)": 0.202812
    },
    {
      "acc": 0.75131006,
      "epoch": 0.41668732410796305,
      "grad_norm": 7.21875,
      "learning_rate": 9.188837149889316e-06,
      "loss": 0.92140636,
      "memory(GiB)": 135.77,
      "step": 17860,
      "train_speed(iter/s)": 0.202873
    },
    {
      "acc": 0.77328281,
      "epoch": 0.41692063168025195,
      "grad_norm": 5.0625,
      "learning_rate": 9.187805344546044e-06,
      "loss": 0.81961975,
      "memory(GiB)": 135.77,
      "step": 17870,
      "train_speed(iter/s)": 0.202932
    },
    {
      "acc": 0.78284941,
      "epoch": 0.41715393925254085,
      "grad_norm": 5.375,
      "learning_rate": 9.186772941394776e-06,
      "loss": 0.78479848,
      "memory(GiB)": 135.77,
      "step": 17880,
      "train_speed(iter/s)": 0.202994
    },
    {
      "acc": 0.77911134,
      "epoch": 0.41738724682482975,
      "grad_norm": 6.78125,
      "learning_rate": 9.185739940582885e-06,
      "loss": 0.82526703,
      "memory(GiB)": 135.77,
      "step": 17890,
      "train_speed(iter/s)": 0.203057
    },
    {
      "acc": 0.77189975,
      "epoch": 0.41762055439711865,
      "grad_norm": 7.34375,
      "learning_rate": 9.184706342257835e-06,
      "loss": 0.80744982,
      "memory(GiB)": 135.77,
      "step": 17900,
      "train_speed(iter/s)": 0.20312
    },
    {
      "acc": 0.77462916,
      "epoch": 0.41785386196940755,
      "grad_norm": 6.46875,
      "learning_rate": 9.183672146567171e-06,
      "loss": 0.81883631,
      "memory(GiB)": 135.77,
      "step": 17910,
      "train_speed(iter/s)": 0.203176
    },
    {
      "acc": 0.75889959,
      "epoch": 0.41808716954169645,
      "grad_norm": 4.78125,
      "learning_rate": 9.182637353658523e-06,
      "loss": 0.85708733,
      "memory(GiB)": 135.77,
      "step": 17920,
      "train_speed(iter/s)": 0.203236
    },
    {
      "acc": 0.77901268,
      "epoch": 0.41832047711398535,
      "grad_norm": 11.125,
      "learning_rate": 9.181601963679607e-06,
      "loss": 0.80118008,
      "memory(GiB)": 135.77,
      "step": 17930,
      "train_speed(iter/s)": 0.203298
    },
    {
      "acc": 0.76691198,
      "epoch": 0.41855378468627424,
      "grad_norm": 6.09375,
      "learning_rate": 9.180565976778226e-06,
      "loss": 0.82437811,
      "memory(GiB)": 135.77,
      "step": 17940,
      "train_speed(iter/s)": 0.203358
    },
    {
      "acc": 0.79058146,
      "epoch": 0.41878709225856314,
      "grad_norm": 5.59375,
      "learning_rate": 9.179529393102265e-06,
      "loss": 0.75918694,
      "memory(GiB)": 135.77,
      "step": 17950,
      "train_speed(iter/s)": 0.203415
    },
    {
      "acc": 0.78140459,
      "epoch": 0.419020399830852,
      "grad_norm": 4.46875,
      "learning_rate": 9.1784922127997e-06,
      "loss": 0.77219858,
      "memory(GiB)": 135.77,
      "step": 17960,
      "train_speed(iter/s)": 0.203476
    },
    {
      "acc": 0.76335707,
      "epoch": 0.4192537074031409,
      "grad_norm": 4.625,
      "learning_rate": 9.177454436018584e-06,
      "loss": 0.84384384,
      "memory(GiB)": 135.77,
      "step": 17970,
      "train_speed(iter/s)": 0.203534
    },
    {
      "acc": 0.77697759,
      "epoch": 0.4194870149754298,
      "grad_norm": 6.5625,
      "learning_rate": 9.17641606290706e-06,
      "loss": 0.85184278,
      "memory(GiB)": 135.77,
      "step": 17980,
      "train_speed(iter/s)": 0.203595
    },
    {
      "acc": 0.75867949,
      "epoch": 0.4197203225477187,
      "grad_norm": 7.40625,
      "learning_rate": 9.175377093613359e-06,
      "loss": 0.84790859,
      "memory(GiB)": 135.77,
      "step": 17990,
      "train_speed(iter/s)": 0.203654
    },
    {
      "acc": 0.77024422,
      "epoch": 0.4199536301200076,
      "grad_norm": 6.71875,
      "learning_rate": 9.174337528285787e-06,
      "loss": 0.83740625,
      "memory(GiB)": 135.77,
      "step": 18000,
      "train_speed(iter/s)": 0.203712
    },
    {
      "epoch": 0.4199536301200076,
      "eval_acc": 0.7375597958545385,
      "eval_loss": 0.8316662907600403,
      "eval_runtime": 1270.1578,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 18000
    },
    {
      "acc": 0.75532227,
      "epoch": 0.4201869376922965,
      "grad_norm": 4.65625,
      "learning_rate": 9.173297367072748e-06,
      "loss": 0.87728586,
      "memory(GiB)": 135.77,
      "step": 18010,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.78155527,
      "epoch": 0.4204202452645854,
      "grad_norm": 6.9375,
      "learning_rate": 9.172256610122721e-06,
      "loss": 0.79191427,
      "memory(GiB)": 135.77,
      "step": 18020,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.7721571,
      "epoch": 0.4206535528368743,
      "grad_norm": 7.78125,
      "learning_rate": 9.171215257584277e-06,
      "loss": 0.82012405,
      "memory(GiB)": 135.77,
      "step": 18030,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.74781427,
      "epoch": 0.4208868604091632,
      "grad_norm": 5.1875,
      "learning_rate": 9.170173309606063e-06,
      "loss": 0.92548695,
      "memory(GiB)": 135.77,
      "step": 18040,
      "train_speed(iter/s)": 0.201023
    },
    {
      "acc": 0.77438364,
      "epoch": 0.421120167981452,
      "grad_norm": 5.3125,
      "learning_rate": 9.169130766336824e-06,
      "loss": 0.8166708,
      "memory(GiB)": 135.77,
      "step": 18050,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.76485076,
      "epoch": 0.4213534755537409,
      "grad_norm": 4.90625,
      "learning_rate": 9.168087627925377e-06,
      "loss": 0.85983076,
      "memory(GiB)": 135.77,
      "step": 18060,
      "train_speed(iter/s)": 0.201133
    },
    {
      "acc": 0.77190371,
      "epoch": 0.4215867831260298,
      "grad_norm": 6.40625,
      "learning_rate": 9.167043894520633e-06,
      "loss": 0.81317501,
      "memory(GiB)": 135.77,
      "step": 18070,
      "train_speed(iter/s)": 0.201191
    },
    {
      "acc": 0.77723899,
      "epoch": 0.4218200906983187,
      "grad_norm": 4.6875,
      "learning_rate": 9.165999566271584e-06,
      "loss": 0.81175861,
      "memory(GiB)": 135.77,
      "step": 18080,
      "train_speed(iter/s)": 0.201248
    },
    {
      "acc": 0.77668056,
      "epoch": 0.4220533982706076,
      "grad_norm": 5.8125,
      "learning_rate": 9.164954643327306e-06,
      "loss": 0.79176364,
      "memory(GiB)": 135.77,
      "step": 18090,
      "train_speed(iter/s)": 0.201302
    },
    {
      "acc": 0.78481464,
      "epoch": 0.4222867058428965,
      "grad_norm": 4.5625,
      "learning_rate": 9.163909125836965e-06,
      "loss": 0.7773984,
      "memory(GiB)": 135.77,
      "step": 18100,
      "train_speed(iter/s)": 0.201359
    },
    {
      "acc": 0.77743702,
      "epoch": 0.4225200134151854,
      "grad_norm": 6.1875,
      "learning_rate": 9.162863013949803e-06,
      "loss": 0.7793623,
      "memory(GiB)": 135.77,
      "step": 18110,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.75463772,
      "epoch": 0.4227533209874743,
      "grad_norm": 5.96875,
      "learning_rate": 9.161816307815157e-06,
      "loss": 0.88482685,
      "memory(GiB)": 135.77,
      "step": 18120,
      "train_speed(iter/s)": 0.201472
    },
    {
      "acc": 0.78212109,
      "epoch": 0.4229866285597632,
      "grad_norm": 5.53125,
      "learning_rate": 9.160769007582441e-06,
      "loss": 0.78088417,
      "memory(GiB)": 135.77,
      "step": 18130,
      "train_speed(iter/s)": 0.201529
    },
    {
      "acc": 0.77992072,
      "epoch": 0.4232199361320521,
      "grad_norm": 5.21875,
      "learning_rate": 9.15972111340116e-06,
      "loss": 0.80578728,
      "memory(GiB)": 135.77,
      "step": 18140,
      "train_speed(iter/s)": 0.201586
    },
    {
      "acc": 0.75823526,
      "epoch": 0.42345324370434095,
      "grad_norm": 7.90625,
      "learning_rate": 9.158672625420894e-06,
      "loss": 0.88358002,
      "memory(GiB)": 135.77,
      "step": 18150,
      "train_speed(iter/s)": 0.201641
    },
    {
      "acc": 0.78155651,
      "epoch": 0.42368655127662985,
      "grad_norm": 6.40625,
      "learning_rate": 9.157623543791323e-06,
      "loss": 0.7862711,
      "memory(GiB)": 135.77,
      "step": 18160,
      "train_speed(iter/s)": 0.201697
    },
    {
      "acc": 0.77112303,
      "epoch": 0.42391985884891875,
      "grad_norm": 6.96875,
      "learning_rate": 9.156573868662197e-06,
      "loss": 0.83190708,
      "memory(GiB)": 135.77,
      "step": 18170,
      "train_speed(iter/s)": 0.201751
    },
    {
      "acc": 0.78510017,
      "epoch": 0.42415316642120765,
      "grad_norm": 5.3125,
      "learning_rate": 9.155523600183359e-06,
      "loss": 0.75841799,
      "memory(GiB)": 135.77,
      "step": 18180,
      "train_speed(iter/s)": 0.201805
    },
    {
      "acc": 0.76126904,
      "epoch": 0.42438647399349655,
      "grad_norm": 6.5625,
      "learning_rate": 9.154472738504735e-06,
      "loss": 0.84513931,
      "memory(GiB)": 135.77,
      "step": 18190,
      "train_speed(iter/s)": 0.201858
    },
    {
      "acc": 0.77650328,
      "epoch": 0.42461978156578545,
      "grad_norm": 5.09375,
      "learning_rate": 9.153421283776334e-06,
      "loss": 0.80113697,
      "memory(GiB)": 135.77,
      "step": 18200,
      "train_speed(iter/s)": 0.201916
    },
    {
      "acc": 0.78944378,
      "epoch": 0.42485308913807435,
      "grad_norm": 4.0625,
      "learning_rate": 9.152369236148252e-06,
      "loss": 0.76571293,
      "memory(GiB)": 135.77,
      "step": 18210,
      "train_speed(iter/s)": 0.201971
    },
    {
      "acc": 0.77950716,
      "epoch": 0.42508639671036325,
      "grad_norm": 5.1875,
      "learning_rate": 9.151316595770665e-06,
      "loss": 0.78431196,
      "memory(GiB)": 135.77,
      "step": 18220,
      "train_speed(iter/s)": 0.202028
    },
    {
      "acc": 0.75776768,
      "epoch": 0.42531970428265214,
      "grad_norm": 6.96875,
      "learning_rate": 9.150263362793844e-06,
      "loss": 0.88814754,
      "memory(GiB)": 135.77,
      "step": 18230,
      "train_speed(iter/s)": 0.20209
    },
    {
      "acc": 0.7760685,
      "epoch": 0.42555301185494104,
      "grad_norm": 4.65625,
      "learning_rate": 9.14920953736813e-06,
      "loss": 0.81063766,
      "memory(GiB)": 135.77,
      "step": 18240,
      "train_speed(iter/s)": 0.202147
    },
    {
      "acc": 0.77113094,
      "epoch": 0.4257863194272299,
      "grad_norm": 5.03125,
      "learning_rate": 9.148155119643963e-06,
      "loss": 0.82983465,
      "memory(GiB)": 135.77,
      "step": 18250,
      "train_speed(iter/s)": 0.202206
    },
    {
      "acc": 0.76979413,
      "epoch": 0.4260196269995188,
      "grad_norm": 4.9375,
      "learning_rate": 9.147100109771856e-06,
      "loss": 0.83212032,
      "memory(GiB)": 135.77,
      "step": 18260,
      "train_speed(iter/s)": 0.202259
    },
    {
      "acc": 0.7747787,
      "epoch": 0.4262529345718077,
      "grad_norm": 4.78125,
      "learning_rate": 9.146044507902411e-06,
      "loss": 0.80783243,
      "memory(GiB)": 135.77,
      "step": 18270,
      "train_speed(iter/s)": 0.202315
    },
    {
      "acc": 0.78634143,
      "epoch": 0.4264862421440966,
      "grad_norm": 7.71875,
      "learning_rate": 9.144988314186321e-06,
      "loss": 0.76813116,
      "memory(GiB)": 135.77,
      "step": 18280,
      "train_speed(iter/s)": 0.202366
    },
    {
      "acc": 0.76954098,
      "epoch": 0.4267195497163855,
      "grad_norm": 5.75,
      "learning_rate": 9.143931528774351e-06,
      "loss": 0.82664013,
      "memory(GiB)": 135.77,
      "step": 18290,
      "train_speed(iter/s)": 0.202423
    },
    {
      "acc": 0.74711275,
      "epoch": 0.4269528572886744,
      "grad_norm": 5.65625,
      "learning_rate": 9.14287415181736e-06,
      "loss": 0.90971508,
      "memory(GiB)": 135.77,
      "step": 18300,
      "train_speed(iter/s)": 0.202481
    },
    {
      "acc": 0.75965652,
      "epoch": 0.4271861648609633,
      "grad_norm": 5.6875,
      "learning_rate": 9.141816183466286e-06,
      "loss": 0.85946274,
      "memory(GiB)": 135.77,
      "step": 18310,
      "train_speed(iter/s)": 0.202539
    },
    {
      "acc": 0.76663742,
      "epoch": 0.4274194724332522,
      "grad_norm": 5.5625,
      "learning_rate": 9.140757623872156e-06,
      "loss": 0.86080341,
      "memory(GiB)": 135.77,
      "step": 18320,
      "train_speed(iter/s)": 0.202596
    },
    {
      "acc": 0.76160979,
      "epoch": 0.4276527800055411,
      "grad_norm": 4.34375,
      "learning_rate": 9.139698473186079e-06,
      "loss": 0.85901699,
      "memory(GiB)": 135.77,
      "step": 18330,
      "train_speed(iter/s)": 0.202652
    },
    {
      "acc": 0.77596149,
      "epoch": 0.4278860875778299,
      "grad_norm": 5.03125,
      "learning_rate": 9.138638731559246e-06,
      "loss": 0.80930233,
      "memory(GiB)": 135.77,
      "step": 18340,
      "train_speed(iter/s)": 0.20271
    },
    {
      "acc": 0.78887205,
      "epoch": 0.4281193951501188,
      "grad_norm": 5.0,
      "learning_rate": 9.137578399142936e-06,
      "loss": 0.73909388,
      "memory(GiB)": 135.77,
      "step": 18350,
      "train_speed(iter/s)": 0.202766
    },
    {
      "acc": 0.77916021,
      "epoch": 0.4283527027224077,
      "grad_norm": 5.0,
      "learning_rate": 9.136517476088513e-06,
      "loss": 0.80833597,
      "memory(GiB)": 135.77,
      "step": 18360,
      "train_speed(iter/s)": 0.202821
    },
    {
      "acc": 0.77610526,
      "epoch": 0.4285860102946966,
      "grad_norm": 6.5625,
      "learning_rate": 9.135455962547422e-06,
      "loss": 0.80960493,
      "memory(GiB)": 135.77,
      "step": 18370,
      "train_speed(iter/s)": 0.202876
    },
    {
      "acc": 0.75241766,
      "epoch": 0.4288193178669855,
      "grad_norm": 6.25,
      "learning_rate": 9.134393858671193e-06,
      "loss": 0.87778807,
      "memory(GiB)": 135.77,
      "step": 18380,
      "train_speed(iter/s)": 0.202933
    },
    {
      "acc": 0.7702713,
      "epoch": 0.4290526254392744,
      "grad_norm": 7.0,
      "learning_rate": 9.13333116461144e-06,
      "loss": 0.82612343,
      "memory(GiB)": 135.77,
      "step": 18390,
      "train_speed(iter/s)": 0.20299
    },
    {
      "acc": 0.78864408,
      "epoch": 0.4292859330115633,
      "grad_norm": 5.15625,
      "learning_rate": 9.132267880519867e-06,
      "loss": 0.75824766,
      "memory(GiB)": 135.77,
      "step": 18400,
      "train_speed(iter/s)": 0.203043
    },
    {
      "acc": 0.76866798,
      "epoch": 0.4295192405838522,
      "grad_norm": 6.1875,
      "learning_rate": 9.131204006548253e-06,
      "loss": 0.83031731,
      "memory(GiB)": 135.77,
      "step": 18410,
      "train_speed(iter/s)": 0.203101
    },
    {
      "acc": 0.7711411,
      "epoch": 0.4297525481561411,
      "grad_norm": 6.90625,
      "learning_rate": 9.130139542848468e-06,
      "loss": 0.83066397,
      "memory(GiB)": 135.77,
      "step": 18420,
      "train_speed(iter/s)": 0.203161
    },
    {
      "acc": 0.76493034,
      "epoch": 0.42998585572843,
      "grad_norm": 5.75,
      "learning_rate": 9.129074489572463e-06,
      "loss": 0.85996132,
      "memory(GiB)": 135.77,
      "step": 18430,
      "train_speed(iter/s)": 0.203214
    },
    {
      "acc": 0.76129556,
      "epoch": 0.43021916330071885,
      "grad_norm": 4.15625,
      "learning_rate": 9.128008846872273e-06,
      "loss": 0.87012196,
      "memory(GiB)": 135.77,
      "step": 18440,
      "train_speed(iter/s)": 0.203266
    },
    {
      "acc": 0.77166157,
      "epoch": 0.43045247087300775,
      "grad_norm": 9.5625,
      "learning_rate": 9.126942614900021e-06,
      "loss": 0.80364647,
      "memory(GiB)": 135.77,
      "step": 18450,
      "train_speed(iter/s)": 0.203324
    },
    {
      "acc": 0.76030464,
      "epoch": 0.43068577844529665,
      "grad_norm": 4.1875,
      "learning_rate": 9.125875793807908e-06,
      "loss": 0.88488598,
      "memory(GiB)": 135.77,
      "step": 18460,
      "train_speed(iter/s)": 0.203381
    },
    {
      "acc": 0.77922316,
      "epoch": 0.43091908601758555,
      "grad_norm": 9.4375,
      "learning_rate": 9.124808383748226e-06,
      "loss": 0.79430113,
      "memory(GiB)": 135.77,
      "step": 18470,
      "train_speed(iter/s)": 0.203433
    },
    {
      "acc": 0.76354847,
      "epoch": 0.43115239358987445,
      "grad_norm": 7.46875,
      "learning_rate": 9.123740384873343e-06,
      "loss": 0.87230701,
      "memory(GiB)": 135.77,
      "step": 18480,
      "train_speed(iter/s)": 0.203489
    },
    {
      "acc": 0.76531634,
      "epoch": 0.43138570116216335,
      "grad_norm": 6.3125,
      "learning_rate": 9.122671797335719e-06,
      "loss": 0.86186409,
      "memory(GiB)": 135.77,
      "step": 18490,
      "train_speed(iter/s)": 0.203546
    },
    {
      "acc": 0.76710591,
      "epoch": 0.43161900873445225,
      "grad_norm": 5.96875,
      "learning_rate": 9.121602621287892e-06,
      "loss": 0.85632124,
      "memory(GiB)": 135.77,
      "step": 18500,
      "train_speed(iter/s)": 0.203601
    },
    {
      "epoch": 0.43161900873445225,
      "eval_acc": 0.737777431866236,
      "eval_loss": 0.8302583694458008,
      "eval_runtime": 1269.9297,
      "eval_samples_per_second": 28.341,
      "eval_steps_per_second": 14.171,
      "step": 18500
    },
    {
      "acc": 0.79308367,
      "epoch": 0.43185231630674115,
      "grad_norm": 5.53125,
      "learning_rate": 9.120532856882491e-06,
      "loss": 0.74874711,
      "memory(GiB)": 135.77,
      "step": 18510,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.77777281,
      "epoch": 0.43208562387903005,
      "grad_norm": 5.375,
      "learning_rate": 9.119462504272221e-06,
      "loss": 0.8051053,
      "memory(GiB)": 135.77,
      "step": 18520,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.74930582,
      "epoch": 0.43231893145131894,
      "grad_norm": 5.9375,
      "learning_rate": 9.118391563609875e-06,
      "loss": 0.8819828,
      "memory(GiB)": 135.77,
      "step": 18530,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.77309723,
      "epoch": 0.4325522390236078,
      "grad_norm": 6.34375,
      "learning_rate": 9.117320035048329e-06,
      "loss": 0.82186155,
      "memory(GiB)": 135.77,
      "step": 18540,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.77171998,
      "epoch": 0.4327855465958967,
      "grad_norm": 6.65625,
      "learning_rate": 9.116247918740544e-06,
      "loss": 0.80551882,
      "memory(GiB)": 135.77,
      "step": 18550,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.76656771,
      "epoch": 0.4330188541681856,
      "grad_norm": 4.59375,
      "learning_rate": 9.115175214839565e-06,
      "loss": 0.87863846,
      "memory(GiB)": 135.77,
      "step": 18560,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.78267555,
      "epoch": 0.4332521617404745,
      "grad_norm": 5.46875,
      "learning_rate": 9.114101923498519e-06,
      "loss": 0.77156582,
      "memory(GiB)": 135.77,
      "step": 18570,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.76703129,
      "epoch": 0.4334854693127634,
      "grad_norm": 8.625,
      "learning_rate": 9.113028044870619e-06,
      "loss": 0.84351616,
      "memory(GiB)": 135.77,
      "step": 18580,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.76306887,
      "epoch": 0.4337187768850523,
      "grad_norm": 5.625,
      "learning_rate": 9.11195357910916e-06,
      "loss": 0.86529598,
      "memory(GiB)": 135.77,
      "step": 18590,
      "train_speed(iter/s)": 0.201258
    },
    {
      "acc": 0.77179461,
      "epoch": 0.4339520844573412,
      "grad_norm": 5.03125,
      "learning_rate": 9.110878526367523e-06,
      "loss": 0.8225605,
      "memory(GiB)": 135.77,
      "step": 18600,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.77156925,
      "epoch": 0.4341853920296301,
      "grad_norm": 6.03125,
      "learning_rate": 9.10980288679917e-06,
      "loss": 0.81905994,
      "memory(GiB)": 135.77,
      "step": 18610,
      "train_speed(iter/s)": 0.201374
    },
    {
      "acc": 0.76532478,
      "epoch": 0.434418699601919,
      "grad_norm": 8.8125,
      "learning_rate": 9.10872666055765e-06,
      "loss": 0.85075331,
      "memory(GiB)": 135.77,
      "step": 18620,
      "train_speed(iter/s)": 0.201432
    },
    {
      "acc": 0.76773767,
      "epoch": 0.4346520071742078,
      "grad_norm": 6.34375,
      "learning_rate": 9.107649847796591e-06,
      "loss": 0.8390007,
      "memory(GiB)": 135.77,
      "step": 18630,
      "train_speed(iter/s)": 0.201484
    },
    {
      "acc": 0.76637259,
      "epoch": 0.4348853147464967,
      "grad_norm": 7.53125,
      "learning_rate": 9.10657244866971e-06,
      "loss": 0.83106155,
      "memory(GiB)": 135.77,
      "step": 18640,
      "train_speed(iter/s)": 0.201536
    },
    {
      "acc": 0.76556454,
      "epoch": 0.4351186223187856,
      "grad_norm": 6.5625,
      "learning_rate": 9.105494463330805e-06,
      "loss": 0.82924232,
      "memory(GiB)": 135.77,
      "step": 18650,
      "train_speed(iter/s)": 0.201594
    },
    {
      "acc": 0.76742105,
      "epoch": 0.4353519298910745,
      "grad_norm": 4.0,
      "learning_rate": 9.104415891933757e-06,
      "loss": 0.83789377,
      "memory(GiB)": 135.77,
      "step": 18660,
      "train_speed(iter/s)": 0.20165
    },
    {
      "acc": 0.77874622,
      "epoch": 0.4355852374633634,
      "grad_norm": 5.1875,
      "learning_rate": 9.103336734632536e-06,
      "loss": 0.78959827,
      "memory(GiB)": 135.77,
      "step": 18670,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.7700861,
      "epoch": 0.4358185450356523,
      "grad_norm": 4.5,
      "learning_rate": 9.102256991581185e-06,
      "loss": 0.82032681,
      "memory(GiB)": 135.77,
      "step": 18680,
      "train_speed(iter/s)": 0.201755
    },
    {
      "acc": 0.76898155,
      "epoch": 0.4360518526079412,
      "grad_norm": 5.65625,
      "learning_rate": 9.101176662933842e-06,
      "loss": 0.85672741,
      "memory(GiB)": 135.77,
      "step": 18690,
      "train_speed(iter/s)": 0.201813
    },
    {
      "acc": 0.76275616,
      "epoch": 0.4362851601802301,
      "grad_norm": 5.09375,
      "learning_rate": 9.10009574884472e-06,
      "loss": 0.86024952,
      "memory(GiB)": 135.77,
      "step": 18700,
      "train_speed(iter/s)": 0.201867
    },
    {
      "acc": 0.76601181,
      "epoch": 0.436518467752519,
      "grad_norm": 6.375,
      "learning_rate": 9.099014249468124e-06,
      "loss": 0.86094589,
      "memory(GiB)": 135.77,
      "step": 18710,
      "train_speed(iter/s)": 0.201923
    },
    {
      "acc": 0.77703285,
      "epoch": 0.4367517753248079,
      "grad_norm": 7.375,
      "learning_rate": 9.097932164958432e-06,
      "loss": 0.79216099,
      "memory(GiB)": 135.77,
      "step": 18720,
      "train_speed(iter/s)": 0.201982
    },
    {
      "acc": 0.76565948,
      "epoch": 0.43698508289709675,
      "grad_norm": 5.28125,
      "learning_rate": 9.096849495470113e-06,
      "loss": 0.86634254,
      "memory(GiB)": 135.77,
      "step": 18730,
      "train_speed(iter/s)": 0.202039
    },
    {
      "acc": 0.77021027,
      "epoch": 0.43721839046938565,
      "grad_norm": 5.5,
      "learning_rate": 9.095766241157721e-06,
      "loss": 0.81364603,
      "memory(GiB)": 135.77,
      "step": 18740,
      "train_speed(iter/s)": 0.202094
    },
    {
      "acc": 0.78136158,
      "epoch": 0.43745169804167455,
      "grad_norm": 4.8125,
      "learning_rate": 9.094682402175887e-06,
      "loss": 0.79765949,
      "memory(GiB)": 135.77,
      "step": 18750,
      "train_speed(iter/s)": 0.202146
    },
    {
      "acc": 0.76939726,
      "epoch": 0.43768500561396345,
      "grad_norm": 9.6875,
      "learning_rate": 9.093597978679329e-06,
      "loss": 0.84548798,
      "memory(GiB)": 135.77,
      "step": 18760,
      "train_speed(iter/s)": 0.202202
    },
    {
      "acc": 0.77349901,
      "epoch": 0.43791831318625235,
      "grad_norm": 7.65625,
      "learning_rate": 9.09251297082285e-06,
      "loss": 0.81046124,
      "memory(GiB)": 135.77,
      "step": 18770,
      "train_speed(iter/s)": 0.202258
    },
    {
      "acc": 0.76422777,
      "epoch": 0.43815162075854125,
      "grad_norm": 4.5,
      "learning_rate": 9.091427378761333e-06,
      "loss": 0.86868763,
      "memory(GiB)": 135.77,
      "step": 18780,
      "train_speed(iter/s)": 0.202317
    },
    {
      "acc": 0.77027721,
      "epoch": 0.43838492833083015,
      "grad_norm": 5.15625,
      "learning_rate": 9.090341202649746e-06,
      "loss": 0.84670544,
      "memory(GiB)": 135.77,
      "step": 18790,
      "train_speed(iter/s)": 0.202371
    },
    {
      "acc": 0.76263161,
      "epoch": 0.43861823590311905,
      "grad_norm": 5.9375,
      "learning_rate": 9.08925444264314e-06,
      "loss": 0.86874142,
      "memory(GiB)": 135.77,
      "step": 18800,
      "train_speed(iter/s)": 0.202424
    },
    {
      "acc": 0.75490875,
      "epoch": 0.43885154347540795,
      "grad_norm": 6.78125,
      "learning_rate": 9.088167098896652e-06,
      "loss": 0.89412689,
      "memory(GiB)": 135.77,
      "step": 18810,
      "train_speed(iter/s)": 0.202481
    },
    {
      "acc": 0.77074194,
      "epoch": 0.4390848510476968,
      "grad_norm": 6.15625,
      "learning_rate": 9.087079171565496e-06,
      "loss": 0.84905615,
      "memory(GiB)": 135.77,
      "step": 18820,
      "train_speed(iter/s)": 0.202537
    },
    {
      "acc": 0.76013765,
      "epoch": 0.4393181586199857,
      "grad_norm": 4.59375,
      "learning_rate": 9.085990660804976e-06,
      "loss": 0.84425535,
      "memory(GiB)": 135.77,
      "step": 18830,
      "train_speed(iter/s)": 0.202595
    },
    {
      "acc": 0.78072777,
      "epoch": 0.4395514661922746,
      "grad_norm": 5.96875,
      "learning_rate": 9.084901566770476e-06,
      "loss": 0.78052444,
      "memory(GiB)": 135.77,
      "step": 18840,
      "train_speed(iter/s)": 0.20265
    },
    {
      "acc": 0.78506041,
      "epoch": 0.4397847737645635,
      "grad_norm": 7.21875,
      "learning_rate": 9.083811889617467e-06,
      "loss": 0.76543379,
      "memory(GiB)": 135.77,
      "step": 18850,
      "train_speed(iter/s)": 0.202704
    },
    {
      "acc": 0.7516593,
      "epoch": 0.4400180813368524,
      "grad_norm": 5.46875,
      "learning_rate": 9.082721629501494e-06,
      "loss": 0.88883514,
      "memory(GiB)": 135.77,
      "step": 18860,
      "train_speed(iter/s)": 0.202756
    },
    {
      "acc": 0.77115822,
      "epoch": 0.4402513889091413,
      "grad_norm": 8.1875,
      "learning_rate": 9.081630786578195e-06,
      "loss": 0.8167778,
      "memory(GiB)": 135.77,
      "step": 18870,
      "train_speed(iter/s)": 0.202811
    },
    {
      "acc": 0.76743846,
      "epoch": 0.4404846964814302,
      "grad_norm": 6.125,
      "learning_rate": 9.080539361003288e-06,
      "loss": 0.85453176,
      "memory(GiB)": 135.77,
      "step": 18880,
      "train_speed(iter/s)": 0.202866
    },
    {
      "acc": 0.75334711,
      "epoch": 0.4407180040537191,
      "grad_norm": 4.375,
      "learning_rate": 9.079447352932571e-06,
      "loss": 0.89333754,
      "memory(GiB)": 135.77,
      "step": 18890,
      "train_speed(iter/s)": 0.202919
    },
    {
      "acc": 0.78106527,
      "epoch": 0.440951311626008,
      "grad_norm": 6.21875,
      "learning_rate": 9.078354762521931e-06,
      "loss": 0.80243101,
      "memory(GiB)": 135.77,
      "step": 18900,
      "train_speed(iter/s)": 0.202975
    },
    {
      "acc": 0.76703472,
      "epoch": 0.4411846191982969,
      "grad_norm": 5.8125,
      "learning_rate": 9.077261589927333e-06,
      "loss": 0.84479599,
      "memory(GiB)": 135.77,
      "step": 18910,
      "train_speed(iter/s)": 0.203031
    },
    {
      "acc": 0.77735329,
      "epoch": 0.4414179267705857,
      "grad_norm": 11.625,
      "learning_rate": 9.076167835304828e-06,
      "loss": 0.80349712,
      "memory(GiB)": 135.77,
      "step": 18920,
      "train_speed(iter/s)": 0.203086
    },
    {
      "acc": 0.79136057,
      "epoch": 0.4416512343428746,
      "grad_norm": 5.25,
      "learning_rate": 9.075073498810547e-06,
      "loss": 0.76573734,
      "memory(GiB)": 135.77,
      "step": 18930,
      "train_speed(iter/s)": 0.203143
    },
    {
      "acc": 0.78350835,
      "epoch": 0.4418845419151635,
      "grad_norm": 5.09375,
      "learning_rate": 9.073978580600709e-06,
      "loss": 0.78551674,
      "memory(GiB)": 135.77,
      "step": 18940,
      "train_speed(iter/s)": 0.203194
    },
    {
      "acc": 0.76640816,
      "epoch": 0.4421178494874524,
      "grad_norm": 5.5625,
      "learning_rate": 9.072883080831611e-06,
      "loss": 0.8355978,
      "memory(GiB)": 135.77,
      "step": 18950,
      "train_speed(iter/s)": 0.203252
    },
    {
      "acc": 0.75378032,
      "epoch": 0.4423511570597413,
      "grad_norm": 6.25,
      "learning_rate": 9.071786999659638e-06,
      "loss": 0.88689022,
      "memory(GiB)": 135.77,
      "step": 18960,
      "train_speed(iter/s)": 0.203313
    },
    {
      "acc": 0.74607992,
      "epoch": 0.4425844646320302,
      "grad_norm": 31.375,
      "learning_rate": 9.070690337241252e-06,
      "loss": 0.98085604,
      "memory(GiB)": 135.77,
      "step": 18970,
      "train_speed(iter/s)": 0.203368
    },
    {
      "acc": 0.78964634,
      "epoch": 0.4428177722043191,
      "grad_norm": 5.71875,
      "learning_rate": 9.069593093733004e-06,
      "loss": 0.76366062,
      "memory(GiB)": 135.77,
      "step": 18980,
      "train_speed(iter/s)": 0.203426
    },
    {
      "acc": 0.77851391,
      "epoch": 0.443051079776608,
      "grad_norm": 7.03125,
      "learning_rate": 9.068495269291524e-06,
      "loss": 0.81780148,
      "memory(GiB)": 135.77,
      "step": 18990,
      "train_speed(iter/s)": 0.203481
    },
    {
      "acc": 0.75752182,
      "epoch": 0.4432843873488969,
      "grad_norm": 6.84375,
      "learning_rate": 9.067396864073527e-06,
      "loss": 0.87670937,
      "memory(GiB)": 135.77,
      "step": 19000,
      "train_speed(iter/s)": 0.203539
    },
    {
      "epoch": 0.4432843873488969,
      "eval_acc": 0.7378800843191412,
      "eval_loss": 0.8296611905097961,
      "eval_runtime": 1269.212,
      "eval_samples_per_second": 28.357,
      "eval_steps_per_second": 14.179,
      "step": 19000
    },
    {
      "acc": 0.77129135,
      "epoch": 0.4435176949211858,
      "grad_norm": 4.09375,
      "learning_rate": 9.066297878235808e-06,
      "loss": 0.82974644,
      "memory(GiB)": 135.77,
      "step": 19010,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.77290907,
      "epoch": 0.44375100249347466,
      "grad_norm": 3.796875,
      "learning_rate": 9.065198311935248e-06,
      "loss": 0.82878008,
      "memory(GiB)": 135.77,
      "step": 19020,
      "train_speed(iter/s)": 0.200885
    },
    {
      "acc": 0.77206917,
      "epoch": 0.44398431006576355,
      "grad_norm": 7.4375,
      "learning_rate": 9.06409816532881e-06,
      "loss": 0.84106894,
      "memory(GiB)": 135.77,
      "step": 19030,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.75648594,
      "epoch": 0.44421761763805245,
      "grad_norm": 4.9375,
      "learning_rate": 9.06299743857354e-06,
      "loss": 0.90824041,
      "memory(GiB)": 135.77,
      "step": 19040,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.76354513,
      "epoch": 0.44445092521034135,
      "grad_norm": 4.8125,
      "learning_rate": 9.061896131826566e-06,
      "loss": 0.86085148,
      "memory(GiB)": 135.77,
      "step": 19050,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.76192708,
      "epoch": 0.44468423278263025,
      "grad_norm": 6.8125,
      "learning_rate": 9.0607942452451e-06,
      "loss": 0.8651083,
      "memory(GiB)": 135.77,
      "step": 19060,
      "train_speed(iter/s)": 0.201098
    },
    {
      "acc": 0.76140442,
      "epoch": 0.44491754035491915,
      "grad_norm": 6.28125,
      "learning_rate": 9.059691778986433e-06,
      "loss": 0.85103846,
      "memory(GiB)": 135.77,
      "step": 19070,
      "train_speed(iter/s)": 0.201155
    },
    {
      "acc": 0.76223292,
      "epoch": 0.44515084792720805,
      "grad_norm": 8.5,
      "learning_rate": 9.058588733207945e-06,
      "loss": 0.83704605,
      "memory(GiB)": 135.77,
      "step": 19080,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.76429772,
      "epoch": 0.44538415549949695,
      "grad_norm": 6.09375,
      "learning_rate": 9.057485108067094e-06,
      "loss": 0.86491184,
      "memory(GiB)": 135.77,
      "step": 19090,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.77062263,
      "epoch": 0.44561746307178585,
      "grad_norm": 4.875,
      "learning_rate": 9.056380903721424e-06,
      "loss": 0.81978521,
      "memory(GiB)": 135.77,
      "step": 19100,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.77301741,
      "epoch": 0.4458507706440747,
      "grad_norm": 5.78125,
      "learning_rate": 9.055276120328557e-06,
      "loss": 0.81939163,
      "memory(GiB)": 135.77,
      "step": 19110,
      "train_speed(iter/s)": 0.201375
    },
    {
      "acc": 0.77150536,
      "epoch": 0.4460840782163636,
      "grad_norm": 4.4375,
      "learning_rate": 9.054170758046204e-06,
      "loss": 0.81701164,
      "memory(GiB)": 135.77,
      "step": 19120,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.77243052,
      "epoch": 0.4463173857886525,
      "grad_norm": 5.375,
      "learning_rate": 9.05306481703215e-06,
      "loss": 0.83213739,
      "memory(GiB)": 135.77,
      "step": 19130,
      "train_speed(iter/s)": 0.201484
    },
    {
      "acc": 0.76942472,
      "epoch": 0.4465506933609414,
      "grad_norm": 6.0625,
      "learning_rate": 9.051958297444272e-06,
      "loss": 0.8267148,
      "memory(GiB)": 135.77,
      "step": 19140,
      "train_speed(iter/s)": 0.201538
    },
    {
      "acc": 0.76962104,
      "epoch": 0.4467840009332303,
      "grad_norm": 5.28125,
      "learning_rate": 9.050851199440524e-06,
      "loss": 0.81351061,
      "memory(GiB)": 135.77,
      "step": 19150,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.7752852,
      "epoch": 0.4470173085055192,
      "grad_norm": 5.25,
      "learning_rate": 9.049743523178945e-06,
      "loss": 0.81729002,
      "memory(GiB)": 135.77,
      "step": 19160,
      "train_speed(iter/s)": 0.201644
    },
    {
      "acc": 0.76699119,
      "epoch": 0.4472506160778081,
      "grad_norm": 4.6875,
      "learning_rate": 9.048635268817653e-06,
      "loss": 0.85345955,
      "memory(GiB)": 135.77,
      "step": 19170,
      "train_speed(iter/s)": 0.201701
    },
    {
      "acc": 0.76262298,
      "epoch": 0.447483923650097,
      "grad_norm": 8.9375,
      "learning_rate": 9.047526436514854e-06,
      "loss": 0.85194921,
      "memory(GiB)": 135.77,
      "step": 19180,
      "train_speed(iter/s)": 0.201754
    },
    {
      "acc": 0.76009088,
      "epoch": 0.4477172312223859,
      "grad_norm": 4.90625,
      "learning_rate": 9.04641702642883e-06,
      "loss": 0.86009045,
      "memory(GiB)": 135.77,
      "step": 19190,
      "train_speed(iter/s)": 0.201813
    },
    {
      "acc": 0.77364101,
      "epoch": 0.4479505387946748,
      "grad_norm": 4.90625,
      "learning_rate": 9.045307038717954e-06,
      "loss": 0.81192675,
      "memory(GiB)": 135.77,
      "step": 19200,
      "train_speed(iter/s)": 0.201865
    },
    {
      "acc": 0.77397594,
      "epoch": 0.4481838463669636,
      "grad_norm": 6.625,
      "learning_rate": 9.044196473540672e-06,
      "loss": 0.8263483,
      "memory(GiB)": 135.77,
      "step": 19210,
      "train_speed(iter/s)": 0.201919
    },
    {
      "acc": 0.753967,
      "epoch": 0.4484171539392525,
      "grad_norm": 5.0625,
      "learning_rate": 9.043085331055516e-06,
      "loss": 0.89315586,
      "memory(GiB)": 135.77,
      "step": 19220,
      "train_speed(iter/s)": 0.201975
    },
    {
      "acc": 0.78117127,
      "epoch": 0.4486504615115414,
      "grad_norm": 3.390625,
      "learning_rate": 9.041973611421106e-06,
      "loss": 0.77627773,
      "memory(GiB)": 135.77,
      "step": 19230,
      "train_speed(iter/s)": 0.20203
    },
    {
      "acc": 0.76547561,
      "epoch": 0.4488837690838303,
      "grad_norm": 5.5625,
      "learning_rate": 9.040861314796137e-06,
      "loss": 0.84824905,
      "memory(GiB)": 135.77,
      "step": 19240,
      "train_speed(iter/s)": 0.20209
    },
    {
      "acc": 0.79242058,
      "epoch": 0.4491170766561192,
      "grad_norm": 4.40625,
      "learning_rate": 9.039748441339389e-06,
      "loss": 0.73243713,
      "memory(GiB)": 135.77,
      "step": 19250,
      "train_speed(iter/s)": 0.202146
    },
    {
      "acc": 0.74393578,
      "epoch": 0.4493503842284081,
      "grad_norm": 7.0625,
      "learning_rate": 9.038634991209725e-06,
      "loss": 0.93040276,
      "memory(GiB)": 135.77,
      "step": 19260,
      "train_speed(iter/s)": 0.202197
    },
    {
      "acc": 0.7861661,
      "epoch": 0.449583691800697,
      "grad_norm": 7.78125,
      "learning_rate": 9.03752096456609e-06,
      "loss": 0.77835784,
      "memory(GiB)": 135.77,
      "step": 19270,
      "train_speed(iter/s)": 0.20225
    },
    {
      "acc": 0.77567,
      "epoch": 0.4498169993729859,
      "grad_norm": 5.0,
      "learning_rate": 9.036406361567506e-06,
      "loss": 0.7994051,
      "memory(GiB)": 135.77,
      "step": 19280,
      "train_speed(iter/s)": 0.202304
    },
    {
      "acc": 0.76911907,
      "epoch": 0.4500503069452748,
      "grad_norm": 5.5,
      "learning_rate": 9.035291182373092e-06,
      "loss": 0.82964163,
      "memory(GiB)": 135.77,
      "step": 19290,
      "train_speed(iter/s)": 0.202358
    },
    {
      "acc": 0.76818333,
      "epoch": 0.4502836145175637,
      "grad_norm": 5.65625,
      "learning_rate": 9.03417542714203e-06,
      "loss": 0.8365284,
      "memory(GiB)": 135.77,
      "step": 19300,
      "train_speed(iter/s)": 0.202412
    },
    {
      "acc": 0.77055397,
      "epoch": 0.45051692208985256,
      "grad_norm": 5.53125,
      "learning_rate": 9.033059096033598e-06,
      "loss": 0.82133121,
      "memory(GiB)": 135.77,
      "step": 19310,
      "train_speed(iter/s)": 0.202465
    },
    {
      "acc": 0.769734,
      "epoch": 0.45075022966214146,
      "grad_norm": 4.65625,
      "learning_rate": 9.031942189207154e-06,
      "loss": 0.81667995,
      "memory(GiB)": 135.77,
      "step": 19320,
      "train_speed(iter/s)": 0.202515
    },
    {
      "acc": 0.7695981,
      "epoch": 0.45098353723443035,
      "grad_norm": 5.25,
      "learning_rate": 9.030824706822132e-06,
      "loss": 0.83406506,
      "memory(GiB)": 135.77,
      "step": 19330,
      "train_speed(iter/s)": 0.202565
    },
    {
      "acc": 0.77217684,
      "epoch": 0.45121684480671925,
      "grad_norm": 5.0625,
      "learning_rate": 9.029706649038055e-06,
      "loss": 0.80703802,
      "memory(GiB)": 135.77,
      "step": 19340,
      "train_speed(iter/s)": 0.202615
    },
    {
      "acc": 0.76901155,
      "epoch": 0.45145015237900815,
      "grad_norm": 6.9375,
      "learning_rate": 9.028588016014524e-06,
      "loss": 0.8353384,
      "memory(GiB)": 135.77,
      "step": 19350,
      "train_speed(iter/s)": 0.202672
    },
    {
      "acc": 0.78265429,
      "epoch": 0.45168345995129705,
      "grad_norm": 5.53125,
      "learning_rate": 9.027468807911223e-06,
      "loss": 0.78769207,
      "memory(GiB)": 135.77,
      "step": 19360,
      "train_speed(iter/s)": 0.202725
    },
    {
      "acc": 0.79214249,
      "epoch": 0.45191676752358595,
      "grad_norm": 11.1875,
      "learning_rate": 9.026349024887921e-06,
      "loss": 0.72728148,
      "memory(GiB)": 135.77,
      "step": 19370,
      "train_speed(iter/s)": 0.202779
    },
    {
      "acc": 0.76945453,
      "epoch": 0.45215007509587485,
      "grad_norm": 5.875,
      "learning_rate": 9.025228667104465e-06,
      "loss": 0.83437386,
      "memory(GiB)": 135.77,
      "step": 19380,
      "train_speed(iter/s)": 0.202835
    },
    {
      "acc": 0.76589389,
      "epoch": 0.45238338266816375,
      "grad_norm": 5.5,
      "learning_rate": 9.024107734720786e-06,
      "loss": 0.85786552,
      "memory(GiB)": 135.77,
      "step": 19390,
      "train_speed(iter/s)": 0.20289
    },
    {
      "acc": 0.76294546,
      "epoch": 0.4526166902404526,
      "grad_norm": 6.5,
      "learning_rate": 9.022986227896898e-06,
      "loss": 0.8357048,
      "memory(GiB)": 135.77,
      "step": 19400,
      "train_speed(iter/s)": 0.202945
    },
    {
      "acc": 0.78395748,
      "epoch": 0.4528499978127415,
      "grad_norm": 7.28125,
      "learning_rate": 9.021864146792894e-06,
      "loss": 0.74856625,
      "memory(GiB)": 135.77,
      "step": 19410,
      "train_speed(iter/s)": 0.203001
    },
    {
      "acc": 0.78489389,
      "epoch": 0.4530833053850304,
      "grad_norm": 15.25,
      "learning_rate": 9.02074149156895e-06,
      "loss": 0.76024318,
      "memory(GiB)": 135.77,
      "step": 19420,
      "train_speed(iter/s)": 0.203057
    },
    {
      "acc": 0.75939164,
      "epoch": 0.4533166129573193,
      "grad_norm": 7.75,
      "learning_rate": 9.019618262385328e-06,
      "loss": 0.85180931,
      "memory(GiB)": 135.77,
      "step": 19430,
      "train_speed(iter/s)": 0.203111
    },
    {
      "acc": 0.78605642,
      "epoch": 0.4535499205296082,
      "grad_norm": 5.4375,
      "learning_rate": 9.018494459402365e-06,
      "loss": 0.76549592,
      "memory(GiB)": 135.77,
      "step": 19440,
      "train_speed(iter/s)": 0.203163
    },
    {
      "acc": 0.76347561,
      "epoch": 0.4537832281018971,
      "grad_norm": 7.0625,
      "learning_rate": 9.017370082780485e-06,
      "loss": 0.85260391,
      "memory(GiB)": 135.77,
      "step": 19450,
      "train_speed(iter/s)": 0.203218
    },
    {
      "acc": 0.78153772,
      "epoch": 0.454016535674186,
      "grad_norm": 4.8125,
      "learning_rate": 9.016245132680195e-06,
      "loss": 0.77672606,
      "memory(GiB)": 135.77,
      "step": 19460,
      "train_speed(iter/s)": 0.203268
    },
    {
      "acc": 0.78809853,
      "epoch": 0.4542498432464749,
      "grad_norm": 5.4375,
      "learning_rate": 9.015119609262078e-06,
      "loss": 0.74799476,
      "memory(GiB)": 135.77,
      "step": 19470,
      "train_speed(iter/s)": 0.203321
    },
    {
      "acc": 0.78086472,
      "epoch": 0.4544831508187638,
      "grad_norm": 6.03125,
      "learning_rate": 9.013993512686803e-06,
      "loss": 0.7939446,
      "memory(GiB)": 135.77,
      "step": 19480,
      "train_speed(iter/s)": 0.203372
    },
    {
      "acc": 0.77702465,
      "epoch": 0.4547164583910527,
      "grad_norm": 5.78125,
      "learning_rate": 9.01286684311512e-06,
      "loss": 0.81716118,
      "memory(GiB)": 135.77,
      "step": 19490,
      "train_speed(iter/s)": 0.203422
    },
    {
      "acc": 0.75863905,
      "epoch": 0.4549497659633415,
      "grad_norm": 6.78125,
      "learning_rate": 9.011739600707862e-06,
      "loss": 0.89850245,
      "memory(GiB)": 135.77,
      "step": 19500,
      "train_speed(iter/s)": 0.203476
    },
    {
      "epoch": 0.4549497659633415,
      "eval_acc": 0.7380889123980622,
      "eval_loss": 0.829414427280426,
      "eval_runtime": 1270.491,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.165,
      "step": 19500
    },
    {
      "acc": 0.7731576,
      "epoch": 0.4551830735356304,
      "grad_norm": 5.09375,
      "learning_rate": 9.01061178562594e-06,
      "loss": 0.81027994,
      "memory(GiB)": 135.77,
      "step": 19510,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.76880264,
      "epoch": 0.4554163811079193,
      "grad_norm": 5.0625,
      "learning_rate": 9.009483398030353e-06,
      "loss": 0.81138697,
      "memory(GiB)": 135.77,
      "step": 19520,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.78158731,
      "epoch": 0.4556496886802082,
      "grad_norm": 5.8125,
      "learning_rate": 9.008354438082173e-06,
      "loss": 0.79674578,
      "memory(GiB)": 135.77,
      "step": 19530,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.75377884,
      "epoch": 0.4558829962524971,
      "grad_norm": 7.09375,
      "learning_rate": 9.007224905942562e-06,
      "loss": 0.89388762,
      "memory(GiB)": 135.77,
      "step": 19540,
      "train_speed(iter/s)": 0.200988
    },
    {
      "acc": 0.77003942,
      "epoch": 0.456116303824786,
      "grad_norm": 6.125,
      "learning_rate": 9.00609480177276e-06,
      "loss": 0.83401814,
      "memory(GiB)": 135.77,
      "step": 19550,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.76566763,
      "epoch": 0.4563496113970749,
      "grad_norm": 4.6875,
      "learning_rate": 9.00496412573409e-06,
      "loss": 0.86685047,
      "memory(GiB)": 135.77,
      "step": 19560,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.79012947,
      "epoch": 0.4565829189693638,
      "grad_norm": 4.8125,
      "learning_rate": 9.003832877987952e-06,
      "loss": 0.74963512,
      "memory(GiB)": 135.77,
      "step": 19570,
      "train_speed(iter/s)": 0.201142
    },
    {
      "acc": 0.76897526,
      "epoch": 0.4568162265416527,
      "grad_norm": 3.921875,
      "learning_rate": 9.002701058695836e-06,
      "loss": 0.82385035,
      "memory(GiB)": 135.77,
      "step": 19580,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.76111898,
      "epoch": 0.45704953411394156,
      "grad_norm": 5.375,
      "learning_rate": 9.001568668019306e-06,
      "loss": 0.85867405,
      "memory(GiB)": 135.77,
      "step": 19590,
      "train_speed(iter/s)": 0.201252
    },
    {
      "acc": 0.78389206,
      "epoch": 0.45728284168623046,
      "grad_norm": 6.125,
      "learning_rate": 9.000435706120011e-06,
      "loss": 0.78628263,
      "memory(GiB)": 135.77,
      "step": 19600,
      "train_speed(iter/s)": 0.201305
    },
    {
      "acc": 0.77073708,
      "epoch": 0.45751614925851936,
      "grad_norm": 6.28125,
      "learning_rate": 8.999302173159681e-06,
      "loss": 0.83480806,
      "memory(GiB)": 135.77,
      "step": 19610,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.76435275,
      "epoch": 0.45774945683080825,
      "grad_norm": 5.15625,
      "learning_rate": 8.998168069300128e-06,
      "loss": 0.86908226,
      "memory(GiB)": 135.77,
      "step": 19620,
      "train_speed(iter/s)": 0.20141
    },
    {
      "acc": 0.77386503,
      "epoch": 0.45798276440309715,
      "grad_norm": 4.78125,
      "learning_rate": 8.997033394703246e-06,
      "loss": 0.82412939,
      "memory(GiB)": 135.77,
      "step": 19630,
      "train_speed(iter/s)": 0.201461
    },
    {
      "acc": 0.75808735,
      "epoch": 0.45821607197538605,
      "grad_norm": 7.875,
      "learning_rate": 8.995898149531005e-06,
      "loss": 0.87254372,
      "memory(GiB)": 135.77,
      "step": 19640,
      "train_speed(iter/s)": 0.201514
    },
    {
      "acc": 0.77585478,
      "epoch": 0.45844937954767495,
      "grad_norm": 6.125,
      "learning_rate": 8.994762333945465e-06,
      "loss": 0.8103075,
      "memory(GiB)": 135.77,
      "step": 19650,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.76372824,
      "epoch": 0.45868268711996385,
      "grad_norm": 5.21875,
      "learning_rate": 8.993625948108764e-06,
      "loss": 0.86949158,
      "memory(GiB)": 135.77,
      "step": 19660,
      "train_speed(iter/s)": 0.201625
    },
    {
      "acc": 0.76708436,
      "epoch": 0.45891599469225275,
      "grad_norm": 5.65625,
      "learning_rate": 8.992488992183116e-06,
      "loss": 0.84779406,
      "memory(GiB)": 135.77,
      "step": 19670,
      "train_speed(iter/s)": 0.201678
    },
    {
      "acc": 0.7899591,
      "epoch": 0.45914930226454165,
      "grad_norm": 5.03125,
      "learning_rate": 8.991351466330827e-06,
      "loss": 0.7441288,
      "memory(GiB)": 135.77,
      "step": 19680,
      "train_speed(iter/s)": 0.201731
    },
    {
      "acc": 0.78236446,
      "epoch": 0.4593826098368305,
      "grad_norm": 4.96875,
      "learning_rate": 8.990213370714274e-06,
      "loss": 0.79019766,
      "memory(GiB)": 135.77,
      "step": 19690,
      "train_speed(iter/s)": 0.201785
    },
    {
      "acc": 0.7680593,
      "epoch": 0.4596159174091194,
      "grad_norm": 7.71875,
      "learning_rate": 8.989074705495921e-06,
      "loss": 0.8316782,
      "memory(GiB)": 135.77,
      "step": 19700,
      "train_speed(iter/s)": 0.201834
    },
    {
      "acc": 0.77131128,
      "epoch": 0.4598492249814083,
      "grad_norm": 8.875,
      "learning_rate": 8.987935470838315e-06,
      "loss": 0.82049313,
      "memory(GiB)": 135.77,
      "step": 19710,
      "train_speed(iter/s)": 0.201885
    },
    {
      "acc": 0.78124671,
      "epoch": 0.4600825325536972,
      "grad_norm": 5.71875,
      "learning_rate": 8.986795666904077e-06,
      "loss": 0.80133085,
      "memory(GiB)": 135.77,
      "step": 19720,
      "train_speed(iter/s)": 0.201938
    },
    {
      "acc": 0.75701828,
      "epoch": 0.4603158401259861,
      "grad_norm": 9.125,
      "learning_rate": 8.985655293855917e-06,
      "loss": 0.89116259,
      "memory(GiB)": 135.77,
      "step": 19730,
      "train_speed(iter/s)": 0.201995
    },
    {
      "acc": 0.76774426,
      "epoch": 0.460549147698275,
      "grad_norm": 5.34375,
      "learning_rate": 8.98451435185662e-06,
      "loss": 0.87877426,
      "memory(GiB)": 135.77,
      "step": 19740,
      "train_speed(iter/s)": 0.202049
    },
    {
      "acc": 0.77983117,
      "epoch": 0.4607824552705639,
      "grad_norm": 5.34375,
      "learning_rate": 8.983372841069059e-06,
      "loss": 0.79148512,
      "memory(GiB)": 135.77,
      "step": 19750,
      "train_speed(iter/s)": 0.202102
    },
    {
      "acc": 0.74972019,
      "epoch": 0.4610157628428528,
      "grad_norm": 5.84375,
      "learning_rate": 8.98223076165618e-06,
      "loss": 0.89190731,
      "memory(GiB)": 135.77,
      "step": 19760,
      "train_speed(iter/s)": 0.202153
    },
    {
      "acc": 0.78662281,
      "epoch": 0.4612490704151417,
      "grad_norm": 5.71875,
      "learning_rate": 8.981088113781018e-06,
      "loss": 0.75874639,
      "memory(GiB)": 135.77,
      "step": 19770,
      "train_speed(iter/s)": 0.202207
    },
    {
      "acc": 0.76131806,
      "epoch": 0.4614823779874306,
      "grad_norm": 7.21875,
      "learning_rate": 8.979944897606685e-06,
      "loss": 0.87373924,
      "memory(GiB)": 135.77,
      "step": 19780,
      "train_speed(iter/s)": 0.20226
    },
    {
      "acc": 0.76455545,
      "epoch": 0.4617156855597194,
      "grad_norm": 6.15625,
      "learning_rate": 8.978801113296371e-06,
      "loss": 0.87849846,
      "memory(GiB)": 135.77,
      "step": 19790,
      "train_speed(iter/s)": 0.20231
    },
    {
      "acc": 0.76590109,
      "epoch": 0.4619489931320083,
      "grad_norm": 5.15625,
      "learning_rate": 8.977656761013357e-06,
      "loss": 0.83625021,
      "memory(GiB)": 135.77,
      "step": 19800,
      "train_speed(iter/s)": 0.202362
    },
    {
      "acc": 0.75346403,
      "epoch": 0.4621823007042972,
      "grad_norm": 15.125,
      "learning_rate": 8.976511840920994e-06,
      "loss": 0.90793972,
      "memory(GiB)": 135.77,
      "step": 19810,
      "train_speed(iter/s)": 0.202415
    },
    {
      "acc": 0.77614479,
      "epoch": 0.4624156082765861,
      "grad_norm": 5.9375,
      "learning_rate": 8.975366353182721e-06,
      "loss": 0.81695967,
      "memory(GiB)": 135.77,
      "step": 19820,
      "train_speed(iter/s)": 0.202469
    },
    {
      "acc": 0.77849016,
      "epoch": 0.462648915848875,
      "grad_norm": 5.46875,
      "learning_rate": 8.974220297962058e-06,
      "loss": 0.79997635,
      "memory(GiB)": 135.77,
      "step": 19830,
      "train_speed(iter/s)": 0.202525
    },
    {
      "acc": 0.77346191,
      "epoch": 0.4628822234211639,
      "grad_norm": 6.0625,
      "learning_rate": 8.973073675422602e-06,
      "loss": 0.80886145,
      "memory(GiB)": 135.77,
      "step": 19840,
      "train_speed(iter/s)": 0.202581
    },
    {
      "acc": 0.81118984,
      "epoch": 0.4631155309934528,
      "grad_norm": 5.6875,
      "learning_rate": 8.97192648572803e-06,
      "loss": 0.68877487,
      "memory(GiB)": 135.77,
      "step": 19850,
      "train_speed(iter/s)": 0.202632
    },
    {
      "acc": 0.75902133,
      "epoch": 0.4633488385657417,
      "grad_norm": 4.53125,
      "learning_rate": 8.970778729042109e-06,
      "loss": 0.86830082,
      "memory(GiB)": 135.77,
      "step": 19860,
      "train_speed(iter/s)": 0.202682
    },
    {
      "acc": 0.75023003,
      "epoch": 0.4635821461380306,
      "grad_norm": 6.3125,
      "learning_rate": 8.969630405528675e-06,
      "loss": 0.90686493,
      "memory(GiB)": 135.77,
      "step": 19870,
      "train_speed(iter/s)": 0.202737
    },
    {
      "acc": 0.78217545,
      "epoch": 0.46381545371031946,
      "grad_norm": 5.875,
      "learning_rate": 8.968481515351656e-06,
      "loss": 0.77547541,
      "memory(GiB)": 135.77,
      "step": 19880,
      "train_speed(iter/s)": 0.202788
    },
    {
      "acc": 0.77224998,
      "epoch": 0.46404876128260836,
      "grad_norm": 5.4375,
      "learning_rate": 8.967332058675054e-06,
      "loss": 0.81753349,
      "memory(GiB)": 135.77,
      "step": 19890,
      "train_speed(iter/s)": 0.20284
    },
    {
      "acc": 0.76978784,
      "epoch": 0.46428206885489726,
      "grad_norm": 5.40625,
      "learning_rate": 8.96618203566295e-06,
      "loss": 0.83251038,
      "memory(GiB)": 135.77,
      "step": 19900,
      "train_speed(iter/s)": 0.202894
    },
    {
      "acc": 0.78358245,
      "epoch": 0.46451537642718616,
      "grad_norm": 5.125,
      "learning_rate": 8.965031446479516e-06,
      "loss": 0.77494287,
      "memory(GiB)": 135.77,
      "step": 19910,
      "train_speed(iter/s)": 0.202944
    },
    {
      "acc": 0.75092211,
      "epoch": 0.46474868399947505,
      "grad_norm": 6.6875,
      "learning_rate": 8.963880291288992e-06,
      "loss": 0.90174236,
      "memory(GiB)": 135.77,
      "step": 19920,
      "train_speed(iter/s)": 0.202996
    },
    {
      "acc": 0.76736884,
      "epoch": 0.46498199157176395,
      "grad_norm": 5.8125,
      "learning_rate": 8.96272857025571e-06,
      "loss": 0.833918,
      "memory(GiB)": 135.77,
      "step": 19930,
      "train_speed(iter/s)": 0.203051
    },
    {
      "acc": 0.75787697,
      "epoch": 0.46521529914405285,
      "grad_norm": 5.375,
      "learning_rate": 8.961576283544076e-06,
      "loss": 0.89234486,
      "memory(GiB)": 135.77,
      "step": 19940,
      "train_speed(iter/s)": 0.203107
    },
    {
      "acc": 0.77328529,
      "epoch": 0.46544860671634175,
      "grad_norm": 6.15625,
      "learning_rate": 8.960423431318576e-06,
      "loss": 0.8312933,
      "memory(GiB)": 135.77,
      "step": 19950,
      "train_speed(iter/s)": 0.20316
    },
    {
      "acc": 0.77509122,
      "epoch": 0.46568191428863065,
      "grad_norm": 6.59375,
      "learning_rate": 8.959270013743784e-06,
      "loss": 0.81753225,
      "memory(GiB)": 135.77,
      "step": 19960,
      "train_speed(iter/s)": 0.203211
    },
    {
      "acc": 0.76919918,
      "epoch": 0.46591522186091955,
      "grad_norm": 5.25,
      "learning_rate": 8.958116030984347e-06,
      "loss": 0.84067249,
      "memory(GiB)": 135.77,
      "step": 19970,
      "train_speed(iter/s)": 0.203265
    },
    {
      "acc": 0.77842779,
      "epoch": 0.4661485294332084,
      "grad_norm": 4.125,
      "learning_rate": 8.956961483204996e-06,
      "loss": 0.78261929,
      "memory(GiB)": 135.77,
      "step": 19980,
      "train_speed(iter/s)": 0.203316
    },
    {
      "acc": 0.75851555,
      "epoch": 0.4663818370054973,
      "grad_norm": 6.78125,
      "learning_rate": 8.955806370570543e-06,
      "loss": 0.87037487,
      "memory(GiB)": 135.77,
      "step": 19990,
      "train_speed(iter/s)": 0.203366
    },
    {
      "acc": 0.77085943,
      "epoch": 0.4666151445777862,
      "grad_norm": 4.84375,
      "learning_rate": 8.954650693245882e-06,
      "loss": 0.81232281,
      "memory(GiB)": 135.77,
      "step": 20000,
      "train_speed(iter/s)": 0.203415
    },
    {
      "epoch": 0.4666151445777862,
      "eval_acc": 0.7384622086035567,
      "eval_loss": 0.8278706669807434,
      "eval_runtime": 1271.2712,
      "eval_samples_per_second": 28.311,
      "eval_steps_per_second": 14.156,
      "step": 20000
    },
    {
      "acc": 0.77572637,
      "epoch": 0.4668484521500751,
      "grad_norm": 5.65625,
      "learning_rate": 8.953494451395979e-06,
      "loss": 0.8144228,
      "memory(GiB)": 135.77,
      "step": 20010,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.79318457,
      "epoch": 0.467081759722364,
      "grad_norm": 9.25,
      "learning_rate": 8.952337645185894e-06,
      "loss": 0.74365864,
      "memory(GiB)": 135.77,
      "step": 20020,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.7678813,
      "epoch": 0.4673150672946529,
      "grad_norm": 4.6875,
      "learning_rate": 8.951180274780758e-06,
      "loss": 0.8354866,
      "memory(GiB)": 135.77,
      "step": 20030,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.76604643,
      "epoch": 0.4675483748669418,
      "grad_norm": 9.0,
      "learning_rate": 8.950022340345786e-06,
      "loss": 0.85739231,
      "memory(GiB)": 135.77,
      "step": 20040,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.76447287,
      "epoch": 0.4677816824392307,
      "grad_norm": 5.75,
      "learning_rate": 8.948863842046272e-06,
      "loss": 0.84905338,
      "memory(GiB)": 135.77,
      "step": 20050,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.77429447,
      "epoch": 0.4680149900115196,
      "grad_norm": 3.671875,
      "learning_rate": 8.947704780047593e-06,
      "loss": 0.81440744,
      "memory(GiB)": 135.77,
      "step": 20060,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.78105116,
      "epoch": 0.4682482975838085,
      "grad_norm": 5.40625,
      "learning_rate": 8.946545154515201e-06,
      "loss": 0.78914485,
      "memory(GiB)": 135.77,
      "step": 20070,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.7645546,
      "epoch": 0.4684816051560973,
      "grad_norm": 7.3125,
      "learning_rate": 8.945384965614636e-06,
      "loss": 0.87174253,
      "memory(GiB)": 135.77,
      "step": 20080,
      "train_speed(iter/s)": 0.201205
    },
    {
      "acc": 0.79997892,
      "epoch": 0.4687149127283862,
      "grad_norm": 5.6875,
      "learning_rate": 8.944224213511514e-06,
      "loss": 0.71954641,
      "memory(GiB)": 135.77,
      "step": 20090,
      "train_speed(iter/s)": 0.20126
    },
    {
      "acc": 0.762815,
      "epoch": 0.4689482203006751,
      "grad_norm": 8.1875,
      "learning_rate": 8.943062898371531e-06,
      "loss": 0.85633907,
      "memory(GiB)": 135.77,
      "step": 20100,
      "train_speed(iter/s)": 0.201312
    },
    {
      "acc": 0.76686854,
      "epoch": 0.469181527872964,
      "grad_norm": 13.0,
      "learning_rate": 8.941901020360464e-06,
      "loss": 0.82543678,
      "memory(GiB)": 135.77,
      "step": 20110,
      "train_speed(iter/s)": 0.201361
    },
    {
      "acc": 0.77254,
      "epoch": 0.4694148354452529,
      "grad_norm": 5.09375,
      "learning_rate": 8.940738579644171e-06,
      "loss": 0.80971737,
      "memory(GiB)": 135.77,
      "step": 20120,
      "train_speed(iter/s)": 0.201415
    },
    {
      "acc": 0.78281898,
      "epoch": 0.4696481430175418,
      "grad_norm": 6.0,
      "learning_rate": 8.939575576388592e-06,
      "loss": 0.78121605,
      "memory(GiB)": 135.77,
      "step": 20130,
      "train_speed(iter/s)": 0.201468
    },
    {
      "acc": 0.77683525,
      "epoch": 0.4698814505898307,
      "grad_norm": 5.0625,
      "learning_rate": 8.938412010759743e-06,
      "loss": 0.82987223,
      "memory(GiB)": 135.77,
      "step": 20140,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.78844047,
      "epoch": 0.4701147581621196,
      "grad_norm": 8.6875,
      "learning_rate": 8.937247882923724e-06,
      "loss": 0.76155491,
      "memory(GiB)": 135.77,
      "step": 20150,
      "train_speed(iter/s)": 0.201573
    },
    {
      "acc": 0.78030567,
      "epoch": 0.4703480657344085,
      "grad_norm": 5.9375,
      "learning_rate": 8.936083193046712e-06,
      "loss": 0.78899822,
      "memory(GiB)": 135.77,
      "step": 20160,
      "train_speed(iter/s)": 0.201621
    },
    {
      "acc": 0.76187115,
      "epoch": 0.47058137330669736,
      "grad_norm": 7.34375,
      "learning_rate": 8.93491794129497e-06,
      "loss": 0.83778381,
      "memory(GiB)": 135.77,
      "step": 20170,
      "train_speed(iter/s)": 0.201673
    },
    {
      "acc": 0.76804895,
      "epoch": 0.47081468087898626,
      "grad_norm": 4.1875,
      "learning_rate": 8.933752127834834e-06,
      "loss": 0.85201845,
      "memory(GiB)": 135.77,
      "step": 20180,
      "train_speed(iter/s)": 0.201725
    },
    {
      "acc": 0.764326,
      "epoch": 0.47104798845127516,
      "grad_norm": 6.3125,
      "learning_rate": 8.932585752832725e-06,
      "loss": 0.83452625,
      "memory(GiB)": 135.77,
      "step": 20190,
      "train_speed(iter/s)": 0.201778
    },
    {
      "acc": 0.76837258,
      "epoch": 0.47128129602356406,
      "grad_norm": 5.0,
      "learning_rate": 8.931418816455142e-06,
      "loss": 0.83890285,
      "memory(GiB)": 135.77,
      "step": 20200,
      "train_speed(iter/s)": 0.201827
    },
    {
      "acc": 0.76820593,
      "epoch": 0.47151460359585295,
      "grad_norm": 5.03125,
      "learning_rate": 8.930251318868664e-06,
      "loss": 0.82331924,
      "memory(GiB)": 135.77,
      "step": 20210,
      "train_speed(iter/s)": 0.201878
    },
    {
      "acc": 0.78207231,
      "epoch": 0.47174791116814185,
      "grad_norm": 4.96875,
      "learning_rate": 8.929083260239952e-06,
      "loss": 0.78689303,
      "memory(GiB)": 135.77,
      "step": 20220,
      "train_speed(iter/s)": 0.201928
    },
    {
      "acc": 0.7715971,
      "epoch": 0.47198121874043075,
      "grad_norm": 5.625,
      "learning_rate": 8.927914640735748e-06,
      "loss": 0.82081528,
      "memory(GiB)": 135.77,
      "step": 20230,
      "train_speed(iter/s)": 0.201983
    },
    {
      "acc": 0.76612329,
      "epoch": 0.47221452631271965,
      "grad_norm": 5.90625,
      "learning_rate": 8.926745460522867e-06,
      "loss": 0.85213652,
      "memory(GiB)": 135.77,
      "step": 20240,
      "train_speed(iter/s)": 0.202032
    },
    {
      "acc": 0.77809095,
      "epoch": 0.47244783388500855,
      "grad_norm": 6.9375,
      "learning_rate": 8.925575719768215e-06,
      "loss": 0.80060663,
      "memory(GiB)": 135.77,
      "step": 20250,
      "train_speed(iter/s)": 0.202087
    },
    {
      "acc": 0.77375336,
      "epoch": 0.47268114145729745,
      "grad_norm": 5.4375,
      "learning_rate": 8.92440541863877e-06,
      "loss": 0.81226845,
      "memory(GiB)": 135.77,
      "step": 20260,
      "train_speed(iter/s)": 0.202137
    },
    {
      "acc": 0.80193977,
      "epoch": 0.4729144490295863,
      "grad_norm": 6.34375,
      "learning_rate": 8.923234557301588e-06,
      "loss": 0.72059889,
      "memory(GiB)": 135.77,
      "step": 20270,
      "train_speed(iter/s)": 0.202186
    },
    {
      "acc": 0.79480009,
      "epoch": 0.4731477566018752,
      "grad_norm": 5.21875,
      "learning_rate": 8.922063135923815e-06,
      "loss": 0.72272043,
      "memory(GiB)": 135.77,
      "step": 20280,
      "train_speed(iter/s)": 0.202237
    },
    {
      "acc": 0.79387074,
      "epoch": 0.4733810641741641,
      "grad_norm": 4.9375,
      "learning_rate": 8.920891154672668e-06,
      "loss": 0.75309858,
      "memory(GiB)": 135.77,
      "step": 20290,
      "train_speed(iter/s)": 0.202289
    },
    {
      "acc": 0.7726409,
      "epoch": 0.473614371746453,
      "grad_norm": 8.4375,
      "learning_rate": 8.91971861371545e-06,
      "loss": 0.83473721,
      "memory(GiB)": 135.77,
      "step": 20300,
      "train_speed(iter/s)": 0.202341
    },
    {
      "acc": 0.77393756,
      "epoch": 0.4738476793187419,
      "grad_norm": 3.65625,
      "learning_rate": 8.918545513219535e-06,
      "loss": 0.79616671,
      "memory(GiB)": 135.77,
      "step": 20310,
      "train_speed(iter/s)": 0.202391
    },
    {
      "acc": 0.76672659,
      "epoch": 0.4740809868910308,
      "grad_norm": 5.90625,
      "learning_rate": 8.917371853352388e-06,
      "loss": 0.85512495,
      "memory(GiB)": 135.77,
      "step": 20320,
      "train_speed(iter/s)": 0.202442
    },
    {
      "acc": 0.765168,
      "epoch": 0.4743142944633197,
      "grad_norm": 8.625,
      "learning_rate": 8.916197634281547e-06,
      "loss": 0.82549314,
      "memory(GiB)": 135.77,
      "step": 20330,
      "train_speed(iter/s)": 0.202492
    },
    {
      "acc": 0.77795401,
      "epoch": 0.4745476020356086,
      "grad_norm": 5.28125,
      "learning_rate": 8.91502285617463e-06,
      "loss": 0.79356041,
      "memory(GiB)": 135.77,
      "step": 20340,
      "train_speed(iter/s)": 0.202544
    },
    {
      "acc": 0.75690231,
      "epoch": 0.4747809096078975,
      "grad_norm": 5.875,
      "learning_rate": 8.913847519199341e-06,
      "loss": 0.89064808,
      "memory(GiB)": 135.77,
      "step": 20350,
      "train_speed(iter/s)": 0.202595
    },
    {
      "acc": 0.76705751,
      "epoch": 0.4750142171801863,
      "grad_norm": 6.65625,
      "learning_rate": 8.912671623523452e-06,
      "loss": 0.84812698,
      "memory(GiB)": 135.77,
      "step": 20360,
      "train_speed(iter/s)": 0.202641
    },
    {
      "acc": 0.78243647,
      "epoch": 0.4752475247524752,
      "grad_norm": 6.40625,
      "learning_rate": 8.911495169314828e-06,
      "loss": 0.77283859,
      "memory(GiB)": 135.77,
      "step": 20370,
      "train_speed(iter/s)": 0.20269
    },
    {
      "acc": 0.78674965,
      "epoch": 0.4754808323247641,
      "grad_norm": 5.59375,
      "learning_rate": 8.910318156741401e-06,
      "loss": 0.76450891,
      "memory(GiB)": 135.77,
      "step": 20380,
      "train_speed(iter/s)": 0.20274
    },
    {
      "acc": 0.7955699,
      "epoch": 0.475714139897053,
      "grad_norm": 4.5,
      "learning_rate": 8.909140585971198e-06,
      "loss": 0.73192253,
      "memory(GiB)": 135.77,
      "step": 20390,
      "train_speed(iter/s)": 0.202788
    },
    {
      "acc": 0.78440504,
      "epoch": 0.4759474474693419,
      "grad_norm": 5.28125,
      "learning_rate": 8.90796245717231e-06,
      "loss": 0.77477818,
      "memory(GiB)": 135.77,
      "step": 20400,
      "train_speed(iter/s)": 0.202837
    },
    {
      "acc": 0.76490378,
      "epoch": 0.4761807550416308,
      "grad_norm": 8.9375,
      "learning_rate": 8.906783770512915e-06,
      "loss": 0.86861458,
      "memory(GiB)": 135.77,
      "step": 20410,
      "train_speed(iter/s)": 0.202886
    },
    {
      "acc": 0.76525488,
      "epoch": 0.4764140626139197,
      "grad_norm": 5.4375,
      "learning_rate": 8.905604526161274e-06,
      "loss": 0.82954245,
      "memory(GiB)": 135.77,
      "step": 20420,
      "train_speed(iter/s)": 0.20294
    },
    {
      "acc": 0.7579958,
      "epoch": 0.4766473701862086,
      "grad_norm": 6.25,
      "learning_rate": 8.904424724285721e-06,
      "loss": 0.87198391,
      "memory(GiB)": 135.77,
      "step": 20430,
      "train_speed(iter/s)": 0.202991
    },
    {
      "acc": 0.76998606,
      "epoch": 0.4768806777584975,
      "grad_norm": 6.78125,
      "learning_rate": 8.903244365054671e-06,
      "loss": 0.81940556,
      "memory(GiB)": 135.77,
      "step": 20440,
      "train_speed(iter/s)": 0.20304
    },
    {
      "acc": 0.76503363,
      "epoch": 0.4771139853307864,
      "grad_norm": 5.8125,
      "learning_rate": 8.902063448636624e-06,
      "loss": 0.86698151,
      "memory(GiB)": 135.77,
      "step": 20450,
      "train_speed(iter/s)": 0.203093
    },
    {
      "acc": 0.77026491,
      "epoch": 0.47734729290307526,
      "grad_norm": 5.71875,
      "learning_rate": 8.900881975200151e-06,
      "loss": 0.82462835,
      "memory(GiB)": 135.77,
      "step": 20460,
      "train_speed(iter/s)": 0.203142
    },
    {
      "acc": 0.76255193,
      "epoch": 0.47758060047536416,
      "grad_norm": 4.3125,
      "learning_rate": 8.89969994491391e-06,
      "loss": 0.84601803,
      "memory(GiB)": 135.77,
      "step": 20470,
      "train_speed(iter/s)": 0.203189
    },
    {
      "acc": 0.75619602,
      "epoch": 0.47781390804765306,
      "grad_norm": 6.28125,
      "learning_rate": 8.898517357946636e-06,
      "loss": 0.88812332,
      "memory(GiB)": 135.77,
      "step": 20480,
      "train_speed(iter/s)": 0.203241
    },
    {
      "acc": 0.77179279,
      "epoch": 0.47804721561994196,
      "grad_norm": 6.3125,
      "learning_rate": 8.897334214467141e-06,
      "loss": 0.82135019,
      "memory(GiB)": 135.77,
      "step": 20490,
      "train_speed(iter/s)": 0.203288
    },
    {
      "acc": 0.77499766,
      "epoch": 0.47828052319223086,
      "grad_norm": 5.53125,
      "learning_rate": 8.89615051464432e-06,
      "loss": 0.79335108,
      "memory(GiB)": 135.77,
      "step": 20500,
      "train_speed(iter/s)": 0.203339
    },
    {
      "epoch": 0.47828052319223086,
      "eval_acc": 0.7387652014910709,
      "eval_loss": 0.827155590057373,
      "eval_runtime": 1271.1109,
      "eval_samples_per_second": 28.315,
      "eval_steps_per_second": 14.158,
      "step": 20500
    },
    {
      "acc": 0.76934633,
      "epoch": 0.47851383076451975,
      "grad_norm": 5.34375,
      "learning_rate": 8.894966258647144e-06,
      "loss": 0.82157841,
      "memory(GiB)": 135.77,
      "step": 20510,
      "train_speed(iter/s)": 0.200826
    },
    {
      "acc": 0.76049023,
      "epoch": 0.47874713833680865,
      "grad_norm": 6.0625,
      "learning_rate": 8.893781446644667e-06,
      "loss": 0.87835636,
      "memory(GiB)": 135.77,
      "step": 20520,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.77881689,
      "epoch": 0.47898044590909755,
      "grad_norm": 4.375,
      "learning_rate": 8.892596078806017e-06,
      "loss": 0.81451454,
      "memory(GiB)": 135.77,
      "step": 20530,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.77435408,
      "epoch": 0.47921375348138645,
      "grad_norm": 4.8125,
      "learning_rate": 8.89141015530041e-06,
      "loss": 0.8198391,
      "memory(GiB)": 135.77,
      "step": 20540,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.77629027,
      "epoch": 0.47944706105367535,
      "grad_norm": 4.34375,
      "learning_rate": 8.890223676297132e-06,
      "loss": 0.80017614,
      "memory(GiB)": 135.77,
      "step": 20550,
      "train_speed(iter/s)": 0.201028
    },
    {
      "acc": 0.78644643,
      "epoch": 0.4796803686259642,
      "grad_norm": 6.3125,
      "learning_rate": 8.889036641965557e-06,
      "loss": 0.77309122,
      "memory(GiB)": 135.77,
      "step": 20560,
      "train_speed(iter/s)": 0.201078
    },
    {
      "acc": 0.77234049,
      "epoch": 0.4799136761982531,
      "grad_norm": 10.125,
      "learning_rate": 8.887849052475128e-06,
      "loss": 0.82501678,
      "memory(GiB)": 135.77,
      "step": 20570,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.78163843,
      "epoch": 0.480146983770542,
      "grad_norm": 4.6875,
      "learning_rate": 8.886660907995379e-06,
      "loss": 0.80244398,
      "memory(GiB)": 135.77,
      "step": 20580,
      "train_speed(iter/s)": 0.20118
    },
    {
      "acc": 0.77460861,
      "epoch": 0.4803802913428309,
      "grad_norm": 6.0,
      "learning_rate": 8.885472208695911e-06,
      "loss": 0.82104893,
      "memory(GiB)": 135.77,
      "step": 20590,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.7811389,
      "epoch": 0.4806135989151198,
      "grad_norm": 6.78125,
      "learning_rate": 8.884282954746417e-06,
      "loss": 0.76911659,
      "memory(GiB)": 135.77,
      "step": 20600,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.77763205,
      "epoch": 0.4808469064874087,
      "grad_norm": 5.1875,
      "learning_rate": 8.88309314631666e-06,
      "loss": 0.78176918,
      "memory(GiB)": 135.77,
      "step": 20610,
      "train_speed(iter/s)": 0.201334
    },
    {
      "acc": 0.76166954,
      "epoch": 0.4810802140596976,
      "grad_norm": 3.90625,
      "learning_rate": 8.881902783576482e-06,
      "loss": 0.86578398,
      "memory(GiB)": 135.77,
      "step": 20620,
      "train_speed(iter/s)": 0.201383
    },
    {
      "acc": 0.7865201,
      "epoch": 0.4813135216319865,
      "grad_norm": 4.96875,
      "learning_rate": 8.88071186669581e-06,
      "loss": 0.77439818,
      "memory(GiB)": 135.77,
      "step": 20630,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.76671553,
      "epoch": 0.4815468292042754,
      "grad_norm": 5.03125,
      "learning_rate": 8.879520395844648e-06,
      "loss": 0.84458523,
      "memory(GiB)": 135.77,
      "step": 20640,
      "train_speed(iter/s)": 0.201488
    },
    {
      "acc": 0.75628233,
      "epoch": 0.4817801367765642,
      "grad_norm": 5.25,
      "learning_rate": 8.878328371193074e-06,
      "loss": 0.86866341,
      "memory(GiB)": 135.77,
      "step": 20650,
      "train_speed(iter/s)": 0.201541
    },
    {
      "acc": 0.76081595,
      "epoch": 0.4820134443488531,
      "grad_norm": 5.09375,
      "learning_rate": 8.877135792911253e-06,
      "loss": 0.86255798,
      "memory(GiB)": 135.77,
      "step": 20660,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.75712156,
      "epoch": 0.482246751921142,
      "grad_norm": 6.03125,
      "learning_rate": 8.875942661169423e-06,
      "loss": 0.89423008,
      "memory(GiB)": 135.77,
      "step": 20670,
      "train_speed(iter/s)": 0.201641
    },
    {
      "acc": 0.77293806,
      "epoch": 0.4824800594934309,
      "grad_norm": 5.5,
      "learning_rate": 8.874748976137905e-06,
      "loss": 0.79875817,
      "memory(GiB)": 135.77,
      "step": 20680,
      "train_speed(iter/s)": 0.201693
    },
    {
      "acc": 0.74118772,
      "epoch": 0.4827133670657198,
      "grad_norm": 5.5625,
      "learning_rate": 8.873554737987098e-06,
      "loss": 0.91008282,
      "memory(GiB)": 135.77,
      "step": 20690,
      "train_speed(iter/s)": 0.201745
    },
    {
      "acc": 0.7834497,
      "epoch": 0.4829466746380087,
      "grad_norm": 7.0625,
      "learning_rate": 8.872359946887474e-06,
      "loss": 0.76405153,
      "memory(GiB)": 135.77,
      "step": 20700,
      "train_speed(iter/s)": 0.201796
    },
    {
      "acc": 0.73779564,
      "epoch": 0.4831799822102976,
      "grad_norm": 5.78125,
      "learning_rate": 8.871164603009595e-06,
      "loss": 0.97299786,
      "memory(GiB)": 135.77,
      "step": 20710,
      "train_speed(iter/s)": 0.201848
    },
    {
      "acc": 0.76881113,
      "epoch": 0.4834132897825865,
      "grad_norm": 5.6875,
      "learning_rate": 8.869968706524092e-06,
      "loss": 0.92142582,
      "memory(GiB)": 135.77,
      "step": 20720,
      "train_speed(iter/s)": 0.201901
    },
    {
      "acc": 0.78435059,
      "epoch": 0.4836465973548754,
      "grad_norm": 4.375,
      "learning_rate": 8.868772257601682e-06,
      "loss": 0.79667149,
      "memory(GiB)": 135.77,
      "step": 20730,
      "train_speed(iter/s)": 0.201952
    },
    {
      "acc": 0.75337543,
      "epoch": 0.4838799049271643,
      "grad_norm": 7.15625,
      "learning_rate": 8.867575256413154e-06,
      "loss": 0.91704159,
      "memory(GiB)": 135.77,
      "step": 20740,
      "train_speed(iter/s)": 0.202003
    },
    {
      "acc": 0.76679363,
      "epoch": 0.48411321249945316,
      "grad_norm": 6.15625,
      "learning_rate": 8.866377703129382e-06,
      "loss": 0.82663794,
      "memory(GiB)": 135.77,
      "step": 20750,
      "train_speed(iter/s)": 0.202056
    },
    {
      "acc": 0.77863007,
      "epoch": 0.48434652007174206,
      "grad_norm": 4.8125,
      "learning_rate": 8.865179597921318e-06,
      "loss": 0.79373541,
      "memory(GiB)": 135.77,
      "step": 20760,
      "train_speed(iter/s)": 0.202104
    },
    {
      "acc": 0.75620103,
      "epoch": 0.48457982764403096,
      "grad_norm": 6.0625,
      "learning_rate": 8.863980940959989e-06,
      "loss": 0.92305183,
      "memory(GiB)": 135.77,
      "step": 20770,
      "train_speed(iter/s)": 0.20215
    },
    {
      "acc": 0.7790472,
      "epoch": 0.48481313521631986,
      "grad_norm": 4.78125,
      "learning_rate": 8.862781732416502e-06,
      "loss": 0.77640362,
      "memory(GiB)": 135.77,
      "step": 20780,
      "train_speed(iter/s)": 0.2022
    },
    {
      "acc": 0.77119627,
      "epoch": 0.48504644278860876,
      "grad_norm": 6.0625,
      "learning_rate": 8.861581972462045e-06,
      "loss": 0.82055054,
      "memory(GiB)": 135.77,
      "step": 20790,
      "train_speed(iter/s)": 0.202249
    },
    {
      "acc": 0.76551666,
      "epoch": 0.48527975036089765,
      "grad_norm": 5.5,
      "learning_rate": 8.860381661267882e-06,
      "loss": 0.82655621,
      "memory(GiB)": 135.77,
      "step": 20800,
      "train_speed(iter/s)": 0.202298
    },
    {
      "acc": 0.78621545,
      "epoch": 0.48551305793318655,
      "grad_norm": 8.9375,
      "learning_rate": 8.859180799005361e-06,
      "loss": 0.76219521,
      "memory(GiB)": 135.77,
      "step": 20810,
      "train_speed(iter/s)": 0.202346
    },
    {
      "acc": 0.77965999,
      "epoch": 0.48574636550547545,
      "grad_norm": 5.34375,
      "learning_rate": 8.857979385845901e-06,
      "loss": 0.781145,
      "memory(GiB)": 135.77,
      "step": 20820,
      "train_speed(iter/s)": 0.202396
    },
    {
      "acc": 0.76239419,
      "epoch": 0.48597967307776435,
      "grad_norm": 4.75,
      "learning_rate": 8.856777421961004e-06,
      "loss": 0.87815866,
      "memory(GiB)": 135.77,
      "step": 20830,
      "train_speed(iter/s)": 0.202448
    },
    {
      "acc": 0.77660542,
      "epoch": 0.48621298065005325,
      "grad_norm": 5.1875,
      "learning_rate": 8.855574907522251e-06,
      "loss": 0.79042578,
      "memory(GiB)": 135.77,
      "step": 20840,
      "train_speed(iter/s)": 0.202497
    },
    {
      "acc": 0.7720644,
      "epoch": 0.4864462882223421,
      "grad_norm": 6.375,
      "learning_rate": 8.854371842701299e-06,
      "loss": 0.82205925,
      "memory(GiB)": 135.77,
      "step": 20850,
      "train_speed(iter/s)": 0.202548
    },
    {
      "acc": 0.77594128,
      "epoch": 0.486679595794631,
      "grad_norm": 5.625,
      "learning_rate": 8.853168227669886e-06,
      "loss": 0.81534863,
      "memory(GiB)": 135.77,
      "step": 20860,
      "train_speed(iter/s)": 0.202599
    },
    {
      "acc": 0.77423177,
      "epoch": 0.4869129033669199,
      "grad_norm": 5.0,
      "learning_rate": 8.851964062599828e-06,
      "loss": 0.79051189,
      "memory(GiB)": 135.77,
      "step": 20870,
      "train_speed(iter/s)": 0.202649
    },
    {
      "acc": 0.77537422,
      "epoch": 0.4871462109392088,
      "grad_norm": 15.1875,
      "learning_rate": 8.850759347663021e-06,
      "loss": 0.80190439,
      "memory(GiB)": 135.77,
      "step": 20880,
      "train_speed(iter/s)": 0.202703
    },
    {
      "acc": 0.76119142,
      "epoch": 0.4873795185114977,
      "grad_norm": 6.40625,
      "learning_rate": 8.849554083031435e-06,
      "loss": 0.86152458,
      "memory(GiB)": 135.77,
      "step": 20890,
      "train_speed(iter/s)": 0.202756
    },
    {
      "acc": 0.75670996,
      "epoch": 0.4876128260837866,
      "grad_norm": 5.1875,
      "learning_rate": 8.84834826887712e-06,
      "loss": 0.87847652,
      "memory(GiB)": 135.77,
      "step": 20900,
      "train_speed(iter/s)": 0.20281
    },
    {
      "acc": 0.76710215,
      "epoch": 0.4878461336560755,
      "grad_norm": 6.84375,
      "learning_rate": 8.84714190537221e-06,
      "loss": 0.84475441,
      "memory(GiB)": 135.77,
      "step": 20910,
      "train_speed(iter/s)": 0.20286
    },
    {
      "acc": 0.77066274,
      "epoch": 0.4880794412283644,
      "grad_norm": 5.6875,
      "learning_rate": 8.84593499268891e-06,
      "loss": 0.82889204,
      "memory(GiB)": 135.77,
      "step": 20920,
      "train_speed(iter/s)": 0.202908
    },
    {
      "acc": 0.76561365,
      "epoch": 0.4883127488006533,
      "grad_norm": 4.75,
      "learning_rate": 8.844727530999506e-06,
      "loss": 0.8475029,
      "memory(GiB)": 135.77,
      "step": 20930,
      "train_speed(iter/s)": 0.202954
    },
    {
      "acc": 0.77417688,
      "epoch": 0.48854605637294213,
      "grad_norm": 4.59375,
      "learning_rate": 8.843519520476365e-06,
      "loss": 0.82181873,
      "memory(GiB)": 135.77,
      "step": 20940,
      "train_speed(iter/s)": 0.203003
    },
    {
      "acc": 0.78241935,
      "epoch": 0.488779363945231,
      "grad_norm": 5.1875,
      "learning_rate": 8.842310961291926e-06,
      "loss": 0.80962143,
      "memory(GiB)": 135.77,
      "step": 20950,
      "train_speed(iter/s)": 0.203051
    },
    {
      "acc": 0.79182229,
      "epoch": 0.4890126715175199,
      "grad_norm": 7.59375,
      "learning_rate": 8.841101853618717e-06,
      "loss": 0.74582729,
      "memory(GiB)": 135.77,
      "step": 20960,
      "train_speed(iter/s)": 0.203103
    },
    {
      "acc": 0.78107166,
      "epoch": 0.4892459790898088,
      "grad_norm": 4.21875,
      "learning_rate": 8.839892197629334e-06,
      "loss": 0.78661366,
      "memory(GiB)": 135.77,
      "step": 20970,
      "train_speed(iter/s)": 0.203151
    },
    {
      "acc": 0.77803307,
      "epoch": 0.4894792866620977,
      "grad_norm": 6.875,
      "learning_rate": 8.838681993496454e-06,
      "loss": 0.79262047,
      "memory(GiB)": 135.77,
      "step": 20980,
      "train_speed(iter/s)": 0.203205
    },
    {
      "acc": 0.74206686,
      "epoch": 0.4897125942343866,
      "grad_norm": 7.4375,
      "learning_rate": 8.837471241392835e-06,
      "loss": 0.94257679,
      "memory(GiB)": 135.77,
      "step": 20990,
      "train_speed(iter/s)": 0.203256
    },
    {
      "acc": 0.76296959,
      "epoch": 0.4899459018066755,
      "grad_norm": 6.03125,
      "learning_rate": 8.83625994149131e-06,
      "loss": 0.87966471,
      "memory(GiB)": 135.77,
      "step": 21000,
      "train_speed(iter/s)": 0.203307
    },
    {
      "epoch": 0.4899459018066755,
      "eval_acc": 0.7386233137012519,
      "eval_loss": 0.8264919519424438,
      "eval_runtime": 1269.5678,
      "eval_samples_per_second": 28.349,
      "eval_steps_per_second": 14.175,
      "step": 21000
    },
    {
      "acc": 0.76228848,
      "epoch": 0.4901792093789644,
      "grad_norm": 5.1875,
      "learning_rate": 8.835048093964796e-06,
      "loss": 0.86440468,
      "memory(GiB)": 135.77,
      "step": 21010,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.7821867,
      "epoch": 0.4904125169512533,
      "grad_norm": 4.21875,
      "learning_rate": 8.833835698986276e-06,
      "loss": 0.78023453,
      "memory(GiB)": 135.77,
      "step": 21020,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.77446461,
      "epoch": 0.4906458245235422,
      "grad_norm": 5.3125,
      "learning_rate": 8.832622756728828e-06,
      "loss": 0.81071396,
      "memory(GiB)": 135.77,
      "step": 21030,
      "train_speed(iter/s)": 0.200954
    },
    {
      "acc": 0.75736957,
      "epoch": 0.49087913209583106,
      "grad_norm": 5.9375,
      "learning_rate": 8.831409267365594e-06,
      "loss": 0.87871695,
      "memory(GiB)": 135.77,
      "step": 21040,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.7824193,
      "epoch": 0.49111243966811996,
      "grad_norm": 5.21875,
      "learning_rate": 8.830195231069799e-06,
      "loss": 0.77914681,
      "memory(GiB)": 135.77,
      "step": 21050,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.77138987,
      "epoch": 0.49134574724040886,
      "grad_norm": 7.03125,
      "learning_rate": 8.828980648014747e-06,
      "loss": 0.82984867,
      "memory(GiB)": 135.77,
      "step": 21060,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.76764355,
      "epoch": 0.49157905481269776,
      "grad_norm": 6.15625,
      "learning_rate": 8.82776551837382e-06,
      "loss": 0.83743019,
      "memory(GiB)": 135.77,
      "step": 21070,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.7618494,
      "epoch": 0.49181236238498666,
      "grad_norm": 5.40625,
      "learning_rate": 8.826549842320478e-06,
      "loss": 0.87587261,
      "memory(GiB)": 135.77,
      "step": 21080,
      "train_speed(iter/s)": 0.201209
    },
    {
      "acc": 0.77934647,
      "epoch": 0.49204566995727556,
      "grad_norm": 4.4375,
      "learning_rate": 8.825333620028257e-06,
      "loss": 0.80535469,
      "memory(GiB)": 135.77,
      "step": 21090,
      "train_speed(iter/s)": 0.201258
    },
    {
      "acc": 0.76986637,
      "epoch": 0.49227897752956445,
      "grad_norm": 4.375,
      "learning_rate": 8.824116851670772e-06,
      "loss": 0.83490791,
      "memory(GiB)": 135.77,
      "step": 21100,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.7731987,
      "epoch": 0.49251228510185335,
      "grad_norm": 6.5625,
      "learning_rate": 8.822899537421721e-06,
      "loss": 0.83892517,
      "memory(GiB)": 135.77,
      "step": 21110,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.76397038,
      "epoch": 0.49274559267414225,
      "grad_norm": 6.46875,
      "learning_rate": 8.821681677454868e-06,
      "loss": 0.84116879,
      "memory(GiB)": 135.77,
      "step": 21120,
      "train_speed(iter/s)": 0.201407
    },
    {
      "acc": 0.76126223,
      "epoch": 0.49297890024643115,
      "grad_norm": 4.8125,
      "learning_rate": 8.820463271944066e-06,
      "loss": 0.85854349,
      "memory(GiB)": 135.77,
      "step": 21130,
      "train_speed(iter/s)": 0.201458
    },
    {
      "acc": 0.77129207,
      "epoch": 0.49321220781872,
      "grad_norm": 8.4375,
      "learning_rate": 8.819244321063243e-06,
      "loss": 0.82659998,
      "memory(GiB)": 135.77,
      "step": 21140,
      "train_speed(iter/s)": 0.201505
    },
    {
      "acc": 0.76816425,
      "epoch": 0.4934455153910089,
      "grad_norm": 7.4375,
      "learning_rate": 8.818024824986404e-06,
      "loss": 0.84741154,
      "memory(GiB)": 135.77,
      "step": 21150,
      "train_speed(iter/s)": 0.201552
    },
    {
      "acc": 0.76245174,
      "epoch": 0.4936788229632978,
      "grad_norm": 8.6875,
      "learning_rate": 8.816804783887628e-06,
      "loss": 0.86243572,
      "memory(GiB)": 135.77,
      "step": 21160,
      "train_speed(iter/s)": 0.2016
    },
    {
      "acc": 0.78625603,
      "epoch": 0.4939121305355867,
      "grad_norm": 6.09375,
      "learning_rate": 8.815584197941078e-06,
      "loss": 0.76783695,
      "memory(GiB)": 135.77,
      "step": 21170,
      "train_speed(iter/s)": 0.201648
    },
    {
      "acc": 0.79220161,
      "epoch": 0.4941454381078756,
      "grad_norm": 6.96875,
      "learning_rate": 8.814363067320995e-06,
      "loss": 0.75065899,
      "memory(GiB)": 135.77,
      "step": 21180,
      "train_speed(iter/s)": 0.201696
    },
    {
      "acc": 0.78739977,
      "epoch": 0.4943787456801645,
      "grad_norm": 5.5,
      "learning_rate": 8.81314139220169e-06,
      "loss": 0.77436671,
      "memory(GiB)": 135.77,
      "step": 21190,
      "train_speed(iter/s)": 0.201746
    },
    {
      "acc": 0.75706043,
      "epoch": 0.4946120532524534,
      "grad_norm": 4.375,
      "learning_rate": 8.811919172757558e-06,
      "loss": 0.86744957,
      "memory(GiB)": 135.77,
      "step": 21200,
      "train_speed(iter/s)": 0.201796
    },
    {
      "acc": 0.77850995,
      "epoch": 0.4948453608247423,
      "grad_norm": 5.03125,
      "learning_rate": 8.810696409163073e-06,
      "loss": 0.81761723,
      "memory(GiB)": 135.77,
      "step": 21210,
      "train_speed(iter/s)": 0.201845
    },
    {
      "acc": 0.79036131,
      "epoch": 0.4950786683970312,
      "grad_norm": 4.65625,
      "learning_rate": 8.809473101592783e-06,
      "loss": 0.77943802,
      "memory(GiB)": 135.77,
      "step": 21220,
      "train_speed(iter/s)": 0.201896
    },
    {
      "acc": 0.77873907,
      "epoch": 0.49531197596932003,
      "grad_norm": 5.75,
      "learning_rate": 8.808249250221312e-06,
      "loss": 0.79133549,
      "memory(GiB)": 135.77,
      "step": 21230,
      "train_speed(iter/s)": 0.201945
    },
    {
      "acc": 0.77417302,
      "epoch": 0.4955452835416089,
      "grad_norm": 5.84375,
      "learning_rate": 8.807024855223369e-06,
      "loss": 0.81216822,
      "memory(GiB)": 135.77,
      "step": 21240,
      "train_speed(iter/s)": 0.201996
    },
    {
      "acc": 0.76840887,
      "epoch": 0.4957785911138978,
      "grad_norm": 7.0625,
      "learning_rate": 8.805799916773734e-06,
      "loss": 0.82953939,
      "memory(GiB)": 135.77,
      "step": 21250,
      "train_speed(iter/s)": 0.202046
    },
    {
      "acc": 0.74933271,
      "epoch": 0.4960118986861867,
      "grad_norm": 6.4375,
      "learning_rate": 8.804574435047265e-06,
      "loss": 0.90532026,
      "memory(GiB)": 135.77,
      "step": 21260,
      "train_speed(iter/s)": 0.202093
    },
    {
      "acc": 0.77972379,
      "epoch": 0.4962452062584756,
      "grad_norm": 6.4375,
      "learning_rate": 8.803348410218902e-06,
      "loss": 0.80596933,
      "memory(GiB)": 135.77,
      "step": 21270,
      "train_speed(iter/s)": 0.202143
    },
    {
      "acc": 0.77269158,
      "epoch": 0.4964785138307645,
      "grad_norm": 4.53125,
      "learning_rate": 8.802121842463658e-06,
      "loss": 0.81263514,
      "memory(GiB)": 135.77,
      "step": 21280,
      "train_speed(iter/s)": 0.202195
    },
    {
      "acc": 0.77480335,
      "epoch": 0.4967118214030534,
      "grad_norm": 6.21875,
      "learning_rate": 8.800894731956624e-06,
      "loss": 0.81107597,
      "memory(GiB)": 135.77,
      "step": 21290,
      "train_speed(iter/s)": 0.202243
    },
    {
      "acc": 0.75720205,
      "epoch": 0.4969451289753423,
      "grad_norm": 8.5625,
      "learning_rate": 8.799667078872973e-06,
      "loss": 0.88299694,
      "memory(GiB)": 135.77,
      "step": 21300,
      "train_speed(iter/s)": 0.202292
    },
    {
      "acc": 0.75752621,
      "epoch": 0.4971784365476312,
      "grad_norm": 6.65625,
      "learning_rate": 8.79843888338795e-06,
      "loss": 0.90188789,
      "memory(GiB)": 135.77,
      "step": 21310,
      "train_speed(iter/s)": 0.202342
    },
    {
      "acc": 0.77910862,
      "epoch": 0.4974117441199201,
      "grad_norm": 4.84375,
      "learning_rate": 8.797210145676879e-06,
      "loss": 0.78877602,
      "memory(GiB)": 135.77,
      "step": 21320,
      "train_speed(iter/s)": 0.202388
    },
    {
      "acc": 0.78843384,
      "epoch": 0.49764505169220896,
      "grad_norm": 5.15625,
      "learning_rate": 8.795980865915164e-06,
      "loss": 0.75855894,
      "memory(GiB)": 135.77,
      "step": 21330,
      "train_speed(iter/s)": 0.202436
    },
    {
      "acc": 0.78700695,
      "epoch": 0.49787835926449786,
      "grad_norm": 6.125,
      "learning_rate": 8.794751044278282e-06,
      "loss": 0.76552219,
      "memory(GiB)": 135.77,
      "step": 21340,
      "train_speed(iter/s)": 0.202487
    },
    {
      "acc": 0.77964926,
      "epoch": 0.49811166683678676,
      "grad_norm": 5.53125,
      "learning_rate": 8.793520680941792e-06,
      "loss": 0.78910999,
      "memory(GiB)": 135.77,
      "step": 21350,
      "train_speed(iter/s)": 0.202533
    },
    {
      "acc": 0.76680984,
      "epoch": 0.49834497440907566,
      "grad_norm": 6.96875,
      "learning_rate": 8.792289776081326e-06,
      "loss": 0.84402256,
      "memory(GiB)": 135.77,
      "step": 21360,
      "train_speed(iter/s)": 0.20258
    },
    {
      "acc": 0.7667099,
      "epoch": 0.49857828198136456,
      "grad_norm": 4.25,
      "learning_rate": 8.791058329872595e-06,
      "loss": 0.85665445,
      "memory(GiB)": 135.77,
      "step": 21370,
      "train_speed(iter/s)": 0.20263
    },
    {
      "acc": 0.77208252,
      "epoch": 0.49881158955365346,
      "grad_norm": 4.71875,
      "learning_rate": 8.78982634249139e-06,
      "loss": 0.82627926,
      "memory(GiB)": 135.77,
      "step": 21380,
      "train_speed(iter/s)": 0.202672
    },
    {
      "acc": 0.76971836,
      "epoch": 0.49904489712594235,
      "grad_norm": 5.03125,
      "learning_rate": 8.788593814113576e-06,
      "loss": 0.82828751,
      "memory(GiB)": 135.77,
      "step": 21390,
      "train_speed(iter/s)": 0.20272
    },
    {
      "acc": 0.77777624,
      "epoch": 0.49927820469823125,
      "grad_norm": 7.25,
      "learning_rate": 8.787360744915096e-06,
      "loss": 0.78845906,
      "memory(GiB)": 135.77,
      "step": 21400,
      "train_speed(iter/s)": 0.202769
    },
    {
      "acc": 0.76814055,
      "epoch": 0.49951151227052015,
      "grad_norm": 4.9375,
      "learning_rate": 8.786127135071968e-06,
      "loss": 0.82526455,
      "memory(GiB)": 135.77,
      "step": 21410,
      "train_speed(iter/s)": 0.202816
    },
    {
      "acc": 0.76935434,
      "epoch": 0.499744819842809,
      "grad_norm": 4.84375,
      "learning_rate": 8.784892984760292e-06,
      "loss": 0.82196589,
      "memory(GiB)": 135.77,
      "step": 21420,
      "train_speed(iter/s)": 0.20286
    },
    {
      "acc": 0.76840973,
      "epoch": 0.4999781274150979,
      "grad_norm": 4.65625,
      "learning_rate": 8.783658294156244e-06,
      "loss": 0.84509468,
      "memory(GiB)": 135.77,
      "step": 21430,
      "train_speed(iter/s)": 0.202908
    },
    {
      "acc": 0.78127012,
      "epoch": 0.5002114349873868,
      "grad_norm": 5.0,
      "learning_rate": 8.782423063436072e-06,
      "loss": 0.79705486,
      "memory(GiB)": 135.77,
      "step": 21440,
      "train_speed(iter/s)": 0.202956
    },
    {
      "acc": 0.79307652,
      "epoch": 0.5004447425596757,
      "grad_norm": 6.28125,
      "learning_rate": 8.781187292776106e-06,
      "loss": 0.75553842,
      "memory(GiB)": 135.77,
      "step": 21450,
      "train_speed(iter/s)": 0.203003
    },
    {
      "acc": 0.77026339,
      "epoch": 0.5006780501319646,
      "grad_norm": 5.28125,
      "learning_rate": 8.779950982352751e-06,
      "loss": 0.82962399,
      "memory(GiB)": 135.77,
      "step": 21460,
      "train_speed(iter/s)": 0.203051
    },
    {
      "acc": 0.78340578,
      "epoch": 0.5009113577042534,
      "grad_norm": 4.40625,
      "learning_rate": 8.778714132342494e-06,
      "loss": 0.76500807,
      "memory(GiB)": 135.77,
      "step": 21470,
      "train_speed(iter/s)": 0.203099
    },
    {
      "acc": 0.76358261,
      "epoch": 0.5011446652765423,
      "grad_norm": 5.3125,
      "learning_rate": 8.777476742921893e-06,
      "loss": 0.86745148,
      "memory(GiB)": 135.77,
      "step": 21480,
      "train_speed(iter/s)": 0.203144
    },
    {
      "acc": 0.76443934,
      "epoch": 0.5013779728488312,
      "grad_norm": 6.3125,
      "learning_rate": 8.776238814267581e-06,
      "loss": 0.85055971,
      "memory(GiB)": 135.77,
      "step": 21490,
      "train_speed(iter/s)": 0.203195
    },
    {
      "acc": 0.7856565,
      "epoch": 0.5016112804211201,
      "grad_norm": 4.84375,
      "learning_rate": 8.775000346556278e-06,
      "loss": 0.76976533,
      "memory(GiB)": 135.77,
      "step": 21500,
      "train_speed(iter/s)": 0.203245
    },
    {
      "epoch": 0.5016112804211201,
      "eval_acc": 0.7388905944249629,
      "eval_loss": 0.8257662653923035,
      "eval_runtime": 1270.2669,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 21500
    },
    {
      "acc": 0.77753158,
      "epoch": 0.501844587993409,
      "grad_norm": 5.96875,
      "learning_rate": 8.773761339964773e-06,
      "loss": 0.79236035,
      "memory(GiB)": 135.77,
      "step": 21510,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.77459497,
      "epoch": 0.5020778955656979,
      "grad_norm": 6.34375,
      "learning_rate": 8.77252179466993e-06,
      "loss": 0.80510149,
      "memory(GiB)": 135.77,
      "step": 21520,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.79065723,
      "epoch": 0.5023112031379868,
      "grad_norm": 5.875,
      "learning_rate": 8.771281710848697e-06,
      "loss": 0.75441141,
      "memory(GiB)": 135.77,
      "step": 21530,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.78152294,
      "epoch": 0.5025445107102757,
      "grad_norm": 5.6875,
      "learning_rate": 8.770041088678098e-06,
      "loss": 0.78967323,
      "memory(GiB)": 135.77,
      "step": 21540,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.76808047,
      "epoch": 0.5027778182825646,
      "grad_norm": 4.875,
      "learning_rate": 8.768799928335227e-06,
      "loss": 0.83531723,
      "memory(GiB)": 135.77,
      "step": 21550,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.78148546,
      "epoch": 0.5030111258548535,
      "grad_norm": 6.8125,
      "learning_rate": 8.76755822999726e-06,
      "loss": 0.77231493,
      "memory(GiB)": 135.77,
      "step": 21560,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.78652892,
      "epoch": 0.5032444334271424,
      "grad_norm": 6.5,
      "learning_rate": 8.766315993841452e-06,
      "loss": 0.76357327,
      "memory(GiB)": 135.77,
      "step": 21570,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.74624329,
      "epoch": 0.5034777409994313,
      "grad_norm": 4.90625,
      "learning_rate": 8.76507322004513e-06,
      "loss": 0.91759214,
      "memory(GiB)": 135.77,
      "step": 21580,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.77871332,
      "epoch": 0.5037110485717202,
      "grad_norm": 5.46875,
      "learning_rate": 8.7638299087857e-06,
      "loss": 0.78241935,
      "memory(GiB)": 135.77,
      "step": 21590,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.7683857,
      "epoch": 0.5039443561440091,
      "grad_norm": 6.59375,
      "learning_rate": 8.762586060240642e-06,
      "loss": 0.83219185,
      "memory(GiB)": 135.77,
      "step": 21600,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.78917575,
      "epoch": 0.504177663716298,
      "grad_norm": 10.875,
      "learning_rate": 8.761341674587518e-06,
      "loss": 0.76239634,
      "memory(GiB)": 135.77,
      "step": 21610,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.78974929,
      "epoch": 0.5044109712885869,
      "grad_norm": 4.96875,
      "learning_rate": 8.760096752003962e-06,
      "loss": 0.77535329,
      "memory(GiB)": 135.77,
      "step": 21620,
      "train_speed(iter/s)": 0.201378
    },
    {
      "acc": 0.77866402,
      "epoch": 0.5046442788608758,
      "grad_norm": 4.9375,
      "learning_rate": 8.758851292667687e-06,
      "loss": 0.78650284,
      "memory(GiB)": 135.77,
      "step": 21630,
      "train_speed(iter/s)": 0.201426
    },
    {
      "acc": 0.77446556,
      "epoch": 0.5048775864331647,
      "grad_norm": 6.6875,
      "learning_rate": 8.757605296756483e-06,
      "loss": 0.80399456,
      "memory(GiB)": 135.77,
      "step": 21640,
      "train_speed(iter/s)": 0.201474
    },
    {
      "acc": 0.76340103,
      "epoch": 0.5051108940054536,
      "grad_norm": 5.0,
      "learning_rate": 8.756358764448214e-06,
      "loss": 0.85637417,
      "memory(GiB)": 135.77,
      "step": 21650,
      "train_speed(iter/s)": 0.201523
    },
    {
      "acc": 0.78347025,
      "epoch": 0.5053442015777424,
      "grad_norm": 4.875,
      "learning_rate": 8.755111695920823e-06,
      "loss": 0.79445,
      "memory(GiB)": 135.77,
      "step": 21660,
      "train_speed(iter/s)": 0.201571
    },
    {
      "acc": 0.77764778,
      "epoch": 0.5055775091500313,
      "grad_norm": 4.125,
      "learning_rate": 8.753864091352326e-06,
      "loss": 0.80275316,
      "memory(GiB)": 135.77,
      "step": 21670,
      "train_speed(iter/s)": 0.201619
    },
    {
      "acc": 0.78915906,
      "epoch": 0.5058108167223202,
      "grad_norm": 3.84375,
      "learning_rate": 8.752615950920824e-06,
      "loss": 0.75332527,
      "memory(GiB)": 135.77,
      "step": 21680,
      "train_speed(iter/s)": 0.201664
    },
    {
      "acc": 0.77775726,
      "epoch": 0.5060441242946091,
      "grad_norm": 10.625,
      "learning_rate": 8.751367274804483e-06,
      "loss": 0.79482059,
      "memory(GiB)": 135.77,
      "step": 21690,
      "train_speed(iter/s)": 0.201709
    },
    {
      "acc": 0.77287965,
      "epoch": 0.506277431866898,
      "grad_norm": 4.53125,
      "learning_rate": 8.750118063181553e-06,
      "loss": 0.8031868,
      "memory(GiB)": 135.77,
      "step": 21700,
      "train_speed(iter/s)": 0.201762
    },
    {
      "acc": 0.77395134,
      "epoch": 0.5065107394391869,
      "grad_norm": 4.8125,
      "learning_rate": 8.74886831623036e-06,
      "loss": 0.79606371,
      "memory(GiB)": 135.77,
      "step": 21710,
      "train_speed(iter/s)": 0.20181
    },
    {
      "acc": 0.75434523,
      "epoch": 0.5067440470114758,
      "grad_norm": 5.375,
      "learning_rate": 8.747618034129304e-06,
      "loss": 0.90953197,
      "memory(GiB)": 135.77,
      "step": 21720,
      "train_speed(iter/s)": 0.20186
    },
    {
      "acc": 0.78446827,
      "epoch": 0.5069773545837647,
      "grad_norm": 6.5625,
      "learning_rate": 8.746367217056861e-06,
      "loss": 0.79035029,
      "memory(GiB)": 135.77,
      "step": 21730,
      "train_speed(iter/s)": 0.201906
    },
    {
      "acc": 0.76340156,
      "epoch": 0.5072106621560536,
      "grad_norm": 6.40625,
      "learning_rate": 8.745115865191587e-06,
      "loss": 0.8527956,
      "memory(GiB)": 135.77,
      "step": 21740,
      "train_speed(iter/s)": 0.201955
    },
    {
      "acc": 0.77388973,
      "epoch": 0.5074439697283425,
      "grad_norm": 5.46875,
      "learning_rate": 8.743863978712111e-06,
      "loss": 0.81416397,
      "memory(GiB)": 135.77,
      "step": 21750,
      "train_speed(iter/s)": 0.202002
    },
    {
      "acc": 0.77853451,
      "epoch": 0.5076772773006314,
      "grad_norm": 5.0,
      "learning_rate": 8.74261155779714e-06,
      "loss": 0.78336382,
      "memory(GiB)": 135.77,
      "step": 21760,
      "train_speed(iter/s)": 0.202047
    },
    {
      "acc": 0.75856724,
      "epoch": 0.5079105848729203,
      "grad_norm": 7.21875,
      "learning_rate": 8.741358602625455e-06,
      "loss": 0.87376547,
      "memory(GiB)": 135.77,
      "step": 21770,
      "train_speed(iter/s)": 0.202089
    },
    {
      "acc": 0.78273201,
      "epoch": 0.5081438924452092,
      "grad_norm": 5.65625,
      "learning_rate": 8.740105113375919e-06,
      "loss": 0.76739621,
      "memory(GiB)": 135.77,
      "step": 21780,
      "train_speed(iter/s)": 0.202136
    },
    {
      "acc": 0.77240195,
      "epoch": 0.5083772000174981,
      "grad_norm": 4.21875,
      "learning_rate": 8.738851090227462e-06,
      "loss": 0.80845757,
      "memory(GiB)": 135.77,
      "step": 21790,
      "train_speed(iter/s)": 0.202185
    },
    {
      "acc": 0.78104792,
      "epoch": 0.508610507589787,
      "grad_norm": 5.625,
      "learning_rate": 8.737596533359101e-06,
      "loss": 0.78938627,
      "memory(GiB)": 135.77,
      "step": 21800,
      "train_speed(iter/s)": 0.202235
    },
    {
      "acc": 0.76249051,
      "epoch": 0.5088438151620759,
      "grad_norm": 5.96875,
      "learning_rate": 8.736341442949919e-06,
      "loss": 0.8694706,
      "memory(GiB)": 135.77,
      "step": 21810,
      "train_speed(iter/s)": 0.202282
    },
    {
      "acc": 0.77197785,
      "epoch": 0.5090771227343648,
      "grad_norm": 6.6875,
      "learning_rate": 8.73508581917908e-06,
      "loss": 0.80897655,
      "memory(GiB)": 135.77,
      "step": 21820,
      "train_speed(iter/s)": 0.202328
    },
    {
      "acc": 0.75979204,
      "epoch": 0.5093104303066537,
      "grad_norm": 7.5625,
      "learning_rate": 8.733829662225825e-06,
      "loss": 0.86723404,
      "memory(GiB)": 135.77,
      "step": 21830,
      "train_speed(iter/s)": 0.202377
    },
    {
      "acc": 0.77098808,
      "epoch": 0.5095437378789426,
      "grad_norm": 7.1875,
      "learning_rate": 8.732572972269472e-06,
      "loss": 0.84290762,
      "memory(GiB)": 135.77,
      "step": 21840,
      "train_speed(iter/s)": 0.202422
    },
    {
      "acc": 0.78516512,
      "epoch": 0.5097770454512315,
      "grad_norm": 5.5,
      "learning_rate": 8.731315749489412e-06,
      "loss": 0.77529469,
      "memory(GiB)": 135.77,
      "step": 21850,
      "train_speed(iter/s)": 0.202471
    },
    {
      "acc": 0.77012315,
      "epoch": 0.5100103530235203,
      "grad_norm": 5.625,
      "learning_rate": 8.730057994065113e-06,
      "loss": 0.81857462,
      "memory(GiB)": 135.77,
      "step": 21860,
      "train_speed(iter/s)": 0.202521
    },
    {
      "acc": 0.76024265,
      "epoch": 0.5102436605958092,
      "grad_norm": 7.84375,
      "learning_rate": 8.728799706176117e-06,
      "loss": 0.89396629,
      "memory(GiB)": 135.77,
      "step": 21870,
      "train_speed(iter/s)": 0.20257
    },
    {
      "acc": 0.79580889,
      "epoch": 0.5104769681680981,
      "grad_norm": 4.96875,
      "learning_rate": 8.727540886002048e-06,
      "loss": 0.71652207,
      "memory(GiB)": 135.77,
      "step": 21880,
      "train_speed(iter/s)": 0.20262
    },
    {
      "acc": 0.7760314,
      "epoch": 0.510710275740387,
      "grad_norm": 4.96875,
      "learning_rate": 8.7262815337226e-06,
      "loss": 0.81335468,
      "memory(GiB)": 135.77,
      "step": 21890,
      "train_speed(iter/s)": 0.202666
    },
    {
      "acc": 0.7670917,
      "epoch": 0.5109435833126759,
      "grad_norm": 5.21875,
      "learning_rate": 8.725021649517545e-06,
      "loss": 0.82741623,
      "memory(GiB)": 135.77,
      "step": 21900,
      "train_speed(iter/s)": 0.202713
    },
    {
      "acc": 0.76540327,
      "epoch": 0.5111768908849648,
      "grad_norm": 7.03125,
      "learning_rate": 8.723761233566732e-06,
      "loss": 0.85327883,
      "memory(GiB)": 135.77,
      "step": 21910,
      "train_speed(iter/s)": 0.202761
    },
    {
      "acc": 0.76294174,
      "epoch": 0.5114101984572537,
      "grad_norm": 7.59375,
      "learning_rate": 8.722500286050084e-06,
      "loss": 0.8585103,
      "memory(GiB)": 135.77,
      "step": 21920,
      "train_speed(iter/s)": 0.202807
    },
    {
      "acc": 0.77434635,
      "epoch": 0.5116435060295426,
      "grad_norm": 6.40625,
      "learning_rate": 8.721238807147602e-06,
      "loss": 0.83060961,
      "memory(GiB)": 135.77,
      "step": 21930,
      "train_speed(iter/s)": 0.202851
    },
    {
      "acc": 0.76418648,
      "epoch": 0.5118768136018315,
      "grad_norm": 10.4375,
      "learning_rate": 8.71997679703936e-06,
      "loss": 0.84364891,
      "memory(GiB)": 135.77,
      "step": 21940,
      "train_speed(iter/s)": 0.202899
    },
    {
      "acc": 0.77589755,
      "epoch": 0.5121101211741204,
      "grad_norm": 6.53125,
      "learning_rate": 8.718714255905514e-06,
      "loss": 0.80891504,
      "memory(GiB)": 135.77,
      "step": 21950,
      "train_speed(iter/s)": 0.202942
    },
    {
      "acc": 0.78276296,
      "epoch": 0.5123434287464093,
      "grad_norm": 4.96875,
      "learning_rate": 8.717451183926286e-06,
      "loss": 0.78558817,
      "memory(GiB)": 135.77,
      "step": 21960,
      "train_speed(iter/s)": 0.20299
    },
    {
      "acc": 0.76813164,
      "epoch": 0.5125767363186982,
      "grad_norm": 5.0625,
      "learning_rate": 8.716187581281982e-06,
      "loss": 0.82764816,
      "memory(GiB)": 135.77,
      "step": 21970,
      "train_speed(iter/s)": 0.203035
    },
    {
      "acc": 0.76752729,
      "epoch": 0.512810043890987,
      "grad_norm": 10.5625,
      "learning_rate": 8.71492344815298e-06,
      "loss": 0.84671097,
      "memory(GiB)": 135.77,
      "step": 21980,
      "train_speed(iter/s)": 0.203084
    },
    {
      "acc": 0.7599195,
      "epoch": 0.513043351463276,
      "grad_norm": 5.5625,
      "learning_rate": 8.713658784719735e-06,
      "loss": 0.85972528,
      "memory(GiB)": 135.77,
      "step": 21990,
      "train_speed(iter/s)": 0.203127
    },
    {
      "acc": 0.77313404,
      "epoch": 0.5132766590355649,
      "grad_norm": 5.90625,
      "learning_rate": 8.712393591162779e-06,
      "loss": 0.82830524,
      "memory(GiB)": 135.77,
      "step": 22000,
      "train_speed(iter/s)": 0.203176
    },
    {
      "epoch": 0.5132766590355649,
      "eval_acc": 0.739051699522658,
      "eval_loss": 0.8252963423728943,
      "eval_runtime": 1269.798,
      "eval_samples_per_second": 28.344,
      "eval_steps_per_second": 14.172,
      "step": 22000
    },
    {
      "acc": 0.76468811,
      "epoch": 0.5135099666078538,
      "grad_norm": 6.0,
      "learning_rate": 8.711127867662715e-06,
      "loss": 0.85957546,
      "memory(GiB)": 135.77,
      "step": 22010,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.77129335,
      "epoch": 0.5137432741801427,
      "grad_norm": 4.25,
      "learning_rate": 8.709861614400223e-06,
      "loss": 0.83853226,
      "memory(GiB)": 135.77,
      "step": 22020,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78872757,
      "epoch": 0.5139765817524315,
      "grad_norm": 4.9375,
      "learning_rate": 8.708594831556068e-06,
      "loss": 0.76557589,
      "memory(GiB)": 135.77,
      "step": 22030,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.76171837,
      "epoch": 0.5142098893247204,
      "grad_norm": 5.75,
      "learning_rate": 8.707327519311075e-06,
      "loss": 0.8551321,
      "memory(GiB)": 135.77,
      "step": 22040,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.76743393,
      "epoch": 0.5144431968970092,
      "grad_norm": 8.3125,
      "learning_rate": 8.706059677846157e-06,
      "loss": 0.85651779,
      "memory(GiB)": 135.77,
      "step": 22050,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.77318702,
      "epoch": 0.5146765044692981,
      "grad_norm": 4.8125,
      "learning_rate": 8.704791307342297e-06,
      "loss": 0.79379597,
      "memory(GiB)": 135.77,
      "step": 22060,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.7719686,
      "epoch": 0.514909812041587,
      "grad_norm": 5.8125,
      "learning_rate": 8.703522407980554e-06,
      "loss": 0.81712446,
      "memory(GiB)": 135.77,
      "step": 22070,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.77638397,
      "epoch": 0.5151431196138759,
      "grad_norm": 10.875,
      "learning_rate": 8.702252979942063e-06,
      "loss": 0.80130062,
      "memory(GiB)": 135.77,
      "step": 22080,
      "train_speed(iter/s)": 0.201174
    },
    {
      "acc": 0.77209868,
      "epoch": 0.5153764271861648,
      "grad_norm": 5.21875,
      "learning_rate": 8.700983023408034e-06,
      "loss": 0.8194252,
      "memory(GiB)": 135.77,
      "step": 22090,
      "train_speed(iter/s)": 0.201224
    },
    {
      "acc": 0.76339512,
      "epoch": 0.5156097347584537,
      "grad_norm": 6.71875,
      "learning_rate": 8.699712538559752e-06,
      "loss": 0.85726337,
      "memory(GiB)": 135.77,
      "step": 22100,
      "train_speed(iter/s)": 0.201271
    },
    {
      "acc": 0.7645709,
      "epoch": 0.5158430423307426,
      "grad_norm": 4.9375,
      "learning_rate": 8.698441525578582e-06,
      "loss": 0.85991373,
      "memory(GiB)": 135.77,
      "step": 22110,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.79457126,
      "epoch": 0.5160763499030315,
      "grad_norm": 5.125,
      "learning_rate": 8.697169984645959e-06,
      "loss": 0.72133851,
      "memory(GiB)": 135.77,
      "step": 22120,
      "train_speed(iter/s)": 0.201368
    },
    {
      "acc": 0.75647445,
      "epoch": 0.5163096574753204,
      "grad_norm": 4.0625,
      "learning_rate": 8.695897915943395e-06,
      "loss": 0.86364326,
      "memory(GiB)": 135.77,
      "step": 22130,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.76443992,
      "epoch": 0.5165429650476093,
      "grad_norm": 5.15625,
      "learning_rate": 8.694625319652477e-06,
      "loss": 0.85923405,
      "memory(GiB)": 135.77,
      "step": 22140,
      "train_speed(iter/s)": 0.201461
    },
    {
      "acc": 0.79320621,
      "epoch": 0.5167762726198982,
      "grad_norm": 5.46875,
      "learning_rate": 8.693352195954866e-06,
      "loss": 0.74455805,
      "memory(GiB)": 135.77,
      "step": 22150,
      "train_speed(iter/s)": 0.201504
    },
    {
      "acc": 0.75500603,
      "epoch": 0.5170095801921871,
      "grad_norm": 4.90625,
      "learning_rate": 8.692078545032304e-06,
      "loss": 0.89729595,
      "memory(GiB)": 135.77,
      "step": 22160,
      "train_speed(iter/s)": 0.201548
    },
    {
      "acc": 0.78628874,
      "epoch": 0.517242887764476,
      "grad_norm": 6.65625,
      "learning_rate": 8.6908043670666e-06,
      "loss": 0.77559862,
      "memory(GiB)": 135.77,
      "step": 22170,
      "train_speed(iter/s)": 0.201596
    },
    {
      "acc": 0.7568099,
      "epoch": 0.5174761953367649,
      "grad_norm": 6.21875,
      "learning_rate": 8.689529662239647e-06,
      "loss": 0.91406498,
      "memory(GiB)": 135.77,
      "step": 22180,
      "train_speed(iter/s)": 0.20164
    },
    {
      "acc": 0.76510105,
      "epoch": 0.5177095029090538,
      "grad_norm": 4.40625,
      "learning_rate": 8.688254430733405e-06,
      "loss": 0.85134945,
      "memory(GiB)": 135.77,
      "step": 22190,
      "train_speed(iter/s)": 0.201688
    },
    {
      "acc": 0.77507906,
      "epoch": 0.5179428104813427,
      "grad_norm": 5.5,
      "learning_rate": 8.686978672729916e-06,
      "loss": 0.8101469,
      "memory(GiB)": 135.77,
      "step": 22200,
      "train_speed(iter/s)": 0.201735
    },
    {
      "acc": 0.75599632,
      "epoch": 0.5181761180536316,
      "grad_norm": 4.78125,
      "learning_rate": 8.68570238841129e-06,
      "loss": 0.86438799,
      "memory(GiB)": 135.77,
      "step": 22210,
      "train_speed(iter/s)": 0.201783
    },
    {
      "acc": 0.76995659,
      "epoch": 0.5184094256259205,
      "grad_norm": 6.125,
      "learning_rate": 8.684425577959722e-06,
      "loss": 0.82093,
      "memory(GiB)": 135.77,
      "step": 22220,
      "train_speed(iter/s)": 0.201832
    },
    {
      "acc": 0.77510152,
      "epoch": 0.5186427331982094,
      "grad_norm": 4.3125,
      "learning_rate": 8.683148241557472e-06,
      "loss": 0.80338688,
      "memory(GiB)": 135.77,
      "step": 22230,
      "train_speed(iter/s)": 0.201878
    },
    {
      "acc": 0.76440411,
      "epoch": 0.5188760407704982,
      "grad_norm": 6.125,
      "learning_rate": 8.681870379386879e-06,
      "loss": 0.83043871,
      "memory(GiB)": 135.77,
      "step": 22240,
      "train_speed(iter/s)": 0.201929
    },
    {
      "acc": 0.76180897,
      "epoch": 0.5191093483427871,
      "grad_norm": 5.78125,
      "learning_rate": 8.68059199163036e-06,
      "loss": 0.86176853,
      "memory(GiB)": 135.77,
      "step": 22250,
      "train_speed(iter/s)": 0.201964
    },
    {
      "acc": 0.77140579,
      "epoch": 0.519342655915076,
      "grad_norm": 6.4375,
      "learning_rate": 8.679313078470403e-06,
      "loss": 0.80798788,
      "memory(GiB)": 135.77,
      "step": 22260,
      "train_speed(iter/s)": 0.202008
    },
    {
      "acc": 0.76855969,
      "epoch": 0.5195759634873649,
      "grad_norm": 4.875,
      "learning_rate": 8.678033640089574e-06,
      "loss": 0.83115768,
      "memory(GiB)": 135.77,
      "step": 22270,
      "train_speed(iter/s)": 0.202055
    },
    {
      "acc": 0.78249931,
      "epoch": 0.5198092710596538,
      "grad_norm": 4.65625,
      "learning_rate": 8.676753676670511e-06,
      "loss": 0.78415298,
      "memory(GiB)": 135.77,
      "step": 22280,
      "train_speed(iter/s)": 0.2021
    },
    {
      "acc": 0.78204756,
      "epoch": 0.5200425786319427,
      "grad_norm": 4.15625,
      "learning_rate": 8.67547318839593e-06,
      "loss": 0.78528633,
      "memory(GiB)": 135.77,
      "step": 22290,
      "train_speed(iter/s)": 0.202146
    },
    {
      "acc": 0.76204596,
      "epoch": 0.5202758862042316,
      "grad_norm": 6.75,
      "learning_rate": 8.674192175448617e-06,
      "loss": 0.8319747,
      "memory(GiB)": 135.77,
      "step": 22300,
      "train_speed(iter/s)": 0.202193
    },
    {
      "acc": 0.76219587,
      "epoch": 0.5205091937765205,
      "grad_norm": 6.53125,
      "learning_rate": 8.672910638011439e-06,
      "loss": 0.84944124,
      "memory(GiB)": 135.77,
      "step": 22310,
      "train_speed(iter/s)": 0.202241
    },
    {
      "acc": 0.76580586,
      "epoch": 0.5207425013488094,
      "grad_norm": 5.25,
      "learning_rate": 8.671628576267333e-06,
      "loss": 0.83932171,
      "memory(GiB)": 135.77,
      "step": 22320,
      "train_speed(iter/s)": 0.20229
    },
    {
      "acc": 0.76935239,
      "epoch": 0.5209758089210983,
      "grad_norm": 5.34375,
      "learning_rate": 8.670345990399317e-06,
      "loss": 0.82982864,
      "memory(GiB)": 135.77,
      "step": 22330,
      "train_speed(iter/s)": 0.202333
    },
    {
      "acc": 0.79416113,
      "epoch": 0.5212091164933872,
      "grad_norm": 7.59375,
      "learning_rate": 8.669062880590474e-06,
      "loss": 0.73947349,
      "memory(GiB)": 135.77,
      "step": 22340,
      "train_speed(iter/s)": 0.202381
    },
    {
      "acc": 0.77279081,
      "epoch": 0.5214424240656761,
      "grad_norm": 4.71875,
      "learning_rate": 8.667779247023974e-06,
      "loss": 0.81007175,
      "memory(GiB)": 135.77,
      "step": 22350,
      "train_speed(iter/s)": 0.202429
    },
    {
      "acc": 0.7789588,
      "epoch": 0.521675731637965,
      "grad_norm": 5.03125,
      "learning_rate": 8.666495089883049e-06,
      "loss": 0.76399951,
      "memory(GiB)": 135.77,
      "step": 22360,
      "train_speed(iter/s)": 0.202475
    },
    {
      "acc": 0.78549161,
      "epoch": 0.5219090392102539,
      "grad_norm": 4.65625,
      "learning_rate": 8.665210409351015e-06,
      "loss": 0.77918396,
      "memory(GiB)": 135.77,
      "step": 22370,
      "train_speed(iter/s)": 0.202519
    },
    {
      "acc": 0.76314993,
      "epoch": 0.5221423467825428,
      "grad_norm": 6.125,
      "learning_rate": 8.663925205611261e-06,
      "loss": 0.87036915,
      "memory(GiB)": 135.77,
      "step": 22380,
      "train_speed(iter/s)": 0.202566
    },
    {
      "acc": 0.79042592,
      "epoch": 0.5223756543548317,
      "grad_norm": 4.875,
      "learning_rate": 8.66263947884725e-06,
      "loss": 0.77454262,
      "memory(GiB)": 135.77,
      "step": 22390,
      "train_speed(iter/s)": 0.202605
    },
    {
      "acc": 0.76517191,
      "epoch": 0.5226089619271206,
      "grad_norm": 3.859375,
      "learning_rate": 8.661353229242514e-06,
      "loss": 0.84707232,
      "memory(GiB)": 135.77,
      "step": 22400,
      "train_speed(iter/s)": 0.202646
    },
    {
      "acc": 0.77440958,
      "epoch": 0.5228422694994095,
      "grad_norm": 5.46875,
      "learning_rate": 8.66006645698067e-06,
      "loss": 0.80511379,
      "memory(GiB)": 135.77,
      "step": 22410,
      "train_speed(iter/s)": 0.202689
    },
    {
      "acc": 0.77160006,
      "epoch": 0.5230755770716984,
      "grad_norm": 6.71875,
      "learning_rate": 8.658779162245404e-06,
      "loss": 0.82624683,
      "memory(GiB)": 135.77,
      "step": 22420,
      "train_speed(iter/s)": 0.202737
    },
    {
      "acc": 0.79068604,
      "epoch": 0.5233088846439872,
      "grad_norm": 6.1875,
      "learning_rate": 8.657491345220475e-06,
      "loss": 0.74397564,
      "memory(GiB)": 135.77,
      "step": 22430,
      "train_speed(iter/s)": 0.202783
    },
    {
      "acc": 0.79531755,
      "epoch": 0.5235421922162761,
      "grad_norm": 4.34375,
      "learning_rate": 8.656203006089716e-06,
      "loss": 0.72965755,
      "memory(GiB)": 135.77,
      "step": 22440,
      "train_speed(iter/s)": 0.202829
    },
    {
      "acc": 0.78984456,
      "epoch": 0.523775499788565,
      "grad_norm": 6.46875,
      "learning_rate": 8.654914145037044e-06,
      "loss": 0.74049387,
      "memory(GiB)": 135.77,
      "step": 22450,
      "train_speed(iter/s)": 0.202874
    },
    {
      "acc": 0.76976719,
      "epoch": 0.5240088073608539,
      "grad_norm": 4.875,
      "learning_rate": 8.653624762246437e-06,
      "loss": 0.82113333,
      "memory(GiB)": 135.77,
      "step": 22460,
      "train_speed(iter/s)": 0.202921
    },
    {
      "acc": 0.76820931,
      "epoch": 0.5242421149331428,
      "grad_norm": 4.90625,
      "learning_rate": 8.652334857901957e-06,
      "loss": 0.8498045,
      "memory(GiB)": 135.77,
      "step": 22470,
      "train_speed(iter/s)": 0.202959
    },
    {
      "acc": 0.75914822,
      "epoch": 0.5244754225054317,
      "grad_norm": 6.53125,
      "learning_rate": 8.651044432187736e-06,
      "loss": 0.895998,
      "memory(GiB)": 135.77,
      "step": 22480,
      "train_speed(iter/s)": 0.203006
    },
    {
      "acc": 0.77738414,
      "epoch": 0.5247087300777206,
      "grad_norm": 4.75,
      "learning_rate": 8.649753485287986e-06,
      "loss": 0.79222326,
      "memory(GiB)": 135.77,
      "step": 22490,
      "train_speed(iter/s)": 0.203052
    },
    {
      "acc": 0.78186607,
      "epoch": 0.5249420376500095,
      "grad_norm": 7.0625,
      "learning_rate": 8.648462017386982e-06,
      "loss": 0.80105238,
      "memory(GiB)": 135.77,
      "step": 22500,
      "train_speed(iter/s)": 0.203094
    },
    {
      "epoch": 0.5249420376500095,
      "eval_acc": 0.7395784139026973,
      "eval_loss": 0.8238633871078491,
      "eval_runtime": 1268.9883,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "step": 22500
    },
    {
      "acc": 0.78160195,
      "epoch": 0.5251753452222984,
      "grad_norm": 4.875,
      "learning_rate": 8.64717002866909e-06,
      "loss": 0.77151155,
      "memory(GiB)": 135.77,
      "step": 22510,
      "train_speed(iter/s)": 0.200805
    },
    {
      "acc": 0.78638101,
      "epoch": 0.5254086527945873,
      "grad_norm": 4.71875,
      "learning_rate": 8.64587751931873e-06,
      "loss": 0.77090092,
      "memory(GiB)": 135.77,
      "step": 22520,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.78011241,
      "epoch": 0.5256419603668762,
      "grad_norm": 5.03125,
      "learning_rate": 8.644584489520418e-06,
      "loss": 0.78862772,
      "memory(GiB)": 135.77,
      "step": 22530,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.77731066,
      "epoch": 0.5258752679391651,
      "grad_norm": 4.8125,
      "learning_rate": 8.643290939458728e-06,
      "loss": 0.79162712,
      "memory(GiB)": 135.77,
      "step": 22540,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.77135563,
      "epoch": 0.526108575511454,
      "grad_norm": 4.84375,
      "learning_rate": 8.641996869318313e-06,
      "loss": 0.83071518,
      "memory(GiB)": 135.77,
      "step": 22550,
      "train_speed(iter/s)": 0.200987
    },
    {
      "acc": 0.78698025,
      "epoch": 0.5263418830837429,
      "grad_norm": 8.125,
      "learning_rate": 8.640702279283904e-06,
      "loss": 0.76378479,
      "memory(GiB)": 135.77,
      "step": 22560,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.75245285,
      "epoch": 0.5265751906560318,
      "grad_norm": 6.875,
      "learning_rate": 8.639407169540302e-06,
      "loss": 0.90592232,
      "memory(GiB)": 135.77,
      "step": 22570,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.76512051,
      "epoch": 0.5268084982283207,
      "grad_norm": 6.40625,
      "learning_rate": 8.638111540272384e-06,
      "loss": 0.84115753,
      "memory(GiB)": 135.77,
      "step": 22580,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.76233463,
      "epoch": 0.5270418058006096,
      "grad_norm": 5.09375,
      "learning_rate": 8.636815391665102e-06,
      "loss": 0.87490129,
      "memory(GiB)": 135.77,
      "step": 22590,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.77505226,
      "epoch": 0.5272751133728985,
      "grad_norm": 5.5625,
      "learning_rate": 8.635518723903478e-06,
      "loss": 0.82525368,
      "memory(GiB)": 135.77,
      "step": 22600,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.77524152,
      "epoch": 0.5275084209451874,
      "grad_norm": 5.375,
      "learning_rate": 8.634221537172612e-06,
      "loss": 0.81201534,
      "memory(GiB)": 135.77,
      "step": 22610,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.76617479,
      "epoch": 0.5277417285174762,
      "grad_norm": 6.625,
      "learning_rate": 8.632923831657678e-06,
      "loss": 0.85680828,
      "memory(GiB)": 135.77,
      "step": 22620,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.77629232,
      "epoch": 0.527975036089765,
      "grad_norm": 5.4375,
      "learning_rate": 8.631625607543921e-06,
      "loss": 0.7853322,
      "memory(GiB)": 135.77,
      "step": 22630,
      "train_speed(iter/s)": 0.201353
    },
    {
      "acc": 0.78849845,
      "epoch": 0.5282083436620539,
      "grad_norm": 4.1875,
      "learning_rate": 8.630326865016663e-06,
      "loss": 0.74947386,
      "memory(GiB)": 135.77,
      "step": 22640,
      "train_speed(iter/s)": 0.201397
    },
    {
      "acc": 0.76544561,
      "epoch": 0.5284416512343428,
      "grad_norm": 5.0625,
      "learning_rate": 8.629027604261303e-06,
      "loss": 0.84232635,
      "memory(GiB)": 135.77,
      "step": 22650,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.77632599,
      "epoch": 0.5286749588066317,
      "grad_norm": 6.21875,
      "learning_rate": 8.627727825463303e-06,
      "loss": 0.79433575,
      "memory(GiB)": 135.77,
      "step": 22660,
      "train_speed(iter/s)": 0.201489
    },
    {
      "acc": 0.78881607,
      "epoch": 0.5289082663789206,
      "grad_norm": 6.40625,
      "learning_rate": 8.626427528808212e-06,
      "loss": 0.75886679,
      "memory(GiB)": 135.77,
      "step": 22670,
      "train_speed(iter/s)": 0.201535
    },
    {
      "acc": 0.78701801,
      "epoch": 0.5291415739512095,
      "grad_norm": 14.4375,
      "learning_rate": 8.625126714481645e-06,
      "loss": 0.79751778,
      "memory(GiB)": 135.77,
      "step": 22680,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.76863942,
      "epoch": 0.5293748815234984,
      "grad_norm": 5.28125,
      "learning_rate": 8.623825382669291e-06,
      "loss": 0.82617855,
      "memory(GiB)": 135.77,
      "step": 22690,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.77472267,
      "epoch": 0.5296081890957873,
      "grad_norm": 6.40625,
      "learning_rate": 8.622523533556916e-06,
      "loss": 0.84323826,
      "memory(GiB)": 135.77,
      "step": 22700,
      "train_speed(iter/s)": 0.201672
    },
    {
      "acc": 0.76703949,
      "epoch": 0.5298414966680762,
      "grad_norm": 6.875,
      "learning_rate": 8.621221167330363e-06,
      "loss": 0.85303726,
      "memory(GiB)": 135.77,
      "step": 22710,
      "train_speed(iter/s)": 0.201718
    },
    {
      "acc": 0.76639805,
      "epoch": 0.5300748042403651,
      "grad_norm": 4.15625,
      "learning_rate": 8.619918284175537e-06,
      "loss": 0.82314072,
      "memory(GiB)": 135.77,
      "step": 22720,
      "train_speed(iter/s)": 0.201763
    },
    {
      "acc": 0.79025326,
      "epoch": 0.530308111812654,
      "grad_norm": 3.8125,
      "learning_rate": 8.618614884278427e-06,
      "loss": 0.73413811,
      "memory(GiB)": 135.77,
      "step": 22730,
      "train_speed(iter/s)": 0.201808
    },
    {
      "acc": 0.76510487,
      "epoch": 0.5305414193849429,
      "grad_norm": 4.625,
      "learning_rate": 8.617310967825094e-06,
      "loss": 0.82598991,
      "memory(GiB)": 135.77,
      "step": 22740,
      "train_speed(iter/s)": 0.201853
    },
    {
      "acc": 0.76874247,
      "epoch": 0.5307747269572318,
      "grad_norm": 4.9375,
      "learning_rate": 8.616006535001673e-06,
      "loss": 0.83377934,
      "memory(GiB)": 135.77,
      "step": 22750,
      "train_speed(iter/s)": 0.201897
    },
    {
      "acc": 0.76021204,
      "epoch": 0.5310080345295207,
      "grad_norm": 5.15625,
      "learning_rate": 8.614701585994368e-06,
      "loss": 0.87588558,
      "memory(GiB)": 135.77,
      "step": 22760,
      "train_speed(iter/s)": 0.201943
    },
    {
      "acc": 0.77022514,
      "epoch": 0.5312413421018096,
      "grad_norm": 5.1875,
      "learning_rate": 8.613396120989463e-06,
      "loss": 0.82545309,
      "memory(GiB)": 135.77,
      "step": 22770,
      "train_speed(iter/s)": 0.201989
    },
    {
      "acc": 0.7779541,
      "epoch": 0.5314746496740985,
      "grad_norm": 4.75,
      "learning_rate": 8.61209014017331e-06,
      "loss": 0.7887279,
      "memory(GiB)": 135.77,
      "step": 22780,
      "train_speed(iter/s)": 0.202035
    },
    {
      "acc": 0.77761946,
      "epoch": 0.5317079572463874,
      "grad_norm": 4.40625,
      "learning_rate": 8.610783643732339e-06,
      "loss": 0.78312531,
      "memory(GiB)": 135.77,
      "step": 22790,
      "train_speed(iter/s)": 0.202078
    },
    {
      "acc": 0.76519241,
      "epoch": 0.5319412648186763,
      "grad_norm": 9.9375,
      "learning_rate": 8.60947663185305e-06,
      "loss": 0.85003548,
      "memory(GiB)": 135.77,
      "step": 22800,
      "train_speed(iter/s)": 0.202127
    },
    {
      "acc": 0.770683,
      "epoch": 0.5321745723909652,
      "grad_norm": 4.90625,
      "learning_rate": 8.608169104722024e-06,
      "loss": 0.83203659,
      "memory(GiB)": 135.77,
      "step": 22810,
      "train_speed(iter/s)": 0.202169
    },
    {
      "acc": 0.75626473,
      "epoch": 0.532407879963254,
      "grad_norm": 6.0,
      "learning_rate": 8.606861062525904e-06,
      "loss": 0.88525591,
      "memory(GiB)": 135.77,
      "step": 22820,
      "train_speed(iter/s)": 0.20221
    },
    {
      "acc": 0.7737916,
      "epoch": 0.5326411875355429,
      "grad_norm": 4.59375,
      "learning_rate": 8.605552505451417e-06,
      "loss": 0.81642294,
      "memory(GiB)": 135.77,
      "step": 22830,
      "train_speed(iter/s)": 0.202255
    },
    {
      "acc": 0.77903752,
      "epoch": 0.5328744951078318,
      "grad_norm": 5.9375,
      "learning_rate": 8.604243433685356e-06,
      "loss": 0.8116827,
      "memory(GiB)": 135.77,
      "step": 22840,
      "train_speed(iter/s)": 0.202302
    },
    {
      "acc": 0.79416924,
      "epoch": 0.5331078026801207,
      "grad_norm": 6.96875,
      "learning_rate": 8.602933847414592e-06,
      "loss": 0.74621596,
      "memory(GiB)": 135.77,
      "step": 22850,
      "train_speed(iter/s)": 0.202346
    },
    {
      "acc": 0.77615061,
      "epoch": 0.5333411102524096,
      "grad_norm": 6.375,
      "learning_rate": 8.601623746826068e-06,
      "loss": 0.80405464,
      "memory(GiB)": 135.77,
      "step": 22860,
      "train_speed(iter/s)": 0.20239
    },
    {
      "acc": 0.75500526,
      "epoch": 0.5335744178246985,
      "grad_norm": 5.4375,
      "learning_rate": 8.600313132106801e-06,
      "loss": 0.92301064,
      "memory(GiB)": 135.77,
      "step": 22870,
      "train_speed(iter/s)": 0.202435
    },
    {
      "acc": 0.74417582,
      "epoch": 0.5338077253969874,
      "grad_norm": 14.125,
      "learning_rate": 8.599002003443879e-06,
      "loss": 0.93030882,
      "memory(GiB)": 135.77,
      "step": 22880,
      "train_speed(iter/s)": 0.202484
    },
    {
      "acc": 0.76454968,
      "epoch": 0.5340410329692763,
      "grad_norm": 5.21875,
      "learning_rate": 8.597690361024468e-06,
      "loss": 0.85199795,
      "memory(GiB)": 135.77,
      "step": 22890,
      "train_speed(iter/s)": 0.202531
    },
    {
      "acc": 0.77811437,
      "epoch": 0.5342743405415652,
      "grad_norm": 7.15625,
      "learning_rate": 8.596378205035803e-06,
      "loss": 0.80663748,
      "memory(GiB)": 135.77,
      "step": 22900,
      "train_speed(iter/s)": 0.202578
    },
    {
      "acc": 0.77453089,
      "epoch": 0.5345076481138541,
      "grad_norm": 5.21875,
      "learning_rate": 8.595065535665192e-06,
      "loss": 0.81512251,
      "memory(GiB)": 135.77,
      "step": 22910,
      "train_speed(iter/s)": 0.202625
    },
    {
      "acc": 0.77443113,
      "epoch": 0.534740955686143,
      "grad_norm": 5.53125,
      "learning_rate": 8.593752353100022e-06,
      "loss": 0.82128239,
      "memory(GiB)": 135.77,
      "step": 22920,
      "train_speed(iter/s)": 0.202669
    },
    {
      "acc": 0.78365164,
      "epoch": 0.5349742632584319,
      "grad_norm": 5.5,
      "learning_rate": 8.592438657527746e-06,
      "loss": 0.77799377,
      "memory(GiB)": 135.77,
      "step": 22930,
      "train_speed(iter/s)": 0.202714
    },
    {
      "acc": 0.76975231,
      "epoch": 0.5352075708307208,
      "grad_norm": 6.40625,
      "learning_rate": 8.591124449135897e-06,
      "loss": 0.81138725,
      "memory(GiB)": 135.77,
      "step": 22940,
      "train_speed(iter/s)": 0.202762
    },
    {
      "acc": 0.79545565,
      "epoch": 0.5354408784030097,
      "grad_norm": 13.875,
      "learning_rate": 8.589809728112076e-06,
      "loss": 0.72523451,
      "memory(GiB)": 135.77,
      "step": 22950,
      "train_speed(iter/s)": 0.202809
    },
    {
      "acc": 0.76595306,
      "epoch": 0.5356741859752986,
      "grad_norm": 11.1875,
      "learning_rate": 8.588494494643959e-06,
      "loss": 0.84729843,
      "memory(GiB)": 135.77,
      "step": 22960,
      "train_speed(iter/s)": 0.202854
    },
    {
      "acc": 0.77811699,
      "epoch": 0.5359074935475875,
      "grad_norm": 5.125,
      "learning_rate": 8.587178748919294e-06,
      "loss": 0.8056942,
      "memory(GiB)": 135.77,
      "step": 22970,
      "train_speed(iter/s)": 0.202899
    },
    {
      "acc": 0.77778416,
      "epoch": 0.5361408011198764,
      "grad_norm": 4.875,
      "learning_rate": 8.585862491125906e-06,
      "loss": 0.79754281,
      "memory(GiB)": 135.77,
      "step": 22980,
      "train_speed(iter/s)": 0.202945
    },
    {
      "acc": 0.77270217,
      "epoch": 0.5363741086921653,
      "grad_norm": 4.84375,
      "learning_rate": 8.584545721451689e-06,
      "loss": 0.81794977,
      "memory(GiB)": 135.77,
      "step": 22990,
      "train_speed(iter/s)": 0.202991
    },
    {
      "acc": 0.75620012,
      "epoch": 0.5366074162644542,
      "grad_norm": 5.03125,
      "learning_rate": 8.583228440084612e-06,
      "loss": 0.89408569,
      "memory(GiB)": 135.77,
      "step": 23000,
      "train_speed(iter/s)": 0.203038
    },
    {
      "epoch": 0.5366074162644542,
      "eval_acc": 0.7395305307772392,
      "eval_loss": 0.8233683705329895,
      "eval_runtime": 1269.3275,
      "eval_samples_per_second": 28.354,
      "eval_steps_per_second": 14.178,
      "step": 23000
    },
    {
      "acc": 0.76377153,
      "epoch": 0.536840723836743,
      "grad_norm": 22.625,
      "learning_rate": 8.581910647212714e-06,
      "loss": 0.84084702,
      "memory(GiB)": 135.77,
      "step": 23010,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.77578378,
      "epoch": 0.5370740314090319,
      "grad_norm": 6.3125,
      "learning_rate": 8.580592343024114e-06,
      "loss": 0.81817398,
      "memory(GiB)": 135.77,
      "step": 23020,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.78007755,
      "epoch": 0.5373073389813208,
      "grad_norm": 4.375,
      "learning_rate": 8.579273527706997e-06,
      "loss": 0.79391298,
      "memory(GiB)": 135.77,
      "step": 23030,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.7673666,
      "epoch": 0.5375406465536097,
      "grad_norm": 5.0625,
      "learning_rate": 8.577954201449621e-06,
      "loss": 0.8341753,
      "memory(GiB)": 135.77,
      "step": 23040,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.75479908,
      "epoch": 0.5377739541258986,
      "grad_norm": 6.9375,
      "learning_rate": 8.576634364440327e-06,
      "loss": 0.87442303,
      "memory(GiB)": 135.77,
      "step": 23050,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.77252378,
      "epoch": 0.5380072616981875,
      "grad_norm": 6.65625,
      "learning_rate": 8.575314016867512e-06,
      "loss": 0.84094563,
      "memory(GiB)": 135.77,
      "step": 23060,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.76748762,
      "epoch": 0.5382405692704764,
      "grad_norm": 7.25,
      "learning_rate": 8.573993158919661e-06,
      "loss": 0.82697706,
      "memory(GiB)": 135.77,
      "step": 23070,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.77766724,
      "epoch": 0.5384738768427653,
      "grad_norm": 3.890625,
      "learning_rate": 8.572671790785325e-06,
      "loss": 0.78724775,
      "memory(GiB)": 135.77,
      "step": 23080,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.76558957,
      "epoch": 0.5387071844150542,
      "grad_norm": 4.8125,
      "learning_rate": 8.57134991265313e-06,
      "loss": 0.844098,
      "memory(GiB)": 135.77,
      "step": 23090,
      "train_speed(iter/s)": 0.201148
    },
    {
      "acc": 0.78299785,
      "epoch": 0.5389404919873431,
      "grad_norm": 5.90625,
      "learning_rate": 8.57002752471177e-06,
      "loss": 0.77799244,
      "memory(GiB)": 135.77,
      "step": 23100,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.78844042,
      "epoch": 0.539173799559632,
      "grad_norm": 4.78125,
      "learning_rate": 8.56870462715002e-06,
      "loss": 0.73825145,
      "memory(GiB)": 135.77,
      "step": 23110,
      "train_speed(iter/s)": 0.201237
    },
    {
      "acc": 0.78577042,
      "epoch": 0.5394071071319209,
      "grad_norm": 8.75,
      "learning_rate": 8.567381220156721e-06,
      "loss": 0.76423135,
      "memory(GiB)": 135.77,
      "step": 23120,
      "train_speed(iter/s)": 0.201279
    },
    {
      "acc": 0.76853571,
      "epoch": 0.5396404147042098,
      "grad_norm": 5.59375,
      "learning_rate": 8.566057303920788e-06,
      "loss": 0.82382536,
      "memory(GiB)": 135.77,
      "step": 23130,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.79181309,
      "epoch": 0.5398737222764987,
      "grad_norm": 3.703125,
      "learning_rate": 8.564732878631212e-06,
      "loss": 0.76008949,
      "memory(GiB)": 135.77,
      "step": 23140,
      "train_speed(iter/s)": 0.20137
    },
    {
      "acc": 0.77274303,
      "epoch": 0.5401070298487876,
      "grad_norm": 5.6875,
      "learning_rate": 8.563407944477052e-06,
      "loss": 0.82584839,
      "memory(GiB)": 135.77,
      "step": 23150,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.76535187,
      "epoch": 0.5403403374210765,
      "grad_norm": 5.8125,
      "learning_rate": 8.562082501647445e-06,
      "loss": 0.82021761,
      "memory(GiB)": 135.77,
      "step": 23160,
      "train_speed(iter/s)": 0.201455
    },
    {
      "acc": 0.76622362,
      "epoch": 0.5405736449933654,
      "grad_norm": 6.09375,
      "learning_rate": 8.560756550331594e-06,
      "loss": 0.82352085,
      "memory(GiB)": 135.77,
      "step": 23170,
      "train_speed(iter/s)": 0.201499
    },
    {
      "acc": 0.78566408,
      "epoch": 0.5408069525656543,
      "grad_norm": 5.34375,
      "learning_rate": 8.55943009071878e-06,
      "loss": 0.78319578,
      "memory(GiB)": 135.77,
      "step": 23180,
      "train_speed(iter/s)": 0.201539
    },
    {
      "acc": 0.79545965,
      "epoch": 0.5410402601379432,
      "grad_norm": 5.90625,
      "learning_rate": 8.558103122998354e-06,
      "loss": 0.74384508,
      "memory(GiB)": 135.77,
      "step": 23190,
      "train_speed(iter/s)": 0.201582
    },
    {
      "acc": 0.76623678,
      "epoch": 0.5412735677102319,
      "grad_norm": 6.3125,
      "learning_rate": 8.556775647359744e-06,
      "loss": 0.8515913,
      "memory(GiB)": 135.77,
      "step": 23200,
      "train_speed(iter/s)": 0.201629
    },
    {
      "acc": 0.76652737,
      "epoch": 0.5415068752825208,
      "grad_norm": 18.25,
      "learning_rate": 8.55544766399244e-06,
      "loss": 0.84666195,
      "memory(GiB)": 135.77,
      "step": 23210,
      "train_speed(iter/s)": 0.201675
    },
    {
      "acc": 0.77143021,
      "epoch": 0.5417401828548097,
      "grad_norm": 4.84375,
      "learning_rate": 8.554119173086014e-06,
      "loss": 0.82694244,
      "memory(GiB)": 135.77,
      "step": 23220,
      "train_speed(iter/s)": 0.201722
    },
    {
      "acc": 0.78573523,
      "epoch": 0.5419734904270986,
      "grad_norm": 4.4375,
      "learning_rate": 8.552790174830112e-06,
      "loss": 0.76147246,
      "memory(GiB)": 135.77,
      "step": 23230,
      "train_speed(iter/s)": 0.201769
    },
    {
      "acc": 0.77535591,
      "epoch": 0.5422067979993875,
      "grad_norm": 5.53125,
      "learning_rate": 8.551460669414444e-06,
      "loss": 0.80129585,
      "memory(GiB)": 135.77,
      "step": 23240,
      "train_speed(iter/s)": 0.201817
    },
    {
      "acc": 0.77153683,
      "epoch": 0.5424401055716764,
      "grad_norm": 5.84375,
      "learning_rate": 8.550130657028797e-06,
      "loss": 0.80181551,
      "memory(GiB)": 135.77,
      "step": 23250,
      "train_speed(iter/s)": 0.201864
    },
    {
      "acc": 0.78025579,
      "epoch": 0.5426734131439653,
      "grad_norm": 4.53125,
      "learning_rate": 8.548800137863028e-06,
      "loss": 0.78459616,
      "memory(GiB)": 135.77,
      "step": 23260,
      "train_speed(iter/s)": 0.20191
    },
    {
      "acc": 0.7832684,
      "epoch": 0.5429067207162542,
      "grad_norm": 5.25,
      "learning_rate": 8.547469112107071e-06,
      "loss": 0.79565368,
      "memory(GiB)": 135.77,
      "step": 23270,
      "train_speed(iter/s)": 0.201959
    },
    {
      "acc": 0.76917963,
      "epoch": 0.5431400282885431,
      "grad_norm": 5.96875,
      "learning_rate": 8.54613757995093e-06,
      "loss": 0.83835936,
      "memory(GiB)": 135.77,
      "step": 23280,
      "train_speed(iter/s)": 0.202004
    },
    {
      "acc": 0.76447086,
      "epoch": 0.543373335860832,
      "grad_norm": 7.375,
      "learning_rate": 8.54480554158468e-06,
      "loss": 0.83568411,
      "memory(GiB)": 135.77,
      "step": 23290,
      "train_speed(iter/s)": 0.202048
    },
    {
      "acc": 0.757164,
      "epoch": 0.5436066434331209,
      "grad_norm": 5.34375,
      "learning_rate": 8.543472997198467e-06,
      "loss": 0.87914925,
      "memory(GiB)": 135.77,
      "step": 23300,
      "train_speed(iter/s)": 0.202092
    },
    {
      "acc": 0.76810994,
      "epoch": 0.5438399510054098,
      "grad_norm": 5.96875,
      "learning_rate": 8.542139946982516e-06,
      "loss": 0.81503553,
      "memory(GiB)": 135.77,
      "step": 23310,
      "train_speed(iter/s)": 0.202138
    },
    {
      "acc": 0.7738739,
      "epoch": 0.5440732585776987,
      "grad_norm": 7.0625,
      "learning_rate": 8.540806391127112e-06,
      "loss": 0.82891731,
      "memory(GiB)": 135.77,
      "step": 23320,
      "train_speed(iter/s)": 0.202183
    },
    {
      "acc": 0.76221933,
      "epoch": 0.5443065661499876,
      "grad_norm": 6.1875,
      "learning_rate": 8.539472329822627e-06,
      "loss": 0.86402712,
      "memory(GiB)": 135.77,
      "step": 23330,
      "train_speed(iter/s)": 0.20223
    },
    {
      "acc": 0.79112253,
      "epoch": 0.5445398737222765,
      "grad_norm": 6.09375,
      "learning_rate": 8.538137763259495e-06,
      "loss": 0.75483222,
      "memory(GiB)": 135.77,
      "step": 23340,
      "train_speed(iter/s)": 0.202277
    },
    {
      "acc": 0.78928189,
      "epoch": 0.5447731812945654,
      "grad_norm": 5.4375,
      "learning_rate": 8.536802691628226e-06,
      "loss": 0.77416606,
      "memory(GiB)": 135.77,
      "step": 23350,
      "train_speed(iter/s)": 0.202318
    },
    {
      "acc": 0.77431383,
      "epoch": 0.5450064888668543,
      "grad_norm": 7.90625,
      "learning_rate": 8.535467115119399e-06,
      "loss": 0.79211545,
      "memory(GiB)": 135.77,
      "step": 23360,
      "train_speed(iter/s)": 0.202362
    },
    {
      "acc": 0.78563662,
      "epoch": 0.5452397964391432,
      "grad_norm": 4.40625,
      "learning_rate": 8.534131033923668e-06,
      "loss": 0.76018214,
      "memory(GiB)": 135.77,
      "step": 23370,
      "train_speed(iter/s)": 0.202406
    },
    {
      "acc": 0.7482007,
      "epoch": 0.5454731040114321,
      "grad_norm": 5.15625,
      "learning_rate": 8.53279444823176e-06,
      "loss": 0.90958967,
      "memory(GiB)": 135.77,
      "step": 23380,
      "train_speed(iter/s)": 0.202452
    },
    {
      "acc": 0.77748098,
      "epoch": 0.545706411583721,
      "grad_norm": 4.75,
      "learning_rate": 8.531457358234469e-06,
      "loss": 0.81523991,
      "memory(GiB)": 135.77,
      "step": 23390,
      "train_speed(iter/s)": 0.202498
    },
    {
      "acc": 0.78067293,
      "epoch": 0.5459397191560098,
      "grad_norm": 4.65625,
      "learning_rate": 8.530119764122666e-06,
      "loss": 0.80296059,
      "memory(GiB)": 135.77,
      "step": 23400,
      "train_speed(iter/s)": 0.202546
    },
    {
      "acc": 0.77213707,
      "epoch": 0.5461730267282987,
      "grad_norm": 6.3125,
      "learning_rate": 8.528781666087294e-06,
      "loss": 0.809834,
      "memory(GiB)": 135.77,
      "step": 23410,
      "train_speed(iter/s)": 0.202589
    },
    {
      "acc": 0.76574154,
      "epoch": 0.5464063343005876,
      "grad_norm": 5.75,
      "learning_rate": 8.527443064319362e-06,
      "loss": 0.82819872,
      "memory(GiB)": 135.77,
      "step": 23420,
      "train_speed(iter/s)": 0.202637
    },
    {
      "acc": 0.78312383,
      "epoch": 0.5466396418728765,
      "grad_norm": 5.4375,
      "learning_rate": 8.526103959009959e-06,
      "loss": 0.75177727,
      "memory(GiB)": 135.77,
      "step": 23430,
      "train_speed(iter/s)": 0.202685
    },
    {
      "acc": 0.77955656,
      "epoch": 0.5468729494451654,
      "grad_norm": 4.34375,
      "learning_rate": 8.52476435035024e-06,
      "loss": 0.78664408,
      "memory(GiB)": 135.77,
      "step": 23440,
      "train_speed(iter/s)": 0.202731
    },
    {
      "acc": 0.76553507,
      "epoch": 0.5471062570174543,
      "grad_norm": 5.8125,
      "learning_rate": 8.523424238531435e-06,
      "loss": 0.84986553,
      "memory(GiB)": 135.77,
      "step": 23450,
      "train_speed(iter/s)": 0.202773
    },
    {
      "acc": 0.77595615,
      "epoch": 0.5473395645897432,
      "grad_norm": 5.1875,
      "learning_rate": 8.522083623744841e-06,
      "loss": 0.80186443,
      "memory(GiB)": 135.77,
      "step": 23460,
      "train_speed(iter/s)": 0.202818
    },
    {
      "acc": 0.76666412,
      "epoch": 0.5475728721620321,
      "grad_norm": 7.65625,
      "learning_rate": 8.520742506181834e-06,
      "loss": 0.85667477,
      "memory(GiB)": 135.77,
      "step": 23470,
      "train_speed(iter/s)": 0.202861
    },
    {
      "acc": 0.78935471,
      "epoch": 0.547806179734321,
      "grad_norm": 5.625,
      "learning_rate": 8.519400886033858e-06,
      "loss": 0.77769156,
      "memory(GiB)": 135.77,
      "step": 23480,
      "train_speed(iter/s)": 0.202904
    },
    {
      "acc": 0.78004494,
      "epoch": 0.5480394873066099,
      "grad_norm": 5.46875,
      "learning_rate": 8.518058763492428e-06,
      "loss": 0.78160458,
      "memory(GiB)": 135.77,
      "step": 23490,
      "train_speed(iter/s)": 0.202951
    },
    {
      "acc": 0.7742156,
      "epoch": 0.5482727948788988,
      "grad_norm": 5.71875,
      "learning_rate": 8.516716138749131e-06,
      "loss": 0.82459059,
      "memory(GiB)": 135.77,
      "step": 23500,
      "train_speed(iter/s)": 0.202994
    },
    {
      "epoch": 0.5482727948788988,
      "eval_acc": 0.7398569047226694,
      "eval_loss": 0.8228998184204102,
      "eval_runtime": 1269.4084,
      "eval_samples_per_second": 28.353,
      "eval_steps_per_second": 14.177,
      "step": 23500
    },
    {
      "acc": 0.76532078,
      "epoch": 0.5485061024511877,
      "grad_norm": 6.0,
      "learning_rate": 8.515373011995624e-06,
      "loss": 0.84984226,
      "memory(GiB)": 135.77,
      "step": 23510,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.78477125,
      "epoch": 0.5487394100234766,
      "grad_norm": 6.78125,
      "learning_rate": 8.514029383423644e-06,
      "loss": 0.75276413,
      "memory(GiB)": 135.77,
      "step": 23520,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.79451346,
      "epoch": 0.5489727175957655,
      "grad_norm": 6.03125,
      "learning_rate": 8.51268525322499e-06,
      "loss": 0.74686522,
      "memory(GiB)": 135.77,
      "step": 23530,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.76671562,
      "epoch": 0.5492060251680544,
      "grad_norm": 6.0625,
      "learning_rate": 8.511340621591536e-06,
      "loss": 0.84974279,
      "memory(GiB)": 135.77,
      "step": 23540,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.77346573,
      "epoch": 0.5494393327403433,
      "grad_norm": 5.78125,
      "learning_rate": 8.509995488715228e-06,
      "loss": 0.83321466,
      "memory(GiB)": 135.77,
      "step": 23550,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.7776125,
      "epoch": 0.5496726403126322,
      "grad_norm": 3.34375,
      "learning_rate": 8.508649854788085e-06,
      "loss": 0.80355291,
      "memory(GiB)": 135.77,
      "step": 23560,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.76173635,
      "epoch": 0.5499059478849211,
      "grad_norm": 7.90625,
      "learning_rate": 8.507303720002194e-06,
      "loss": 0.87492666,
      "memory(GiB)": 135.77,
      "step": 23570,
      "train_speed(iter/s)": 0.201065
    },
    {
      "acc": 0.77865009,
      "epoch": 0.55013925545721,
      "grad_norm": 5.0,
      "learning_rate": 8.505957084549714e-06,
      "loss": 0.82421484,
      "memory(GiB)": 135.77,
      "step": 23580,
      "train_speed(iter/s)": 0.201109
    },
    {
      "acc": 0.77421579,
      "epoch": 0.5503725630294988,
      "grad_norm": 5.1875,
      "learning_rate": 8.50460994862288e-06,
      "loss": 0.80633135,
      "memory(GiB)": 135.77,
      "step": 23590,
      "train_speed(iter/s)": 0.201151
    },
    {
      "acc": 0.77418938,
      "epoch": 0.5506058706017877,
      "grad_norm": 4.65625,
      "learning_rate": 8.503262312413994e-06,
      "loss": 0.82369556,
      "memory(GiB)": 135.77,
      "step": 23600,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.775385,
      "epoch": 0.5508391781740766,
      "grad_norm": 4.96875,
      "learning_rate": 8.501914176115432e-06,
      "loss": 0.81491432,
      "memory(GiB)": 135.77,
      "step": 23610,
      "train_speed(iter/s)": 0.201234
    },
    {
      "acc": 0.76755171,
      "epoch": 0.5510724857463655,
      "grad_norm": 7.46875,
      "learning_rate": 8.500565539919636e-06,
      "loss": 0.84395332,
      "memory(GiB)": 135.77,
      "step": 23620,
      "train_speed(iter/s)": 0.201278
    },
    {
      "acc": 0.77598143,
      "epoch": 0.5513057933186544,
      "grad_norm": 6.75,
      "learning_rate": 8.499216404019129e-06,
      "loss": 0.80111885,
      "memory(GiB)": 135.77,
      "step": 23630,
      "train_speed(iter/s)": 0.201323
    },
    {
      "acc": 0.7583106,
      "epoch": 0.5515391008909433,
      "grad_norm": 7.28125,
      "learning_rate": 8.497866768606493e-06,
      "loss": 0.86915989,
      "memory(GiB)": 135.77,
      "step": 23640,
      "train_speed(iter/s)": 0.201369
    },
    {
      "acc": 0.77197981,
      "epoch": 0.5517724084632322,
      "grad_norm": 5.4375,
      "learning_rate": 8.496516633874395e-06,
      "loss": 0.81613064,
      "memory(GiB)": 135.77,
      "step": 23650,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.7657783,
      "epoch": 0.5520057160355211,
      "grad_norm": 5.625,
      "learning_rate": 8.495166000015562e-06,
      "loss": 0.84468031,
      "memory(GiB)": 135.77,
      "step": 23660,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.77991076,
      "epoch": 0.55223902360781,
      "grad_norm": 6.15625,
      "learning_rate": 8.493814867222799e-06,
      "loss": 0.80958157,
      "memory(GiB)": 135.77,
      "step": 23670,
      "train_speed(iter/s)": 0.201502
    },
    {
      "acc": 0.77264977,
      "epoch": 0.5524723311800989,
      "grad_norm": 5.5625,
      "learning_rate": 8.492463235688977e-06,
      "loss": 0.79891062,
      "memory(GiB)": 135.77,
      "step": 23680,
      "train_speed(iter/s)": 0.201547
    },
    {
      "acc": 0.7838623,
      "epoch": 0.5527056387523878,
      "grad_norm": 5.65625,
      "learning_rate": 8.491111105607044e-06,
      "loss": 0.77505665,
      "memory(GiB)": 135.77,
      "step": 23690,
      "train_speed(iter/s)": 0.20159
    },
    {
      "acc": 0.77503614,
      "epoch": 0.5529389463246767,
      "grad_norm": 4.59375,
      "learning_rate": 8.489758477170015e-06,
      "loss": 0.80072899,
      "memory(GiB)": 135.77,
      "step": 23700,
      "train_speed(iter/s)": 0.201631
    },
    {
      "acc": 0.76735268,
      "epoch": 0.5531722538969656,
      "grad_norm": 6.40625,
      "learning_rate": 8.488405350570976e-06,
      "loss": 0.84666109,
      "memory(GiB)": 135.77,
      "step": 23710,
      "train_speed(iter/s)": 0.201675
    },
    {
      "acc": 0.76926999,
      "epoch": 0.5534055614692545,
      "grad_norm": 7.34375,
      "learning_rate": 8.487051726003087e-06,
      "loss": 0.81005325,
      "memory(GiB)": 135.77,
      "step": 23720,
      "train_speed(iter/s)": 0.201719
    },
    {
      "acc": 0.7819787,
      "epoch": 0.5536388690415434,
      "grad_norm": 6.34375,
      "learning_rate": 8.485697603659578e-06,
      "loss": 0.77697315,
      "memory(GiB)": 135.77,
      "step": 23730,
      "train_speed(iter/s)": 0.201765
    },
    {
      "acc": 0.76455793,
      "epoch": 0.5538721766138323,
      "grad_norm": 6.1875,
      "learning_rate": 8.484342983733747e-06,
      "loss": 0.8478878,
      "memory(GiB)": 135.77,
      "step": 23740,
      "train_speed(iter/s)": 0.201811
    },
    {
      "acc": 0.77278852,
      "epoch": 0.5541054841861212,
      "grad_norm": 5.375,
      "learning_rate": 8.482987866418968e-06,
      "loss": 0.82200966,
      "memory(GiB)": 135.77,
      "step": 23750,
      "train_speed(iter/s)": 0.201851
    },
    {
      "acc": 0.79056635,
      "epoch": 0.55433879175841,
      "grad_norm": 7.78125,
      "learning_rate": 8.481632251908684e-06,
      "loss": 0.75850916,
      "memory(GiB)": 135.77,
      "step": 23760,
      "train_speed(iter/s)": 0.201893
    },
    {
      "acc": 0.78154702,
      "epoch": 0.554572099330699,
      "grad_norm": 7.5,
      "learning_rate": 8.480276140396406e-06,
      "loss": 0.78442221,
      "memory(GiB)": 135.77,
      "step": 23770,
      "train_speed(iter/s)": 0.201934
    },
    {
      "acc": 0.77974281,
      "epoch": 0.5548054069029877,
      "grad_norm": 6.03125,
      "learning_rate": 8.478919532075723e-06,
      "loss": 0.80161047,
      "memory(GiB)": 135.77,
      "step": 23780,
      "train_speed(iter/s)": 0.201977
    },
    {
      "acc": 0.77860785,
      "epoch": 0.5550387144752766,
      "grad_norm": 5.53125,
      "learning_rate": 8.477562427140283e-06,
      "loss": 0.80894966,
      "memory(GiB)": 135.77,
      "step": 23790,
      "train_speed(iter/s)": 0.20202
    },
    {
      "acc": 0.77603893,
      "epoch": 0.5552720220475655,
      "grad_norm": 4.3125,
      "learning_rate": 8.47620482578382e-06,
      "loss": 0.81759472,
      "memory(GiB)": 135.77,
      "step": 23800,
      "train_speed(iter/s)": 0.202062
    },
    {
      "acc": 0.77046776,
      "epoch": 0.5555053296198544,
      "grad_norm": 4.90625,
      "learning_rate": 8.474846728200125e-06,
      "loss": 0.82772932,
      "memory(GiB)": 135.77,
      "step": 23810,
      "train_speed(iter/s)": 0.202107
    },
    {
      "acc": 0.77512856,
      "epoch": 0.5557386371921433,
      "grad_norm": 3.6875,
      "learning_rate": 8.473488134583071e-06,
      "loss": 0.80098133,
      "memory(GiB)": 135.77,
      "step": 23820,
      "train_speed(iter/s)": 0.20215
    },
    {
      "acc": 0.7914567,
      "epoch": 0.5559719447644322,
      "grad_norm": 7.59375,
      "learning_rate": 8.472129045126596e-06,
      "loss": 0.74370127,
      "memory(GiB)": 135.77,
      "step": 23830,
      "train_speed(iter/s)": 0.202193
    },
    {
      "acc": 0.78856716,
      "epoch": 0.5562052523367211,
      "grad_norm": 6.0,
      "learning_rate": 8.470769460024705e-06,
      "loss": 0.75198135,
      "memory(GiB)": 135.77,
      "step": 23840,
      "train_speed(iter/s)": 0.202238
    },
    {
      "acc": 0.76632452,
      "epoch": 0.55643855990901,
      "grad_norm": 5.9375,
      "learning_rate": 8.469409379471486e-06,
      "loss": 0.86098537,
      "memory(GiB)": 135.77,
      "step": 23850,
      "train_speed(iter/s)": 0.202281
    },
    {
      "acc": 0.77517405,
      "epoch": 0.5566718674812989,
      "grad_norm": 4.625,
      "learning_rate": 8.468048803661083e-06,
      "loss": 0.79920835,
      "memory(GiB)": 135.77,
      "step": 23860,
      "train_speed(iter/s)": 0.202325
    },
    {
      "acc": 0.78587723,
      "epoch": 0.5569051750535878,
      "grad_norm": 7.875,
      "learning_rate": 8.466687732787721e-06,
      "loss": 0.77098179,
      "memory(GiB)": 135.77,
      "step": 23870,
      "train_speed(iter/s)": 0.202368
    },
    {
      "acc": 0.76620317,
      "epoch": 0.5571384826258767,
      "grad_norm": 10.5625,
      "learning_rate": 8.465326167045693e-06,
      "loss": 0.86184235,
      "memory(GiB)": 135.77,
      "step": 23880,
      "train_speed(iter/s)": 0.202409
    },
    {
      "acc": 0.77027483,
      "epoch": 0.5573717901981656,
      "grad_norm": 5.875,
      "learning_rate": 8.463964106629361e-06,
      "loss": 0.81875496,
      "memory(GiB)": 135.77,
      "step": 23890,
      "train_speed(iter/s)": 0.20245
    },
    {
      "acc": 0.78449135,
      "epoch": 0.5576050977704545,
      "grad_norm": 4.625,
      "learning_rate": 8.46260155173316e-06,
      "loss": 0.77044706,
      "memory(GiB)": 135.77,
      "step": 23900,
      "train_speed(iter/s)": 0.202491
    },
    {
      "acc": 0.77500076,
      "epoch": 0.5578384053427434,
      "grad_norm": 4.90625,
      "learning_rate": 8.461238502551592e-06,
      "loss": 0.81618023,
      "memory(GiB)": 135.77,
      "step": 23910,
      "train_speed(iter/s)": 0.202537
    },
    {
      "acc": 0.7681963,
      "epoch": 0.5580717129150323,
      "grad_norm": 6.3125,
      "learning_rate": 8.459874959279235e-06,
      "loss": 0.85666189,
      "memory(GiB)": 135.77,
      "step": 23920,
      "train_speed(iter/s)": 0.202581
    },
    {
      "acc": 0.75990124,
      "epoch": 0.5583050204873212,
      "grad_norm": 4.40625,
      "learning_rate": 8.45851092211073e-06,
      "loss": 0.86498947,
      "memory(GiB)": 135.77,
      "step": 23930,
      "train_speed(iter/s)": 0.202626
    },
    {
      "acc": 0.77422037,
      "epoch": 0.5585383280596101,
      "grad_norm": 4.59375,
      "learning_rate": 8.457146391240798e-06,
      "loss": 0.81769314,
      "memory(GiB)": 135.77,
      "step": 23940,
      "train_speed(iter/s)": 0.202668
    },
    {
      "acc": 0.78469434,
      "epoch": 0.558771635631899,
      "grad_norm": 5.125,
      "learning_rate": 8.455781366864223e-06,
      "loss": 0.77262511,
      "memory(GiB)": 135.77,
      "step": 23950,
      "train_speed(iter/s)": 0.202713
    },
    {
      "acc": 0.77914438,
      "epoch": 0.5590049432041879,
      "grad_norm": 6.5,
      "learning_rate": 8.45441584917586e-06,
      "loss": 0.7662066,
      "memory(GiB)": 135.77,
      "step": 23960,
      "train_speed(iter/s)": 0.202752
    },
    {
      "acc": 0.78139782,
      "epoch": 0.5592382507764767,
      "grad_norm": 4.90625,
      "learning_rate": 8.453049838370639e-06,
      "loss": 0.78332119,
      "memory(GiB)": 135.77,
      "step": 23970,
      "train_speed(iter/s)": 0.202796
    },
    {
      "acc": 0.75784979,
      "epoch": 0.5594715583487656,
      "grad_norm": 7.40625,
      "learning_rate": 8.451683334643557e-06,
      "loss": 0.88706684,
      "memory(GiB)": 135.77,
      "step": 23980,
      "train_speed(iter/s)": 0.20284
    },
    {
      "acc": 0.79134464,
      "epoch": 0.5597048659210545,
      "grad_norm": 8.8125,
      "learning_rate": 8.45031633818968e-06,
      "loss": 0.73851681,
      "memory(GiB)": 135.77,
      "step": 23990,
      "train_speed(iter/s)": 0.202884
    },
    {
      "acc": 0.76034174,
      "epoch": 0.5599381734933434,
      "grad_norm": 5.25,
      "learning_rate": 8.44894884920415e-06,
      "loss": 0.87289333,
      "memory(GiB)": 135.77,
      "step": 24000,
      "train_speed(iter/s)": 0.20293
    },
    {
      "epoch": 0.5599381734933434,
      "eval_acc": 0.7398836488094638,
      "eval_loss": 0.822571873664856,
      "eval_runtime": 1269.3247,
      "eval_samples_per_second": 28.354,
      "eval_steps_per_second": 14.178,
      "step": 24000
    },
    {
      "acc": 0.76281738,
      "epoch": 0.5601714810656323,
      "grad_norm": 19.375,
      "learning_rate": 8.447580867882172e-06,
      "loss": 0.85142126,
      "memory(GiB)": 135.77,
      "step": 24010,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.79376831,
      "epoch": 0.5604047886379212,
      "grad_norm": 4.65625,
      "learning_rate": 8.446212394419028e-06,
      "loss": 0.74882054,
      "memory(GiB)": 135.77,
      "step": 24020,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.78502007,
      "epoch": 0.5606380962102101,
      "grad_norm": 4.5625,
      "learning_rate": 8.444843429010065e-06,
      "loss": 0.7658041,
      "memory(GiB)": 135.77,
      "step": 24030,
      "train_speed(iter/s)": 0.200876
    },
    {
      "acc": 0.79705825,
      "epoch": 0.560871403782499,
      "grad_norm": 4.75,
      "learning_rate": 8.443473971850703e-06,
      "loss": 0.72210217,
      "memory(GiB)": 135.77,
      "step": 24040,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.76908588,
      "epoch": 0.5611047113547879,
      "grad_norm": 7.78125,
      "learning_rate": 8.442104023136435e-06,
      "loss": 0.83583965,
      "memory(GiB)": 135.77,
      "step": 24050,
      "train_speed(iter/s)": 0.20096
    },
    {
      "acc": 0.78580241,
      "epoch": 0.5613380189270768,
      "grad_norm": 6.5625,
      "learning_rate": 8.440733583062814e-06,
      "loss": 0.76494713,
      "memory(GiB)": 135.77,
      "step": 24060,
      "train_speed(iter/s)": 0.201002
    },
    {
      "acc": 0.75843534,
      "epoch": 0.5615713264993657,
      "grad_norm": 6.8125,
      "learning_rate": 8.439362651825475e-06,
      "loss": 0.8755847,
      "memory(GiB)": 135.77,
      "step": 24070,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.77627125,
      "epoch": 0.5618046340716546,
      "grad_norm": 9.9375,
      "learning_rate": 8.437991229620117e-06,
      "loss": 0.80955582,
      "memory(GiB)": 135.77,
      "step": 24080,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.77059631,
      "epoch": 0.5620379416439435,
      "grad_norm": 4.46875,
      "learning_rate": 8.436619316642508e-06,
      "loss": 0.82761564,
      "memory(GiB)": 135.77,
      "step": 24090,
      "train_speed(iter/s)": 0.20113
    },
    {
      "acc": 0.77916279,
      "epoch": 0.5622712492162324,
      "grad_norm": 4.625,
      "learning_rate": 8.435246913088492e-06,
      "loss": 0.80605087,
      "memory(GiB)": 135.77,
      "step": 24100,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.78393345,
      "epoch": 0.5625045567885213,
      "grad_norm": 5.9375,
      "learning_rate": 8.433874019153976e-06,
      "loss": 0.78385267,
      "memory(GiB)": 135.77,
      "step": 24110,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.78179646,
      "epoch": 0.5627378643608102,
      "grad_norm": 5.28125,
      "learning_rate": 8.432500635034942e-06,
      "loss": 0.77540627,
      "memory(GiB)": 135.77,
      "step": 24120,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.76452451,
      "epoch": 0.5629711719330991,
      "grad_norm": 6.65625,
      "learning_rate": 8.43112676092744e-06,
      "loss": 0.88243942,
      "memory(GiB)": 135.77,
      "step": 24130,
      "train_speed(iter/s)": 0.201301
    },
    {
      "acc": 0.77689223,
      "epoch": 0.563204479505388,
      "grad_norm": 5.375,
      "learning_rate": 8.429752397027585e-06,
      "loss": 0.79288564,
      "memory(GiB)": 135.77,
      "step": 24140,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.77824707,
      "epoch": 0.5634377870776769,
      "grad_norm": 7.28125,
      "learning_rate": 8.428377543531577e-06,
      "loss": 0.79980693,
      "memory(GiB)": 135.77,
      "step": 24150,
      "train_speed(iter/s)": 0.201385
    },
    {
      "acc": 0.77147899,
      "epoch": 0.5636710946499658,
      "grad_norm": 9.625,
      "learning_rate": 8.427002200635669e-06,
      "loss": 0.82787066,
      "memory(GiB)": 135.77,
      "step": 24160,
      "train_speed(iter/s)": 0.201428
    },
    {
      "acc": 0.75589886,
      "epoch": 0.5639044022222546,
      "grad_norm": 5.8125,
      "learning_rate": 8.425626368536192e-06,
      "loss": 0.88443241,
      "memory(GiB)": 135.77,
      "step": 24170,
      "train_speed(iter/s)": 0.201471
    },
    {
      "acc": 0.7736052,
      "epoch": 0.5641377097945435,
      "grad_norm": 6.0625,
      "learning_rate": 8.424250047429547e-06,
      "loss": 0.82636681,
      "memory(GiB)": 135.77,
      "step": 24180,
      "train_speed(iter/s)": 0.201514
    },
    {
      "acc": 0.76566911,
      "epoch": 0.5643710173668324,
      "grad_norm": 5.8125,
      "learning_rate": 8.4228732375122e-06,
      "loss": 0.85641251,
      "memory(GiB)": 135.77,
      "step": 24190,
      "train_speed(iter/s)": 0.201557
    },
    {
      "acc": 0.77742729,
      "epoch": 0.5646043249391213,
      "grad_norm": 5.375,
      "learning_rate": 8.421495938980695e-06,
      "loss": 0.78332605,
      "memory(GiB)": 135.77,
      "step": 24200,
      "train_speed(iter/s)": 0.201598
    },
    {
      "acc": 0.77955341,
      "epoch": 0.5648376325114102,
      "grad_norm": 4.40625,
      "learning_rate": 8.420118152031638e-06,
      "loss": 0.79840794,
      "memory(GiB)": 135.77,
      "step": 24210,
      "train_speed(iter/s)": 0.201641
    },
    {
      "acc": 0.79004874,
      "epoch": 0.5650709400836991,
      "grad_norm": 6.71875,
      "learning_rate": 8.418739876861708e-06,
      "loss": 0.75119066,
      "memory(GiB)": 135.77,
      "step": 24220,
      "train_speed(iter/s)": 0.201681
    },
    {
      "acc": 0.77495551,
      "epoch": 0.565304247655988,
      "grad_norm": 5.875,
      "learning_rate": 8.417361113667654e-06,
      "loss": 0.80294151,
      "memory(GiB)": 135.77,
      "step": 24230,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.77338634,
      "epoch": 0.5655375552282769,
      "grad_norm": 6.34375,
      "learning_rate": 8.415981862646295e-06,
      "loss": 0.80396366,
      "memory(GiB)": 135.77,
      "step": 24240,
      "train_speed(iter/s)": 0.201765
    },
    {
      "acc": 0.76142254,
      "epoch": 0.5657708628005658,
      "grad_norm": 5.53125,
      "learning_rate": 8.414602123994517e-06,
      "loss": 0.8642437,
      "memory(GiB)": 135.77,
      "step": 24250,
      "train_speed(iter/s)": 0.201808
    },
    {
      "acc": 0.77578001,
      "epoch": 0.5660041703728547,
      "grad_norm": 7.84375,
      "learning_rate": 8.413221897909277e-06,
      "loss": 0.80832329,
      "memory(GiB)": 135.77,
      "step": 24260,
      "train_speed(iter/s)": 0.201849
    },
    {
      "acc": 0.75002069,
      "epoch": 0.5662374779451436,
      "grad_norm": 27.875,
      "learning_rate": 8.411841184587602e-06,
      "loss": 0.94861403,
      "memory(GiB)": 135.77,
      "step": 24270,
      "train_speed(iter/s)": 0.201892
    },
    {
      "acc": 0.79898376,
      "epoch": 0.5664707855174325,
      "grad_norm": 6.53125,
      "learning_rate": 8.41045998422659e-06,
      "loss": 0.73172636,
      "memory(GiB)": 135.77,
      "step": 24280,
      "train_speed(iter/s)": 0.201931
    },
    {
      "acc": 0.79310303,
      "epoch": 0.5667040930897214,
      "grad_norm": 6.03125,
      "learning_rate": 8.409078297023406e-06,
      "loss": 0.71440797,
      "memory(GiB)": 135.77,
      "step": 24290,
      "train_speed(iter/s)": 0.201972
    },
    {
      "acc": 0.77959661,
      "epoch": 0.5669374006620103,
      "grad_norm": 4.5625,
      "learning_rate": 8.407696123175285e-06,
      "loss": 0.78759675,
      "memory(GiB)": 135.77,
      "step": 24300,
      "train_speed(iter/s)": 0.202013
    },
    {
      "acc": 0.79495506,
      "epoch": 0.5671707082342992,
      "grad_norm": 4.53125,
      "learning_rate": 8.406313462879533e-06,
      "loss": 0.73519535,
      "memory(GiB)": 135.77,
      "step": 24310,
      "train_speed(iter/s)": 0.202053
    },
    {
      "acc": 0.77956157,
      "epoch": 0.5674040158065881,
      "grad_norm": 4.875,
      "learning_rate": 8.404930316333524e-06,
      "loss": 0.80541306,
      "memory(GiB)": 135.77,
      "step": 24320,
      "train_speed(iter/s)": 0.202096
    },
    {
      "acc": 0.76412115,
      "epoch": 0.567637323378877,
      "grad_norm": 5.8125,
      "learning_rate": 8.4035466837347e-06,
      "loss": 0.86554079,
      "memory(GiB)": 135.77,
      "step": 24330,
      "train_speed(iter/s)": 0.202139
    },
    {
      "acc": 0.76175427,
      "epoch": 0.5678706309511659,
      "grad_norm": 4.46875,
      "learning_rate": 8.402162565280577e-06,
      "loss": 0.92497549,
      "memory(GiB)": 135.77,
      "step": 24340,
      "train_speed(iter/s)": 0.202181
    },
    {
      "acc": 0.77174816,
      "epoch": 0.5681039385234548,
      "grad_norm": 5.34375,
      "learning_rate": 8.400777961168736e-06,
      "loss": 0.83751183,
      "memory(GiB)": 135.77,
      "step": 24350,
      "train_speed(iter/s)": 0.202224
    },
    {
      "acc": 0.76675367,
      "epoch": 0.5683372460957435,
      "grad_norm": 4.5,
      "learning_rate": 8.399392871596828e-06,
      "loss": 0.82344303,
      "memory(GiB)": 135.77,
      "step": 24360,
      "train_speed(iter/s)": 0.20227
    },
    {
      "acc": 0.79096088,
      "epoch": 0.5685705536680324,
      "grad_norm": 6.0625,
      "learning_rate": 8.398007296762576e-06,
      "loss": 0.74808559,
      "memory(GiB)": 135.77,
      "step": 24370,
      "train_speed(iter/s)": 0.202311
    },
    {
      "acc": 0.78298712,
      "epoch": 0.5688038612403213,
      "grad_norm": 5.96875,
      "learning_rate": 8.39662123686377e-06,
      "loss": 0.7767374,
      "memory(GiB)": 135.77,
      "step": 24380,
      "train_speed(iter/s)": 0.202355
    },
    {
      "acc": 0.77403417,
      "epoch": 0.5690371688126102,
      "grad_norm": 5.65625,
      "learning_rate": 8.395234692098267e-06,
      "loss": 0.80425043,
      "memory(GiB)": 135.77,
      "step": 24390,
      "train_speed(iter/s)": 0.202399
    },
    {
      "acc": 0.76454625,
      "epoch": 0.5692704763848991,
      "grad_norm": 4.96875,
      "learning_rate": 8.393847662663998e-06,
      "loss": 0.83601761,
      "memory(GiB)": 135.77,
      "step": 24400,
      "train_speed(iter/s)": 0.20244
    },
    {
      "acc": 0.78817453,
      "epoch": 0.569503783957188,
      "grad_norm": 5.625,
      "learning_rate": 8.392460148758962e-06,
      "loss": 0.75539885,
      "memory(GiB)": 135.77,
      "step": 24410,
      "train_speed(iter/s)": 0.202484
    },
    {
      "acc": 0.77908316,
      "epoch": 0.5697370915294769,
      "grad_norm": 5.25,
      "learning_rate": 8.391072150581228e-06,
      "loss": 0.78923807,
      "memory(GiB)": 135.77,
      "step": 24420,
      "train_speed(iter/s)": 0.202523
    },
    {
      "acc": 0.77952948,
      "epoch": 0.5699703991017658,
      "grad_norm": 7.96875,
      "learning_rate": 8.389683668328927e-06,
      "loss": 0.79888625,
      "memory(GiB)": 135.77,
      "step": 24430,
      "train_speed(iter/s)": 0.202564
    },
    {
      "acc": 0.7668036,
      "epoch": 0.5702037066740547,
      "grad_norm": 5.21875,
      "learning_rate": 8.388294702200267e-06,
      "loss": 0.83973408,
      "memory(GiB)": 135.77,
      "step": 24440,
      "train_speed(iter/s)": 0.202607
    },
    {
      "acc": 0.78203883,
      "epoch": 0.5704370142463436,
      "grad_norm": 6.28125,
      "learning_rate": 8.386905252393522e-06,
      "loss": 0.7930686,
      "memory(GiB)": 135.77,
      "step": 24450,
      "train_speed(iter/s)": 0.20265
    },
    {
      "acc": 0.78412566,
      "epoch": 0.5706703218186325,
      "grad_norm": 10.0,
      "learning_rate": 8.385515319107038e-06,
      "loss": 0.77225199,
      "memory(GiB)": 135.77,
      "step": 24460,
      "train_speed(iter/s)": 0.202691
    },
    {
      "acc": 0.76364594,
      "epoch": 0.5709036293909214,
      "grad_norm": 6.0,
      "learning_rate": 8.384124902539225e-06,
      "loss": 0.83564758,
      "memory(GiB)": 135.77,
      "step": 24470,
      "train_speed(iter/s)": 0.202734
    },
    {
      "acc": 0.76822014,
      "epoch": 0.5711369369632103,
      "grad_norm": 4.6875,
      "learning_rate": 8.382734002888565e-06,
      "loss": 0.81515923,
      "memory(GiB)": 135.77,
      "step": 24480,
      "train_speed(iter/s)": 0.202778
    },
    {
      "acc": 0.78435955,
      "epoch": 0.5713702445354992,
      "grad_norm": 6.0625,
      "learning_rate": 8.381342620353609e-06,
      "loss": 0.7858181,
      "memory(GiB)": 135.77,
      "step": 24490,
      "train_speed(iter/s)": 0.202821
    },
    {
      "acc": 0.7721489,
      "epoch": 0.5716035521077881,
      "grad_norm": 6.1875,
      "learning_rate": 8.379950755132975e-06,
      "loss": 0.81357956,
      "memory(GiB)": 135.77,
      "step": 24500,
      "train_speed(iter/s)": 0.202865
    },
    {
      "epoch": 0.5716035521077881,
      "eval_acc": 0.739892296598008,
      "eval_loss": 0.8211784958839417,
      "eval_runtime": 1271.0226,
      "eval_samples_per_second": 28.317,
      "eval_steps_per_second": 14.159,
      "step": 24500
    },
    {
      "acc": 0.79300923,
      "epoch": 0.571836859680077,
      "grad_norm": 6.03125,
      "learning_rate": 8.378558407425355e-06,
      "loss": 0.75731974,
      "memory(GiB)": 135.77,
      "step": 24510,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.7801857,
      "epoch": 0.5720701672523659,
      "grad_norm": 5.375,
      "learning_rate": 8.377165577429502e-06,
      "loss": 0.79804745,
      "memory(GiB)": 135.77,
      "step": 24520,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.78895264,
      "epoch": 0.5723034748246548,
      "grad_norm": 6.4375,
      "learning_rate": 8.375772265344244e-06,
      "loss": 0.74042835,
      "memory(GiB)": 135.77,
      "step": 24530,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.78116617,
      "epoch": 0.5725367823969437,
      "grad_norm": 6.0625,
      "learning_rate": 8.374378471368476e-06,
      "loss": 0.78446503,
      "memory(GiB)": 135.77,
      "step": 24540,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.78846378,
      "epoch": 0.5727700899692325,
      "grad_norm": 7.96875,
      "learning_rate": 8.37298419570116e-06,
      "loss": 0.76507263,
      "memory(GiB)": 135.77,
      "step": 24550,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.75134902,
      "epoch": 0.5730033975415214,
      "grad_norm": 8.1875,
      "learning_rate": 8.371589438541333e-06,
      "loss": 0.91664543,
      "memory(GiB)": 135.77,
      "step": 24560,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.77958789,
      "epoch": 0.5732367051138103,
      "grad_norm": 5.6875,
      "learning_rate": 8.370194200088091e-06,
      "loss": 0.78155866,
      "memory(GiB)": 135.77,
      "step": 24570,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.78482103,
      "epoch": 0.5734700126860992,
      "grad_norm": 5.375,
      "learning_rate": 8.368798480540607e-06,
      "loss": 0.76826344,
      "memory(GiB)": 135.77,
      "step": 24580,
      "train_speed(iter/s)": 0.201055
    },
    {
      "acc": 0.7931962,
      "epoch": 0.5737033202583881,
      "grad_norm": 5.125,
      "learning_rate": 8.367402280098118e-06,
      "loss": 0.71536484,
      "memory(GiB)": 135.77,
      "step": 24590,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.77020044,
      "epoch": 0.573936627830677,
      "grad_norm": 6.3125,
      "learning_rate": 8.366005598959932e-06,
      "loss": 0.83158503,
      "memory(GiB)": 135.77,
      "step": 24600,
      "train_speed(iter/s)": 0.201134
    },
    {
      "acc": 0.79495125,
      "epoch": 0.5741699354029659,
      "grad_norm": 5.09375,
      "learning_rate": 8.364608437325426e-06,
      "loss": 0.74491563,
      "memory(GiB)": 135.77,
      "step": 24610,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.75548391,
      "epoch": 0.5744032429752548,
      "grad_norm": 5.09375,
      "learning_rate": 8.363210795394042e-06,
      "loss": 0.88298302,
      "memory(GiB)": 135.77,
      "step": 24620,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.7842412,
      "epoch": 0.5746365505475437,
      "grad_norm": 6.53125,
      "learning_rate": 8.361812673365292e-06,
      "loss": 0.76854534,
      "memory(GiB)": 135.77,
      "step": 24630,
      "train_speed(iter/s)": 0.201263
    },
    {
      "acc": 0.7659914,
      "epoch": 0.5748698581198326,
      "grad_norm": 5.125,
      "learning_rate": 8.360414071438761e-06,
      "loss": 0.84088774,
      "memory(GiB)": 135.77,
      "step": 24640,
      "train_speed(iter/s)": 0.201306
    },
    {
      "acc": 0.78436427,
      "epoch": 0.5751031656921215,
      "grad_norm": 5.0625,
      "learning_rate": 8.359014989814099e-06,
      "loss": 0.79782057,
      "memory(GiB)": 135.77,
      "step": 24650,
      "train_speed(iter/s)": 0.201352
    },
    {
      "acc": 0.76118832,
      "epoch": 0.5753364732644104,
      "grad_norm": 8.0625,
      "learning_rate": 8.35761542869102e-06,
      "loss": 0.86918697,
      "memory(GiB)": 135.77,
      "step": 24660,
      "train_speed(iter/s)": 0.201396
    },
    {
      "acc": 0.76398478,
      "epoch": 0.5755697808366993,
      "grad_norm": 6.6875,
      "learning_rate": 8.356215388269316e-06,
      "loss": 0.84599447,
      "memory(GiB)": 135.77,
      "step": 24670,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.78095145,
      "epoch": 0.5758030884089882,
      "grad_norm": 5.3125,
      "learning_rate": 8.354814868748839e-06,
      "loss": 0.77294083,
      "memory(GiB)": 135.77,
      "step": 24680,
      "train_speed(iter/s)": 0.201476
    },
    {
      "acc": 0.76520529,
      "epoch": 0.5760363959812771,
      "grad_norm": 7.0625,
      "learning_rate": 8.353413870329514e-06,
      "loss": 0.86367378,
      "memory(GiB)": 135.77,
      "step": 24690,
      "train_speed(iter/s)": 0.201518
    },
    {
      "acc": 0.77567444,
      "epoch": 0.576269703553566,
      "grad_norm": 4.25,
      "learning_rate": 8.352012393211336e-06,
      "loss": 0.79126658,
      "memory(GiB)": 135.77,
      "step": 24700,
      "train_speed(iter/s)": 0.20156
    },
    {
      "acc": 0.7710907,
      "epoch": 0.5765030111258549,
      "grad_norm": 6.28125,
      "learning_rate": 8.35061043759436e-06,
      "loss": 0.83803148,
      "memory(GiB)": 135.77,
      "step": 24710,
      "train_speed(iter/s)": 0.201599
    },
    {
      "acc": 0.76869202,
      "epoch": 0.5767363186981438,
      "grad_norm": 5.21875,
      "learning_rate": 8.349208003678716e-06,
      "loss": 0.83444061,
      "memory(GiB)": 135.77,
      "step": 24720,
      "train_speed(iter/s)": 0.201641
    },
    {
      "acc": 0.77576461,
      "epoch": 0.5769696262704327,
      "grad_norm": 5.375,
      "learning_rate": 8.347805091664606e-06,
      "loss": 0.82614613,
      "memory(GiB)": 135.77,
      "step": 24730,
      "train_speed(iter/s)": 0.201681
    },
    {
      "acc": 0.77615957,
      "epoch": 0.5772029338427216,
      "grad_norm": 5.46875,
      "learning_rate": 8.34640170175229e-06,
      "loss": 0.79450846,
      "memory(GiB)": 135.77,
      "step": 24740,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.77073669,
      "epoch": 0.5774362414150104,
      "grad_norm": 7.9375,
      "learning_rate": 8.344997834142103e-06,
      "loss": 0.82609434,
      "memory(GiB)": 135.77,
      "step": 24750,
      "train_speed(iter/s)": 0.201762
    },
    {
      "acc": 0.79335299,
      "epoch": 0.5776695489872993,
      "grad_norm": 5.59375,
      "learning_rate": 8.343593489034447e-06,
      "loss": 0.74660645,
      "memory(GiB)": 135.77,
      "step": 24760,
      "train_speed(iter/s)": 0.201803
    },
    {
      "acc": 0.77887115,
      "epoch": 0.5779028565595882,
      "grad_norm": 6.34375,
      "learning_rate": 8.342188666629793e-06,
      "loss": 0.82571802,
      "memory(GiB)": 135.77,
      "step": 24770,
      "train_speed(iter/s)": 0.201846
    },
    {
      "acc": 0.78749199,
      "epoch": 0.5781361641318771,
      "grad_norm": 5.53125,
      "learning_rate": 8.340783367128677e-06,
      "loss": 0.7465394,
      "memory(GiB)": 135.77,
      "step": 24780,
      "train_speed(iter/s)": 0.20189
    },
    {
      "acc": 0.7764246,
      "epoch": 0.578369471704166,
      "grad_norm": 3.9375,
      "learning_rate": 8.339377590731705e-06,
      "loss": 0.78514347,
      "memory(GiB)": 135.77,
      "step": 24790,
      "train_speed(iter/s)": 0.201931
    },
    {
      "acc": 0.76679988,
      "epoch": 0.5786027792764549,
      "grad_norm": 6.8125,
      "learning_rate": 8.337971337639552e-06,
      "loss": 0.84274111,
      "memory(GiB)": 135.77,
      "step": 24800,
      "train_speed(iter/s)": 0.201972
    },
    {
      "acc": 0.77209625,
      "epoch": 0.5788360868487438,
      "grad_norm": 5.65625,
      "learning_rate": 8.336564608052961e-06,
      "loss": 0.80813961,
      "memory(GiB)": 135.77,
      "step": 24810,
      "train_speed(iter/s)": 0.202017
    },
    {
      "acc": 0.77081928,
      "epoch": 0.5790693944210327,
      "grad_norm": 5.0625,
      "learning_rate": 8.335157402172743e-06,
      "loss": 0.83020134,
      "memory(GiB)": 135.77,
      "step": 24820,
      "train_speed(iter/s)": 0.20206
    },
    {
      "acc": 0.78011541,
      "epoch": 0.5793027019933216,
      "grad_norm": 5.46875,
      "learning_rate": 8.333749720199772e-06,
      "loss": 0.80131474,
      "memory(GiB)": 135.77,
      "step": 24830,
      "train_speed(iter/s)": 0.202103
    },
    {
      "acc": 0.78027191,
      "epoch": 0.5795360095656105,
      "grad_norm": 7.375,
      "learning_rate": 8.332341562334998e-06,
      "loss": 0.78537641,
      "memory(GiB)": 135.77,
      "step": 24840,
      "train_speed(iter/s)": 0.202141
    },
    {
      "acc": 0.78687563,
      "epoch": 0.5797693171378994,
      "grad_norm": 4.90625,
      "learning_rate": 8.330932928779434e-06,
      "loss": 0.75612831,
      "memory(GiB)": 135.77,
      "step": 24850,
      "train_speed(iter/s)": 0.202183
    },
    {
      "acc": 0.78615999,
      "epoch": 0.5800026247101883,
      "grad_norm": 5.5625,
      "learning_rate": 8.329523819734161e-06,
      "loss": 0.74080048,
      "memory(GiB)": 135.77,
      "step": 24860,
      "train_speed(iter/s)": 0.202226
    },
    {
      "acc": 0.77790847,
      "epoch": 0.5802359322824772,
      "grad_norm": 7.1875,
      "learning_rate": 8.328114235400331e-06,
      "loss": 0.79633036,
      "memory(GiB)": 135.77,
      "step": 24870,
      "train_speed(iter/s)": 0.202268
    },
    {
      "acc": 0.76988811,
      "epoch": 0.5804692398547661,
      "grad_norm": 5.625,
      "learning_rate": 8.326704175979162e-06,
      "loss": 0.8448287,
      "memory(GiB)": 135.77,
      "step": 24880,
      "train_speed(iter/s)": 0.20231
    },
    {
      "acc": 0.78928442,
      "epoch": 0.580702547427055,
      "grad_norm": 4.65625,
      "learning_rate": 8.325293641671936e-06,
      "loss": 0.76444321,
      "memory(GiB)": 135.77,
      "step": 24890,
      "train_speed(iter/s)": 0.202353
    },
    {
      "acc": 0.77171555,
      "epoch": 0.5809358549993439,
      "grad_norm": 4.65625,
      "learning_rate": 8.32388263268001e-06,
      "loss": 0.83708153,
      "memory(GiB)": 135.77,
      "step": 24900,
      "train_speed(iter/s)": 0.202393
    },
    {
      "acc": 0.76470823,
      "epoch": 0.5811691625716328,
      "grad_norm": 6.5625,
      "learning_rate": 8.322471149204804e-06,
      "loss": 0.86168737,
      "memory(GiB)": 135.77,
      "step": 24910,
      "train_speed(iter/s)": 0.202435
    },
    {
      "acc": 0.76229396,
      "epoch": 0.5814024701439217,
      "grad_norm": 6.3125,
      "learning_rate": 8.321059191447807e-06,
      "loss": 0.8397192,
      "memory(GiB)": 135.77,
      "step": 24920,
      "train_speed(iter/s)": 0.202476
    },
    {
      "acc": 0.76328716,
      "epoch": 0.5816357777162106,
      "grad_norm": 4.65625,
      "learning_rate": 8.319646759610573e-06,
      "loss": 0.85694113,
      "memory(GiB)": 135.77,
      "step": 24930,
      "train_speed(iter/s)": 0.202516
    },
    {
      "acc": 0.76525183,
      "epoch": 0.5818690852884993,
      "grad_norm": 18.875,
      "learning_rate": 8.31823385389473e-06,
      "loss": 0.8547142,
      "memory(GiB)": 135.77,
      "step": 24940,
      "train_speed(iter/s)": 0.202558
    },
    {
      "acc": 0.79167852,
      "epoch": 0.5821023928607882,
      "grad_norm": 4.8125,
      "learning_rate": 8.316820474501968e-06,
      "loss": 0.74602165,
      "memory(GiB)": 135.77,
      "step": 24950,
      "train_speed(iter/s)": 0.202601
    },
    {
      "acc": 0.79028225,
      "epoch": 0.5823357004330771,
      "grad_norm": 14.0,
      "learning_rate": 8.315406621634048e-06,
      "loss": 0.76107903,
      "memory(GiB)": 135.77,
      "step": 24960,
      "train_speed(iter/s)": 0.202643
    },
    {
      "acc": 0.76247883,
      "epoch": 0.582569008005366,
      "grad_norm": 4.78125,
      "learning_rate": 8.313992295492794e-06,
      "loss": 0.87208462,
      "memory(GiB)": 135.77,
      "step": 24970,
      "train_speed(iter/s)": 0.202687
    },
    {
      "acc": 0.77712317,
      "epoch": 0.5828023155776549,
      "grad_norm": 4.59375,
      "learning_rate": 8.312577496280103e-06,
      "loss": 0.79026947,
      "memory(GiB)": 135.77,
      "step": 24980,
      "train_speed(iter/s)": 0.202728
    },
    {
      "acc": 0.76572886,
      "epoch": 0.5830356231499438,
      "grad_norm": 5.15625,
      "learning_rate": 8.311162224197938e-06,
      "loss": 0.84107924,
      "memory(GiB)": 135.77,
      "step": 24990,
      "train_speed(iter/s)": 0.202769
    },
    {
      "acc": 0.77434468,
      "epoch": 0.5832689307222327,
      "grad_norm": 5.53125,
      "learning_rate": 8.309746479448324e-06,
      "loss": 0.8128541,
      "memory(GiB)": 135.77,
      "step": 25000,
      "train_speed(iter/s)": 0.202812
    },
    {
      "epoch": 0.5832689307222327,
      "eval_acc": 0.7402269980435179,
      "eval_loss": 0.8210588693618774,
      "eval_runtime": 1270.1769,
      "eval_samples_per_second": 28.335,
      "eval_steps_per_second": 14.168,
      "step": 25000
    },
    {
      "acc": 0.78035855,
      "epoch": 0.5835022382945216,
      "grad_norm": 4.78125,
      "learning_rate": 8.308330262233366e-06,
      "loss": 0.78389635,
      "memory(GiB)": 135.77,
      "step": 25010,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.77793059,
      "epoch": 0.5837355458668105,
      "grad_norm": 4.03125,
      "learning_rate": 8.306913572755221e-06,
      "loss": 0.77305503,
      "memory(GiB)": 135.77,
      "step": 25020,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.77444048,
      "epoch": 0.5839688534390994,
      "grad_norm": 12.3125,
      "learning_rate": 8.305496411216125e-06,
      "loss": 0.82593975,
      "memory(GiB)": 135.77,
      "step": 25030,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.79068418,
      "epoch": 0.5842021610113883,
      "grad_norm": 4.96875,
      "learning_rate": 8.304078777818377e-06,
      "loss": 0.76902728,
      "memory(GiB)": 135.77,
      "step": 25040,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.78544216,
      "epoch": 0.5844354685836772,
      "grad_norm": 6.4375,
      "learning_rate": 8.302660672764343e-06,
      "loss": 0.77539878,
      "memory(GiB)": 135.77,
      "step": 25050,
      "train_speed(iter/s)": 0.200921
    },
    {
      "acc": 0.7708941,
      "epoch": 0.5846687761559661,
      "grad_norm": 6.03125,
      "learning_rate": 8.301242096256457e-06,
      "loss": 0.82189264,
      "memory(GiB)": 135.77,
      "step": 25060,
      "train_speed(iter/s)": 0.20096
    },
    {
      "acc": 0.76837645,
      "epoch": 0.584902083728255,
      "grad_norm": 5.75,
      "learning_rate": 8.299823048497221e-06,
      "loss": 0.8290554,
      "memory(GiB)": 135.77,
      "step": 25070,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.77567482,
      "epoch": 0.5851353913005439,
      "grad_norm": 5.28125,
      "learning_rate": 8.298403529689204e-06,
      "loss": 0.81522255,
      "memory(GiB)": 135.77,
      "step": 25080,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.78811507,
      "epoch": 0.5853686988728328,
      "grad_norm": 5.15625,
      "learning_rate": 8.296983540035041e-06,
      "loss": 0.73775659,
      "memory(GiB)": 135.77,
      "step": 25090,
      "train_speed(iter/s)": 0.201087
    },
    {
      "acc": 0.74818373,
      "epoch": 0.5856020064451217,
      "grad_norm": 4.75,
      "learning_rate": 8.295563079737436e-06,
      "loss": 0.91113939,
      "memory(GiB)": 135.77,
      "step": 25100,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.7771193,
      "epoch": 0.5858353140174106,
      "grad_norm": 5.46875,
      "learning_rate": 8.294142148999157e-06,
      "loss": 0.78849659,
      "memory(GiB)": 135.77,
      "step": 25110,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.77626028,
      "epoch": 0.5860686215896995,
      "grad_norm": 4.9375,
      "learning_rate": 8.292720748023045e-06,
      "loss": 0.78736038,
      "memory(GiB)": 135.77,
      "step": 25120,
      "train_speed(iter/s)": 0.201212
    },
    {
      "acc": 0.78426456,
      "epoch": 0.5863019291619883,
      "grad_norm": 5.71875,
      "learning_rate": 8.291298877012002e-06,
      "loss": 0.78405647,
      "memory(GiB)": 135.77,
      "step": 25130,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.77909927,
      "epoch": 0.5865352367342772,
      "grad_norm": 5.9375,
      "learning_rate": 8.289876536169002e-06,
      "loss": 0.80689306,
      "memory(GiB)": 135.77,
      "step": 25140,
      "train_speed(iter/s)": 0.201293
    },
    {
      "acc": 0.77817984,
      "epoch": 0.5867685443065661,
      "grad_norm": 5.46875,
      "learning_rate": 8.28845372569708e-06,
      "loss": 0.80842056,
      "memory(GiB)": 135.77,
      "step": 25150,
      "train_speed(iter/s)": 0.201335
    },
    {
      "acc": 0.76585131,
      "epoch": 0.587001851878855,
      "grad_norm": 5.3125,
      "learning_rate": 8.287030445799345e-06,
      "loss": 0.84683504,
      "memory(GiB)": 135.77,
      "step": 25160,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.76378751,
      "epoch": 0.5872351594511439,
      "grad_norm": 5.9375,
      "learning_rate": 8.285606696678969e-06,
      "loss": 0.85097218,
      "memory(GiB)": 135.77,
      "step": 25170,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.76524734,
      "epoch": 0.5874684670234328,
      "grad_norm": 6.75,
      "learning_rate": 8.28418247853919e-06,
      "loss": 0.82923918,
      "memory(GiB)": 135.77,
      "step": 25180,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.7815011,
      "epoch": 0.5877017745957217,
      "grad_norm": 5.5,
      "learning_rate": 8.282757791583316e-06,
      "loss": 0.78978734,
      "memory(GiB)": 135.77,
      "step": 25190,
      "train_speed(iter/s)": 0.201499
    },
    {
      "acc": 0.75179167,
      "epoch": 0.5879350821680106,
      "grad_norm": 5.09375,
      "learning_rate": 8.281332636014723e-06,
      "loss": 0.89616728,
      "memory(GiB)": 135.77,
      "step": 25200,
      "train_speed(iter/s)": 0.201541
    },
    {
      "acc": 0.77656918,
      "epoch": 0.5881683897402995,
      "grad_norm": 5.09375,
      "learning_rate": 8.279907012036849e-06,
      "loss": 0.80306749,
      "memory(GiB)": 135.77,
      "step": 25210,
      "train_speed(iter/s)": 0.201579
    },
    {
      "acc": 0.77237129,
      "epoch": 0.5884016973125884,
      "grad_norm": 5.5625,
      "learning_rate": 8.2784809198532e-06,
      "loss": 0.8121336,
      "memory(GiB)": 135.77,
      "step": 25220,
      "train_speed(iter/s)": 0.20162
    },
    {
      "acc": 0.7861938,
      "epoch": 0.5886350048848773,
      "grad_norm": 5.125,
      "learning_rate": 8.277054359667355e-06,
      "loss": 0.76683536,
      "memory(GiB)": 135.77,
      "step": 25230,
      "train_speed(iter/s)": 0.201662
    },
    {
      "acc": 0.75934463,
      "epoch": 0.5888683124571662,
      "grad_norm": 5.8125,
      "learning_rate": 8.27562733168295e-06,
      "loss": 0.87756844,
      "memory(GiB)": 135.77,
      "step": 25240,
      "train_speed(iter/s)": 0.201704
    },
    {
      "acc": 0.76303396,
      "epoch": 0.5891016200294551,
      "grad_norm": 5.34375,
      "learning_rate": 8.274199836103696e-06,
      "loss": 0.84503393,
      "memory(GiB)": 135.77,
      "step": 25250,
      "train_speed(iter/s)": 0.201746
    },
    {
      "acc": 0.77054796,
      "epoch": 0.589334927601744,
      "grad_norm": 11.0,
      "learning_rate": 8.272771873133365e-06,
      "loss": 0.83606682,
      "memory(GiB)": 135.77,
      "step": 25260,
      "train_speed(iter/s)": 0.201789
    },
    {
      "acc": 0.78724136,
      "epoch": 0.5895682351740329,
      "grad_norm": 5.46875,
      "learning_rate": 8.271343442975803e-06,
      "loss": 0.76417646,
      "memory(GiB)": 135.77,
      "step": 25270,
      "train_speed(iter/s)": 0.201829
    },
    {
      "acc": 0.75213413,
      "epoch": 0.5898015427463218,
      "grad_norm": 5.375,
      "learning_rate": 8.269914545834911e-06,
      "loss": 0.90432196,
      "memory(GiB)": 135.77,
      "step": 25280,
      "train_speed(iter/s)": 0.201869
    },
    {
      "acc": 0.76026382,
      "epoch": 0.5900348503186107,
      "grad_norm": 5.34375,
      "learning_rate": 8.26848518191467e-06,
      "loss": 0.88067017,
      "memory(GiB)": 135.77,
      "step": 25290,
      "train_speed(iter/s)": 0.201909
    },
    {
      "acc": 0.77908268,
      "epoch": 0.5902681578908996,
      "grad_norm": 3.953125,
      "learning_rate": 8.267055351419117e-06,
      "loss": 0.78281631,
      "memory(GiB)": 135.77,
      "step": 25300,
      "train_speed(iter/s)": 0.201951
    },
    {
      "acc": 0.78081703,
      "epoch": 0.5905014654631885,
      "grad_norm": 5.6875,
      "learning_rate": 8.265625054552363e-06,
      "loss": 0.7916995,
      "memory(GiB)": 135.77,
      "step": 25310,
      "train_speed(iter/s)": 0.201991
    },
    {
      "acc": 0.78177204,
      "epoch": 0.5907347730354773,
      "grad_norm": 8.9375,
      "learning_rate": 8.264194291518583e-06,
      "loss": 0.79910288,
      "memory(GiB)": 135.77,
      "step": 25320,
      "train_speed(iter/s)": 0.202033
    },
    {
      "acc": 0.75857491,
      "epoch": 0.5909680806077662,
      "grad_norm": 7.46875,
      "learning_rate": 8.262763062522013e-06,
      "loss": 0.87688332,
      "memory(GiB)": 135.77,
      "step": 25330,
      "train_speed(iter/s)": 0.202076
    },
    {
      "acc": 0.76642208,
      "epoch": 0.5912013881800551,
      "grad_norm": 6.375,
      "learning_rate": 8.261331367766965e-06,
      "loss": 0.83190136,
      "memory(GiB)": 135.77,
      "step": 25340,
      "train_speed(iter/s)": 0.202117
    },
    {
      "acc": 0.77461624,
      "epoch": 0.591434695752344,
      "grad_norm": 5.3125,
      "learning_rate": 8.25989920745781e-06,
      "loss": 0.80886202,
      "memory(GiB)": 135.77,
      "step": 25350,
      "train_speed(iter/s)": 0.202155
    },
    {
      "acc": 0.75091953,
      "epoch": 0.5916680033246329,
      "grad_norm": 5.25,
      "learning_rate": 8.258466581798992e-06,
      "loss": 0.90273914,
      "memory(GiB)": 135.77,
      "step": 25360,
      "train_speed(iter/s)": 0.202196
    },
    {
      "acc": 0.77260246,
      "epoch": 0.5919013108969218,
      "grad_norm": 5.8125,
      "learning_rate": 8.257033490995017e-06,
      "loss": 0.81039619,
      "memory(GiB)": 135.77,
      "step": 25370,
      "train_speed(iter/s)": 0.202235
    },
    {
      "acc": 0.745508,
      "epoch": 0.5921346184692107,
      "grad_norm": 5.40625,
      "learning_rate": 8.255599935250456e-06,
      "loss": 0.90657196,
      "memory(GiB)": 135.77,
      "step": 25380,
      "train_speed(iter/s)": 0.202278
    },
    {
      "acc": 0.76827936,
      "epoch": 0.5923679260414996,
      "grad_norm": 6.0625,
      "learning_rate": 8.254165914769949e-06,
      "loss": 0.85670547,
      "memory(GiB)": 135.77,
      "step": 25390,
      "train_speed(iter/s)": 0.202321
    },
    {
      "acc": 0.76772423,
      "epoch": 0.5926012336137885,
      "grad_norm": 4.84375,
      "learning_rate": 8.252731429758205e-06,
      "loss": 0.83695488,
      "memory(GiB)": 135.77,
      "step": 25400,
      "train_speed(iter/s)": 0.202362
    },
    {
      "acc": 0.78266163,
      "epoch": 0.5928345411860774,
      "grad_norm": 5.65625,
      "learning_rate": 8.251296480419992e-06,
      "loss": 0.7788908,
      "memory(GiB)": 135.77,
      "step": 25410,
      "train_speed(iter/s)": 0.202402
    },
    {
      "acc": 0.77551551,
      "epoch": 0.5930678487583663,
      "grad_norm": 4.5625,
      "learning_rate": 8.249861066960154e-06,
      "loss": 0.80347424,
      "memory(GiB)": 135.77,
      "step": 25420,
      "train_speed(iter/s)": 0.202442
    },
    {
      "acc": 0.75031691,
      "epoch": 0.5933011563306552,
      "grad_norm": 5.25,
      "learning_rate": 8.248425189583589e-06,
      "loss": 0.90826054,
      "memory(GiB)": 135.77,
      "step": 25430,
      "train_speed(iter/s)": 0.202484
    },
    {
      "acc": 0.78029752,
      "epoch": 0.5935344639029441,
      "grad_norm": 6.40625,
      "learning_rate": 8.246988848495275e-06,
      "loss": 0.79855204,
      "memory(GiB)": 135.77,
      "step": 25440,
      "train_speed(iter/s)": 0.202525
    },
    {
      "acc": 0.77923479,
      "epoch": 0.593767771475233,
      "grad_norm": 4.5625,
      "learning_rate": 8.245552043900245e-06,
      "loss": 0.80320797,
      "memory(GiB)": 135.77,
      "step": 25450,
      "train_speed(iter/s)": 0.202567
    },
    {
      "acc": 0.76215806,
      "epoch": 0.5940010790475219,
      "grad_norm": 4.53125,
      "learning_rate": 8.244114776003605e-06,
      "loss": 0.88092327,
      "memory(GiB)": 135.77,
      "step": 25460,
      "train_speed(iter/s)": 0.202603
    },
    {
      "acc": 0.7797924,
      "epoch": 0.5942343866198108,
      "grad_norm": 5.6875,
      "learning_rate": 8.24267704501052e-06,
      "loss": 0.78126807,
      "memory(GiB)": 135.77,
      "step": 25470,
      "train_speed(iter/s)": 0.202644
    },
    {
      "acc": 0.78742337,
      "epoch": 0.5944676941920997,
      "grad_norm": 4.84375,
      "learning_rate": 8.241238851126231e-06,
      "loss": 0.78035583,
      "memory(GiB)": 135.77,
      "step": 25480,
      "train_speed(iter/s)": 0.202687
    },
    {
      "acc": 0.77751136,
      "epoch": 0.5947010017643886,
      "grad_norm": 7.8125,
      "learning_rate": 8.239800194556036e-06,
      "loss": 0.80985718,
      "memory(GiB)": 135.77,
      "step": 25490,
      "train_speed(iter/s)": 0.202728
    },
    {
      "acc": 0.76632509,
      "epoch": 0.5949343093366775,
      "grad_norm": 4.8125,
      "learning_rate": 8.238361075505307e-06,
      "loss": 0.8863204,
      "memory(GiB)": 135.77,
      "step": 25500,
      "train_speed(iter/s)": 0.202766
    },
    {
      "epoch": 0.5949343093366775,
      "eval_acc": 0.7401671041006372,
      "eval_loss": 0.8206676840782166,
      "eval_runtime": 1268.9058,
      "eval_samples_per_second": 28.364,
      "eval_steps_per_second": 14.182,
      "step": 25500
    },
    {
      "acc": 0.77610083,
      "epoch": 0.5951676169089664,
      "grad_norm": 6.28125,
      "learning_rate": 8.236921494179474e-06,
      "loss": 0.79520016,
      "memory(GiB)": 135.77,
      "step": 25510,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.76967053,
      "epoch": 0.5954009244812551,
      "grad_norm": 4.5625,
      "learning_rate": 8.235481450784037e-06,
      "loss": 0.82552261,
      "memory(GiB)": 135.77,
      "step": 25520,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.78743782,
      "epoch": 0.595634232053544,
      "grad_norm": 5.78125,
      "learning_rate": 8.234040945524563e-06,
      "loss": 0.78872375,
      "memory(GiB)": 135.77,
      "step": 25530,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.77408667,
      "epoch": 0.5958675396258329,
      "grad_norm": 5.8125,
      "learning_rate": 8.232599978606683e-06,
      "loss": 0.82129602,
      "memory(GiB)": 135.77,
      "step": 25540,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.76378164,
      "epoch": 0.5961008471981218,
      "grad_norm": 7.125,
      "learning_rate": 8.231158550236098e-06,
      "loss": 0.86470642,
      "memory(GiB)": 135.77,
      "step": 25550,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77116547,
      "epoch": 0.5963341547704107,
      "grad_norm": 4.8125,
      "learning_rate": 8.229716660618567e-06,
      "loss": 0.80563984,
      "memory(GiB)": 135.77,
      "step": 25560,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.78916759,
      "epoch": 0.5965674623426996,
      "grad_norm": 5.0,
      "learning_rate": 8.22827430995992e-06,
      "loss": 0.74210978,
      "memory(GiB)": 135.77,
      "step": 25570,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.76487999,
      "epoch": 0.5968007699149885,
      "grad_norm": 5.75,
      "learning_rate": 8.226831498466054e-06,
      "loss": 0.86971359,
      "memory(GiB)": 135.77,
      "step": 25580,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.77532277,
      "epoch": 0.5970340774872774,
      "grad_norm": 5.53125,
      "learning_rate": 8.22538822634293e-06,
      "loss": 0.79406919,
      "memory(GiB)": 135.77,
      "step": 25590,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.78031688,
      "epoch": 0.5972673850595663,
      "grad_norm": 4.90625,
      "learning_rate": 8.223944493796572e-06,
      "loss": 0.77325659,
      "memory(GiB)": 135.77,
      "step": 25600,
      "train_speed(iter/s)": 0.201122
    },
    {
      "acc": 0.78207927,
      "epoch": 0.5975006926318552,
      "grad_norm": 12.0625,
      "learning_rate": 8.222500301033075e-06,
      "loss": 0.78344164,
      "memory(GiB)": 135.77,
      "step": 25610,
      "train_speed(iter/s)": 0.20116
    },
    {
      "acc": 0.76561384,
      "epoch": 0.5977340002041441,
      "grad_norm": 5.1875,
      "learning_rate": 8.221055648258596e-06,
      "loss": 0.85672531,
      "memory(GiB)": 135.77,
      "step": 25620,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.76766024,
      "epoch": 0.597967307776433,
      "grad_norm": 5.0,
      "learning_rate": 8.21961053567936e-06,
      "loss": 0.82406807,
      "memory(GiB)": 135.77,
      "step": 25630,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.79372106,
      "epoch": 0.5982006153487219,
      "grad_norm": 6.25,
      "learning_rate": 8.218164963501651e-06,
      "loss": 0.71891031,
      "memory(GiB)": 135.77,
      "step": 25640,
      "train_speed(iter/s)": 0.201286
    },
    {
      "acc": 0.76460505,
      "epoch": 0.5984339229210108,
      "grad_norm": 5.375,
      "learning_rate": 8.216718931931832e-06,
      "loss": 0.83040361,
      "memory(GiB)": 135.77,
      "step": 25650,
      "train_speed(iter/s)": 0.201327
    },
    {
      "acc": 0.77256279,
      "epoch": 0.5986672304932997,
      "grad_norm": 10.75,
      "learning_rate": 8.21527244117632e-06,
      "loss": 0.82293129,
      "memory(GiB)": 135.77,
      "step": 25660,
      "train_speed(iter/s)": 0.201372
    },
    {
      "acc": 0.79064188,
      "epoch": 0.5989005380655886,
      "grad_norm": 5.90625,
      "learning_rate": 8.2138254914416e-06,
      "loss": 0.72057476,
      "memory(GiB)": 135.77,
      "step": 25670,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.77393141,
      "epoch": 0.5991338456378775,
      "grad_norm": 5.8125,
      "learning_rate": 8.212378082934225e-06,
      "loss": 0.84098873,
      "memory(GiB)": 135.77,
      "step": 25680,
      "train_speed(iter/s)": 0.201451
    },
    {
      "acc": 0.78354292,
      "epoch": 0.5993671532101664,
      "grad_norm": 4.9375,
      "learning_rate": 8.210930215860812e-06,
      "loss": 0.77515049,
      "memory(GiB)": 135.77,
      "step": 25690,
      "train_speed(iter/s)": 0.201493
    },
    {
      "acc": 0.76722078,
      "epoch": 0.5996004607824553,
      "grad_norm": 5.03125,
      "learning_rate": 8.209481890428044e-06,
      "loss": 0.85564814,
      "memory(GiB)": 135.77,
      "step": 25700,
      "train_speed(iter/s)": 0.201533
    },
    {
      "acc": 0.78583412,
      "epoch": 0.5998337683547441,
      "grad_norm": 4.25,
      "learning_rate": 8.208033106842668e-06,
      "loss": 0.79263468,
      "memory(GiB)": 135.77,
      "step": 25710,
      "train_speed(iter/s)": 0.201569
    },
    {
      "acc": 0.77676077,
      "epoch": 0.600067075927033,
      "grad_norm": 7.96875,
      "learning_rate": 8.206583865311497e-06,
      "loss": 0.79163513,
      "memory(GiB)": 135.77,
      "step": 25720,
      "train_speed(iter/s)": 0.201609
    },
    {
      "acc": 0.78243885,
      "epoch": 0.6003003834993219,
      "grad_norm": 4.96875,
      "learning_rate": 8.205134166041412e-06,
      "loss": 0.79570303,
      "memory(GiB)": 135.77,
      "step": 25730,
      "train_speed(iter/s)": 0.201648
    },
    {
      "acc": 0.7864254,
      "epoch": 0.6005336910716108,
      "grad_norm": 5.4375,
      "learning_rate": 8.203684009239356e-06,
      "loss": 0.76417632,
      "memory(GiB)": 135.77,
      "step": 25740,
      "train_speed(iter/s)": 0.201687
    },
    {
      "acc": 0.76960988,
      "epoch": 0.6007669986438997,
      "grad_norm": 5.75,
      "learning_rate": 8.202233395112338e-06,
      "loss": 0.82158051,
      "memory(GiB)": 135.77,
      "step": 25750,
      "train_speed(iter/s)": 0.201728
    },
    {
      "acc": 0.77994032,
      "epoch": 0.6010003062161886,
      "grad_norm": 6.03125,
      "learning_rate": 8.200782323867432e-06,
      "loss": 0.82229548,
      "memory(GiB)": 135.77,
      "step": 25760,
      "train_speed(iter/s)": 0.201767
    },
    {
      "acc": 0.77866874,
      "epoch": 0.6012336137884775,
      "grad_norm": 5.96875,
      "learning_rate": 8.19933079571178e-06,
      "loss": 0.78451872,
      "memory(GiB)": 135.77,
      "step": 25770,
      "train_speed(iter/s)": 0.201805
    },
    {
      "acc": 0.78407946,
      "epoch": 0.6014669213607664,
      "grad_norm": 4.84375,
      "learning_rate": 8.197878810852587e-06,
      "loss": 0.76278973,
      "memory(GiB)": 135.77,
      "step": 25780,
      "train_speed(iter/s)": 0.201846
    },
    {
      "acc": 0.78200378,
      "epoch": 0.6017002289330553,
      "grad_norm": 7.875,
      "learning_rate": 8.196426369497121e-06,
      "loss": 0.78061686,
      "memory(GiB)": 135.77,
      "step": 25790,
      "train_speed(iter/s)": 0.201885
    },
    {
      "acc": 0.76984148,
      "epoch": 0.6019335365053442,
      "grad_norm": 7.0625,
      "learning_rate": 8.19497347185272e-06,
      "loss": 0.82394571,
      "memory(GiB)": 135.77,
      "step": 25800,
      "train_speed(iter/s)": 0.201928
    },
    {
      "acc": 0.74774513,
      "epoch": 0.6021668440776331,
      "grad_norm": 4.90625,
      "learning_rate": 8.193520118126785e-06,
      "loss": 0.89157257,
      "memory(GiB)": 135.77,
      "step": 25810,
      "train_speed(iter/s)": 0.201969
    },
    {
      "acc": 0.77559814,
      "epoch": 0.602400151649922,
      "grad_norm": 6.15625,
      "learning_rate": 8.19206630852678e-06,
      "loss": 0.78932376,
      "memory(GiB)": 135.77,
      "step": 25820,
      "train_speed(iter/s)": 0.202011
    },
    {
      "acc": 0.76863947,
      "epoch": 0.6026334592222109,
      "grad_norm": 5.9375,
      "learning_rate": 8.190612043260238e-06,
      "loss": 0.83424625,
      "memory(GiB)": 135.77,
      "step": 25830,
      "train_speed(iter/s)": 0.202052
    },
    {
      "acc": 0.73870344,
      "epoch": 0.6028667667944998,
      "grad_norm": 5.15625,
      "learning_rate": 8.189157322534753e-06,
      "loss": 0.97275181,
      "memory(GiB)": 135.77,
      "step": 25840,
      "train_speed(iter/s)": 0.202089
    },
    {
      "acc": 0.77539172,
      "epoch": 0.6031000743667887,
      "grad_norm": 6.1875,
      "learning_rate": 8.187702146557986e-06,
      "loss": 0.78618755,
      "memory(GiB)": 135.77,
      "step": 25850,
      "train_speed(iter/s)": 0.20213
    },
    {
      "acc": 0.76942921,
      "epoch": 0.6033333819390776,
      "grad_norm": 4.21875,
      "learning_rate": 8.186246515537664e-06,
      "loss": 0.83434811,
      "memory(GiB)": 135.77,
      "step": 25860,
      "train_speed(iter/s)": 0.202171
    },
    {
      "acc": 0.78348265,
      "epoch": 0.6035666895113665,
      "grad_norm": 6.8125,
      "learning_rate": 8.184790429681577e-06,
      "loss": 0.77311735,
      "memory(GiB)": 135.77,
      "step": 25870,
      "train_speed(iter/s)": 0.202209
    },
    {
      "acc": 0.76982718,
      "epoch": 0.6037999970836554,
      "grad_norm": 5.125,
      "learning_rate": 8.183333889197582e-06,
      "loss": 0.84020672,
      "memory(GiB)": 135.77,
      "step": 25880,
      "train_speed(iter/s)": 0.202249
    },
    {
      "acc": 0.77111564,
      "epoch": 0.6040333046559443,
      "grad_norm": 5.75,
      "learning_rate": 8.181876894293601e-06,
      "loss": 0.80620203,
      "memory(GiB)": 135.77,
      "step": 25890,
      "train_speed(iter/s)": 0.202291
    },
    {
      "acc": 0.77545085,
      "epoch": 0.6042666122282331,
      "grad_norm": 5.8125,
      "learning_rate": 8.180419445177614e-06,
      "loss": 0.81389084,
      "memory(GiB)": 135.77,
      "step": 25900,
      "train_speed(iter/s)": 0.202331
    },
    {
      "acc": 0.77913561,
      "epoch": 0.604499919800522,
      "grad_norm": 5.875,
      "learning_rate": 8.178961542057677e-06,
      "loss": 0.79586811,
      "memory(GiB)": 135.77,
      "step": 25910,
      "train_speed(iter/s)": 0.20237
    },
    {
      "acc": 0.78647404,
      "epoch": 0.6047332273728109,
      "grad_norm": 6.15625,
      "learning_rate": 8.177503185141904e-06,
      "loss": 0.77673407,
      "memory(GiB)": 135.77,
      "step": 25920,
      "train_speed(iter/s)": 0.202411
    },
    {
      "acc": 0.78315163,
      "epoch": 0.6049665349450998,
      "grad_norm": 11.3125,
      "learning_rate": 8.176044374638473e-06,
      "loss": 0.767239,
      "memory(GiB)": 135.77,
      "step": 25930,
      "train_speed(iter/s)": 0.202451
    },
    {
      "acc": 0.76807127,
      "epoch": 0.6051998425173887,
      "grad_norm": 5.6875,
      "learning_rate": 8.174585110755631e-06,
      "loss": 0.86293449,
      "memory(GiB)": 135.77,
      "step": 25940,
      "train_speed(iter/s)": 0.202491
    },
    {
      "acc": 0.78088841,
      "epoch": 0.6054331500896776,
      "grad_norm": 4.5,
      "learning_rate": 8.173125393701686e-06,
      "loss": 0.7886466,
      "memory(GiB)": 135.77,
      "step": 25950,
      "train_speed(iter/s)": 0.202533
    },
    {
      "acc": 0.77306123,
      "epoch": 0.6056664576619665,
      "grad_norm": 7.28125,
      "learning_rate": 8.171665223685014e-06,
      "loss": 0.81952877,
      "memory(GiB)": 135.77,
      "step": 25960,
      "train_speed(iter/s)": 0.202574
    },
    {
      "acc": 0.78507872,
      "epoch": 0.6058997652342554,
      "grad_norm": 4.78125,
      "learning_rate": 8.170204600914051e-06,
      "loss": 0.76790247,
      "memory(GiB)": 135.77,
      "step": 25970,
      "train_speed(iter/s)": 0.202612
    },
    {
      "acc": 0.76558475,
      "epoch": 0.6061330728065443,
      "grad_norm": 6.09375,
      "learning_rate": 8.168743525597304e-06,
      "loss": 0.86151485,
      "memory(GiB)": 135.77,
      "step": 25980,
      "train_speed(iter/s)": 0.20265
    },
    {
      "acc": 0.75887089,
      "epoch": 0.6063663803788332,
      "grad_norm": 6.46875,
      "learning_rate": 8.167281997943338e-06,
      "loss": 0.87729855,
      "memory(GiB)": 135.77,
      "step": 25990,
      "train_speed(iter/s)": 0.202691
    },
    {
      "acc": 0.7482584,
      "epoch": 0.6065996879511221,
      "grad_norm": 7.3125,
      "learning_rate": 8.165820018160787e-06,
      "loss": 0.91938248,
      "memory(GiB)": 135.77,
      "step": 26000,
      "train_speed(iter/s)": 0.202728
    },
    {
      "epoch": 0.6065996879511221,
      "eval_acc": 0.7404893142960276,
      "eval_loss": 0.8198909163475037,
      "eval_runtime": 1269.2964,
      "eval_samples_per_second": 28.355,
      "eval_steps_per_second": 14.178,
      "step": 26000
    },
    {
      "acc": 0.78726702,
      "epoch": 0.606832995523411,
      "grad_norm": 6.21875,
      "learning_rate": 8.164357586458348e-06,
      "loss": 0.75815754,
      "memory(GiB)": 135.77,
      "step": 26010,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.7841042,
      "epoch": 0.6070663030956999,
      "grad_norm": 20.0,
      "learning_rate": 8.162894703044783e-06,
      "loss": 0.7632184,
      "memory(GiB)": 135.77,
      "step": 26020,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.75316358,
      "epoch": 0.6072996106679888,
      "grad_norm": 4.625,
      "learning_rate": 8.161431368128919e-06,
      "loss": 0.89499702,
      "memory(GiB)": 135.77,
      "step": 26030,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.78723059,
      "epoch": 0.6075329182402777,
      "grad_norm": 6.0625,
      "learning_rate": 8.159967581919644e-06,
      "loss": 0.74924011,
      "memory(GiB)": 135.77,
      "step": 26040,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.79428697,
      "epoch": 0.6077662258125666,
      "grad_norm": 5.90625,
      "learning_rate": 8.158503344625915e-06,
      "loss": 0.73409123,
      "memory(GiB)": 135.77,
      "step": 26050,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.76840878,
      "epoch": 0.6079995333848555,
      "grad_norm": 5.53125,
      "learning_rate": 8.157038656456752e-06,
      "loss": 0.83939495,
      "memory(GiB)": 135.77,
      "step": 26060,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.77537632,
      "epoch": 0.6082328409571444,
      "grad_norm": 4.09375,
      "learning_rate": 8.155573517621238e-06,
      "loss": 0.81753826,
      "memory(GiB)": 135.77,
      "step": 26070,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.7842936,
      "epoch": 0.6084661485294333,
      "grad_norm": 6.59375,
      "learning_rate": 8.154107928328521e-06,
      "loss": 0.79942732,
      "memory(GiB)": 135.77,
      "step": 26080,
      "train_speed(iter/s)": 0.201029
    },
    {
      "acc": 0.77272925,
      "epoch": 0.608699456101722,
      "grad_norm": 6.71875,
      "learning_rate": 8.152641888787812e-06,
      "loss": 0.80734158,
      "memory(GiB)": 135.77,
      "step": 26090,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.76670575,
      "epoch": 0.608932763674011,
      "grad_norm": 7.75,
      "learning_rate": 8.15117539920839e-06,
      "loss": 0.85535202,
      "memory(GiB)": 135.77,
      "step": 26100,
      "train_speed(iter/s)": 0.201109
    },
    {
      "acc": 0.75531483,
      "epoch": 0.6091660712462998,
      "grad_norm": 6.09375,
      "learning_rate": 8.149708459799595e-06,
      "loss": 0.8802372,
      "memory(GiB)": 135.77,
      "step": 26110,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.76685176,
      "epoch": 0.6093993788185887,
      "grad_norm": 6.0,
      "learning_rate": 8.148241070770834e-06,
      "loss": 0.83814058,
      "memory(GiB)": 135.77,
      "step": 26120,
      "train_speed(iter/s)": 0.20119
    },
    {
      "acc": 0.7886631,
      "epoch": 0.6096326863908776,
      "grad_norm": 9.1875,
      "learning_rate": 8.146773232331574e-06,
      "loss": 0.77238207,
      "memory(GiB)": 135.77,
      "step": 26130,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.75037575,
      "epoch": 0.6098659939631665,
      "grad_norm": 5.65625,
      "learning_rate": 8.145304944691347e-06,
      "loss": 0.90184555,
      "memory(GiB)": 135.77,
      "step": 26140,
      "train_speed(iter/s)": 0.201269
    },
    {
      "acc": 0.7713768,
      "epoch": 0.6100993015354554,
      "grad_norm": 7.84375,
      "learning_rate": 8.143836208059754e-06,
      "loss": 0.80705109,
      "memory(GiB)": 135.77,
      "step": 26150,
      "train_speed(iter/s)": 0.20131
    },
    {
      "acc": 0.77182598,
      "epoch": 0.6103326091077443,
      "grad_norm": 4.78125,
      "learning_rate": 8.142367022646457e-06,
      "loss": 0.83266945,
      "memory(GiB)": 135.77,
      "step": 26160,
      "train_speed(iter/s)": 0.201352
    },
    {
      "acc": 0.77788215,
      "epoch": 0.6105659166800332,
      "grad_norm": 5.8125,
      "learning_rate": 8.14089738866118e-06,
      "loss": 0.80510206,
      "memory(GiB)": 135.77,
      "step": 26170,
      "train_speed(iter/s)": 0.201388
    },
    {
      "acc": 0.77358456,
      "epoch": 0.6107992242523221,
      "grad_norm": 4.59375,
      "learning_rate": 8.139427306313713e-06,
      "loss": 0.81454687,
      "memory(GiB)": 135.77,
      "step": 26180,
      "train_speed(iter/s)": 0.201426
    },
    {
      "acc": 0.78963194,
      "epoch": 0.611032531824611,
      "grad_norm": 7.6875,
      "learning_rate": 8.137956775813909e-06,
      "loss": 0.75747194,
      "memory(GiB)": 135.77,
      "step": 26190,
      "train_speed(iter/s)": 0.201467
    },
    {
      "acc": 0.76336627,
      "epoch": 0.6112658393968999,
      "grad_norm": 4.6875,
      "learning_rate": 8.136485797371687e-06,
      "loss": 0.84080048,
      "memory(GiB)": 135.77,
      "step": 26200,
      "train_speed(iter/s)": 0.201506
    },
    {
      "acc": 0.772651,
      "epoch": 0.6114991469691888,
      "grad_norm": 4.78125,
      "learning_rate": 8.13501437119703e-06,
      "loss": 0.81818466,
      "memory(GiB)": 135.77,
      "step": 26210,
      "train_speed(iter/s)": 0.201547
    },
    {
      "acc": 0.75983801,
      "epoch": 0.6117324545414777,
      "grad_norm": 7.03125,
      "learning_rate": 8.133542497499981e-06,
      "loss": 0.85904999,
      "memory(GiB)": 135.77,
      "step": 26220,
      "train_speed(iter/s)": 0.201587
    },
    {
      "acc": 0.77595029,
      "epoch": 0.6119657621137666,
      "grad_norm": 6.8125,
      "learning_rate": 8.132070176490652e-06,
      "loss": 0.79862223,
      "memory(GiB)": 135.77,
      "step": 26230,
      "train_speed(iter/s)": 0.201625
    },
    {
      "acc": 0.76928434,
      "epoch": 0.6121990696860555,
      "grad_norm": 6.28125,
      "learning_rate": 8.130597408379214e-06,
      "loss": 0.80921583,
      "memory(GiB)": 135.77,
      "step": 26240,
      "train_speed(iter/s)": 0.201663
    },
    {
      "acc": 0.76527472,
      "epoch": 0.6124323772583444,
      "grad_norm": 4.28125,
      "learning_rate": 8.129124193375906e-06,
      "loss": 0.83155708,
      "memory(GiB)": 135.77,
      "step": 26250,
      "train_speed(iter/s)": 0.201702
    },
    {
      "acc": 0.79154787,
      "epoch": 0.6126656848306333,
      "grad_norm": 5.53125,
      "learning_rate": 8.127650531691028e-06,
      "loss": 0.74951324,
      "memory(GiB)": 135.77,
      "step": 26260,
      "train_speed(iter/s)": 0.201743
    },
    {
      "acc": 0.78243771,
      "epoch": 0.6128989924029222,
      "grad_norm": 5.8125,
      "learning_rate": 8.126176423534945e-06,
      "loss": 0.76880646,
      "memory(GiB)": 135.77,
      "step": 26270,
      "train_speed(iter/s)": 0.201785
    },
    {
      "acc": 0.77010064,
      "epoch": 0.6131322999752111,
      "grad_norm": 5.34375,
      "learning_rate": 8.124701869118086e-06,
      "loss": 0.84182453,
      "memory(GiB)": 135.77,
      "step": 26280,
      "train_speed(iter/s)": 0.201823
    },
    {
      "acc": 0.7652318,
      "epoch": 0.6133656075474999,
      "grad_norm": 4.8125,
      "learning_rate": 8.123226868650944e-06,
      "loss": 0.8603693,
      "memory(GiB)": 135.77,
      "step": 26290,
      "train_speed(iter/s)": 0.201861
    },
    {
      "acc": 0.77005825,
      "epoch": 0.6135989151197888,
      "grad_norm": 8.875,
      "learning_rate": 8.121751422344072e-06,
      "loss": 0.81795826,
      "memory(GiB)": 135.77,
      "step": 26300,
      "train_speed(iter/s)": 0.201901
    },
    {
      "acc": 0.78708076,
      "epoch": 0.6138322226920777,
      "grad_norm": 6.1875,
      "learning_rate": 8.120275530408092e-06,
      "loss": 0.74994001,
      "memory(GiB)": 135.77,
      "step": 26310,
      "train_speed(iter/s)": 0.201942
    },
    {
      "acc": 0.78978472,
      "epoch": 0.6140655302643666,
      "grad_norm": 6.03125,
      "learning_rate": 8.118799193053686e-06,
      "loss": 0.74331207,
      "memory(GiB)": 135.77,
      "step": 26320,
      "train_speed(iter/s)": 0.201979
    },
    {
      "acc": 0.75865965,
      "epoch": 0.6142988378366555,
      "grad_norm": 5.6875,
      "learning_rate": 8.117322410491602e-06,
      "loss": 0.86435623,
      "memory(GiB)": 135.77,
      "step": 26330,
      "train_speed(iter/s)": 0.202018
    },
    {
      "acc": 0.77246876,
      "epoch": 0.6145321454089444,
      "grad_norm": 6.625,
      "learning_rate": 8.11584518293265e-06,
      "loss": 0.81503639,
      "memory(GiB)": 135.77,
      "step": 26340,
      "train_speed(iter/s)": 0.202058
    },
    {
      "acc": 0.76987753,
      "epoch": 0.6147654529812333,
      "grad_norm": 5.15625,
      "learning_rate": 8.114367510587701e-06,
      "loss": 0.82482548,
      "memory(GiB)": 135.77,
      "step": 26350,
      "train_speed(iter/s)": 0.202097
    },
    {
      "acc": 0.76622534,
      "epoch": 0.6149987605535222,
      "grad_norm": 6.28125,
      "learning_rate": 8.112889393667698e-06,
      "loss": 0.83077908,
      "memory(GiB)": 135.77,
      "step": 26360,
      "train_speed(iter/s)": 0.202137
    },
    {
      "acc": 0.77782397,
      "epoch": 0.6152320681258111,
      "grad_norm": 4.25,
      "learning_rate": 8.111410832383635e-06,
      "loss": 0.77891083,
      "memory(GiB)": 135.77,
      "step": 26370,
      "train_speed(iter/s)": 0.202173
    },
    {
      "acc": 0.79491034,
      "epoch": 0.6154653756981,
      "grad_norm": 4.21875,
      "learning_rate": 8.109931826946582e-06,
      "loss": 0.73813534,
      "memory(GiB)": 135.77,
      "step": 26380,
      "train_speed(iter/s)": 0.202211
    },
    {
      "acc": 0.78529496,
      "epoch": 0.6156986832703889,
      "grad_norm": 4.6875,
      "learning_rate": 8.108452377567663e-06,
      "loss": 0.77241392,
      "memory(GiB)": 135.77,
      "step": 26390,
      "train_speed(iter/s)": 0.202246
    },
    {
      "acc": 0.76942668,
      "epoch": 0.6159319908426778,
      "grad_norm": 6.875,
      "learning_rate": 8.10697248445807e-06,
      "loss": 0.82725449,
      "memory(GiB)": 135.77,
      "step": 26400,
      "train_speed(iter/s)": 0.202286
    },
    {
      "acc": 0.77350941,
      "epoch": 0.6161652984149667,
      "grad_norm": 4.5,
      "learning_rate": 8.105492147829059e-06,
      "loss": 0.81450577,
      "memory(GiB)": 135.77,
      "step": 26410,
      "train_speed(iter/s)": 0.202328
    },
    {
      "acc": 0.7737443,
      "epoch": 0.6163986059872556,
      "grad_norm": 5.6875,
      "learning_rate": 8.104011367891944e-06,
      "loss": 0.81999092,
      "memory(GiB)": 135.77,
      "step": 26420,
      "train_speed(iter/s)": 0.202367
    },
    {
      "acc": 0.78863039,
      "epoch": 0.6166319135595445,
      "grad_norm": 4.96875,
      "learning_rate": 8.102530144858109e-06,
      "loss": 0.77086744,
      "memory(GiB)": 135.77,
      "step": 26430,
      "train_speed(iter/s)": 0.202406
    },
    {
      "acc": 0.77370825,
      "epoch": 0.6168652211318334,
      "grad_norm": 4.46875,
      "learning_rate": 8.101048478938997e-06,
      "loss": 0.78342042,
      "memory(GiB)": 135.77,
      "step": 26440,
      "train_speed(iter/s)": 0.202446
    },
    {
      "acc": 0.78318248,
      "epoch": 0.6170985287041223,
      "grad_norm": 4.65625,
      "learning_rate": 8.099566370346115e-06,
      "loss": 0.78425016,
      "memory(GiB)": 135.77,
      "step": 26450,
      "train_speed(iter/s)": 0.202487
    },
    {
      "acc": 0.78459816,
      "epoch": 0.6173318362764112,
      "grad_norm": 5.4375,
      "learning_rate": 8.098083819291034e-06,
      "loss": 0.78062539,
      "memory(GiB)": 135.77,
      "step": 26460,
      "train_speed(iter/s)": 0.202527
    },
    {
      "acc": 0.77326651,
      "epoch": 0.6175651438487001,
      "grad_norm": 4.875,
      "learning_rate": 8.096600825985388e-06,
      "loss": 0.79871411,
      "memory(GiB)": 135.77,
      "step": 26470,
      "train_speed(iter/s)": 0.202567
    },
    {
      "acc": 0.75785589,
      "epoch": 0.6177984514209889,
      "grad_norm": 4.96875,
      "learning_rate": 8.095117390640875e-06,
      "loss": 0.87607994,
      "memory(GiB)": 135.77,
      "step": 26480,
      "train_speed(iter/s)": 0.202605
    },
    {
      "acc": 0.80151081,
      "epoch": 0.6180317589932778,
      "grad_norm": 7.53125,
      "learning_rate": 8.093633513469252e-06,
      "loss": 0.70373769,
      "memory(GiB)": 135.77,
      "step": 26490,
      "train_speed(iter/s)": 0.202645
    },
    {
      "acc": 0.78222547,
      "epoch": 0.6182650665655667,
      "grad_norm": 6.90625,
      "learning_rate": 8.092149194682343e-06,
      "loss": 0.78799133,
      "memory(GiB)": 135.77,
      "step": 26500,
      "train_speed(iter/s)": 0.202684
    },
    {
      "epoch": 0.6182650665655667,
      "eval_acc": 0.7407008648268977,
      "eval_loss": 0.8189985156059265,
      "eval_runtime": 1271.9101,
      "eval_samples_per_second": 28.297,
      "eval_steps_per_second": 14.149,
      "step": 26500
    },
    {
      "acc": 0.77204461,
      "epoch": 0.6184983741378556,
      "grad_norm": 9.25,
      "learning_rate": 8.090664434492037e-06,
      "loss": 0.81106329,
      "memory(GiB)": 135.77,
      "step": 26510,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.79130363,
      "epoch": 0.6187316817101445,
      "grad_norm": 4.65625,
      "learning_rate": 8.08917923311028e-06,
      "loss": 0.7358408,
      "memory(GiB)": 135.77,
      "step": 26520,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.78020353,
      "epoch": 0.6189649892824334,
      "grad_norm": 6.3125,
      "learning_rate": 8.087693590749083e-06,
      "loss": 0.80418053,
      "memory(GiB)": 135.77,
      "step": 26530,
      "train_speed(iter/s)": 0.200819
    },
    {
      "acc": 0.79271488,
      "epoch": 0.6191982968547223,
      "grad_norm": 17.625,
      "learning_rate": 8.086207507620524e-06,
      "loss": 0.71984701,
      "memory(GiB)": 135.77,
      "step": 26540,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.7886724,
      "epoch": 0.6194316044270112,
      "grad_norm": 6.34375,
      "learning_rate": 8.084720983936742e-06,
      "loss": 0.75640163,
      "memory(GiB)": 135.77,
      "step": 26550,
      "train_speed(iter/s)": 0.200895
    },
    {
      "acc": 0.776264,
      "epoch": 0.6196649119993001,
      "grad_norm": 5.84375,
      "learning_rate": 8.083234019909933e-06,
      "loss": 0.79287157,
      "memory(GiB)": 135.77,
      "step": 26560,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.77901525,
      "epoch": 0.619898219571589,
      "grad_norm": 7.875,
      "learning_rate": 8.081746615752365e-06,
      "loss": 0.80261803,
      "memory(GiB)": 135.77,
      "step": 26570,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.77423306,
      "epoch": 0.6201315271438779,
      "grad_norm": 13.75,
      "learning_rate": 8.080258771676363e-06,
      "loss": 0.78158264,
      "memory(GiB)": 135.77,
      "step": 26580,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.79111242,
      "epoch": 0.6203648347161668,
      "grad_norm": 5.25,
      "learning_rate": 8.078770487894314e-06,
      "loss": 0.75197382,
      "memory(GiB)": 135.77,
      "step": 26590,
      "train_speed(iter/s)": 0.201049
    },
    {
      "acc": 0.76238708,
      "epoch": 0.6205981422884557,
      "grad_norm": 7.65625,
      "learning_rate": 8.077281764618674e-06,
      "loss": 0.85257568,
      "memory(GiB)": 135.77,
      "step": 26600,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.77836542,
      "epoch": 0.6208314498607446,
      "grad_norm": 4.625,
      "learning_rate": 8.075792602061955e-06,
      "loss": 0.79014702,
      "memory(GiB)": 135.77,
      "step": 26610,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.77676468,
      "epoch": 0.6210647574330335,
      "grad_norm": 4.90625,
      "learning_rate": 8.074303000436737e-06,
      "loss": 0.79348154,
      "memory(GiB)": 135.77,
      "step": 26620,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.7470377,
      "epoch": 0.6212980650053224,
      "grad_norm": 6.5625,
      "learning_rate": 8.072812959955657e-06,
      "loss": 0.92623911,
      "memory(GiB)": 135.77,
      "step": 26630,
      "train_speed(iter/s)": 0.201208
    },
    {
      "acc": 0.78787136,
      "epoch": 0.6215313725776113,
      "grad_norm": 5.0625,
      "learning_rate": 8.071322480831422e-06,
      "loss": 0.77648373,
      "memory(GiB)": 135.77,
      "step": 26640,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.7703825,
      "epoch": 0.6217646801499002,
      "grad_norm": 4.25,
      "learning_rate": 8.069831563276793e-06,
      "loss": 0.82653189,
      "memory(GiB)": 135.77,
      "step": 26650,
      "train_speed(iter/s)": 0.201288
    },
    {
      "acc": 0.76768293,
      "epoch": 0.6219979877221891,
      "grad_norm": 4.9375,
      "learning_rate": 8.068340207504601e-06,
      "loss": 0.84485979,
      "memory(GiB)": 135.77,
      "step": 26660,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.78323984,
      "epoch": 0.6222312952944778,
      "grad_norm": 5.9375,
      "learning_rate": 8.066848413727736e-06,
      "loss": 0.77385492,
      "memory(GiB)": 135.77,
      "step": 26670,
      "train_speed(iter/s)": 0.201365
    },
    {
      "acc": 0.77316294,
      "epoch": 0.6224646028667667,
      "grad_norm": 15.5,
      "learning_rate": 8.06535618215915e-06,
      "loss": 0.83223066,
      "memory(GiB)": 135.77,
      "step": 26680,
      "train_speed(iter/s)": 0.201403
    },
    {
      "acc": 0.77781687,
      "epoch": 0.6226979104390556,
      "grad_norm": 6.1875,
      "learning_rate": 8.06386351301186e-06,
      "loss": 0.80505581,
      "memory(GiB)": 135.77,
      "step": 26690,
      "train_speed(iter/s)": 0.201442
    },
    {
      "acc": 0.78201003,
      "epoch": 0.6229312180113445,
      "grad_norm": 4.96875,
      "learning_rate": 8.062370406498944e-06,
      "loss": 0.77910857,
      "memory(GiB)": 135.77,
      "step": 26700,
      "train_speed(iter/s)": 0.20148
    },
    {
      "acc": 0.75401936,
      "epoch": 0.6231645255836334,
      "grad_norm": 4.1875,
      "learning_rate": 8.060876862833543e-06,
      "loss": 0.88524904,
      "memory(GiB)": 135.77,
      "step": 26710,
      "train_speed(iter/s)": 0.201519
    },
    {
      "acc": 0.77211714,
      "epoch": 0.6233978331559223,
      "grad_norm": 4.5625,
      "learning_rate": 8.059382882228857e-06,
      "loss": 0.81682949,
      "memory(GiB)": 135.77,
      "step": 26720,
      "train_speed(iter/s)": 0.201559
    },
    {
      "acc": 0.77653065,
      "epoch": 0.6236311407282112,
      "grad_norm": 3.890625,
      "learning_rate": 8.057888464898153e-06,
      "loss": 0.81049767,
      "memory(GiB)": 135.77,
      "step": 26730,
      "train_speed(iter/s)": 0.201597
    },
    {
      "acc": 0.76580505,
      "epoch": 0.6238644483005001,
      "grad_norm": 6.96875,
      "learning_rate": 8.056393611054761e-06,
      "loss": 0.84048862,
      "memory(GiB)": 135.77,
      "step": 26740,
      "train_speed(iter/s)": 0.201636
    },
    {
      "acc": 0.77059517,
      "epoch": 0.624097755872789,
      "grad_norm": 6.0,
      "learning_rate": 8.054898320912069e-06,
      "loss": 0.84382706,
      "memory(GiB)": 135.77,
      "step": 26750,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.78693314,
      "epoch": 0.6243310634450779,
      "grad_norm": 7.25,
      "learning_rate": 8.053402594683527e-06,
      "loss": 0.76821909,
      "memory(GiB)": 135.77,
      "step": 26760,
      "train_speed(iter/s)": 0.201715
    },
    {
      "acc": 0.78143969,
      "epoch": 0.6245643710173668,
      "grad_norm": 6.21875,
      "learning_rate": 8.051906432582651e-06,
      "loss": 0.77775021,
      "memory(GiB)": 135.77,
      "step": 26770,
      "train_speed(iter/s)": 0.201755
    },
    {
      "acc": 0.78252888,
      "epoch": 0.6247976785896557,
      "grad_norm": 4.625,
      "learning_rate": 8.050409834823021e-06,
      "loss": 0.77935286,
      "memory(GiB)": 135.77,
      "step": 26780,
      "train_speed(iter/s)": 0.201791
    },
    {
      "acc": 0.78484788,
      "epoch": 0.6250309861619446,
      "grad_norm": 5.59375,
      "learning_rate": 8.04891280161827e-06,
      "loss": 0.7744689,
      "memory(GiB)": 135.77,
      "step": 26790,
      "train_speed(iter/s)": 0.201828
    },
    {
      "acc": 0.7690753,
      "epoch": 0.6252642937342335,
      "grad_norm": 4.84375,
      "learning_rate": 8.047415333182105e-06,
      "loss": 0.82785397,
      "memory(GiB)": 135.77,
      "step": 26800,
      "train_speed(iter/s)": 0.201867
    },
    {
      "acc": 0.78505082,
      "epoch": 0.6254976013065224,
      "grad_norm": 4.40625,
      "learning_rate": 8.045917429728286e-06,
      "loss": 0.79035482,
      "memory(GiB)": 135.77,
      "step": 26810,
      "train_speed(iter/s)": 0.201908
    },
    {
      "acc": 0.77882872,
      "epoch": 0.6257309088788113,
      "grad_norm": 5.0625,
      "learning_rate": 8.044419091470638e-06,
      "loss": 0.79798956,
      "memory(GiB)": 135.77,
      "step": 26820,
      "train_speed(iter/s)": 0.201949
    },
    {
      "acc": 0.80624542,
      "epoch": 0.6259642164511002,
      "grad_norm": 4.375,
      "learning_rate": 8.042920318623051e-06,
      "loss": 0.67850709,
      "memory(GiB)": 135.77,
      "step": 26830,
      "train_speed(iter/s)": 0.201989
    },
    {
      "acc": 0.77503967,
      "epoch": 0.6261975240233891,
      "grad_norm": 5.375,
      "learning_rate": 8.04142111139947e-06,
      "loss": 0.8061245,
      "memory(GiB)": 135.77,
      "step": 26840,
      "train_speed(iter/s)": 0.202027
    },
    {
      "acc": 0.76675572,
      "epoch": 0.626430831595678,
      "grad_norm": 4.75,
      "learning_rate": 8.039921470013912e-06,
      "loss": 0.85774889,
      "memory(GiB)": 135.77,
      "step": 26850,
      "train_speed(iter/s)": 0.202064
    },
    {
      "acc": 0.76741781,
      "epoch": 0.6266641391679668,
      "grad_norm": 9.8125,
      "learning_rate": 8.038421394680445e-06,
      "loss": 0.82042427,
      "memory(GiB)": 135.77,
      "step": 26860,
      "train_speed(iter/s)": 0.202104
    },
    {
      "acc": 0.78301311,
      "epoch": 0.6268974467402557,
      "grad_norm": 5.0,
      "learning_rate": 8.036920885613206e-06,
      "loss": 0.77580643,
      "memory(GiB)": 135.77,
      "step": 26870,
      "train_speed(iter/s)": 0.202141
    },
    {
      "acc": 0.78032942,
      "epoch": 0.6271307543125446,
      "grad_norm": 3.984375,
      "learning_rate": 8.035419943026395e-06,
      "loss": 0.7978806,
      "memory(GiB)": 135.77,
      "step": 26880,
      "train_speed(iter/s)": 0.202178
    },
    {
      "acc": 0.77603312,
      "epoch": 0.6273640618848335,
      "grad_norm": 4.28125,
      "learning_rate": 8.033918567134266e-06,
      "loss": 0.81758986,
      "memory(GiB)": 135.77,
      "step": 26890,
      "train_speed(iter/s)": 0.202213
    },
    {
      "acc": 0.76494641,
      "epoch": 0.6275973694571224,
      "grad_norm": 4.21875,
      "learning_rate": 8.032416758151144e-06,
      "loss": 0.83805542,
      "memory(GiB)": 135.77,
      "step": 26900,
      "train_speed(iter/s)": 0.202252
    },
    {
      "acc": 0.78844762,
      "epoch": 0.6278306770294113,
      "grad_norm": 6.625,
      "learning_rate": 8.030914516291413e-06,
      "loss": 0.73256688,
      "memory(GiB)": 135.77,
      "step": 26910,
      "train_speed(iter/s)": 0.202292
    },
    {
      "acc": 0.78790379,
      "epoch": 0.6280639846017002,
      "grad_norm": 4.84375,
      "learning_rate": 8.029411841769515e-06,
      "loss": 0.77325296,
      "memory(GiB)": 135.77,
      "step": 26920,
      "train_speed(iter/s)": 0.20233
    },
    {
      "acc": 0.77131805,
      "epoch": 0.6282972921739891,
      "grad_norm": 5.3125,
      "learning_rate": 8.027908734799954e-06,
      "loss": 0.82325258,
      "memory(GiB)": 135.77,
      "step": 26930,
      "train_speed(iter/s)": 0.202368
    },
    {
      "acc": 0.77684565,
      "epoch": 0.628530599746278,
      "grad_norm": 5.125,
      "learning_rate": 8.026405195597302e-06,
      "loss": 0.77707276,
      "memory(GiB)": 135.77,
      "step": 26940,
      "train_speed(iter/s)": 0.202407
    },
    {
      "acc": 0.78030272,
      "epoch": 0.6287639073185669,
      "grad_norm": 4.625,
      "learning_rate": 8.024901224376186e-06,
      "loss": 0.77970848,
      "memory(GiB)": 135.77,
      "step": 26950,
      "train_speed(iter/s)": 0.202447
    },
    {
      "acc": 0.78028889,
      "epoch": 0.6289972148908558,
      "grad_norm": 5.75,
      "learning_rate": 8.023396821351302e-06,
      "loss": 0.80391293,
      "memory(GiB)": 135.77,
      "step": 26960,
      "train_speed(iter/s)": 0.202486
    },
    {
      "acc": 0.76113973,
      "epoch": 0.6292305224631447,
      "grad_norm": 4.96875,
      "learning_rate": 8.021891986737399e-06,
      "loss": 0.84696064,
      "memory(GiB)": 135.77,
      "step": 26970,
      "train_speed(iter/s)": 0.202525
    },
    {
      "acc": 0.77851915,
      "epoch": 0.6294638300354336,
      "grad_norm": 12.1875,
      "learning_rate": 8.020386720749292e-06,
      "loss": 0.8118988,
      "memory(GiB)": 135.77,
      "step": 26980,
      "train_speed(iter/s)": 0.202564
    },
    {
      "acc": 0.78185272,
      "epoch": 0.6296971376077225,
      "grad_norm": 6.625,
      "learning_rate": 8.018881023601858e-06,
      "loss": 0.77297134,
      "memory(GiB)": 135.77,
      "step": 26990,
      "train_speed(iter/s)": 0.202603
    },
    {
      "acc": 0.77753963,
      "epoch": 0.6299304451800114,
      "grad_norm": 4.90625,
      "learning_rate": 8.017374895510035e-06,
      "loss": 0.80134878,
      "memory(GiB)": 135.77,
      "step": 27000,
      "train_speed(iter/s)": 0.202641
    },
    {
      "epoch": 0.6299304451800114,
      "eval_acc": 0.7406633910765392,
      "eval_loss": 0.8188360929489136,
      "eval_runtime": 1270.4842,
      "eval_samples_per_second": 28.329,
      "eval_steps_per_second": 14.165,
      "step": 27000
    },
    {
      "acc": 0.78660893,
      "epoch": 0.6301637527523003,
      "grad_norm": 4.59375,
      "learning_rate": 8.015868336688822e-06,
      "loss": 0.79268298,
      "memory(GiB)": 135.77,
      "step": 27010,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.78694639,
      "epoch": 0.6303970603245892,
      "grad_norm": 10.5625,
      "learning_rate": 8.01436134735328e-06,
      "loss": 0.76775427,
      "memory(GiB)": 135.77,
      "step": 27020,
      "train_speed(iter/s)": 0.20078
    },
    {
      "acc": 0.79044333,
      "epoch": 0.6306303678968781,
      "grad_norm": 6.0,
      "learning_rate": 8.012853927718532e-06,
      "loss": 0.76151662,
      "memory(GiB)": 135.77,
      "step": 27030,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.77135372,
      "epoch": 0.630863675469167,
      "grad_norm": 5.28125,
      "learning_rate": 8.011346077999762e-06,
      "loss": 0.82462111,
      "memory(GiB)": 135.77,
      "step": 27040,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.76923323,
      "epoch": 0.6310969830414559,
      "grad_norm": 5.6875,
      "learning_rate": 8.009837798412213e-06,
      "loss": 0.81341896,
      "memory(GiB)": 135.77,
      "step": 27050,
      "train_speed(iter/s)": 0.200892
    },
    {
      "acc": 0.77677755,
      "epoch": 0.6313302906137447,
      "grad_norm": 5.9375,
      "learning_rate": 8.008329089171192e-06,
      "loss": 0.80337753,
      "memory(GiB)": 135.77,
      "step": 27060,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.77744932,
      "epoch": 0.6315635981860336,
      "grad_norm": 5.4375,
      "learning_rate": 8.006819950492067e-06,
      "loss": 0.79293747,
      "memory(GiB)": 135.77,
      "step": 27070,
      "train_speed(iter/s)": 0.200969
    },
    {
      "acc": 0.76862764,
      "epoch": 0.6317969057583225,
      "grad_norm": 5.9375,
      "learning_rate": 8.00531038259027e-06,
      "loss": 0.82000647,
      "memory(GiB)": 135.77,
      "step": 27080,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.79620075,
      "epoch": 0.6320302133306114,
      "grad_norm": 6.09375,
      "learning_rate": 8.003800385681287e-06,
      "loss": 0.73623171,
      "memory(GiB)": 135.77,
      "step": 27090,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.77440538,
      "epoch": 0.6322635209029003,
      "grad_norm": 5.75,
      "learning_rate": 8.002289959980672e-06,
      "loss": 0.81353531,
      "memory(GiB)": 135.77,
      "step": 27100,
      "train_speed(iter/s)": 0.201087
    },
    {
      "acc": 0.77269154,
      "epoch": 0.6324968284751892,
      "grad_norm": 7.9375,
      "learning_rate": 8.000779105704037e-06,
      "loss": 0.82167168,
      "memory(GiB)": 135.77,
      "step": 27110,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.78157363,
      "epoch": 0.6327301360474781,
      "grad_norm": 6.1875,
      "learning_rate": 7.999267823067056e-06,
      "loss": 0.79839826,
      "memory(GiB)": 135.77,
      "step": 27120,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.77923985,
      "epoch": 0.632963443619767,
      "grad_norm": 7.96875,
      "learning_rate": 7.997756112285467e-06,
      "loss": 0.77553253,
      "memory(GiB)": 135.77,
      "step": 27130,
      "train_speed(iter/s)": 0.201207
    },
    {
      "acc": 0.77168446,
      "epoch": 0.6331967511920559,
      "grad_norm": 3.703125,
      "learning_rate": 7.996243973575062e-06,
      "loss": 0.82124062,
      "memory(GiB)": 135.77,
      "step": 27140,
      "train_speed(iter/s)": 0.201245
    },
    {
      "acc": 0.77455969,
      "epoch": 0.6334300587643448,
      "grad_norm": 6.1875,
      "learning_rate": 7.994731407151702e-06,
      "loss": 0.80292664,
      "memory(GiB)": 135.77,
      "step": 27150,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.782617,
      "epoch": 0.6336633663366337,
      "grad_norm": 5.03125,
      "learning_rate": 7.9932184132313e-06,
      "loss": 0.77485504,
      "memory(GiB)": 135.77,
      "step": 27160,
      "train_speed(iter/s)": 0.201316
    },
    {
      "acc": 0.75979109,
      "epoch": 0.6338966739089226,
      "grad_norm": 4.46875,
      "learning_rate": 7.99170499202984e-06,
      "loss": 0.8872653,
      "memory(GiB)": 135.77,
      "step": 27170,
      "train_speed(iter/s)": 0.201356
    },
    {
      "acc": 0.7633131,
      "epoch": 0.6341299814812115,
      "grad_norm": 6.46875,
      "learning_rate": 7.990191143763364e-06,
      "loss": 0.84359808,
      "memory(GiB)": 135.77,
      "step": 27180,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.77210131,
      "epoch": 0.6343632890535004,
      "grad_norm": 7.96875,
      "learning_rate": 7.988676868647969e-06,
      "loss": 0.83302078,
      "memory(GiB)": 135.77,
      "step": 27190,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.76380386,
      "epoch": 0.6345965966257893,
      "grad_norm": 6.25,
      "learning_rate": 7.98716216689982e-06,
      "loss": 0.85607281,
      "memory(GiB)": 135.77,
      "step": 27200,
      "train_speed(iter/s)": 0.201476
    },
    {
      "acc": 0.77592916,
      "epoch": 0.6348299041980782,
      "grad_norm": 5.0625,
      "learning_rate": 7.985647038735139e-06,
      "loss": 0.812115,
      "memory(GiB)": 135.77,
      "step": 27210,
      "train_speed(iter/s)": 0.201516
    },
    {
      "acc": 0.77886858,
      "epoch": 0.6350632117703671,
      "grad_norm": 4.71875,
      "learning_rate": 7.98413148437021e-06,
      "loss": 0.79281015,
      "memory(GiB)": 135.77,
      "step": 27220,
      "train_speed(iter/s)": 0.201551
    },
    {
      "acc": 0.78781528,
      "epoch": 0.635296519342656,
      "grad_norm": 7.28125,
      "learning_rate": 7.98261550402138e-06,
      "loss": 0.75333433,
      "memory(GiB)": 135.77,
      "step": 27230,
      "train_speed(iter/s)": 0.201589
    },
    {
      "acc": 0.781744,
      "epoch": 0.6355298269149449,
      "grad_norm": 7.21875,
      "learning_rate": 7.981099097905051e-06,
      "loss": 0.78279748,
      "memory(GiB)": 135.77,
      "step": 27240,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.76686149,
      "epoch": 0.6357631344872336,
      "grad_norm": 6.34375,
      "learning_rate": 7.979582266237695e-06,
      "loss": 0.87493477,
      "memory(GiB)": 135.77,
      "step": 27250,
      "train_speed(iter/s)": 0.201666
    },
    {
      "acc": 0.7830512,
      "epoch": 0.6359964420595225,
      "grad_norm": 4.78125,
      "learning_rate": 7.978065009235834e-06,
      "loss": 0.78735223,
      "memory(GiB)": 135.77,
      "step": 27260,
      "train_speed(iter/s)": 0.201705
    },
    {
      "acc": 0.77184772,
      "epoch": 0.6362297496318114,
      "grad_norm": 5.53125,
      "learning_rate": 7.976547327116058e-06,
      "loss": 0.80568523,
      "memory(GiB)": 135.77,
      "step": 27270,
      "train_speed(iter/s)": 0.201741
    },
    {
      "acc": 0.76909132,
      "epoch": 0.6364630572041003,
      "grad_norm": 8.375,
      "learning_rate": 7.975029220095016e-06,
      "loss": 0.81954041,
      "memory(GiB)": 135.77,
      "step": 27280,
      "train_speed(iter/s)": 0.20178
    },
    {
      "acc": 0.77826443,
      "epoch": 0.6366963647763892,
      "grad_norm": 6.03125,
      "learning_rate": 7.973510688389417e-06,
      "loss": 0.7962677,
      "memory(GiB)": 135.77,
      "step": 27290,
      "train_speed(iter/s)": 0.201818
    },
    {
      "acc": 0.77817726,
      "epoch": 0.6369296723486781,
      "grad_norm": 5.6875,
      "learning_rate": 7.971991732216032e-06,
      "loss": 0.80456171,
      "memory(GiB)": 135.77,
      "step": 27300,
      "train_speed(iter/s)": 0.201857
    },
    {
      "acc": 0.76000805,
      "epoch": 0.637162979920967,
      "grad_norm": 5.375,
      "learning_rate": 7.97047235179169e-06,
      "loss": 0.8421195,
      "memory(GiB)": 135.77,
      "step": 27310,
      "train_speed(iter/s)": 0.201896
    },
    {
      "acc": 0.78686552,
      "epoch": 0.6373962874932559,
      "grad_norm": 5.71875,
      "learning_rate": 7.968952547333281e-06,
      "loss": 0.75428095,
      "memory(GiB)": 135.77,
      "step": 27320,
      "train_speed(iter/s)": 0.201932
    },
    {
      "acc": 0.76575708,
      "epoch": 0.6376295950655448,
      "grad_norm": 6.15625,
      "learning_rate": 7.967432319057762e-06,
      "loss": 0.84427853,
      "memory(GiB)": 135.77,
      "step": 27330,
      "train_speed(iter/s)": 0.201969
    },
    {
      "acc": 0.76280413,
      "epoch": 0.6378629026378337,
      "grad_norm": 4.1875,
      "learning_rate": 7.965911667182138e-06,
      "loss": 0.85895653,
      "memory(GiB)": 135.77,
      "step": 27340,
      "train_speed(iter/s)": 0.202008
    },
    {
      "acc": 0.78277311,
      "epoch": 0.6380962102101226,
      "grad_norm": 4.34375,
      "learning_rate": 7.964390591923487e-06,
      "loss": 0.77046175,
      "memory(GiB)": 135.77,
      "step": 27350,
      "train_speed(iter/s)": 0.202046
    },
    {
      "acc": 0.76807733,
      "epoch": 0.6383295177824115,
      "grad_norm": 4.25,
      "learning_rate": 7.962869093498939e-06,
      "loss": 0.84009323,
      "memory(GiB)": 135.77,
      "step": 27360,
      "train_speed(iter/s)": 0.202083
    },
    {
      "acc": 0.78158169,
      "epoch": 0.6385628253547004,
      "grad_norm": 5.34375,
      "learning_rate": 7.961347172125689e-06,
      "loss": 0.78689456,
      "memory(GiB)": 135.77,
      "step": 27370,
      "train_speed(iter/s)": 0.202122
    },
    {
      "acc": 0.75279794,
      "epoch": 0.6387961329269893,
      "grad_norm": 5.875,
      "learning_rate": 7.959824828020991e-06,
      "loss": 0.91323185,
      "memory(GiB)": 135.77,
      "step": 27380,
      "train_speed(iter/s)": 0.202153
    },
    {
      "acc": 0.78088951,
      "epoch": 0.6390294404992782,
      "grad_norm": 6.28125,
      "learning_rate": 7.958302061402159e-06,
      "loss": 0.76333561,
      "memory(GiB)": 135.77,
      "step": 27390,
      "train_speed(iter/s)": 0.202192
    },
    {
      "acc": 0.76975918,
      "epoch": 0.6392627480715671,
      "grad_norm": 4.5,
      "learning_rate": 7.956778872486566e-06,
      "loss": 0.82985058,
      "memory(GiB)": 135.77,
      "step": 27400,
      "train_speed(iter/s)": 0.202226
    },
    {
      "acc": 0.77449579,
      "epoch": 0.639496055643856,
      "grad_norm": 6.5,
      "learning_rate": 7.955255261491648e-06,
      "loss": 0.80823078,
      "memory(GiB)": 135.77,
      "step": 27410,
      "train_speed(iter/s)": 0.202263
    },
    {
      "acc": 0.78136916,
      "epoch": 0.6397293632161449,
      "grad_norm": 5.5,
      "learning_rate": 7.9537312286349e-06,
      "loss": 0.78419118,
      "memory(GiB)": 135.77,
      "step": 27420,
      "train_speed(iter/s)": 0.202299
    },
    {
      "acc": 0.7795558,
      "epoch": 0.6399626707884338,
      "grad_norm": 5.125,
      "learning_rate": 7.952206774133878e-06,
      "loss": 0.79317288,
      "memory(GiB)": 135.77,
      "step": 27430,
      "train_speed(iter/s)": 0.202337
    },
    {
      "acc": 0.76971769,
      "epoch": 0.6401959783607226,
      "grad_norm": 7.875,
      "learning_rate": 7.950681898206197e-06,
      "loss": 0.81421986,
      "memory(GiB)": 135.77,
      "step": 27440,
      "train_speed(iter/s)": 0.202373
    },
    {
      "acc": 0.76713629,
      "epoch": 0.6404292859330115,
      "grad_norm": 7.46875,
      "learning_rate": 7.949156601069531e-06,
      "loss": 0.86134014,
      "memory(GiB)": 135.77,
      "step": 27450,
      "train_speed(iter/s)": 0.20241
    },
    {
      "acc": 0.78460703,
      "epoch": 0.6406625935053004,
      "grad_norm": 5.40625,
      "learning_rate": 7.947630882941617e-06,
      "loss": 0.7690053,
      "memory(GiB)": 135.77,
      "step": 27460,
      "train_speed(iter/s)": 0.202448
    },
    {
      "acc": 0.77735882,
      "epoch": 0.6408959010775893,
      "grad_norm": 6.96875,
      "learning_rate": 7.94610474404025e-06,
      "loss": 0.82368259,
      "memory(GiB)": 135.77,
      "step": 27470,
      "train_speed(iter/s)": 0.202484
    },
    {
      "acc": 0.77271509,
      "epoch": 0.6411292086498782,
      "grad_norm": 6.15625,
      "learning_rate": 7.944578184583289e-06,
      "loss": 0.82098379,
      "memory(GiB)": 135.77,
      "step": 27480,
      "train_speed(iter/s)": 0.202525
    },
    {
      "acc": 0.77546682,
      "epoch": 0.6413625162221671,
      "grad_norm": 4.0625,
      "learning_rate": 7.943051204788646e-06,
      "loss": 0.81826458,
      "memory(GiB)": 135.77,
      "step": 27490,
      "train_speed(iter/s)": 0.202563
    },
    {
      "acc": 0.79403472,
      "epoch": 0.641595823794456,
      "grad_norm": 5.1875,
      "learning_rate": 7.941523804874298e-06,
      "loss": 0.73600636,
      "memory(GiB)": 135.77,
      "step": 27500,
      "train_speed(iter/s)": 0.202599
    },
    {
      "epoch": 0.641595823794456,
      "eval_acc": 0.7411853011296093,
      "eval_loss": 0.818037211894989,
      "eval_runtime": 1271.0412,
      "eval_samples_per_second": 28.316,
      "eval_steps_per_second": 14.158,
      "step": 27500
    },
    {
      "acc": 0.757687,
      "epoch": 0.6418291313667449,
      "grad_norm": 6.21875,
      "learning_rate": 7.939995985058282e-06,
      "loss": 0.86413383,
      "memory(GiB)": 135.77,
      "step": 27510,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.7694766,
      "epoch": 0.6420624389390338,
      "grad_norm": 7.8125,
      "learning_rate": 7.938467745558693e-06,
      "loss": 0.82724009,
      "memory(GiB)": 135.77,
      "step": 27520,
      "train_speed(iter/s)": 0.200767
    },
    {
      "acc": 0.7876204,
      "epoch": 0.6422957465113227,
      "grad_norm": 8.8125,
      "learning_rate": 7.936939086593688e-06,
      "loss": 0.7523385,
      "memory(GiB)": 135.77,
      "step": 27530,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.81005259,
      "epoch": 0.6425290540836116,
      "grad_norm": 6.96875,
      "learning_rate": 7.935410008381482e-06,
      "loss": 0.67726383,
      "memory(GiB)": 135.77,
      "step": 27540,
      "train_speed(iter/s)": 0.200839
    },
    {
      "acc": 0.78100648,
      "epoch": 0.6427623616559005,
      "grad_norm": 5.625,
      "learning_rate": 7.933880511140349e-06,
      "loss": 0.77527313,
      "memory(GiB)": 135.77,
      "step": 27550,
      "train_speed(iter/s)": 0.200879
    },
    {
      "acc": 0.7760963,
      "epoch": 0.6429956692281894,
      "grad_norm": 5.5625,
      "learning_rate": 7.932350595088623e-06,
      "loss": 0.7997427,
      "memory(GiB)": 135.77,
      "step": 27560,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.78075814,
      "epoch": 0.6432289768004783,
      "grad_norm": 6.3125,
      "learning_rate": 7.930820260444705e-06,
      "loss": 0.78330669,
      "memory(GiB)": 135.77,
      "step": 27570,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.77887297,
      "epoch": 0.6434622843727672,
      "grad_norm": 11.3125,
      "learning_rate": 7.929289507427044e-06,
      "loss": 0.79019032,
      "memory(GiB)": 135.77,
      "step": 27580,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.78241224,
      "epoch": 0.6436955919450561,
      "grad_norm": 4.96875,
      "learning_rate": 7.927758336254156e-06,
      "loss": 0.79271832,
      "memory(GiB)": 135.77,
      "step": 27590,
      "train_speed(iter/s)": 0.201029
    },
    {
      "acc": 0.75557337,
      "epoch": 0.643928899517345,
      "grad_norm": 8.4375,
      "learning_rate": 7.926226747144618e-06,
      "loss": 0.85448341,
      "memory(GiB)": 135.77,
      "step": 27600,
      "train_speed(iter/s)": 0.201065
    },
    {
      "acc": 0.77307668,
      "epoch": 0.6441622070896339,
      "grad_norm": 5.34375,
      "learning_rate": 7.924694740317063e-06,
      "loss": 0.83668203,
      "memory(GiB)": 135.77,
      "step": 27610,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.76995897,
      "epoch": 0.6443955146619228,
      "grad_norm": 6.6875,
      "learning_rate": 7.923162315990181e-06,
      "loss": 0.82851915,
      "memory(GiB)": 135.77,
      "step": 27620,
      "train_speed(iter/s)": 0.201139
    },
    {
      "acc": 0.77117805,
      "epoch": 0.6446288222342116,
      "grad_norm": 5.65625,
      "learning_rate": 7.92162947438273e-06,
      "loss": 0.82947626,
      "memory(GiB)": 135.77,
      "step": 27630,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.76669693,
      "epoch": 0.6448621298065005,
      "grad_norm": 5.3125,
      "learning_rate": 7.920096215713518e-06,
      "loss": 0.83536053,
      "memory(GiB)": 135.77,
      "step": 27640,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.78462639,
      "epoch": 0.6450954373787894,
      "grad_norm": 6.09375,
      "learning_rate": 7.91856254020142e-06,
      "loss": 0.78443203,
      "memory(GiB)": 135.77,
      "step": 27650,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.78006172,
      "epoch": 0.6453287449510783,
      "grad_norm": 6.28125,
      "learning_rate": 7.917028448065368e-06,
      "loss": 0.8074728,
      "memory(GiB)": 135.77,
      "step": 27660,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.76845789,
      "epoch": 0.6455620525233672,
      "grad_norm": 4.1875,
      "learning_rate": 7.915493939524352e-06,
      "loss": 0.8165184,
      "memory(GiB)": 135.77,
      "step": 27670,
      "train_speed(iter/s)": 0.201316
    },
    {
      "acc": 0.78136225,
      "epoch": 0.6457953600956561,
      "grad_norm": 6.53125,
      "learning_rate": 7.913959014797424e-06,
      "loss": 0.77983022,
      "memory(GiB)": 135.77,
      "step": 27680,
      "train_speed(iter/s)": 0.201352
    },
    {
      "acc": 0.75951281,
      "epoch": 0.646028667667945,
      "grad_norm": 5.84375,
      "learning_rate": 7.91242367410369e-06,
      "loss": 0.85881014,
      "memory(GiB)": 135.77,
      "step": 27690,
      "train_speed(iter/s)": 0.20139
    },
    {
      "acc": 0.77766533,
      "epoch": 0.6462619752402339,
      "grad_norm": 5.28125,
      "learning_rate": 7.910887917662326e-06,
      "loss": 0.80150757,
      "memory(GiB)": 135.77,
      "step": 27700,
      "train_speed(iter/s)": 0.201428
    },
    {
      "acc": 0.77703729,
      "epoch": 0.6464952828125228,
      "grad_norm": 5.6875,
      "learning_rate": 7.909351745692557e-06,
      "loss": 0.83589764,
      "memory(GiB)": 135.77,
      "step": 27710,
      "train_speed(iter/s)": 0.201465
    },
    {
      "acc": 0.7554203,
      "epoch": 0.6467285903848117,
      "grad_norm": 4.6875,
      "learning_rate": 7.907815158413669e-06,
      "loss": 0.90056362,
      "memory(GiB)": 135.77,
      "step": 27720,
      "train_speed(iter/s)": 0.201502
    },
    {
      "acc": 0.77123346,
      "epoch": 0.6469618979571006,
      "grad_norm": 6.375,
      "learning_rate": 7.906278156045015e-06,
      "loss": 0.8136116,
      "memory(GiB)": 135.77,
      "step": 27730,
      "train_speed(iter/s)": 0.20154
    },
    {
      "acc": 0.78147049,
      "epoch": 0.6471952055293895,
      "grad_norm": 5.59375,
      "learning_rate": 7.904740738805996e-06,
      "loss": 0.79858246,
      "memory(GiB)": 135.77,
      "step": 27740,
      "train_speed(iter/s)": 0.201576
    },
    {
      "acc": 0.76583729,
      "epoch": 0.6474285131016784,
      "grad_norm": 19.5,
      "learning_rate": 7.90320290691608e-06,
      "loss": 0.8749052,
      "memory(GiB)": 135.77,
      "step": 27750,
      "train_speed(iter/s)": 0.201611
    },
    {
      "acc": 0.77532988,
      "epoch": 0.6476618206739673,
      "grad_norm": 5.3125,
      "learning_rate": 7.901664660594794e-06,
      "loss": 0.81616383,
      "memory(GiB)": 135.77,
      "step": 27760,
      "train_speed(iter/s)": 0.20165
    },
    {
      "acc": 0.78104067,
      "epoch": 0.6478951282462562,
      "grad_norm": 5.3125,
      "learning_rate": 7.90012600006172e-06,
      "loss": 0.76530261,
      "memory(GiB)": 135.77,
      "step": 27770,
      "train_speed(iter/s)": 0.201684
    },
    {
      "acc": 0.78051329,
      "epoch": 0.6481284358185451,
      "grad_norm": 4.1875,
      "learning_rate": 7.898586925536504e-06,
      "loss": 0.78549862,
      "memory(GiB)": 135.77,
      "step": 27780,
      "train_speed(iter/s)": 0.201719
    },
    {
      "acc": 0.77190304,
      "epoch": 0.648361743390834,
      "grad_norm": 5.1875,
      "learning_rate": 7.897047437238845e-06,
      "loss": 0.83245649,
      "memory(GiB)": 135.77,
      "step": 27790,
      "train_speed(iter/s)": 0.201756
    },
    {
      "acc": 0.76675649,
      "epoch": 0.6485950509631229,
      "grad_norm": 6.84375,
      "learning_rate": 7.895507535388506e-06,
      "loss": 0.85838509,
      "memory(GiB)": 135.77,
      "step": 27800,
      "train_speed(iter/s)": 0.201794
    },
    {
      "acc": 0.76727285,
      "epoch": 0.6488283585354118,
      "grad_norm": 4.9375,
      "learning_rate": 7.893967220205307e-06,
      "loss": 0.84388571,
      "memory(GiB)": 135.77,
      "step": 27810,
      "train_speed(iter/s)": 0.201829
    },
    {
      "acc": 0.77514563,
      "epoch": 0.6490616661077007,
      "grad_norm": 4.5625,
      "learning_rate": 7.89242649190913e-06,
      "loss": 0.80860777,
      "memory(GiB)": 135.77,
      "step": 27820,
      "train_speed(iter/s)": 0.201866
    },
    {
      "acc": 0.76715317,
      "epoch": 0.6492949736799895,
      "grad_norm": 5.53125,
      "learning_rate": 7.890885350719907e-06,
      "loss": 0.83706264,
      "memory(GiB)": 135.77,
      "step": 27830,
      "train_speed(iter/s)": 0.201903
    },
    {
      "acc": 0.77728567,
      "epoch": 0.6495282812522783,
      "grad_norm": 5.25,
      "learning_rate": 7.889343796857645e-06,
      "loss": 0.79823217,
      "memory(GiB)": 135.77,
      "step": 27840,
      "train_speed(iter/s)": 0.201939
    },
    {
      "acc": 0.78821392,
      "epoch": 0.6497615888245672,
      "grad_norm": 6.71875,
      "learning_rate": 7.887801830542392e-06,
      "loss": 0.75157042,
      "memory(GiB)": 135.77,
      "step": 27850,
      "train_speed(iter/s)": 0.201976
    },
    {
      "acc": 0.7779901,
      "epoch": 0.6499948963968561,
      "grad_norm": 4.84375,
      "learning_rate": 7.886259451994267e-06,
      "loss": 0.79298458,
      "memory(GiB)": 135.77,
      "step": 27860,
      "train_speed(iter/s)": 0.202013
    },
    {
      "acc": 0.76499109,
      "epoch": 0.650228203969145,
      "grad_norm": 5.25,
      "learning_rate": 7.884716661433444e-06,
      "loss": 0.87161884,
      "memory(GiB)": 135.77,
      "step": 27870,
      "train_speed(iter/s)": 0.202049
    },
    {
      "acc": 0.7676559,
      "epoch": 0.650461511541434,
      "grad_norm": 6.40625,
      "learning_rate": 7.883173459080159e-06,
      "loss": 0.83052654,
      "memory(GiB)": 135.77,
      "step": 27880,
      "train_speed(iter/s)": 0.202088
    },
    {
      "acc": 0.78522291,
      "epoch": 0.6506948191137228,
      "grad_norm": 5.46875,
      "learning_rate": 7.881629845154696e-06,
      "loss": 0.77097383,
      "memory(GiB)": 135.77,
      "step": 27890,
      "train_speed(iter/s)": 0.202123
    },
    {
      "acc": 0.76635356,
      "epoch": 0.6509281266860117,
      "grad_norm": 4.8125,
      "learning_rate": 7.880085819877411e-06,
      "loss": 0.84650917,
      "memory(GiB)": 135.77,
      "step": 27900,
      "train_speed(iter/s)": 0.202162
    },
    {
      "acc": 0.76490736,
      "epoch": 0.6511614342583006,
      "grad_norm": 4.71875,
      "learning_rate": 7.878541383468712e-06,
      "loss": 0.8467804,
      "memory(GiB)": 135.77,
      "step": 27910,
      "train_speed(iter/s)": 0.202199
    },
    {
      "acc": 0.76074095,
      "epoch": 0.6513947418305895,
      "grad_norm": 5.03125,
      "learning_rate": 7.876996536149067e-06,
      "loss": 0.85577497,
      "memory(GiB)": 135.77,
      "step": 27920,
      "train_speed(iter/s)": 0.202233
    },
    {
      "acc": 0.79905081,
      "epoch": 0.6516280494028784,
      "grad_norm": 5.9375,
      "learning_rate": 7.875451278139001e-06,
      "loss": 0.71797628,
      "memory(GiB)": 135.77,
      "step": 27930,
      "train_speed(iter/s)": 0.202268
    },
    {
      "acc": 0.77437353,
      "epoch": 0.6518613569751673,
      "grad_norm": 6.6875,
      "learning_rate": 7.873905609659102e-06,
      "loss": 0.80220547,
      "memory(GiB)": 135.77,
      "step": 27940,
      "train_speed(iter/s)": 0.202306
    },
    {
      "acc": 0.78056316,
      "epoch": 0.6520946645474562,
      "grad_norm": 5.25,
      "learning_rate": 7.872359530930011e-06,
      "loss": 0.76737919,
      "memory(GiB)": 135.77,
      "step": 27950,
      "train_speed(iter/s)": 0.202343
    },
    {
      "acc": 0.77328043,
      "epoch": 0.6523279721197451,
      "grad_norm": 6.5,
      "learning_rate": 7.870813042172432e-06,
      "loss": 0.8179678,
      "memory(GiB)": 135.77,
      "step": 27960,
      "train_speed(iter/s)": 0.202381
    },
    {
      "acc": 0.784305,
      "epoch": 0.652561279692034,
      "grad_norm": 4.0,
      "learning_rate": 7.869266143607124e-06,
      "loss": 0.76675787,
      "memory(GiB)": 135.77,
      "step": 27970,
      "train_speed(iter/s)": 0.202421
    },
    {
      "acc": 0.77683382,
      "epoch": 0.6527945872643229,
      "grad_norm": 4.6875,
      "learning_rate": 7.86771883545491e-06,
      "loss": 0.80360317,
      "memory(GiB)": 135.77,
      "step": 27980,
      "train_speed(iter/s)": 0.202459
    },
    {
      "acc": 0.76954408,
      "epoch": 0.6530278948366118,
      "grad_norm": 6.28125,
      "learning_rate": 7.866171117936663e-06,
      "loss": 0.83105335,
      "memory(GiB)": 135.77,
      "step": 27990,
      "train_speed(iter/s)": 0.202497
    },
    {
      "acc": 0.76160703,
      "epoch": 0.6532612024089007,
      "grad_norm": 6.3125,
      "learning_rate": 7.864622991273322e-06,
      "loss": 0.8450861,
      "memory(GiB)": 135.77,
      "step": 28000,
      "train_speed(iter/s)": 0.202537
    },
    {
      "epoch": 0.6532612024089007,
      "eval_acc": 0.7410069004548256,
      "eval_loss": 0.8178855180740356,
      "eval_runtime": 1271.7089,
      "eval_samples_per_second": 28.301,
      "eval_steps_per_second": 14.151,
      "step": 28000
    },
    {
      "acc": 0.77182312,
      "epoch": 0.6534945099811896,
      "grad_norm": 8.375,
      "learning_rate": 7.863074455685882e-06,
      "loss": 0.83168869,
      "memory(GiB)": 135.77,
      "step": 28010,
      "train_speed(iter/s)": 0.200706
    },
    {
      "acc": 0.78294449,
      "epoch": 0.6537278175534784,
      "grad_norm": 4.375,
      "learning_rate": 7.861525511395394e-06,
      "loss": 0.76020308,
      "memory(GiB)": 135.77,
      "step": 28020,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.76998968,
      "epoch": 0.6539611251257673,
      "grad_norm": 4.71875,
      "learning_rate": 7.859976158622971e-06,
      "loss": 0.82431984,
      "memory(GiB)": 135.77,
      "step": 28030,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.75824432,
      "epoch": 0.6541944326980562,
      "grad_norm": 6.0625,
      "learning_rate": 7.858426397589783e-06,
      "loss": 0.87359171,
      "memory(GiB)": 135.77,
      "step": 28040,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.76430311,
      "epoch": 0.6544277402703451,
      "grad_norm": 10.0,
      "learning_rate": 7.856876228517057e-06,
      "loss": 0.84922371,
      "memory(GiB)": 135.77,
      "step": 28050,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.7763073,
      "epoch": 0.654661047842634,
      "grad_norm": 5.21875,
      "learning_rate": 7.85532565162608e-06,
      "loss": 0.81707954,
      "memory(GiB)": 135.77,
      "step": 28060,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.78287649,
      "epoch": 0.6548943554149229,
      "grad_norm": 3.984375,
      "learning_rate": 7.853774667138192e-06,
      "loss": 0.76915636,
      "memory(GiB)": 135.77,
      "step": 28070,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.78447509,
      "epoch": 0.6551276629872118,
      "grad_norm": 4.34375,
      "learning_rate": 7.852223275274804e-06,
      "loss": 0.77548103,
      "memory(GiB)": 135.77,
      "step": 28080,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.77808089,
      "epoch": 0.6553609705595007,
      "grad_norm": 4.625,
      "learning_rate": 7.85067147625737e-06,
      "loss": 0.79100828,
      "memory(GiB)": 135.77,
      "step": 28090,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.776159,
      "epoch": 0.6555942781317896,
      "grad_norm": 8.3125,
      "learning_rate": 7.84911927030741e-06,
      "loss": 0.79818797,
      "memory(GiB)": 135.77,
      "step": 28100,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.76561847,
      "epoch": 0.6558275857040785,
      "grad_norm": 3.890625,
      "learning_rate": 7.847566657646502e-06,
      "loss": 0.84216652,
      "memory(GiB)": 135.77,
      "step": 28110,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.77506824,
      "epoch": 0.6560608932763674,
      "grad_norm": 5.34375,
      "learning_rate": 7.846013638496281e-06,
      "loss": 0.82243309,
      "memory(GiB)": 135.77,
      "step": 28120,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.78300853,
      "epoch": 0.6562942008486563,
      "grad_norm": 4.75,
      "learning_rate": 7.84446021307844e-06,
      "loss": 0.76296067,
      "memory(GiB)": 135.77,
      "step": 28130,
      "train_speed(iter/s)": 0.20116
    },
    {
      "acc": 0.75912266,
      "epoch": 0.6565275084209452,
      "grad_norm": 4.40625,
      "learning_rate": 7.842906381614732e-06,
      "loss": 0.84529533,
      "memory(GiB)": 135.77,
      "step": 28140,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.78884659,
      "epoch": 0.6567608159932341,
      "grad_norm": 6.75,
      "learning_rate": 7.841352144326962e-06,
      "loss": 0.76649256,
      "memory(GiB)": 135.77,
      "step": 28150,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.78107457,
      "epoch": 0.656994123565523,
      "grad_norm": 3.78125,
      "learning_rate": 7.839797501436999e-06,
      "loss": 0.7952775,
      "memory(GiB)": 135.77,
      "step": 28160,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.7618453,
      "epoch": 0.6572274311378119,
      "grad_norm": 5.28125,
      "learning_rate": 7.838242453166766e-06,
      "loss": 0.86572094,
      "memory(GiB)": 135.77,
      "step": 28170,
      "train_speed(iter/s)": 0.201305
    },
    {
      "acc": 0.77299466,
      "epoch": 0.6574607387101008,
      "grad_norm": 6.59375,
      "learning_rate": 7.83668699973825e-06,
      "loss": 0.83398046,
      "memory(GiB)": 135.77,
      "step": 28180,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.77326827,
      "epoch": 0.6576940462823897,
      "grad_norm": 5.3125,
      "learning_rate": 7.835131141373487e-06,
      "loss": 0.80730629,
      "memory(GiB)": 135.77,
      "step": 28190,
      "train_speed(iter/s)": 0.201379
    },
    {
      "acc": 0.78218737,
      "epoch": 0.6579273538546786,
      "grad_norm": 5.65625,
      "learning_rate": 7.833574878294578e-06,
      "loss": 0.79669929,
      "memory(GiB)": 135.77,
      "step": 28200,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.76740227,
      "epoch": 0.6581606614269674,
      "grad_norm": 5.6875,
      "learning_rate": 7.832018210723679e-06,
      "loss": 0.84285278,
      "memory(GiB)": 135.77,
      "step": 28210,
      "train_speed(iter/s)": 0.201447
    },
    {
      "acc": 0.76748447,
      "epoch": 0.6583939689992563,
      "grad_norm": 4.46875,
      "learning_rate": 7.830461138883e-06,
      "loss": 0.84718819,
      "memory(GiB)": 135.77,
      "step": 28220,
      "train_speed(iter/s)": 0.201482
    },
    {
      "acc": 0.77524757,
      "epoch": 0.6586272765715452,
      "grad_norm": 6.375,
      "learning_rate": 7.82890366299482e-06,
      "loss": 0.81450605,
      "memory(GiB)": 135.77,
      "step": 28230,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.76804585,
      "epoch": 0.6588605841438341,
      "grad_norm": 4.8125,
      "learning_rate": 7.827345783281462e-06,
      "loss": 0.82887039,
      "memory(GiB)": 135.77,
      "step": 28240,
      "train_speed(iter/s)": 0.201558
    },
    {
      "acc": 0.77396336,
      "epoch": 0.659093891716123,
      "grad_norm": 5.5625,
      "learning_rate": 7.825787499965315e-06,
      "loss": 0.83359642,
      "memory(GiB)": 135.77,
      "step": 28250,
      "train_speed(iter/s)": 0.201595
    },
    {
      "acc": 0.74629498,
      "epoch": 0.6593271992884119,
      "grad_norm": 5.96875,
      "learning_rate": 7.824228813268823e-06,
      "loss": 0.92181091,
      "memory(GiB)": 135.77,
      "step": 28260,
      "train_speed(iter/s)": 0.20163
    },
    {
      "acc": 0.78333774,
      "epoch": 0.6595605068607008,
      "grad_norm": 4.71875,
      "learning_rate": 7.822669723414488e-06,
      "loss": 0.78887167,
      "memory(GiB)": 135.77,
      "step": 28270,
      "train_speed(iter/s)": 0.201667
    },
    {
      "acc": 0.75382862,
      "epoch": 0.6597938144329897,
      "grad_norm": 3.953125,
      "learning_rate": 7.82111023062487e-06,
      "loss": 0.89491863,
      "memory(GiB)": 135.77,
      "step": 28280,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.76489215,
      "epoch": 0.6600271220052786,
      "grad_norm": 5.5625,
      "learning_rate": 7.819550335122587e-06,
      "loss": 0.86177435,
      "memory(GiB)": 135.77,
      "step": 28290,
      "train_speed(iter/s)": 0.201738
    },
    {
      "acc": 0.76324902,
      "epoch": 0.6602604295775675,
      "grad_norm": 5.25,
      "learning_rate": 7.817990037130312e-06,
      "loss": 0.85025234,
      "memory(GiB)": 135.77,
      "step": 28300,
      "train_speed(iter/s)": 0.201777
    },
    {
      "acc": 0.77776289,
      "epoch": 0.6604937371498564,
      "grad_norm": 6.0,
      "learning_rate": 7.816429336870778e-06,
      "loss": 0.80502405,
      "memory(GiB)": 135.77,
      "step": 28310,
      "train_speed(iter/s)": 0.201815
    },
    {
      "acc": 0.78841567,
      "epoch": 0.6607270447221453,
      "grad_norm": 10.3125,
      "learning_rate": 7.814868234566775e-06,
      "loss": 0.72700367,
      "memory(GiB)": 135.77,
      "step": 28320,
      "train_speed(iter/s)": 0.201851
    },
    {
      "acc": 0.7645155,
      "epoch": 0.6609603522944342,
      "grad_norm": 4.75,
      "learning_rate": 7.813306730441147e-06,
      "loss": 0.84438887,
      "memory(GiB)": 135.77,
      "step": 28330,
      "train_speed(iter/s)": 0.20189
    },
    {
      "acc": 0.77188535,
      "epoch": 0.6611936598667231,
      "grad_norm": 4.9375,
      "learning_rate": 7.811744824716803e-06,
      "loss": 0.82249565,
      "memory(GiB)": 135.77,
      "step": 28340,
      "train_speed(iter/s)": 0.201927
    },
    {
      "acc": 0.79082394,
      "epoch": 0.661426967439012,
      "grad_norm": 6.4375,
      "learning_rate": 7.810182517616702e-06,
      "loss": 0.73457441,
      "memory(GiB)": 135.77,
      "step": 28350,
      "train_speed(iter/s)": 0.201965
    },
    {
      "acc": 0.77906938,
      "epoch": 0.6616602750113009,
      "grad_norm": 5.625,
      "learning_rate": 7.808619809363863e-06,
      "loss": 0.78538847,
      "memory(GiB)": 135.77,
      "step": 28360,
      "train_speed(iter/s)": 0.202
    },
    {
      "acc": 0.78205838,
      "epoch": 0.6618935825835898,
      "grad_norm": 9.25,
      "learning_rate": 7.80705670018136e-06,
      "loss": 0.79499912,
      "memory(GiB)": 135.77,
      "step": 28370,
      "train_speed(iter/s)": 0.202036
    },
    {
      "acc": 0.78381143,
      "epoch": 0.6621268901558787,
      "grad_norm": 5.96875,
      "learning_rate": 7.805493190292327e-06,
      "loss": 0.78962469,
      "memory(GiB)": 135.77,
      "step": 28380,
      "train_speed(iter/s)": 0.202067
    },
    {
      "acc": 0.77339664,
      "epoch": 0.6623601977281676,
      "grad_norm": 4.15625,
      "learning_rate": 7.80392927991996e-06,
      "loss": 0.79976535,
      "memory(GiB)": 135.77,
      "step": 28390,
      "train_speed(iter/s)": 0.202101
    },
    {
      "acc": 0.77922812,
      "epoch": 0.6625935053004564,
      "grad_norm": 5.25,
      "learning_rate": 7.802364969287501e-06,
      "loss": 0.77621737,
      "memory(GiB)": 135.77,
      "step": 28400,
      "train_speed(iter/s)": 0.202139
    },
    {
      "acc": 0.7975256,
      "epoch": 0.6628268128727453,
      "grad_norm": 5.1875,
      "learning_rate": 7.80080025861826e-06,
      "loss": 0.7247571,
      "memory(GiB)": 135.77,
      "step": 28410,
      "train_speed(iter/s)": 0.202175
    },
    {
      "acc": 0.79595299,
      "epoch": 0.6630601204450342,
      "grad_norm": 5.4375,
      "learning_rate": 7.799235148135592e-06,
      "loss": 0.71732335,
      "memory(GiB)": 135.77,
      "step": 28420,
      "train_speed(iter/s)": 0.20221
    },
    {
      "acc": 0.77594852,
      "epoch": 0.663293428017323,
      "grad_norm": 5.84375,
      "learning_rate": 7.797669638062921e-06,
      "loss": 0.79501762,
      "memory(GiB)": 135.77,
      "step": 28430,
      "train_speed(iter/s)": 0.202245
    },
    {
      "acc": 0.77370367,
      "epoch": 0.663526735589612,
      "grad_norm": 8.25,
      "learning_rate": 7.796103728623723e-06,
      "loss": 0.82125921,
      "memory(GiB)": 135.77,
      "step": 28440,
      "train_speed(iter/s)": 0.202281
    },
    {
      "acc": 0.78018131,
      "epoch": 0.6637600431619008,
      "grad_norm": 5.53125,
      "learning_rate": 7.794537420041527e-06,
      "loss": 0.80163622,
      "memory(GiB)": 135.77,
      "step": 28450,
      "train_speed(iter/s)": 0.202317
    },
    {
      "acc": 0.77699895,
      "epoch": 0.6639933507341897,
      "grad_norm": 7.96875,
      "learning_rate": 7.792970712539929e-06,
      "loss": 0.81302996,
      "memory(GiB)": 135.77,
      "step": 28460,
      "train_speed(iter/s)": 0.202354
    },
    {
      "acc": 0.78811188,
      "epoch": 0.6642266583064786,
      "grad_norm": 5.0625,
      "learning_rate": 7.791403606342572e-06,
      "loss": 0.77973518,
      "memory(GiB)": 135.77,
      "step": 28470,
      "train_speed(iter/s)": 0.202388
    },
    {
      "acc": 0.77782664,
      "epoch": 0.6644599658787675,
      "grad_norm": 7.90625,
      "learning_rate": 7.78983610167316e-06,
      "loss": 0.79536843,
      "memory(GiB)": 135.77,
      "step": 28480,
      "train_speed(iter/s)": 0.202424
    },
    {
      "acc": 0.78967228,
      "epoch": 0.6646932734510564,
      "grad_norm": 7.125,
      "learning_rate": 7.788268198755456e-06,
      "loss": 0.7731019,
      "memory(GiB)": 135.77,
      "step": 28490,
      "train_speed(iter/s)": 0.202461
    },
    {
      "acc": 0.76639204,
      "epoch": 0.6649265810233453,
      "grad_norm": 6.96875,
      "learning_rate": 7.786699897813277e-06,
      "loss": 0.83777695,
      "memory(GiB)": 135.77,
      "step": 28500,
      "train_speed(iter/s)": 0.202497
    },
    {
      "epoch": 0.6649265810233453,
      "eval_acc": 0.7410190714164805,
      "eval_loss": 0.8165611028671265,
      "eval_runtime": 1270.5641,
      "eval_samples_per_second": 28.327,
      "eval_steps_per_second": 14.164,
      "step": 28500
    },
    {
      "acc": 0.75553088,
      "epoch": 0.6651598885956342,
      "grad_norm": 9.125,
      "learning_rate": 7.785131199070497e-06,
      "loss": 0.88834152,
      "memory(GiB)": 135.77,
      "step": 28510,
      "train_speed(iter/s)": 0.200698
    },
    {
      "acc": 0.76838694,
      "epoch": 0.6653931961679231,
      "grad_norm": 3.9375,
      "learning_rate": 7.783562102751048e-06,
      "loss": 0.82640839,
      "memory(GiB)": 135.77,
      "step": 28520,
      "train_speed(iter/s)": 0.200734
    },
    {
      "acc": 0.78102646,
      "epoch": 0.665626503740212,
      "grad_norm": 7.0,
      "learning_rate": 7.781992609078916e-06,
      "loss": 0.77242785,
      "memory(GiB)": 135.77,
      "step": 28530,
      "train_speed(iter/s)": 0.200769
    },
    {
      "acc": 0.78945589,
      "epoch": 0.6658598113125009,
      "grad_norm": 7.03125,
      "learning_rate": 7.780422718278148e-06,
      "loss": 0.77674723,
      "memory(GiB)": 135.77,
      "step": 28540,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.75834866,
      "epoch": 0.6660931188847898,
      "grad_norm": 5.5,
      "learning_rate": 7.778852430572846e-06,
      "loss": 0.89787149,
      "memory(GiB)": 135.77,
      "step": 28550,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.7631052,
      "epoch": 0.6663264264570787,
      "grad_norm": 6.75,
      "learning_rate": 7.777281746187163e-06,
      "loss": 0.86248188,
      "memory(GiB)": 135.77,
      "step": 28560,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.78461332,
      "epoch": 0.6665597340293676,
      "grad_norm": 6.25,
      "learning_rate": 7.775710665345322e-06,
      "loss": 0.78493948,
      "memory(GiB)": 135.77,
      "step": 28570,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.75799332,
      "epoch": 0.6667930416016565,
      "grad_norm": 4.71875,
      "learning_rate": 7.774139188271588e-06,
      "loss": 0.86719799,
      "memory(GiB)": 135.77,
      "step": 28580,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78168793,
      "epoch": 0.6670263491739454,
      "grad_norm": 5.59375,
      "learning_rate": 7.772567315190291e-06,
      "loss": 0.79164333,
      "memory(GiB)": 135.77,
      "step": 28590,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.76213493,
      "epoch": 0.6672596567462342,
      "grad_norm": 4.65625,
      "learning_rate": 7.770995046325813e-06,
      "loss": 0.86090012,
      "memory(GiB)": 135.77,
      "step": 28600,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.76110024,
      "epoch": 0.6674929643185231,
      "grad_norm": 5.53125,
      "learning_rate": 7.769422381902601e-06,
      "loss": 0.84771814,
      "memory(GiB)": 135.77,
      "step": 28610,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.78071322,
      "epoch": 0.667726271890812,
      "grad_norm": 5.8125,
      "learning_rate": 7.767849322145144e-06,
      "loss": 0.79260278,
      "memory(GiB)": 135.77,
      "step": 28620,
      "train_speed(iter/s)": 0.201098
    },
    {
      "acc": 0.76000824,
      "epoch": 0.6679595794631009,
      "grad_norm": 5.96875,
      "learning_rate": 7.766275867278004e-06,
      "loss": 0.8921133,
      "memory(GiB)": 135.77,
      "step": 28630,
      "train_speed(iter/s)": 0.201133
    },
    {
      "acc": 0.77165904,
      "epoch": 0.6681928870353898,
      "grad_norm": 7.3125,
      "learning_rate": 7.764702017525787e-06,
      "loss": 0.80725384,
      "memory(GiB)": 135.77,
      "step": 28640,
      "train_speed(iter/s)": 0.201167
    },
    {
      "acc": 0.77876086,
      "epoch": 0.6684261946076787,
      "grad_norm": 5.15625,
      "learning_rate": 7.763127773113159e-06,
      "loss": 0.79236784,
      "memory(GiB)": 135.77,
      "step": 28650,
      "train_speed(iter/s)": 0.201201
    },
    {
      "acc": 0.79020371,
      "epoch": 0.6686595021799676,
      "grad_norm": 8.625,
      "learning_rate": 7.761553134264844e-06,
      "loss": 0.75915909,
      "memory(GiB)": 135.77,
      "step": 28660,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.76552114,
      "epoch": 0.6688928097522565,
      "grad_norm": 5.90625,
      "learning_rate": 7.759978101205623e-06,
      "loss": 0.84368448,
      "memory(GiB)": 135.77,
      "step": 28670,
      "train_speed(iter/s)": 0.201275
    },
    {
      "acc": 0.78740005,
      "epoch": 0.6691261173245454,
      "grad_norm": 5.71875,
      "learning_rate": 7.758402674160328e-06,
      "loss": 0.75369263,
      "memory(GiB)": 135.77,
      "step": 28680,
      "train_speed(iter/s)": 0.201313
    },
    {
      "acc": 0.76241765,
      "epoch": 0.6693594248968343,
      "grad_norm": 4.96875,
      "learning_rate": 7.756826853353854e-06,
      "loss": 0.85766029,
      "memory(GiB)": 135.77,
      "step": 28690,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.76723084,
      "epoch": 0.6695927324691232,
      "grad_norm": 5.625,
      "learning_rate": 7.755250639011147e-06,
      "loss": 0.84183302,
      "memory(GiB)": 135.77,
      "step": 28700,
      "train_speed(iter/s)": 0.201389
    },
    {
      "acc": 0.77271857,
      "epoch": 0.6698260400414121,
      "grad_norm": 4.65625,
      "learning_rate": 7.75367403135721e-06,
      "loss": 0.80901413,
      "memory(GiB)": 135.77,
      "step": 28710,
      "train_speed(iter/s)": 0.201427
    },
    {
      "acc": 0.75930796,
      "epoch": 0.670059347613701,
      "grad_norm": 6.78125,
      "learning_rate": 7.752097030617107e-06,
      "loss": 0.87077732,
      "memory(GiB)": 135.77,
      "step": 28720,
      "train_speed(iter/s)": 0.201465
    },
    {
      "acc": 0.77311296,
      "epoch": 0.6702926551859899,
      "grad_norm": 7.84375,
      "learning_rate": 7.750519637015953e-06,
      "loss": 0.80176973,
      "memory(GiB)": 135.77,
      "step": 28730,
      "train_speed(iter/s)": 0.2015
    },
    {
      "acc": 0.78030658,
      "epoch": 0.6705259627582788,
      "grad_norm": 4.78125,
      "learning_rate": 7.748941850778917e-06,
      "loss": 0.774928,
      "memory(GiB)": 135.77,
      "step": 28740,
      "train_speed(iter/s)": 0.201535
    },
    {
      "acc": 0.76701117,
      "epoch": 0.6707592703305677,
      "grad_norm": 5.625,
      "learning_rate": 7.747363672131233e-06,
      "loss": 0.83319283,
      "memory(GiB)": 135.77,
      "step": 28750,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.78679342,
      "epoch": 0.6709925779028566,
      "grad_norm": 5.75,
      "learning_rate": 7.745785101298182e-06,
      "loss": 0.75664225,
      "memory(GiB)": 135.77,
      "step": 28760,
      "train_speed(iter/s)": 0.201606
    },
    {
      "acc": 0.76925483,
      "epoch": 0.6712258854751455,
      "grad_norm": 7.3125,
      "learning_rate": 7.744206138505106e-06,
      "loss": 0.83768616,
      "memory(GiB)": 135.77,
      "step": 28770,
      "train_speed(iter/s)": 0.201642
    },
    {
      "acc": 0.80132675,
      "epoch": 0.6714591930474344,
      "grad_norm": 5.6875,
      "learning_rate": 7.7426267839774e-06,
      "loss": 0.70183382,
      "memory(GiB)": 135.77,
      "step": 28780,
      "train_speed(iter/s)": 0.201677
    },
    {
      "acc": 0.77842226,
      "epoch": 0.6716925006197232,
      "grad_norm": 6.09375,
      "learning_rate": 7.741047037940516e-06,
      "loss": 0.79888177,
      "memory(GiB)": 135.77,
      "step": 28790,
      "train_speed(iter/s)": 0.201714
    },
    {
      "acc": 0.77139082,
      "epoch": 0.6719258081920121,
      "grad_norm": 5.15625,
      "learning_rate": 7.739466900619966e-06,
      "loss": 0.81673717,
      "memory(GiB)": 135.77,
      "step": 28800,
      "train_speed(iter/s)": 0.201753
    },
    {
      "acc": 0.76923008,
      "epoch": 0.672159115764301,
      "grad_norm": 8.4375,
      "learning_rate": 7.737886372241311e-06,
      "loss": 0.84230728,
      "memory(GiB)": 135.77,
      "step": 28810,
      "train_speed(iter/s)": 0.201789
    },
    {
      "acc": 0.75902824,
      "epoch": 0.6723924233365899,
      "grad_norm": 8.0625,
      "learning_rate": 7.736305453030172e-06,
      "loss": 0.86554337,
      "memory(GiB)": 135.77,
      "step": 28820,
      "train_speed(iter/s)": 0.201823
    },
    {
      "acc": 0.78267469,
      "epoch": 0.6726257309088788,
      "grad_norm": 9.0625,
      "learning_rate": 7.734724143212224e-06,
      "loss": 0.79381328,
      "memory(GiB)": 135.77,
      "step": 28830,
      "train_speed(iter/s)": 0.201857
    },
    {
      "acc": 0.77059851,
      "epoch": 0.6728590384811677,
      "grad_norm": 5.8125,
      "learning_rate": 7.733142443013199e-06,
      "loss": 0.82365742,
      "memory(GiB)": 135.77,
      "step": 28840,
      "train_speed(iter/s)": 0.201894
    },
    {
      "acc": 0.7740037,
      "epoch": 0.6730923460534566,
      "grad_norm": 5.6875,
      "learning_rate": 7.731560352658886e-06,
      "loss": 0.81236382,
      "memory(GiB)": 135.77,
      "step": 28850,
      "train_speed(iter/s)": 0.20193
    },
    {
      "acc": 0.77031889,
      "epoch": 0.6733256536257455,
      "grad_norm": 8.0,
      "learning_rate": 7.729977872375125e-06,
      "loss": 0.83344927,
      "memory(GiB)": 135.77,
      "step": 28860,
      "train_speed(iter/s)": 0.201964
    },
    {
      "acc": 0.77817774,
      "epoch": 0.6735589611980344,
      "grad_norm": 5.25,
      "learning_rate": 7.728395002387815e-06,
      "loss": 0.79905758,
      "memory(GiB)": 135.77,
      "step": 28870,
      "train_speed(iter/s)": 0.201999
    },
    {
      "acc": 0.76579313,
      "epoch": 0.6737922687703233,
      "grad_norm": 5.4375,
      "learning_rate": 7.726811742922912e-06,
      "loss": 0.84702206,
      "memory(GiB)": 135.77,
      "step": 28880,
      "train_speed(iter/s)": 0.202036
    },
    {
      "acc": 0.77942319,
      "epoch": 0.6740255763426122,
      "grad_norm": 4.8125,
      "learning_rate": 7.725228094206423e-06,
      "loss": 0.78971949,
      "memory(GiB)": 135.77,
      "step": 28890,
      "train_speed(iter/s)": 0.202071
    },
    {
      "acc": 0.76637468,
      "epoch": 0.6742588839149011,
      "grad_norm": 4.9375,
      "learning_rate": 7.723644056464416e-06,
      "loss": 0.83664494,
      "memory(GiB)": 135.77,
      "step": 28900,
      "train_speed(iter/s)": 0.202105
    },
    {
      "acc": 0.77963085,
      "epoch": 0.67449219148719,
      "grad_norm": 4.46875,
      "learning_rate": 7.722059629923014e-06,
      "loss": 0.78608251,
      "memory(GiB)": 135.77,
      "step": 28910,
      "train_speed(iter/s)": 0.202138
    },
    {
      "acc": 0.78031402,
      "epoch": 0.6747254990594789,
      "grad_norm": 5.09375,
      "learning_rate": 7.720474814808387e-06,
      "loss": 0.77677441,
      "memory(GiB)": 135.77,
      "step": 28920,
      "train_speed(iter/s)": 0.202173
    },
    {
      "acc": 0.76079268,
      "epoch": 0.6749588066317678,
      "grad_norm": 5.375,
      "learning_rate": 7.718889611346771e-06,
      "loss": 0.8742177,
      "memory(GiB)": 135.77,
      "step": 28930,
      "train_speed(iter/s)": 0.202208
    },
    {
      "acc": 0.7644505,
      "epoch": 0.6751921142040567,
      "grad_norm": 5.9375,
      "learning_rate": 7.717304019764456e-06,
      "loss": 0.86214409,
      "memory(GiB)": 135.77,
      "step": 28940,
      "train_speed(iter/s)": 0.202241
    },
    {
      "acc": 0.77155561,
      "epoch": 0.6754254217763456,
      "grad_norm": 4.4375,
      "learning_rate": 7.71571804028778e-06,
      "loss": 0.81877155,
      "memory(GiB)": 135.77,
      "step": 28950,
      "train_speed(iter/s)": 0.202275
    },
    {
      "acc": 0.77610378,
      "epoch": 0.6756587293486345,
      "grad_norm": 5.84375,
      "learning_rate": 7.714131673143139e-06,
      "loss": 0.80574827,
      "memory(GiB)": 135.77,
      "step": 28960,
      "train_speed(iter/s)": 0.202311
    },
    {
      "acc": 0.77921677,
      "epoch": 0.6758920369209234,
      "grad_norm": 7.875,
      "learning_rate": 7.712544918556994e-06,
      "loss": 0.77683654,
      "memory(GiB)": 135.77,
      "step": 28970,
      "train_speed(iter/s)": 0.202347
    },
    {
      "acc": 0.76622157,
      "epoch": 0.6761253444932122,
      "grad_norm": 6.09375,
      "learning_rate": 7.71095777675585e-06,
      "loss": 0.87221556,
      "memory(GiB)": 135.77,
      "step": 28980,
      "train_speed(iter/s)": 0.202383
    },
    {
      "acc": 0.79096737,
      "epoch": 0.676358652065501,
      "grad_norm": 6.09375,
      "learning_rate": 7.709370247966269e-06,
      "loss": 0.7472538,
      "memory(GiB)": 135.77,
      "step": 28990,
      "train_speed(iter/s)": 0.202418
    },
    {
      "acc": 0.77579155,
      "epoch": 0.67659195963779,
      "grad_norm": 6.75,
      "learning_rate": 7.707782332414873e-06,
      "loss": 0.80246773,
      "memory(GiB)": 135.77,
      "step": 29000,
      "train_speed(iter/s)": 0.202455
    },
    {
      "epoch": 0.67659195963779,
      "eval_acc": 0.7414119052183158,
      "eval_loss": 0.8165162205696106,
      "eval_runtime": 1270.8006,
      "eval_samples_per_second": 28.322,
      "eval_steps_per_second": 14.161,
      "step": 29000
    },
    {
      "acc": 0.76836233,
      "epoch": 0.6768252672100789,
      "grad_norm": 6.34375,
      "learning_rate": 7.706194030328336e-06,
      "loss": 0.81320305,
      "memory(GiB)": 135.77,
      "step": 29010,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.77826796,
      "epoch": 0.6770585747823678,
      "grad_norm": 4.1875,
      "learning_rate": 7.704605341933385e-06,
      "loss": 0.79997253,
      "memory(GiB)": 135.77,
      "step": 29020,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.76980276,
      "epoch": 0.6772918823546566,
      "grad_norm": 5.34375,
      "learning_rate": 7.70301626745681e-06,
      "loss": 0.82670012,
      "memory(GiB)": 135.77,
      "step": 29030,
      "train_speed(iter/s)": 0.200759
    },
    {
      "acc": 0.7695508,
      "epoch": 0.6775251899269455,
      "grad_norm": 7.53125,
      "learning_rate": 7.701426807125447e-06,
      "loss": 0.83088665,
      "memory(GiB)": 135.77,
      "step": 29040,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.77592602,
      "epoch": 0.6777584974992344,
      "grad_norm": 5.5,
      "learning_rate": 7.699836961166192e-06,
      "loss": 0.82842121,
      "memory(GiB)": 135.77,
      "step": 29050,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.78777199,
      "epoch": 0.6779918050715233,
      "grad_norm": 4.9375,
      "learning_rate": 7.698246729805996e-06,
      "loss": 0.76934967,
      "memory(GiB)": 135.77,
      "step": 29060,
      "train_speed(iter/s)": 0.200867
    },
    {
      "acc": 0.7728559,
      "epoch": 0.6782251126438122,
      "grad_norm": 4.9375,
      "learning_rate": 7.696656113271863e-06,
      "loss": 0.80948372,
      "memory(GiB)": 135.77,
      "step": 29070,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.79076042,
      "epoch": 0.6784584202161011,
      "grad_norm": 6.25,
      "learning_rate": 7.695065111790852e-06,
      "loss": 0.7413765,
      "memory(GiB)": 135.77,
      "step": 29080,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.76918368,
      "epoch": 0.67869172778839,
      "grad_norm": 5.28125,
      "learning_rate": 7.693473725590079e-06,
      "loss": 0.82540627,
      "memory(GiB)": 135.77,
      "step": 29090,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.77715392,
      "epoch": 0.6789250353606789,
      "grad_norm": 5.71875,
      "learning_rate": 7.691881954896716e-06,
      "loss": 0.77763295,
      "memory(GiB)": 135.77,
      "step": 29100,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.78276401,
      "epoch": 0.6791583429329678,
      "grad_norm": 5.0625,
      "learning_rate": 7.690289799937985e-06,
      "loss": 0.78063684,
      "memory(GiB)": 135.77,
      "step": 29110,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.76057873,
      "epoch": 0.6793916505052567,
      "grad_norm": 5.75,
      "learning_rate": 7.688697260941164e-06,
      "loss": 0.87658768,
      "memory(GiB)": 135.77,
      "step": 29120,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78228416,
      "epoch": 0.6796249580775456,
      "grad_norm": 4.75,
      "learning_rate": 7.687104338133595e-06,
      "loss": 0.79393601,
      "memory(GiB)": 135.77,
      "step": 29130,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.7692297,
      "epoch": 0.6798582656498345,
      "grad_norm": 4.8125,
      "learning_rate": 7.68551103174266e-06,
      "loss": 0.82615576,
      "memory(GiB)": 135.77,
      "step": 29140,
      "train_speed(iter/s)": 0.201147
    },
    {
      "acc": 0.7818614,
      "epoch": 0.6800915732221234,
      "grad_norm": 4.3125,
      "learning_rate": 7.683917341995806e-06,
      "loss": 0.76776161,
      "memory(GiB)": 135.77,
      "step": 29150,
      "train_speed(iter/s)": 0.201182
    },
    {
      "acc": 0.75885959,
      "epoch": 0.6803248807944123,
      "grad_norm": 8.5,
      "learning_rate": 7.68232326912053e-06,
      "loss": 0.88169804,
      "memory(GiB)": 135.77,
      "step": 29160,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.77344918,
      "epoch": 0.6805581883667011,
      "grad_norm": 6.78125,
      "learning_rate": 7.680728813344388e-06,
      "loss": 0.81062412,
      "memory(GiB)": 135.77,
      "step": 29170,
      "train_speed(iter/s)": 0.201256
    },
    {
      "acc": 0.75116019,
      "epoch": 0.68079149593899,
      "grad_norm": 10.125,
      "learning_rate": 7.679133974894984e-06,
      "loss": 0.88641338,
      "memory(GiB)": 135.77,
      "step": 29180,
      "train_speed(iter/s)": 0.201294
    },
    {
      "acc": 0.77474976,
      "epoch": 0.6810248035112789,
      "grad_norm": 4.6875,
      "learning_rate": 7.677538753999984e-06,
      "loss": 0.81274004,
      "memory(GiB)": 135.77,
      "step": 29190,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.77214217,
      "epoch": 0.6812581110835678,
      "grad_norm": 5.875,
      "learning_rate": 7.675943150887107e-06,
      "loss": 0.83875799,
      "memory(GiB)": 135.77,
      "step": 29200,
      "train_speed(iter/s)": 0.201368
    },
    {
      "acc": 0.77097301,
      "epoch": 0.6814914186558567,
      "grad_norm": 6.8125,
      "learning_rate": 7.674347165784122e-06,
      "loss": 0.82045498,
      "memory(GiB)": 135.77,
      "step": 29210,
      "train_speed(iter/s)": 0.201403
    },
    {
      "acc": 0.77530422,
      "epoch": 0.6817247262281456,
      "grad_norm": 4.71875,
      "learning_rate": 7.672750798918854e-06,
      "loss": 0.81686249,
      "memory(GiB)": 135.77,
      "step": 29220,
      "train_speed(iter/s)": 0.201439
    },
    {
      "acc": 0.77789793,
      "epoch": 0.6819580338004345,
      "grad_norm": 7.09375,
      "learning_rate": 7.671154050519187e-06,
      "loss": 0.80894594,
      "memory(GiB)": 135.77,
      "step": 29230,
      "train_speed(iter/s)": 0.201476
    },
    {
      "acc": 0.77112041,
      "epoch": 0.6821913413727234,
      "grad_norm": 4.65625,
      "learning_rate": 7.669556920813056e-06,
      "loss": 0.82041979,
      "memory(GiB)": 135.77,
      "step": 29240,
      "train_speed(iter/s)": 0.201512
    },
    {
      "acc": 0.79637423,
      "epoch": 0.6824246489450123,
      "grad_norm": 7.46875,
      "learning_rate": 7.66795941002845e-06,
      "loss": 0.74020443,
      "memory(GiB)": 135.77,
      "step": 29250,
      "train_speed(iter/s)": 0.201548
    },
    {
      "acc": 0.77368479,
      "epoch": 0.6826579565173012,
      "grad_norm": 4.65625,
      "learning_rate": 7.666361518393413e-06,
      "loss": 0.83006983,
      "memory(GiB)": 135.77,
      "step": 29260,
      "train_speed(iter/s)": 0.201584
    },
    {
      "acc": 0.77014456,
      "epoch": 0.6828912640895901,
      "grad_norm": 4.25,
      "learning_rate": 7.664763246136042e-06,
      "loss": 0.830159,
      "memory(GiB)": 135.77,
      "step": 29270,
      "train_speed(iter/s)": 0.201618
    },
    {
      "acc": 0.78274126,
      "epoch": 0.683124571661879,
      "grad_norm": 15.0,
      "learning_rate": 7.663164593484493e-06,
      "loss": 0.80830526,
      "memory(GiB)": 135.77,
      "step": 29280,
      "train_speed(iter/s)": 0.201656
    },
    {
      "acc": 0.7715332,
      "epoch": 0.6833578792341679,
      "grad_norm": 6.09375,
      "learning_rate": 7.661565560666973e-06,
      "loss": 0.81128178,
      "memory(GiB)": 135.77,
      "step": 29290,
      "train_speed(iter/s)": 0.201691
    },
    {
      "acc": 0.78186312,
      "epoch": 0.6835911868064568,
      "grad_norm": 6.5625,
      "learning_rate": 7.65996614791174e-06,
      "loss": 0.76622462,
      "memory(GiB)": 135.77,
      "step": 29300,
      "train_speed(iter/s)": 0.201725
    },
    {
      "acc": 0.77319794,
      "epoch": 0.6838244943787457,
      "grad_norm": 5.09375,
      "learning_rate": 7.658366355447115e-06,
      "loss": 0.85582533,
      "memory(GiB)": 135.77,
      "step": 29310,
      "train_speed(iter/s)": 0.201761
    },
    {
      "acc": 0.74860148,
      "epoch": 0.6840578019510346,
      "grad_norm": 5.375,
      "learning_rate": 7.656766183501465e-06,
      "loss": 0.90369864,
      "memory(GiB)": 135.77,
      "step": 29320,
      "train_speed(iter/s)": 0.201797
    },
    {
      "acc": 0.78511214,
      "epoch": 0.6842911095233235,
      "grad_norm": 5.21875,
      "learning_rate": 7.655165632303212e-06,
      "loss": 0.77512712,
      "memory(GiB)": 135.77,
      "step": 29330,
      "train_speed(iter/s)": 0.201833
    },
    {
      "acc": 0.78140364,
      "epoch": 0.6845244170956124,
      "grad_norm": 6.34375,
      "learning_rate": 7.653564702080837e-06,
      "loss": 0.78134556,
      "memory(GiB)": 135.77,
      "step": 29340,
      "train_speed(iter/s)": 0.201869
    },
    {
      "acc": 0.77730131,
      "epoch": 0.6847577246679013,
      "grad_norm": 4.1875,
      "learning_rate": 7.651963393062872e-06,
      "loss": 0.80187654,
      "memory(GiB)": 135.77,
      "step": 29350,
      "train_speed(iter/s)": 0.201905
    },
    {
      "acc": 0.78949361,
      "epoch": 0.6849910322401902,
      "grad_norm": 7.34375,
      "learning_rate": 7.650361705477903e-06,
      "loss": 0.73667974,
      "memory(GiB)": 135.77,
      "step": 29360,
      "train_speed(iter/s)": 0.201939
    },
    {
      "acc": 0.78830619,
      "epoch": 0.685224339812479,
      "grad_norm": 5.875,
      "learning_rate": 7.648759639554571e-06,
      "loss": 0.74675617,
      "memory(GiB)": 135.77,
      "step": 29370,
      "train_speed(iter/s)": 0.201976
    },
    {
      "acc": 0.77039886,
      "epoch": 0.6854576473847679,
      "grad_norm": 5.34375,
      "learning_rate": 7.647157195521568e-06,
      "loss": 0.82584143,
      "memory(GiB)": 135.77,
      "step": 29380,
      "train_speed(iter/s)": 0.202012
    },
    {
      "acc": 0.78677912,
      "epoch": 0.6856909549570568,
      "grad_norm": 6.34375,
      "learning_rate": 7.645554373607647e-06,
      "loss": 0.74630208,
      "memory(GiB)": 135.77,
      "step": 29390,
      "train_speed(iter/s)": 0.202047
    },
    {
      "acc": 0.76948566,
      "epoch": 0.6859242625293457,
      "grad_norm": 4.5,
      "learning_rate": 7.643951174041606e-06,
      "loss": 0.82137985,
      "memory(GiB)": 135.77,
      "step": 29400,
      "train_speed(iter/s)": 0.20208
    },
    {
      "acc": 0.78999381,
      "epoch": 0.6861575701016346,
      "grad_norm": 4.78125,
      "learning_rate": 7.642347597052303e-06,
      "loss": 0.74593878,
      "memory(GiB)": 135.77,
      "step": 29410,
      "train_speed(iter/s)": 0.202116
    },
    {
      "acc": 0.76763272,
      "epoch": 0.6863908776739235,
      "grad_norm": 7.21875,
      "learning_rate": 7.64074364286865e-06,
      "loss": 0.84214306,
      "memory(GiB)": 135.77,
      "step": 29420,
      "train_speed(iter/s)": 0.202153
    },
    {
      "acc": 0.78076696,
      "epoch": 0.6866241852462124,
      "grad_norm": 6.65625,
      "learning_rate": 7.639139311719605e-06,
      "loss": 0.7823586,
      "memory(GiB)": 135.77,
      "step": 29430,
      "train_speed(iter/s)": 0.202187
    },
    {
      "acc": 0.77389212,
      "epoch": 0.6868574928185013,
      "grad_norm": 4.59375,
      "learning_rate": 7.637534603834193e-06,
      "loss": 0.8016942,
      "memory(GiB)": 135.77,
      "step": 29440,
      "train_speed(iter/s)": 0.202222
    },
    {
      "acc": 0.77874303,
      "epoch": 0.6870908003907902,
      "grad_norm": 6.28125,
      "learning_rate": 7.635929519441483e-06,
      "loss": 0.80179472,
      "memory(GiB)": 135.77,
      "step": 29450,
      "train_speed(iter/s)": 0.202259
    },
    {
      "acc": 0.76265059,
      "epoch": 0.6873241079630791,
      "grad_norm": 5.78125,
      "learning_rate": 7.634324058770598e-06,
      "loss": 0.86517248,
      "memory(GiB)": 135.77,
      "step": 29460,
      "train_speed(iter/s)": 0.202293
    },
    {
      "acc": 0.77614689,
      "epoch": 0.687557415535368,
      "grad_norm": 4.125,
      "learning_rate": 7.632718222050719e-06,
      "loss": 0.80183201,
      "memory(GiB)": 135.77,
      "step": 29470,
      "train_speed(iter/s)": 0.202329
    },
    {
      "acc": 0.76075153,
      "epoch": 0.6877907231076569,
      "grad_norm": 12.5,
      "learning_rate": 7.63111200951108e-06,
      "loss": 0.85887947,
      "memory(GiB)": 135.77,
      "step": 29480,
      "train_speed(iter/s)": 0.202363
    },
    {
      "acc": 0.76486063,
      "epoch": 0.6880240306799458,
      "grad_norm": 6.46875,
      "learning_rate": 7.629505421380965e-06,
      "loss": 0.83933964,
      "memory(GiB)": 135.77,
      "step": 29490,
      "train_speed(iter/s)": 0.2024
    },
    {
      "acc": 0.7917264,
      "epoch": 0.6882573382522347,
      "grad_norm": 5.4375,
      "learning_rate": 7.627898457889717e-06,
      "loss": 0.74840088,
      "memory(GiB)": 135.77,
      "step": 29500,
      "train_speed(iter/s)": 0.202435
    },
    {
      "epoch": 0.6882573382522347,
      "eval_acc": 0.7415806972391614,
      "eval_loss": 0.8159891366958618,
      "eval_runtime": 1269.8855,
      "eval_samples_per_second": 28.342,
      "eval_steps_per_second": 14.171,
      "step": 29500
    },
    {
      "acc": 0.78372993,
      "epoch": 0.6884906458245236,
      "grad_norm": 4.625,
      "learning_rate": 7.6262911192667245e-06,
      "loss": 0.7768353,
      "memory(GiB)": 135.77,
      "step": 29510,
      "train_speed(iter/s)": 0.200698
    },
    {
      "acc": 0.78984094,
      "epoch": 0.6887239533968125,
      "grad_norm": 5.9375,
      "learning_rate": 7.62468340574144e-06,
      "loss": 0.75353394,
      "memory(GiB)": 135.77,
      "step": 29520,
      "train_speed(iter/s)": 0.20073
    },
    {
      "acc": 0.78566389,
      "epoch": 0.6889572609691014,
      "grad_norm": 4.46875,
      "learning_rate": 7.623075317543361e-06,
      "loss": 0.74840536,
      "memory(GiB)": 135.77,
      "step": 29530,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.79994617,
      "epoch": 0.6891905685413903,
      "grad_norm": 6.40625,
      "learning_rate": 7.62146685490204e-06,
      "loss": 0.70613475,
      "memory(GiB)": 135.77,
      "step": 29540,
      "train_speed(iter/s)": 0.200797
    },
    {
      "acc": 0.78054943,
      "epoch": 0.6894238761136792,
      "grad_norm": 5.28125,
      "learning_rate": 7.6198580180470904e-06,
      "loss": 0.80197172,
      "memory(GiB)": 135.77,
      "step": 29550,
      "train_speed(iter/s)": 0.200835
    },
    {
      "acc": 0.75962677,
      "epoch": 0.689657183685968,
      "grad_norm": 5.9375,
      "learning_rate": 7.618248807208169e-06,
      "loss": 0.85815258,
      "memory(GiB)": 135.77,
      "step": 29560,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.76958151,
      "epoch": 0.6898904912582569,
      "grad_norm": 5.65625,
      "learning_rate": 7.61663922261499e-06,
      "loss": 0.82609196,
      "memory(GiB)": 135.77,
      "step": 29570,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.7712954,
      "epoch": 0.6901237988305458,
      "grad_norm": 4.90625,
      "learning_rate": 7.615029264497322e-06,
      "loss": 0.82603159,
      "memory(GiB)": 135.77,
      "step": 29580,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.76029787,
      "epoch": 0.6903571064028347,
      "grad_norm": 5.15625,
      "learning_rate": 7.6134189330849885e-06,
      "loss": 0.86350775,
      "memory(GiB)": 135.77,
      "step": 29590,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.77219863,
      "epoch": 0.6905904139751236,
      "grad_norm": 5.34375,
      "learning_rate": 7.611808228607859e-06,
      "loss": 0.82623463,
      "memory(GiB)": 135.77,
      "step": 29600,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.78632302,
      "epoch": 0.6908237215474125,
      "grad_norm": 6.53125,
      "learning_rate": 7.610197151295865e-06,
      "loss": 0.76113815,
      "memory(GiB)": 135.77,
      "step": 29610,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.78305635,
      "epoch": 0.6910570291197013,
      "grad_norm": 5.375,
      "learning_rate": 7.608585701378985e-06,
      "loss": 0.76118941,
      "memory(GiB)": 135.77,
      "step": 29620,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.7888319,
      "epoch": 0.6912903366919902,
      "grad_norm": 4.03125,
      "learning_rate": 7.6069738790872545e-06,
      "loss": 0.7636095,
      "memory(GiB)": 135.77,
      "step": 29630,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.77132835,
      "epoch": 0.6915236442642791,
      "grad_norm": 4.9375,
      "learning_rate": 7.6053616846507606e-06,
      "loss": 0.79716229,
      "memory(GiB)": 135.77,
      "step": 29640,
      "train_speed(iter/s)": 0.201143
    },
    {
      "acc": 0.7883872,
      "epoch": 0.691756951836568,
      "grad_norm": 3.296875,
      "learning_rate": 7.6037491182996415e-06,
      "loss": 0.77416801,
      "memory(GiB)": 135.77,
      "step": 29650,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.78499637,
      "epoch": 0.691990259408857,
      "grad_norm": 4.3125,
      "learning_rate": 7.602136180264094e-06,
      "loss": 0.77752113,
      "memory(GiB)": 135.77,
      "step": 29660,
      "train_speed(iter/s)": 0.201215
    },
    {
      "acc": 0.76922569,
      "epoch": 0.6922235669811458,
      "grad_norm": 6.0,
      "learning_rate": 7.6005228707743606e-06,
      "loss": 0.83131847,
      "memory(GiB)": 135.77,
      "step": 29670,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.77869415,
      "epoch": 0.6924568745534347,
      "grad_norm": 6.84375,
      "learning_rate": 7.598909190060744e-06,
      "loss": 0.8007268,
      "memory(GiB)": 135.77,
      "step": 29680,
      "train_speed(iter/s)": 0.201276
    },
    {
      "acc": 0.78622375,
      "epoch": 0.6926901821257236,
      "grad_norm": 5.0,
      "learning_rate": 7.597295138353596e-06,
      "loss": 0.75257502,
      "memory(GiB)": 135.77,
      "step": 29690,
      "train_speed(iter/s)": 0.201314
    },
    {
      "acc": 0.77413979,
      "epoch": 0.6929234896980125,
      "grad_norm": 5.6875,
      "learning_rate": 7.595680715883321e-06,
      "loss": 0.80841618,
      "memory(GiB)": 135.77,
      "step": 29700,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.77899332,
      "epoch": 0.6931567972703014,
      "grad_norm": 5.4375,
      "learning_rate": 7.594065922880378e-06,
      "loss": 0.78428011,
      "memory(GiB)": 135.77,
      "step": 29710,
      "train_speed(iter/s)": 0.201382
    },
    {
      "acc": 0.78220787,
      "epoch": 0.6933901048425903,
      "grad_norm": 3.53125,
      "learning_rate": 7.592450759575278e-06,
      "loss": 0.77431474,
      "memory(GiB)": 135.77,
      "step": 29720,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.78128052,
      "epoch": 0.6936234124148792,
      "grad_norm": 4.3125,
      "learning_rate": 7.590835226198585e-06,
      "loss": 0.79154778,
      "memory(GiB)": 135.77,
      "step": 29730,
      "train_speed(iter/s)": 0.201451
    },
    {
      "acc": 0.77904191,
      "epoch": 0.6938567199871681,
      "grad_norm": 5.28125,
      "learning_rate": 7.589219322980916e-06,
      "loss": 0.79745936,
      "memory(GiB)": 135.77,
      "step": 29740,
      "train_speed(iter/s)": 0.201485
    },
    {
      "acc": 0.78388314,
      "epoch": 0.6940900275594569,
      "grad_norm": 5.34375,
      "learning_rate": 7.587603050152941e-06,
      "loss": 0.79330082,
      "memory(GiB)": 135.77,
      "step": 29750,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.77729387,
      "epoch": 0.6943233351317458,
      "grad_norm": 4.75,
      "learning_rate": 7.585986407945383e-06,
      "loss": 0.79987221,
      "memory(GiB)": 135.77,
      "step": 29760,
      "train_speed(iter/s)": 0.201555
    },
    {
      "acc": 0.76541524,
      "epoch": 0.6945566427040347,
      "grad_norm": 5.78125,
      "learning_rate": 7.584369396589015e-06,
      "loss": 0.8547636,
      "memory(GiB)": 135.77,
      "step": 29770,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.75868149,
      "epoch": 0.6947899502763236,
      "grad_norm": 5.78125,
      "learning_rate": 7.582752016314669e-06,
      "loss": 0.84939613,
      "memory(GiB)": 135.77,
      "step": 29780,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.80678043,
      "epoch": 0.6950232578486125,
      "grad_norm": 4.90625,
      "learning_rate": 7.58113426735322e-06,
      "loss": 0.68545928,
      "memory(GiB)": 135.77,
      "step": 29790,
      "train_speed(iter/s)": 0.201662
    },
    {
      "acc": 0.75615764,
      "epoch": 0.6952565654209014,
      "grad_norm": 6.9375,
      "learning_rate": 7.579516149935606e-06,
      "loss": 0.86680984,
      "memory(GiB)": 135.77,
      "step": 29800,
      "train_speed(iter/s)": 0.201699
    },
    {
      "acc": 0.77607203,
      "epoch": 0.6954898729931903,
      "grad_norm": 4.8125,
      "learning_rate": 7.577897664292811e-06,
      "loss": 0.80326805,
      "memory(GiB)": 135.77,
      "step": 29810,
      "train_speed(iter/s)": 0.201734
    },
    {
      "acc": 0.78733912,
      "epoch": 0.6957231805654792,
      "grad_norm": 5.40625,
      "learning_rate": 7.57627881065587e-06,
      "loss": 0.77941236,
      "memory(GiB)": 135.77,
      "step": 29820,
      "train_speed(iter/s)": 0.201768
    },
    {
      "acc": 0.75822802,
      "epoch": 0.6959564881377681,
      "grad_norm": 4.65625,
      "learning_rate": 7.574659589255881e-06,
      "loss": 0.86697035,
      "memory(GiB)": 135.77,
      "step": 29830,
      "train_speed(iter/s)": 0.201804
    },
    {
      "acc": 0.74630752,
      "epoch": 0.696189795710057,
      "grad_norm": 5.25,
      "learning_rate": 7.573040000323984e-06,
      "loss": 0.92794971,
      "memory(GiB)": 135.77,
      "step": 29840,
      "train_speed(iter/s)": 0.201839
    },
    {
      "acc": 0.79665966,
      "epoch": 0.6964231032823459,
      "grad_norm": 5.46875,
      "learning_rate": 7.571420044091372e-06,
      "loss": 0.70622387,
      "memory(GiB)": 135.77,
      "step": 29850,
      "train_speed(iter/s)": 0.201874
    },
    {
      "acc": 0.76552095,
      "epoch": 0.6966564108546348,
      "grad_norm": 3.84375,
      "learning_rate": 7.569799720789297e-06,
      "loss": 0.86479321,
      "memory(GiB)": 135.77,
      "step": 29860,
      "train_speed(iter/s)": 0.201909
    },
    {
      "acc": 0.78278189,
      "epoch": 0.6968897184269237,
      "grad_norm": 5.90625,
      "learning_rate": 7.568179030649057e-06,
      "loss": 0.76828461,
      "memory(GiB)": 135.77,
      "step": 29870,
      "train_speed(iter/s)": 0.201945
    },
    {
      "acc": 0.76996212,
      "epoch": 0.6971230259992126,
      "grad_norm": 4.84375,
      "learning_rate": 7.566557973902007e-06,
      "loss": 0.85869331,
      "memory(GiB)": 135.77,
      "step": 29880,
      "train_speed(iter/s)": 0.201981
    },
    {
      "acc": 0.79064722,
      "epoch": 0.6973563335715015,
      "grad_norm": 5.96875,
      "learning_rate": 7.564936550779553e-06,
      "loss": 0.73188696,
      "memory(GiB)": 135.77,
      "step": 29890,
      "train_speed(iter/s)": 0.202018
    },
    {
      "acc": 0.77210503,
      "epoch": 0.6975896411437904,
      "grad_norm": 5.375,
      "learning_rate": 7.563314761513151e-06,
      "loss": 0.81467781,
      "memory(GiB)": 135.77,
      "step": 29900,
      "train_speed(iter/s)": 0.202053
    },
    {
      "acc": 0.79614601,
      "epoch": 0.6978229487160793,
      "grad_norm": 4.15625,
      "learning_rate": 7.56169260633431e-06,
      "loss": 0.71139755,
      "memory(GiB)": 135.77,
      "step": 29910,
      "train_speed(iter/s)": 0.202089
    },
    {
      "acc": 0.80187359,
      "epoch": 0.6980562562883682,
      "grad_norm": 4.40625,
      "learning_rate": 7.560070085474596e-06,
      "loss": 0.69751916,
      "memory(GiB)": 135.77,
      "step": 29920,
      "train_speed(iter/s)": 0.202123
    },
    {
      "acc": 0.77271166,
      "epoch": 0.6982895638606571,
      "grad_norm": 5.125,
      "learning_rate": 7.55844719916562e-06,
      "loss": 0.82679062,
      "memory(GiB)": 135.77,
      "step": 29930,
      "train_speed(iter/s)": 0.202157
    },
    {
      "acc": 0.77691226,
      "epoch": 0.6985228714329459,
      "grad_norm": 4.6875,
      "learning_rate": 7.556823947639048e-06,
      "loss": 0.79194837,
      "memory(GiB)": 135.77,
      "step": 29940,
      "train_speed(iter/s)": 0.202193
    },
    {
      "acc": 0.77552891,
      "epoch": 0.6987561790052348,
      "grad_norm": 5.84375,
      "learning_rate": 7.555200331126602e-06,
      "loss": 0.82497501,
      "memory(GiB)": 135.77,
      "step": 29950,
      "train_speed(iter/s)": 0.202227
    },
    {
      "acc": 0.77520766,
      "epoch": 0.6989894865775237,
      "grad_norm": 5.6875,
      "learning_rate": 7.55357634986005e-06,
      "loss": 0.81748133,
      "memory(GiB)": 135.77,
      "step": 29960,
      "train_speed(iter/s)": 0.202263
    },
    {
      "acc": 0.78164778,
      "epoch": 0.6992227941498126,
      "grad_norm": 5.75,
      "learning_rate": 7.551952004071217e-06,
      "loss": 0.78227224,
      "memory(GiB)": 135.77,
      "step": 29970,
      "train_speed(iter/s)": 0.202296
    },
    {
      "acc": 0.77848063,
      "epoch": 0.6994561017221015,
      "grad_norm": 6.4375,
      "learning_rate": 7.550327293991976e-06,
      "loss": 0.78909693,
      "memory(GiB)": 135.77,
      "step": 29980,
      "train_speed(iter/s)": 0.202326
    },
    {
      "acc": 0.75997553,
      "epoch": 0.6996894092943904,
      "grad_norm": 5.65625,
      "learning_rate": 7.5487022198542555e-06,
      "loss": 0.8917922,
      "memory(GiB)": 135.77,
      "step": 29990,
      "train_speed(iter/s)": 0.202361
    },
    {
      "acc": 0.78532553,
      "epoch": 0.6999227168666793,
      "grad_norm": 6.84375,
      "learning_rate": 7.547076781890032e-06,
      "loss": 0.7683835,
      "memory(GiB)": 135.77,
      "step": 30000,
      "train_speed(iter/s)": 0.202393
    },
    {
      "epoch": 0.6999227168666793,
      "eval_acc": 0.7416280999319227,
      "eval_loss": 0.8156638741493225,
      "eval_runtime": 1270.1597,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 30000
    },
    {
      "acc": 0.7785594,
      "epoch": 0.7001560244389682,
      "grad_norm": 7.5625,
      "learning_rate": 7.5454509803313394e-06,
      "loss": 0.79886799,
      "memory(GiB)": 135.77,
      "step": 30010,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.76497021,
      "epoch": 0.7003893320112571,
      "grad_norm": 4.71875,
      "learning_rate": 7.543824815410259e-06,
      "loss": 0.85206814,
      "memory(GiB)": 135.77,
      "step": 30020,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.78620939,
      "epoch": 0.700622639583546,
      "grad_norm": 5.46875,
      "learning_rate": 7.542198287358924e-06,
      "loss": 0.78464289,
      "memory(GiB)": 135.77,
      "step": 30030,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76833134,
      "epoch": 0.7008559471558349,
      "grad_norm": 4.5625,
      "learning_rate": 7.540571396409522e-06,
      "loss": 0.8300333,
      "memory(GiB)": 135.77,
      "step": 30040,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.78864598,
      "epoch": 0.7010892547281238,
      "grad_norm": 4.0625,
      "learning_rate": 7.538944142794291e-06,
      "loss": 0.74310522,
      "memory(GiB)": 135.77,
      "step": 30050,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.76193199,
      "epoch": 0.7013225623004127,
      "grad_norm": 4.78125,
      "learning_rate": 7.537316526745522e-06,
      "loss": 0.86004868,
      "memory(GiB)": 135.77,
      "step": 30060,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.79334602,
      "epoch": 0.7015558698727016,
      "grad_norm": 5.125,
      "learning_rate": 7.535688548495557e-06,
      "loss": 0.73706589,
      "memory(GiB)": 135.77,
      "step": 30070,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.76868596,
      "epoch": 0.7017891774449905,
      "grad_norm": 7.25,
      "learning_rate": 7.534060208276786e-06,
      "loss": 0.81477661,
      "memory(GiB)": 135.77,
      "step": 30080,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.74454851,
      "epoch": 0.7020224850172794,
      "grad_norm": 5.78125,
      "learning_rate": 7.532431506321657e-06,
      "loss": 0.95340862,
      "memory(GiB)": 135.77,
      "step": 30090,
      "train_speed(iter/s)": 0.200948
    },
    {
      "acc": 0.76665716,
      "epoch": 0.7022557925895683,
      "grad_norm": 6.03125,
      "learning_rate": 7.530802442862666e-06,
      "loss": 0.84706726,
      "memory(GiB)": 135.77,
      "step": 30100,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.78391695,
      "epoch": 0.7024891001618572,
      "grad_norm": 5.1875,
      "learning_rate": 7.529173018132362e-06,
      "loss": 0.77013426,
      "memory(GiB)": 135.77,
      "step": 30110,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.80045862,
      "epoch": 0.7027224077341461,
      "grad_norm": 6.0,
      "learning_rate": 7.5275432323633446e-06,
      "loss": 0.71537471,
      "memory(GiB)": 135.77,
      "step": 30120,
      "train_speed(iter/s)": 0.20105
    },
    {
      "acc": 0.76974192,
      "epoch": 0.702955715306435,
      "grad_norm": 4.5,
      "learning_rate": 7.525913085788264e-06,
      "loss": 0.81845627,
      "memory(GiB)": 135.77,
      "step": 30130,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.7624217,
      "epoch": 0.7031890228787238,
      "grad_norm": 5.625,
      "learning_rate": 7.524282578639825e-06,
      "loss": 0.85741539,
      "memory(GiB)": 135.77,
      "step": 30140,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.77117977,
      "epoch": 0.7034223304510127,
      "grad_norm": 5.5,
      "learning_rate": 7.522651711150781e-06,
      "loss": 0.82536449,
      "memory(GiB)": 135.77,
      "step": 30150,
      "train_speed(iter/s)": 0.201154
    },
    {
      "acc": 0.76665888,
      "epoch": 0.7036556380233016,
      "grad_norm": 4.8125,
      "learning_rate": 7.521020483553939e-06,
      "loss": 0.82699442,
      "memory(GiB)": 135.77,
      "step": 30160,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.76989846,
      "epoch": 0.7038889455955905,
      "grad_norm": 4.75,
      "learning_rate": 7.519388896082154e-06,
      "loss": 0.83109045,
      "memory(GiB)": 135.77,
      "step": 30170,
      "train_speed(iter/s)": 0.201223
    },
    {
      "acc": 0.77924395,
      "epoch": 0.7041222531678794,
      "grad_norm": 4.78125,
      "learning_rate": 7.517756948968338e-06,
      "loss": 0.7907198,
      "memory(GiB)": 135.77,
      "step": 30180,
      "train_speed(iter/s)": 0.201257
    },
    {
      "acc": 0.76878471,
      "epoch": 0.7043555607401683,
      "grad_norm": 5.25,
      "learning_rate": 7.516124642445447e-06,
      "loss": 0.84302406,
      "memory(GiB)": 135.77,
      "step": 30190,
      "train_speed(iter/s)": 0.201291
    },
    {
      "acc": 0.77351613,
      "epoch": 0.7045888683124572,
      "grad_norm": 5.4375,
      "learning_rate": 7.514491976746494e-06,
      "loss": 0.81995602,
      "memory(GiB)": 135.77,
      "step": 30200,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.76795378,
      "epoch": 0.704822175884746,
      "grad_norm": 5.9375,
      "learning_rate": 7.512858952104544e-06,
      "loss": 0.83630714,
      "memory(GiB)": 135.77,
      "step": 30210,
      "train_speed(iter/s)": 0.201361
    },
    {
      "acc": 0.75973902,
      "epoch": 0.705055483457035,
      "grad_norm": 6.03125,
      "learning_rate": 7.511225568752707e-06,
      "loss": 0.854778,
      "memory(GiB)": 135.77,
      "step": 30220,
      "train_speed(iter/s)": 0.201397
    },
    {
      "acc": 0.7747612,
      "epoch": 0.7052887910293238,
      "grad_norm": 4.21875,
      "learning_rate": 7.50959182692415e-06,
      "loss": 0.81778402,
      "memory(GiB)": 135.77,
      "step": 30230,
      "train_speed(iter/s)": 0.201434
    },
    {
      "acc": 0.77487125,
      "epoch": 0.7055220986016127,
      "grad_norm": 3.984375,
      "learning_rate": 7.507957726852087e-06,
      "loss": 0.78481297,
      "memory(GiB)": 135.77,
      "step": 30240,
      "train_speed(iter/s)": 0.201469
    },
    {
      "acc": 0.77745609,
      "epoch": 0.7057554061739016,
      "grad_norm": 6.03125,
      "learning_rate": 7.506323268769788e-06,
      "loss": 0.7958005,
      "memory(GiB)": 135.77,
      "step": 30250,
      "train_speed(iter/s)": 0.201506
    },
    {
      "acc": 0.77071285,
      "epoch": 0.7059887137461905,
      "grad_norm": 4.8125,
      "learning_rate": 7.504688452910571e-06,
      "loss": 0.81425562,
      "memory(GiB)": 135.77,
      "step": 30260,
      "train_speed(iter/s)": 0.201541
    },
    {
      "acc": 0.77105627,
      "epoch": 0.7062220213184794,
      "grad_norm": 5.78125,
      "learning_rate": 7.503053279507806e-06,
      "loss": 0.82091007,
      "memory(GiB)": 135.77,
      "step": 30270,
      "train_speed(iter/s)": 0.201575
    },
    {
      "acc": 0.79748535,
      "epoch": 0.7064553288907683,
      "grad_norm": 5.21875,
      "learning_rate": 7.501417748794911e-06,
      "loss": 0.73630304,
      "memory(GiB)": 135.77,
      "step": 30280,
      "train_speed(iter/s)": 0.20161
    },
    {
      "acc": 0.78553853,
      "epoch": 0.7066886364630572,
      "grad_norm": 6.0625,
      "learning_rate": 7.49978186100536e-06,
      "loss": 0.77834482,
      "memory(GiB)": 135.77,
      "step": 30290,
      "train_speed(iter/s)": 0.201643
    },
    {
      "acc": 0.77966051,
      "epoch": 0.7069219440353461,
      "grad_norm": 4.90625,
      "learning_rate": 7.498145616372674e-06,
      "loss": 0.77920589,
      "memory(GiB)": 135.77,
      "step": 30300,
      "train_speed(iter/s)": 0.201678
    },
    {
      "acc": 0.78804102,
      "epoch": 0.707155251607635,
      "grad_norm": 4.875,
      "learning_rate": 7.4965090151304265e-06,
      "loss": 0.75938053,
      "memory(GiB)": 135.77,
      "step": 30310,
      "train_speed(iter/s)": 0.201711
    },
    {
      "acc": 0.7741745,
      "epoch": 0.7073885591799239,
      "grad_norm": 4.6875,
      "learning_rate": 7.494872057512242e-06,
      "loss": 0.82201767,
      "memory(GiB)": 135.77,
      "step": 30320,
      "train_speed(iter/s)": 0.201744
    },
    {
      "acc": 0.78020763,
      "epoch": 0.7076218667522127,
      "grad_norm": 4.25,
      "learning_rate": 7.493234743751797e-06,
      "loss": 0.77473249,
      "memory(GiB)": 135.77,
      "step": 30330,
      "train_speed(iter/s)": 0.201778
    },
    {
      "acc": 0.78065357,
      "epoch": 0.7078551743245016,
      "grad_norm": 8.8125,
      "learning_rate": 7.491597074082817e-06,
      "loss": 0.79642534,
      "memory(GiB)": 135.77,
      "step": 30340,
      "train_speed(iter/s)": 0.201811
    },
    {
      "acc": 0.77584829,
      "epoch": 0.7080884818967905,
      "grad_norm": 7.34375,
      "learning_rate": 7.489959048739079e-06,
      "loss": 0.81745863,
      "memory(GiB)": 135.77,
      "step": 30350,
      "train_speed(iter/s)": 0.201842
    },
    {
      "acc": 0.77412333,
      "epoch": 0.7083217894690794,
      "grad_norm": 9.125,
      "learning_rate": 7.488320667954408e-06,
      "loss": 0.81564293,
      "memory(GiB)": 135.77,
      "step": 30360,
      "train_speed(iter/s)": 0.201876
    },
    {
      "acc": 0.76567802,
      "epoch": 0.7085550970413683,
      "grad_norm": 4.96875,
      "learning_rate": 7.486681931962686e-06,
      "loss": 0.84731541,
      "memory(GiB)": 135.77,
      "step": 30370,
      "train_speed(iter/s)": 0.20191
    },
    {
      "acc": 0.76095247,
      "epoch": 0.7087884046136572,
      "grad_norm": 5.65625,
      "learning_rate": 7.48504284099784e-06,
      "loss": 0.87232685,
      "memory(GiB)": 135.77,
      "step": 30380,
      "train_speed(iter/s)": 0.201943
    },
    {
      "acc": 0.77653809,
      "epoch": 0.7090217121859461,
      "grad_norm": 6.53125,
      "learning_rate": 7.48340339529385e-06,
      "loss": 0.78932204,
      "memory(GiB)": 135.77,
      "step": 30390,
      "train_speed(iter/s)": 0.201979
    },
    {
      "acc": 0.78220081,
      "epoch": 0.709255019758235,
      "grad_norm": 5.34375,
      "learning_rate": 7.481763595084747e-06,
      "loss": 0.77936535,
      "memory(GiB)": 135.77,
      "step": 30400,
      "train_speed(iter/s)": 0.202013
    },
    {
      "acc": 0.7953701,
      "epoch": 0.7094883273305239,
      "grad_norm": 3.890625,
      "learning_rate": 7.480123440604613e-06,
      "loss": 0.73328466,
      "memory(GiB)": 135.77,
      "step": 30410,
      "train_speed(iter/s)": 0.202048
    },
    {
      "acc": 0.77866468,
      "epoch": 0.7097216349028128,
      "grad_norm": 6.71875,
      "learning_rate": 7.478482932087577e-06,
      "loss": 0.80838766,
      "memory(GiB)": 135.77,
      "step": 30420,
      "train_speed(iter/s)": 0.202079
    },
    {
      "acc": 0.78703556,
      "epoch": 0.7099549424751017,
      "grad_norm": 4.125,
      "learning_rate": 7.476842069767824e-06,
      "loss": 0.77653255,
      "memory(GiB)": 135.77,
      "step": 30430,
      "train_speed(iter/s)": 0.202113
    },
    {
      "acc": 0.77524529,
      "epoch": 0.7101882500473906,
      "grad_norm": 7.28125,
      "learning_rate": 7.475200853879583e-06,
      "loss": 0.79806619,
      "memory(GiB)": 135.77,
      "step": 30440,
      "train_speed(iter/s)": 0.202148
    },
    {
      "acc": 0.7719944,
      "epoch": 0.7104215576196795,
      "grad_norm": 5.21875,
      "learning_rate": 7.473559284657139e-06,
      "loss": 0.82557144,
      "memory(GiB)": 135.77,
      "step": 30450,
      "train_speed(iter/s)": 0.202185
    },
    {
      "acc": 0.77690635,
      "epoch": 0.7106548651919684,
      "grad_norm": 6.40625,
      "learning_rate": 7.471917362334828e-06,
      "loss": 0.79608321,
      "memory(GiB)": 135.77,
      "step": 30460,
      "train_speed(iter/s)": 0.20222
    },
    {
      "acc": 0.76047997,
      "epoch": 0.7108881727642573,
      "grad_norm": 5.75,
      "learning_rate": 7.47027508714703e-06,
      "loss": 0.8347681,
      "memory(GiB)": 135.77,
      "step": 30470,
      "train_speed(iter/s)": 0.202255
    },
    {
      "acc": 0.78284283,
      "epoch": 0.7111214803365462,
      "grad_norm": 4.53125,
      "learning_rate": 7.468632459328181e-06,
      "loss": 0.76897001,
      "memory(GiB)": 135.77,
      "step": 30480,
      "train_speed(iter/s)": 0.202289
    },
    {
      "acc": 0.77444568,
      "epoch": 0.7113547879088351,
      "grad_norm": 6.71875,
      "learning_rate": 7.466989479112766e-06,
      "loss": 0.81205473,
      "memory(GiB)": 135.77,
      "step": 30490,
      "train_speed(iter/s)": 0.202323
    },
    {
      "acc": 0.76428804,
      "epoch": 0.711588095481124,
      "grad_norm": 5.90625,
      "learning_rate": 7.465346146735319e-06,
      "loss": 0.83355408,
      "memory(GiB)": 135.77,
      "step": 30500,
      "train_speed(iter/s)": 0.202361
    },
    {
      "epoch": 0.711588095481124,
      "eval_acc": 0.7417022467114782,
      "eval_loss": 0.8149046301841736,
      "eval_runtime": 1270.5935,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "step": 30500
    },
    {
      "acc": 0.77408185,
      "epoch": 0.7118214030534129,
      "grad_norm": 7.84375,
      "learning_rate": 7.463702462430427e-06,
      "loss": 0.82522774,
      "memory(GiB)": 135.77,
      "step": 30510,
      "train_speed(iter/s)": 0.20068
    },
    {
      "acc": 0.78299146,
      "epoch": 0.7120547106257017,
      "grad_norm": 4.4375,
      "learning_rate": 7.4620584264327236e-06,
      "loss": 0.79395037,
      "memory(GiB)": 135.77,
      "step": 30520,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.7901659,
      "epoch": 0.7122880181979906,
      "grad_norm": 5.03125,
      "learning_rate": 7.460414038976894e-06,
      "loss": 0.73990803,
      "memory(GiB)": 135.77,
      "step": 30530,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76877952,
      "epoch": 0.7125213257702795,
      "grad_norm": 4.9375,
      "learning_rate": 7.458769300297676e-06,
      "loss": 0.82868195,
      "memory(GiB)": 135.77,
      "step": 30540,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.77149434,
      "epoch": 0.7127546333425684,
      "grad_norm": 6.40625,
      "learning_rate": 7.457124210629853e-06,
      "loss": 0.80213509,
      "memory(GiB)": 135.77,
      "step": 30550,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.76947355,
      "epoch": 0.7129879409148573,
      "grad_norm": 8.5,
      "learning_rate": 7.455478770208267e-06,
      "loss": 0.84286346,
      "memory(GiB)": 135.77,
      "step": 30560,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.77109156,
      "epoch": 0.7132212484871462,
      "grad_norm": 4.3125,
      "learning_rate": 7.453832979267796e-06,
      "loss": 0.81874142,
      "memory(GiB)": 135.77,
      "step": 30570,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.77773933,
      "epoch": 0.7134545560594351,
      "grad_norm": 5.46875,
      "learning_rate": 7.452186838043381e-06,
      "loss": 0.79809036,
      "memory(GiB)": 135.77,
      "step": 30580,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.78017259,
      "epoch": 0.713687863631724,
      "grad_norm": 4.34375,
      "learning_rate": 7.450540346770008e-06,
      "loss": 0.79023871,
      "memory(GiB)": 135.77,
      "step": 30590,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.78224049,
      "epoch": 0.7139211712040129,
      "grad_norm": 5.5625,
      "learning_rate": 7.4488935056827115e-06,
      "loss": 0.79336486,
      "memory(GiB)": 135.77,
      "step": 30600,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.7667264,
      "epoch": 0.7141544787763018,
      "grad_norm": 5.53125,
      "learning_rate": 7.447246315016579e-06,
      "loss": 0.84322777,
      "memory(GiB)": 135.77,
      "step": 30610,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.78506231,
      "epoch": 0.7143877863485907,
      "grad_norm": 6.4375,
      "learning_rate": 7.445598775006745e-06,
      "loss": 0.77431684,
      "memory(GiB)": 135.77,
      "step": 30620,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.76365628,
      "epoch": 0.7146210939208796,
      "grad_norm": 6.59375,
      "learning_rate": 7.443950885888398e-06,
      "loss": 0.85304022,
      "memory(GiB)": 135.77,
      "step": 30630,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.7706419,
      "epoch": 0.7148544014931685,
      "grad_norm": 5.8125,
      "learning_rate": 7.4423026478967706e-06,
      "loss": 0.83546371,
      "memory(GiB)": 135.77,
      "step": 30640,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.76318879,
      "epoch": 0.7150877090654574,
      "grad_norm": 4.78125,
      "learning_rate": 7.440654061267151e-06,
      "loss": 0.84243946,
      "memory(GiB)": 135.77,
      "step": 30650,
      "train_speed(iter/s)": 0.201148
    },
    {
      "acc": 0.7682981,
      "epoch": 0.7153210166377463,
      "grad_norm": 5.5625,
      "learning_rate": 7.439005126234872e-06,
      "loss": 0.82606144,
      "memory(GiB)": 135.77,
      "step": 30660,
      "train_speed(iter/s)": 0.201182
    },
    {
      "acc": 0.76082344,
      "epoch": 0.7155543242100352,
      "grad_norm": 5.3125,
      "learning_rate": 7.43735584303532e-06,
      "loss": 0.86378193,
      "memory(GiB)": 135.77,
      "step": 30670,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.77880592,
      "epoch": 0.7157876317823241,
      "grad_norm": 5.34375,
      "learning_rate": 7.435706211903929e-06,
      "loss": 0.8006321,
      "memory(GiB)": 135.77,
      "step": 30680,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.78893566,
      "epoch": 0.716020939354613,
      "grad_norm": 6.03125,
      "learning_rate": 7.434056233076184e-06,
      "loss": 0.74485593,
      "memory(GiB)": 135.77,
      "step": 30690,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.77817774,
      "epoch": 0.7162542469269019,
      "grad_norm": 6.84375,
      "learning_rate": 7.43240590678762e-06,
      "loss": 0.76789885,
      "memory(GiB)": 135.77,
      "step": 30700,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.77524462,
      "epoch": 0.7164875544991907,
      "grad_norm": 5.9375,
      "learning_rate": 7.4307552332738184e-06,
      "loss": 0.79615641,
      "memory(GiB)": 135.77,
      "step": 30710,
      "train_speed(iter/s)": 0.201355
    },
    {
      "acc": 0.76029038,
      "epoch": 0.7167208620714796,
      "grad_norm": 5.53125,
      "learning_rate": 7.429104212770414e-06,
      "loss": 0.88227682,
      "memory(GiB)": 135.77,
      "step": 30720,
      "train_speed(iter/s)": 0.201386
    },
    {
      "acc": 0.7853837,
      "epoch": 0.7169541696437685,
      "grad_norm": 4.1875,
      "learning_rate": 7.427452845513088e-06,
      "loss": 0.78255091,
      "memory(GiB)": 135.77,
      "step": 30730,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.7675179,
      "epoch": 0.7171874772160574,
      "grad_norm": 10.1875,
      "learning_rate": 7.4258011317375735e-06,
      "loss": 0.83768845,
      "memory(GiB)": 135.77,
      "step": 30740,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.79261498,
      "epoch": 0.7174207847883463,
      "grad_norm": 6.875,
      "learning_rate": 7.424149071679654e-06,
      "loss": 0.74290218,
      "memory(GiB)": 135.77,
      "step": 30750,
      "train_speed(iter/s)": 0.201484
    },
    {
      "acc": 0.78618517,
      "epoch": 0.7176540923606352,
      "grad_norm": 4.46875,
      "learning_rate": 7.422496665575156e-06,
      "loss": 0.75908709,
      "memory(GiB)": 135.77,
      "step": 30760,
      "train_speed(iter/s)": 0.201519
    },
    {
      "acc": 0.79118176,
      "epoch": 0.717887399932924,
      "grad_norm": 4.71875,
      "learning_rate": 7.420843913659965e-06,
      "loss": 0.75808463,
      "memory(GiB)": 135.77,
      "step": 30770,
      "train_speed(iter/s)": 0.201554
    },
    {
      "acc": 0.77106438,
      "epoch": 0.718120707505213,
      "grad_norm": 5.4375,
      "learning_rate": 7.419190816170008e-06,
      "loss": 0.83572817,
      "memory(GiB)": 135.77,
      "step": 30780,
      "train_speed(iter/s)": 0.201588
    },
    {
      "acc": 0.76454983,
      "epoch": 0.7183540150775019,
      "grad_norm": 4.4375,
      "learning_rate": 7.417537373341263e-06,
      "loss": 0.86166372,
      "memory(GiB)": 135.77,
      "step": 30790,
      "train_speed(iter/s)": 0.201623
    },
    {
      "acc": 0.77741642,
      "epoch": 0.7185873226497907,
      "grad_norm": 4.59375,
      "learning_rate": 7.415883585409762e-06,
      "loss": 0.79155254,
      "memory(GiB)": 135.77,
      "step": 30800,
      "train_speed(iter/s)": 0.201659
    },
    {
      "acc": 0.78092527,
      "epoch": 0.7188206302220796,
      "grad_norm": 5.4375,
      "learning_rate": 7.414229452611582e-06,
      "loss": 0.78085165,
      "memory(GiB)": 135.77,
      "step": 30810,
      "train_speed(iter/s)": 0.201694
    },
    {
      "acc": 0.75764637,
      "epoch": 0.7190539377943685,
      "grad_norm": 4.59375,
      "learning_rate": 7.412574975182848e-06,
      "loss": 0.88091917,
      "memory(GiB)": 135.77,
      "step": 30820,
      "train_speed(iter/s)": 0.201727
    },
    {
      "acc": 0.77109008,
      "epoch": 0.7192872453666574,
      "grad_norm": 4.34375,
      "learning_rate": 7.410920153359736e-06,
      "loss": 0.8264492,
      "memory(GiB)": 135.77,
      "step": 30830,
      "train_speed(iter/s)": 0.201761
    },
    {
      "acc": 0.76509042,
      "epoch": 0.7195205529389463,
      "grad_norm": 11.5625,
      "learning_rate": 7.409264987378473e-06,
      "loss": 0.83569603,
      "memory(GiB)": 135.77,
      "step": 30840,
      "train_speed(iter/s)": 0.201795
    },
    {
      "acc": 0.75531335,
      "epoch": 0.7197538605112352,
      "grad_norm": 6.625,
      "learning_rate": 7.407609477475334e-06,
      "loss": 0.87194748,
      "memory(GiB)": 135.77,
      "step": 30850,
      "train_speed(iter/s)": 0.201829
    },
    {
      "acc": 0.78215508,
      "epoch": 0.7199871680835241,
      "grad_norm": 4.90625,
      "learning_rate": 7.405953623886642e-06,
      "loss": 0.78222771,
      "memory(GiB)": 135.77,
      "step": 30860,
      "train_speed(iter/s)": 0.201862
    },
    {
      "acc": 0.77151766,
      "epoch": 0.720220475655813,
      "grad_norm": 9.6875,
      "learning_rate": 7.404297426848768e-06,
      "loss": 0.82328587,
      "memory(GiB)": 135.77,
      "step": 30870,
      "train_speed(iter/s)": 0.201896
    },
    {
      "acc": 0.74816885,
      "epoch": 0.7204537832281019,
      "grad_norm": 7.03125,
      "learning_rate": 7.4026408865981335e-06,
      "loss": 0.90460873,
      "memory(GiB)": 135.77,
      "step": 30880,
      "train_speed(iter/s)": 0.201931
    },
    {
      "acc": 0.77453132,
      "epoch": 0.7206870908003908,
      "grad_norm": 6.71875,
      "learning_rate": 7.400984003371211e-06,
      "loss": 0.83123531,
      "memory(GiB)": 135.77,
      "step": 30890,
      "train_speed(iter/s)": 0.201965
    },
    {
      "acc": 0.78442779,
      "epoch": 0.7209203983726797,
      "grad_norm": 5.75,
      "learning_rate": 7.3993267774045206e-06,
      "loss": 0.7607954,
      "memory(GiB)": 135.77,
      "step": 30900,
      "train_speed(iter/s)": 0.201998
    },
    {
      "acc": 0.79025621,
      "epoch": 0.7211537059449685,
      "grad_norm": 13.8125,
      "learning_rate": 7.397669208934628e-06,
      "loss": 0.77729826,
      "memory(GiB)": 135.77,
      "step": 30910,
      "train_speed(iter/s)": 0.202029
    },
    {
      "acc": 0.7896234,
      "epoch": 0.7213870135172574,
      "grad_norm": 4.25,
      "learning_rate": 7.396011298198155e-06,
      "loss": 0.74766607,
      "memory(GiB)": 135.77,
      "step": 30920,
      "train_speed(iter/s)": 0.202061
    },
    {
      "acc": 0.76902571,
      "epoch": 0.7216203210895463,
      "grad_norm": 5.625,
      "learning_rate": 7.394353045431765e-06,
      "loss": 0.83471451,
      "memory(GiB)": 135.77,
      "step": 30930,
      "train_speed(iter/s)": 0.202094
    },
    {
      "acc": 0.76524734,
      "epoch": 0.7218536286618352,
      "grad_norm": 6.84375,
      "learning_rate": 7.392694450872171e-06,
      "loss": 0.84251699,
      "memory(GiB)": 135.77,
      "step": 30940,
      "train_speed(iter/s)": 0.202129
    },
    {
      "acc": 0.75952892,
      "epoch": 0.7220869362341241,
      "grad_norm": 10.0625,
      "learning_rate": 7.3910355147561394e-06,
      "loss": 0.86127234,
      "memory(GiB)": 135.77,
      "step": 30950,
      "train_speed(iter/s)": 0.202164
    },
    {
      "acc": 0.78890076,
      "epoch": 0.722320243806413,
      "grad_norm": 5.09375,
      "learning_rate": 7.389376237320485e-06,
      "loss": 0.73546257,
      "memory(GiB)": 135.77,
      "step": 30960,
      "train_speed(iter/s)": 0.202198
    },
    {
      "acc": 0.79085941,
      "epoch": 0.7225535513787019,
      "grad_norm": 4.46875,
      "learning_rate": 7.387716618802064e-06,
      "loss": 0.73829441,
      "memory(GiB)": 135.77,
      "step": 30970,
      "train_speed(iter/s)": 0.20223
    },
    {
      "acc": 0.76388206,
      "epoch": 0.7227868589509908,
      "grad_norm": 3.796875,
      "learning_rate": 7.386056659437792e-06,
      "loss": 0.84366646,
      "memory(GiB)": 135.77,
      "step": 30980,
      "train_speed(iter/s)": 0.202265
    },
    {
      "acc": 0.77729826,
      "epoch": 0.7230201665232797,
      "grad_norm": 5.21875,
      "learning_rate": 7.384396359464623e-06,
      "loss": 0.80204477,
      "memory(GiB)": 135.77,
      "step": 30990,
      "train_speed(iter/s)": 0.202296
    },
    {
      "acc": 0.77947817,
      "epoch": 0.7232534740955686,
      "grad_norm": 4.5,
      "learning_rate": 7.382735719119568e-06,
      "loss": 0.77528582,
      "memory(GiB)": 135.77,
      "step": 31000,
      "train_speed(iter/s)": 0.20233
    },
    {
      "epoch": 0.7232534740955686,
      "eval_acc": 0.7419815382526118,
      "eval_loss": 0.8143436312675476,
      "eval_runtime": 1269.6984,
      "eval_samples_per_second": 28.346,
      "eval_steps_per_second": 14.173,
      "step": 31000
    },
    {
      "acc": 0.77952061,
      "epoch": 0.7234867816678575,
      "grad_norm": 5.15625,
      "learning_rate": 7.38107473863968e-06,
      "loss": 0.80758448,
      "memory(GiB)": 135.77,
      "step": 31010,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.76929088,
      "epoch": 0.7237200892401464,
      "grad_norm": 4.75,
      "learning_rate": 7.3794134182620646e-06,
      "loss": 0.84603424,
      "memory(GiB)": 135.77,
      "step": 31020,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.78471966,
      "epoch": 0.7239533968124353,
      "grad_norm": 5.625,
      "learning_rate": 7.377751758223876e-06,
      "loss": 0.77363634,
      "memory(GiB)": 135.77,
      "step": 31030,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.75998449,
      "epoch": 0.7241867043847242,
      "grad_norm": 5.96875,
      "learning_rate": 7.376089758762315e-06,
      "loss": 0.86344814,
      "memory(GiB)": 135.77,
      "step": 31040,
      "train_speed(iter/s)": 0.20078
    },
    {
      "acc": 0.79045277,
      "epoch": 0.7244200119570131,
      "grad_norm": 7.46875,
      "learning_rate": 7.374427420114629e-06,
      "loss": 0.74859829,
      "memory(GiB)": 135.77,
      "step": 31050,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.78292723,
      "epoch": 0.724653319529302,
      "grad_norm": 4.34375,
      "learning_rate": 7.37276474251812e-06,
      "loss": 0.78144426,
      "memory(GiB)": 135.77,
      "step": 31060,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.77971792,
      "epoch": 0.7248866271015909,
      "grad_norm": 4.3125,
      "learning_rate": 7.371101726210135e-06,
      "loss": 0.79532051,
      "memory(GiB)": 135.77,
      "step": 31070,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.7901329,
      "epoch": 0.7251199346738798,
      "grad_norm": 7.375,
      "learning_rate": 7.369438371428065e-06,
      "loss": 0.75919361,
      "memory(GiB)": 135.77,
      "step": 31080,
      "train_speed(iter/s)": 0.200907
    },
    {
      "acc": 0.77284174,
      "epoch": 0.7253532422461687,
      "grad_norm": 7.5625,
      "learning_rate": 7.367774678409357e-06,
      "loss": 0.81404877,
      "memory(GiB)": 135.77,
      "step": 31090,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.78015275,
      "epoch": 0.7255865498184575,
      "grad_norm": 4.8125,
      "learning_rate": 7.366110647391501e-06,
      "loss": 0.78611984,
      "memory(GiB)": 135.77,
      "step": 31100,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.79383612,
      "epoch": 0.7258198573907464,
      "grad_norm": 4.65625,
      "learning_rate": 7.364446278612036e-06,
      "loss": 0.7279952,
      "memory(GiB)": 135.77,
      "step": 31110,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.7822289,
      "epoch": 0.7260531649630353,
      "grad_norm": 4.59375,
      "learning_rate": 7.3627815723085535e-06,
      "loss": 0.77796512,
      "memory(GiB)": 135.77,
      "step": 31120,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.77322626,
      "epoch": 0.7262864725353242,
      "grad_norm": 8.75,
      "learning_rate": 7.361116528718688e-06,
      "loss": 0.80861044,
      "memory(GiB)": 135.77,
      "step": 31130,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.77036018,
      "epoch": 0.7265197801076131,
      "grad_norm": 8.0,
      "learning_rate": 7.359451148080123e-06,
      "loss": 0.83087883,
      "memory(GiB)": 135.77,
      "step": 31140,
      "train_speed(iter/s)": 0.201104
    },
    {
      "acc": 0.77294779,
      "epoch": 0.726753087679902,
      "grad_norm": 5.96875,
      "learning_rate": 7.357785430630593e-06,
      "loss": 0.81432848,
      "memory(GiB)": 135.77,
      "step": 31150,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.78335571,
      "epoch": 0.7269863952521909,
      "grad_norm": 4.0625,
      "learning_rate": 7.356119376607877e-06,
      "loss": 0.78180637,
      "memory(GiB)": 135.77,
      "step": 31160,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.75630546,
      "epoch": 0.7272197028244798,
      "grad_norm": 6.03125,
      "learning_rate": 7.354452986249805e-06,
      "loss": 0.88844528,
      "memory(GiB)": 135.77,
      "step": 31170,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.76719637,
      "epoch": 0.7274530103967687,
      "grad_norm": 9.9375,
      "learning_rate": 7.352786259794252e-06,
      "loss": 0.84311905,
      "memory(GiB)": 135.77,
      "step": 31180,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.78604798,
      "epoch": 0.7276863179690576,
      "grad_norm": 6.09375,
      "learning_rate": 7.351119197479144e-06,
      "loss": 0.7596261,
      "memory(GiB)": 135.77,
      "step": 31190,
      "train_speed(iter/s)": 0.201273
    },
    {
      "acc": 0.78300457,
      "epoch": 0.7279196255413465,
      "grad_norm": 6.0,
      "learning_rate": 7.349451799542455e-06,
      "loss": 0.75782876,
      "memory(GiB)": 135.77,
      "step": 31200,
      "train_speed(iter/s)": 0.201308
    },
    {
      "acc": 0.79482107,
      "epoch": 0.7281529331136354,
      "grad_norm": 6.25,
      "learning_rate": 7.3477840662222045e-06,
      "loss": 0.73927917,
      "memory(GiB)": 135.77,
      "step": 31210,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.7678937,
      "epoch": 0.7283862406859243,
      "grad_norm": 4.4375,
      "learning_rate": 7.346115997756459e-06,
      "loss": 0.82241888,
      "memory(GiB)": 135.77,
      "step": 31220,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.78195829,
      "epoch": 0.7286195482582132,
      "grad_norm": 4.65625,
      "learning_rate": 7.3444475943833375e-06,
      "loss": 0.78644238,
      "memory(GiB)": 135.77,
      "step": 31230,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.78667397,
      "epoch": 0.7288528558305021,
      "grad_norm": 6.96875,
      "learning_rate": 7.342778856341002e-06,
      "loss": 0.78144002,
      "memory(GiB)": 135.77,
      "step": 31240,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.75145693,
      "epoch": 0.729086163402791,
      "grad_norm": 4.65625,
      "learning_rate": 7.3411097838676645e-06,
      "loss": 0.92597151,
      "memory(GiB)": 135.77,
      "step": 31250,
      "train_speed(iter/s)": 0.201478
    },
    {
      "acc": 0.78834095,
      "epoch": 0.7293194709750799,
      "grad_norm": 6.09375,
      "learning_rate": 7.339440377201588e-06,
      "loss": 0.76001892,
      "memory(GiB)": 135.77,
      "step": 31260,
      "train_speed(iter/s)": 0.201512
    },
    {
      "acc": 0.76082201,
      "epoch": 0.7295527785473688,
      "grad_norm": 7.3125,
      "learning_rate": 7.337770636581075e-06,
      "loss": 0.87978334,
      "memory(GiB)": 135.77,
      "step": 31270,
      "train_speed(iter/s)": 0.201545
    },
    {
      "acc": 0.78952765,
      "epoch": 0.7297860861196577,
      "grad_norm": 4.5,
      "learning_rate": 7.3361005622444834e-06,
      "loss": 0.76534405,
      "memory(GiB)": 135.77,
      "step": 31280,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.78288898,
      "epoch": 0.7300193936919465,
      "grad_norm": 6.5,
      "learning_rate": 7.334430154430217e-06,
      "loss": 0.79545956,
      "memory(GiB)": 135.77,
      "step": 31290,
      "train_speed(iter/s)": 0.201612
    },
    {
      "acc": 0.76784639,
      "epoch": 0.7302527012642354,
      "grad_norm": 8.0625,
      "learning_rate": 7.332759413376721e-06,
      "loss": 0.83361969,
      "memory(GiB)": 135.77,
      "step": 31300,
      "train_speed(iter/s)": 0.201644
    },
    {
      "acc": 0.78950572,
      "epoch": 0.7304860088365243,
      "grad_norm": 4.5,
      "learning_rate": 7.331088339322499e-06,
      "loss": 0.7704586,
      "memory(GiB)": 135.77,
      "step": 31310,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.77479763,
      "epoch": 0.7307193164088132,
      "grad_norm": 5.15625,
      "learning_rate": 7.3294169325060925e-06,
      "loss": 0.81133862,
      "memory(GiB)": 135.77,
      "step": 31320,
      "train_speed(iter/s)": 0.20171
    },
    {
      "acc": 0.76270666,
      "epoch": 0.7309526239811021,
      "grad_norm": 5.875,
      "learning_rate": 7.327745193166096e-06,
      "loss": 0.86329832,
      "memory(GiB)": 135.77,
      "step": 31330,
      "train_speed(iter/s)": 0.201744
    },
    {
      "acc": 0.77042761,
      "epoch": 0.731185931553391,
      "grad_norm": 8.1875,
      "learning_rate": 7.3260731215411484e-06,
      "loss": 0.84471989,
      "memory(GiB)": 135.77,
      "step": 31340,
      "train_speed(iter/s)": 0.201777
    },
    {
      "acc": 0.76804247,
      "epoch": 0.7314192391256799,
      "grad_norm": 7.75,
      "learning_rate": 7.32440071786994e-06,
      "loss": 0.84253273,
      "memory(GiB)": 135.77,
      "step": 31350,
      "train_speed(iter/s)": 0.20181
    },
    {
      "acc": 0.77282157,
      "epoch": 0.7316525466979688,
      "grad_norm": 6.4375,
      "learning_rate": 7.322727982391203e-06,
      "loss": 0.82727652,
      "memory(GiB)": 135.77,
      "step": 31360,
      "train_speed(iter/s)": 0.201842
    },
    {
      "acc": 0.7756979,
      "epoch": 0.7318858542702577,
      "grad_norm": 6.59375,
      "learning_rate": 7.321054915343722e-06,
      "loss": 0.82158718,
      "memory(GiB)": 135.77,
      "step": 31370,
      "train_speed(iter/s)": 0.201877
    },
    {
      "acc": 0.75797024,
      "epoch": 0.7321191618425466,
      "grad_norm": 6.5625,
      "learning_rate": 7.3193815169663266e-06,
      "loss": 0.87084713,
      "memory(GiB)": 135.77,
      "step": 31380,
      "train_speed(iter/s)": 0.201911
    },
    {
      "acc": 0.7460907,
      "epoch": 0.7323524694148354,
      "grad_norm": 6.90625,
      "learning_rate": 7.317707787497892e-06,
      "loss": 0.90985165,
      "memory(GiB)": 135.77,
      "step": 31390,
      "train_speed(iter/s)": 0.201946
    },
    {
      "acc": 0.79898553,
      "epoch": 0.7325857769871243,
      "grad_norm": 6.5,
      "learning_rate": 7.316033727177345e-06,
      "loss": 0.72486515,
      "memory(GiB)": 135.77,
      "step": 31400,
      "train_speed(iter/s)": 0.20198
    },
    {
      "acc": 0.78949633,
      "epoch": 0.7328190845594132,
      "grad_norm": 4.9375,
      "learning_rate": 7.314359336243656e-06,
      "loss": 0.77411165,
      "memory(GiB)": 135.77,
      "step": 31410,
      "train_speed(iter/s)": 0.202011
    },
    {
      "acc": 0.77078295,
      "epoch": 0.7330523921317021,
      "grad_norm": 5.53125,
      "learning_rate": 7.312684614935846e-06,
      "loss": 0.80220585,
      "memory(GiB)": 135.77,
      "step": 31420,
      "train_speed(iter/s)": 0.20204
    },
    {
      "acc": 0.78742967,
      "epoch": 0.733285699703991,
      "grad_norm": 6.0625,
      "learning_rate": 7.311009563492977e-06,
      "loss": 0.77101741,
      "memory(GiB)": 135.77,
      "step": 31430,
      "train_speed(iter/s)": 0.202071
    },
    {
      "acc": 0.78280091,
      "epoch": 0.7335190072762799,
      "grad_norm": 5.125,
      "learning_rate": 7.309334182154164e-06,
      "loss": 0.78739457,
      "memory(GiB)": 135.77,
      "step": 31440,
      "train_speed(iter/s)": 0.202105
    },
    {
      "acc": 0.78376379,
      "epoch": 0.7337523148485688,
      "grad_norm": 8.9375,
      "learning_rate": 7.307658471158567e-06,
      "loss": 0.78450098,
      "memory(GiB)": 135.77,
      "step": 31450,
      "train_speed(iter/s)": 0.202138
    },
    {
      "acc": 0.79066029,
      "epoch": 0.7339856224208577,
      "grad_norm": 4.875,
      "learning_rate": 7.305982430745395e-06,
      "loss": 0.75622902,
      "memory(GiB)": 135.77,
      "step": 31460,
      "train_speed(iter/s)": 0.202173
    },
    {
      "acc": 0.769873,
      "epoch": 0.7342189299931466,
      "grad_norm": 3.984375,
      "learning_rate": 7.3043060611538995e-06,
      "loss": 0.82575455,
      "memory(GiB)": 135.77,
      "step": 31470,
      "train_speed(iter/s)": 0.202204
    },
    {
      "acc": 0.77303362,
      "epoch": 0.7344522375654354,
      "grad_norm": 6.5,
      "learning_rate": 7.302629362623384e-06,
      "loss": 0.81925964,
      "memory(GiB)": 135.77,
      "step": 31480,
      "train_speed(iter/s)": 0.202236
    },
    {
      "acc": 0.76517944,
      "epoch": 0.7346855451377243,
      "grad_norm": 5.125,
      "learning_rate": 7.3009523353931966e-06,
      "loss": 0.83144541,
      "memory(GiB)": 135.77,
      "step": 31490,
      "train_speed(iter/s)": 0.202271
    },
    {
      "acc": 0.78074522,
      "epoch": 0.7349188527100132,
      "grad_norm": 5.03125,
      "learning_rate": 7.299274979702732e-06,
      "loss": 0.79223261,
      "memory(GiB)": 135.77,
      "step": 31500,
      "train_speed(iter/s)": 0.202305
    },
    {
      "epoch": 0.7349188527100132,
      "eval_acc": 0.741985381714187,
      "eval_loss": 0.8141977787017822,
      "eval_runtime": 1269.7118,
      "eval_samples_per_second": 28.346,
      "eval_steps_per_second": 14.173,
      "step": 31500
    },
    {
      "acc": 0.78632488,
      "epoch": 0.7351521602823021,
      "grad_norm": 3.75,
      "learning_rate": 7.29759729579143e-06,
      "loss": 0.78005762,
      "memory(GiB)": 135.77,
      "step": 31510,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.77185574,
      "epoch": 0.735385467854591,
      "grad_norm": 7.78125,
      "learning_rate": 7.295919283898782e-06,
      "loss": 0.82155762,
      "memory(GiB)": 135.77,
      "step": 31520,
      "train_speed(iter/s)": 0.200713
    },
    {
      "acc": 0.77478409,
      "epoch": 0.7356187754268799,
      "grad_norm": 3.828125,
      "learning_rate": 7.294240944264323e-06,
      "loss": 0.80740013,
      "memory(GiB)": 135.77,
      "step": 31530,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.79292769,
      "epoch": 0.7358520829991688,
      "grad_norm": 4.78125,
      "learning_rate": 7.292562277127637e-06,
      "loss": 0.7431262,
      "memory(GiB)": 135.77,
      "step": 31540,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.79306731,
      "epoch": 0.7360853905714577,
      "grad_norm": 4.6875,
      "learning_rate": 7.290883282728352e-06,
      "loss": 0.74995842,
      "memory(GiB)": 135.77,
      "step": 31550,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.77480698,
      "epoch": 0.7363186981437466,
      "grad_norm": 5.6875,
      "learning_rate": 7.289203961306143e-06,
      "loss": 0.82969999,
      "memory(GiB)": 135.77,
      "step": 31560,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.76988225,
      "epoch": 0.7365520057160355,
      "grad_norm": 6.90625,
      "learning_rate": 7.287524313100735e-06,
      "loss": 0.82661495,
      "memory(GiB)": 135.77,
      "step": 31570,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.75901461,
      "epoch": 0.7367853132883244,
      "grad_norm": 5.15625,
      "learning_rate": 7.285844338351894e-06,
      "loss": 0.87335548,
      "memory(GiB)": 135.77,
      "step": 31580,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.77605467,
      "epoch": 0.7370186208606133,
      "grad_norm": 7.65625,
      "learning_rate": 7.284164037299438e-06,
      "loss": 0.80913677,
      "memory(GiB)": 135.77,
      "step": 31590,
      "train_speed(iter/s)": 0.200947
    },
    {
      "acc": 0.79126759,
      "epoch": 0.7372519284329022,
      "grad_norm": 7.125,
      "learning_rate": 7.28248341018323e-06,
      "loss": 0.72892237,
      "memory(GiB)": 135.77,
      "step": 31600,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.78675766,
      "epoch": 0.7374852360051911,
      "grad_norm": 5.375,
      "learning_rate": 7.280802457243178e-06,
      "loss": 0.76203346,
      "memory(GiB)": 135.77,
      "step": 31610,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.77029533,
      "epoch": 0.73771854357748,
      "grad_norm": 5.84375,
      "learning_rate": 7.2791211787192376e-06,
      "loss": 0.8017519,
      "memory(GiB)": 135.77,
      "step": 31620,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.76799793,
      "epoch": 0.7379518511497689,
      "grad_norm": 5.09375,
      "learning_rate": 7.27743957485141e-06,
      "loss": 0.83303394,
      "memory(GiB)": 135.77,
      "step": 31630,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.77150707,
      "epoch": 0.7381851587220578,
      "grad_norm": 4.5,
      "learning_rate": 7.2757576458797465e-06,
      "loss": 0.83672428,
      "memory(GiB)": 135.77,
      "step": 31640,
      "train_speed(iter/s)": 0.201114
    },
    {
      "acc": 0.77037902,
      "epoch": 0.7384184662943467,
      "grad_norm": 6.65625,
      "learning_rate": 7.27407539204434e-06,
      "loss": 0.83148661,
      "memory(GiB)": 135.77,
      "step": 31650,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.76977768,
      "epoch": 0.7386517738666356,
      "grad_norm": 6.0,
      "learning_rate": 7.272392813585332e-06,
      "loss": 0.81001329,
      "memory(GiB)": 135.77,
      "step": 31660,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.80247803,
      "epoch": 0.7388850814389245,
      "grad_norm": 3.875,
      "learning_rate": 7.270709910742908e-06,
      "loss": 0.70846043,
      "memory(GiB)": 135.77,
      "step": 31670,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.78375702,
      "epoch": 0.7391183890112133,
      "grad_norm": 4.9375,
      "learning_rate": 7.269026683757306e-06,
      "loss": 0.78587623,
      "memory(GiB)": 135.77,
      "step": 31680,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.76821456,
      "epoch": 0.7393516965835022,
      "grad_norm": 4.84375,
      "learning_rate": 7.267343132868803e-06,
      "loss": 0.82991257,
      "memory(GiB)": 135.77,
      "step": 31690,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.78773546,
      "epoch": 0.7395850041557911,
      "grad_norm": 8.5,
      "learning_rate": 7.265659258317725e-06,
      "loss": 0.77818766,
      "memory(GiB)": 135.77,
      "step": 31700,
      "train_speed(iter/s)": 0.201314
    },
    {
      "acc": 0.77224665,
      "epoch": 0.73981831172808,
      "grad_norm": 5.375,
      "learning_rate": 7.263975060344449e-06,
      "loss": 0.81635828,
      "memory(GiB)": 135.77,
      "step": 31710,
      "train_speed(iter/s)": 0.201345
    },
    {
      "acc": 0.79697518,
      "epoch": 0.7400516193003689,
      "grad_norm": 5.46875,
      "learning_rate": 7.26229053918939e-06,
      "loss": 0.74089785,
      "memory(GiB)": 135.77,
      "step": 31720,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.79544964,
      "epoch": 0.7402849268726578,
      "grad_norm": 4.78125,
      "learning_rate": 7.260605695093014e-06,
      "loss": 0.69942369,
      "memory(GiB)": 135.77,
      "step": 31730,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.77880449,
      "epoch": 0.7405182344449467,
      "grad_norm": 5.40625,
      "learning_rate": 7.25892052829583e-06,
      "loss": 0.79678488,
      "memory(GiB)": 135.77,
      "step": 31740,
      "train_speed(iter/s)": 0.201442
    },
    {
      "acc": 0.78319736,
      "epoch": 0.7407515420172356,
      "grad_norm": 5.3125,
      "learning_rate": 7.257235039038397e-06,
      "loss": 0.76695404,
      "memory(GiB)": 135.77,
      "step": 31750,
      "train_speed(iter/s)": 0.201476
    },
    {
      "acc": 0.78057632,
      "epoch": 0.7409848495895245,
      "grad_norm": 6.59375,
      "learning_rate": 7.25554922756132e-06,
      "loss": 0.77434168,
      "memory(GiB)": 135.77,
      "step": 31760,
      "train_speed(iter/s)": 0.20151
    },
    {
      "acc": 0.78480177,
      "epoch": 0.7412181571618134,
      "grad_norm": 4.59375,
      "learning_rate": 7.253863094105243e-06,
      "loss": 0.76221313,
      "memory(GiB)": 135.77,
      "step": 31770,
      "train_speed(iter/s)": 0.201541
    },
    {
      "acc": 0.78093719,
      "epoch": 0.7414514647341023,
      "grad_norm": 5.46875,
      "learning_rate": 7.252176638910867e-06,
      "loss": 0.79842443,
      "memory(GiB)": 135.77,
      "step": 31780,
      "train_speed(iter/s)": 0.201573
    },
    {
      "acc": 0.78112631,
      "epoch": 0.7416847723063912,
      "grad_norm": 7.46875,
      "learning_rate": 7.25048986221893e-06,
      "loss": 0.79473867,
      "memory(GiB)": 135.77,
      "step": 31790,
      "train_speed(iter/s)": 0.201606
    },
    {
      "acc": 0.76295033,
      "epoch": 0.7419180798786801,
      "grad_norm": 5.5625,
      "learning_rate": 7.248802764270217e-06,
      "loss": 0.86861839,
      "memory(GiB)": 135.77,
      "step": 31800,
      "train_speed(iter/s)": 0.201638
    },
    {
      "acc": 0.7788949,
      "epoch": 0.742151387450969,
      "grad_norm": 5.34375,
      "learning_rate": 7.247115345305564e-06,
      "loss": 0.79273958,
      "memory(GiB)": 135.77,
      "step": 31810,
      "train_speed(iter/s)": 0.201671
    },
    {
      "acc": 0.78000889,
      "epoch": 0.7423846950232579,
      "grad_norm": 6.15625,
      "learning_rate": 7.245427605565847e-06,
      "loss": 0.77720737,
      "memory(GiB)": 135.77,
      "step": 31820,
      "train_speed(iter/s)": 0.201705
    },
    {
      "acc": 0.77433429,
      "epoch": 0.7426180025955468,
      "grad_norm": 5.8125,
      "learning_rate": 7.243739545291994e-06,
      "loss": 0.83012581,
      "memory(GiB)": 135.77,
      "step": 31830,
      "train_speed(iter/s)": 0.201738
    },
    {
      "acc": 0.77186079,
      "epoch": 0.7428513101678357,
      "grad_norm": 7.53125,
      "learning_rate": 7.24205116472497e-06,
      "loss": 0.81709404,
      "memory(GiB)": 135.77,
      "step": 31840,
      "train_speed(iter/s)": 0.20177
    },
    {
      "acc": 0.77884254,
      "epoch": 0.7430846177401246,
      "grad_norm": 8.1875,
      "learning_rate": 7.240362464105795e-06,
      "loss": 0.79129939,
      "memory(GiB)": 135.77,
      "step": 31850,
      "train_speed(iter/s)": 0.201803
    },
    {
      "acc": 0.77921705,
      "epoch": 0.7433179253124135,
      "grad_norm": 4.96875,
      "learning_rate": 7.238673443675529e-06,
      "loss": 0.78770342,
      "memory(GiB)": 135.77,
      "step": 31860,
      "train_speed(iter/s)": 0.201836
    },
    {
      "acc": 0.79026484,
      "epoch": 0.7435512328847023,
      "grad_norm": 5.09375,
      "learning_rate": 7.236984103675278e-06,
      "loss": 0.73446693,
      "memory(GiB)": 135.77,
      "step": 31870,
      "train_speed(iter/s)": 0.20187
    },
    {
      "acc": 0.76663461,
      "epoch": 0.7437845404569912,
      "grad_norm": 4.71875,
      "learning_rate": 7.235294444346197e-06,
      "loss": 0.85680876,
      "memory(GiB)": 135.77,
      "step": 31880,
      "train_speed(iter/s)": 0.201904
    },
    {
      "acc": 0.76660204,
      "epoch": 0.7440178480292801,
      "grad_norm": 7.84375,
      "learning_rate": 7.233604465929485e-06,
      "loss": 0.83537083,
      "memory(GiB)": 135.77,
      "step": 31890,
      "train_speed(iter/s)": 0.201937
    },
    {
      "acc": 0.78674612,
      "epoch": 0.744251155601569,
      "grad_norm": 4.59375,
      "learning_rate": 7.231914168666382e-06,
      "loss": 0.73860254,
      "memory(GiB)": 135.77,
      "step": 31900,
      "train_speed(iter/s)": 0.201973
    },
    {
      "acc": 0.76229892,
      "epoch": 0.7444844631738579,
      "grad_norm": 7.0,
      "learning_rate": 7.23022355279818e-06,
      "loss": 0.86733685,
      "memory(GiB)": 135.77,
      "step": 31910,
      "train_speed(iter/s)": 0.202005
    },
    {
      "acc": 0.76368475,
      "epoch": 0.7447177707461468,
      "grad_norm": 5.0625,
      "learning_rate": 7.228532618566214e-06,
      "loss": 0.8780262,
      "memory(GiB)": 135.77,
      "step": 31920,
      "train_speed(iter/s)": 0.202034
    },
    {
      "acc": 0.78525524,
      "epoch": 0.7449510783184357,
      "grad_norm": 6.34375,
      "learning_rate": 7.226841366211865e-06,
      "loss": 0.77032347,
      "memory(GiB)": 135.77,
      "step": 31930,
      "train_speed(iter/s)": 0.202065
    },
    {
      "acc": 0.77381964,
      "epoch": 0.7451843858907246,
      "grad_norm": 5.3125,
      "learning_rate": 7.225149795976558e-06,
      "loss": 0.79736581,
      "memory(GiB)": 135.77,
      "step": 31940,
      "train_speed(iter/s)": 0.202097
    },
    {
      "acc": 0.78271856,
      "epoch": 0.7454176934630135,
      "grad_norm": 6.1875,
      "learning_rate": 7.223457908101763e-06,
      "loss": 0.769701,
      "memory(GiB)": 135.77,
      "step": 31950,
      "train_speed(iter/s)": 0.202131
    },
    {
      "acc": 0.77590184,
      "epoch": 0.7456510010353024,
      "grad_norm": 7.28125,
      "learning_rate": 7.2217657028289974e-06,
      "loss": 0.81713409,
      "memory(GiB)": 135.77,
      "step": 31960,
      "train_speed(iter/s)": 0.202164
    },
    {
      "acc": 0.79009981,
      "epoch": 0.7458843086075913,
      "grad_norm": 5.46875,
      "learning_rate": 7.220073180399824e-06,
      "loss": 0.76196728,
      "memory(GiB)": 135.77,
      "step": 31970,
      "train_speed(iter/s)": 0.202199
    },
    {
      "acc": 0.79176197,
      "epoch": 0.7461176161798801,
      "grad_norm": 5.0,
      "learning_rate": 7.218380341055848e-06,
      "loss": 0.74932814,
      "memory(GiB)": 135.77,
      "step": 31980,
      "train_speed(iter/s)": 0.202231
    },
    {
      "acc": 0.75970087,
      "epoch": 0.746350923752169,
      "grad_norm": 4.625,
      "learning_rate": 7.216687185038724e-06,
      "loss": 0.87884674,
      "memory(GiB)": 135.77,
      "step": 31990,
      "train_speed(iter/s)": 0.202261
    },
    {
      "acc": 0.77163577,
      "epoch": 0.746584231324458,
      "grad_norm": 6.3125,
      "learning_rate": 7.214993712590148e-06,
      "loss": 0.85210857,
      "memory(GiB)": 135.77,
      "step": 32000,
      "train_speed(iter/s)": 0.202296
    },
    {
      "epoch": 0.746584231324458,
      "eval_acc": 0.7420050794547601,
      "eval_loss": 0.8138281106948853,
      "eval_runtime": 1270.8326,
      "eval_samples_per_second": 28.321,
      "eval_steps_per_second": 14.161,
      "step": 32000
    },
    {
      "acc": 0.76468716,
      "epoch": 0.7468175388967468,
      "grad_norm": 100.5,
      "learning_rate": 7.213299923951863e-06,
      "loss": 0.88969822,
      "memory(GiB)": 135.77,
      "step": 32010,
      "train_speed(iter/s)": 0.200694
    },
    {
      "acc": 0.77527003,
      "epoch": 0.7470508464690357,
      "grad_norm": 5.59375,
      "learning_rate": 7.211605819365657e-06,
      "loss": 0.82085714,
      "memory(GiB)": 135.77,
      "step": 32020,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.77617159,
      "epoch": 0.7472841540413246,
      "grad_norm": 6.375,
      "learning_rate": 7.209911399073361e-06,
      "loss": 0.83005333,
      "memory(GiB)": 135.77,
      "step": 32030,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.77418909,
      "epoch": 0.7475174616136135,
      "grad_norm": 6.0,
      "learning_rate": 7.208216663316856e-06,
      "loss": 0.79054193,
      "memory(GiB)": 135.77,
      "step": 32040,
      "train_speed(iter/s)": 0.200791
    },
    {
      "acc": 0.78965578,
      "epoch": 0.7477507691859024,
      "grad_norm": 4.4375,
      "learning_rate": 7.206521612338064e-06,
      "loss": 0.74522982,
      "memory(GiB)": 135.77,
      "step": 32050,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.76647468,
      "epoch": 0.7479840767581912,
      "grad_norm": 6.28125,
      "learning_rate": 7.204826246378953e-06,
      "loss": 0.81988811,
      "memory(GiB)": 135.77,
      "step": 32060,
      "train_speed(iter/s)": 0.200858
    },
    {
      "acc": 0.77853847,
      "epoch": 0.7482173843304801,
      "grad_norm": 4.9375,
      "learning_rate": 7.203130565681537e-06,
      "loss": 0.79634762,
      "memory(GiB)": 135.77,
      "step": 32070,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.7814064,
      "epoch": 0.748450691902769,
      "grad_norm": 7.0,
      "learning_rate": 7.201434570487871e-06,
      "loss": 0.78645287,
      "memory(GiB)": 135.77,
      "step": 32080,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77312088,
      "epoch": 0.7486839994750579,
      "grad_norm": 4.78125,
      "learning_rate": 7.199738261040059e-06,
      "loss": 0.83004112,
      "memory(GiB)": 135.77,
      "step": 32090,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.79272485,
      "epoch": 0.7489173070473468,
      "grad_norm": 5.9375,
      "learning_rate": 7.1980416375802494e-06,
      "loss": 0.7410665,
      "memory(GiB)": 135.77,
      "step": 32100,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.76613655,
      "epoch": 0.7491506146196357,
      "grad_norm": 5.46875,
      "learning_rate": 7.196344700350635e-06,
      "loss": 0.86439323,
      "memory(GiB)": 135.77,
      "step": 32110,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.78310966,
      "epoch": 0.7493839221919246,
      "grad_norm": 5.375,
      "learning_rate": 7.1946474495934535e-06,
      "loss": 0.765874,
      "memory(GiB)": 135.77,
      "step": 32120,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.79054403,
      "epoch": 0.7496172297642135,
      "grad_norm": 4.5,
      "learning_rate": 7.192949885550986e-06,
      "loss": 0.73549843,
      "memory(GiB)": 135.77,
      "step": 32130,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.77819924,
      "epoch": 0.7498505373365024,
      "grad_norm": 4.625,
      "learning_rate": 7.1912520084655594e-06,
      "loss": 0.79788404,
      "memory(GiB)": 135.77,
      "step": 32140,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.79777637,
      "epoch": 0.7500838449087913,
      "grad_norm": 6.8125,
      "learning_rate": 7.189553818579545e-06,
      "loss": 0.70898933,
      "memory(GiB)": 135.77,
      "step": 32150,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.77831688,
      "epoch": 0.7503171524810802,
      "grad_norm": 4.9375,
      "learning_rate": 7.187855316135358e-06,
      "loss": 0.79211369,
      "memory(GiB)": 135.77,
      "step": 32160,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.77343702,
      "epoch": 0.7505504600533691,
      "grad_norm": 5.1875,
      "learning_rate": 7.1861565013754605e-06,
      "loss": 0.81101351,
      "memory(GiB)": 135.77,
      "step": 32170,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.77341728,
      "epoch": 0.750783767625658,
      "grad_norm": 6.625,
      "learning_rate": 7.18445737454236e-06,
      "loss": 0.81564074,
      "memory(GiB)": 135.77,
      "step": 32180,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.78846521,
      "epoch": 0.7510170751979469,
      "grad_norm": 4.5625,
      "learning_rate": 7.182757935878601e-06,
      "loss": 0.75658441,
      "memory(GiB)": 135.77,
      "step": 32190,
      "train_speed(iter/s)": 0.201264
    },
    {
      "acc": 0.77531843,
      "epoch": 0.7512503827702358,
      "grad_norm": 4.90625,
      "learning_rate": 7.1810581856267815e-06,
      "loss": 0.82540045,
      "memory(GiB)": 135.77,
      "step": 32200,
      "train_speed(iter/s)": 0.201295
    },
    {
      "acc": 0.78034782,
      "epoch": 0.7514836903425247,
      "grad_norm": 14.75,
      "learning_rate": 7.17935812402954e-06,
      "loss": 0.78024654,
      "memory(GiB)": 135.77,
      "step": 32210,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.76428576,
      "epoch": 0.7517169979148136,
      "grad_norm": 5.0,
      "learning_rate": 7.177657751329559e-06,
      "loss": 0.84508133,
      "memory(GiB)": 135.77,
      "step": 32220,
      "train_speed(iter/s)": 0.201356
    },
    {
      "acc": 0.77687807,
      "epoch": 0.7519503054871025,
      "grad_norm": 4.6875,
      "learning_rate": 7.1759570677695665e-06,
      "loss": 0.80794582,
      "memory(GiB)": 135.77,
      "step": 32230,
      "train_speed(iter/s)": 0.201387
    },
    {
      "acc": 0.75994444,
      "epoch": 0.7521836130593914,
      "grad_norm": 6.78125,
      "learning_rate": 7.174256073592335e-06,
      "loss": 0.86417484,
      "memory(GiB)": 135.77,
      "step": 32240,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.79446583,
      "epoch": 0.7524169206316802,
      "grad_norm": 6.5625,
      "learning_rate": 7.172554769040681e-06,
      "loss": 0.72564263,
      "memory(GiB)": 135.77,
      "step": 32250,
      "train_speed(iter/s)": 0.20145
    },
    {
      "acc": 0.77692814,
      "epoch": 0.7526502282039691,
      "grad_norm": 5.1875,
      "learning_rate": 7.1708531543574635e-06,
      "loss": 0.79970336,
      "memory(GiB)": 135.77,
      "step": 32260,
      "train_speed(iter/s)": 0.201483
    },
    {
      "acc": 0.79846258,
      "epoch": 0.752883535776258,
      "grad_norm": 4.75,
      "learning_rate": 7.169151229785589e-06,
      "loss": 0.72652612,
      "memory(GiB)": 135.77,
      "step": 32270,
      "train_speed(iter/s)": 0.201517
    },
    {
      "acc": 0.79553976,
      "epoch": 0.7531168433485469,
      "grad_norm": 5.90625,
      "learning_rate": 7.167448995568009e-06,
      "loss": 0.73974028,
      "memory(GiB)": 135.77,
      "step": 32280,
      "train_speed(iter/s)": 0.201549
    },
    {
      "acc": 0.77979794,
      "epoch": 0.7533501509208358,
      "grad_norm": 9.8125,
      "learning_rate": 7.165746451947713e-06,
      "loss": 0.79384131,
      "memory(GiB)": 135.77,
      "step": 32290,
      "train_speed(iter/s)": 0.201579
    },
    {
      "acc": 0.76877642,
      "epoch": 0.7535834584931247,
      "grad_norm": 5.5625,
      "learning_rate": 7.16404359916774e-06,
      "loss": 0.86544285,
      "memory(GiB)": 135.77,
      "step": 32300,
      "train_speed(iter/s)": 0.20161
    },
    {
      "acc": 0.76505909,
      "epoch": 0.7538167660654136,
      "grad_norm": 10.9375,
      "learning_rate": 7.1623404374711715e-06,
      "loss": 0.85449657,
      "memory(GiB)": 135.77,
      "step": 32310,
      "train_speed(iter/s)": 0.201643
    },
    {
      "acc": 0.78930197,
      "epoch": 0.7540500736377025,
      "grad_norm": 4.125,
      "learning_rate": 7.160636967101134e-06,
      "loss": 0.75111027,
      "memory(GiB)": 135.77,
      "step": 32320,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.79103222,
      "epoch": 0.7542833812099914,
      "grad_norm": 5.9375,
      "learning_rate": 7.1589331883007965e-06,
      "loss": 0.75493755,
      "memory(GiB)": 135.77,
      "step": 32330,
      "train_speed(iter/s)": 0.201708
    },
    {
      "acc": 0.76673937,
      "epoch": 0.7545166887822803,
      "grad_norm": 5.3125,
      "learning_rate": 7.1572291013133745e-06,
      "loss": 0.858673,
      "memory(GiB)": 135.77,
      "step": 32340,
      "train_speed(iter/s)": 0.201739
    },
    {
      "acc": 0.77912583,
      "epoch": 0.7547499963545692,
      "grad_norm": 5.5625,
      "learning_rate": 7.155524706382125e-06,
      "loss": 0.79643412,
      "memory(GiB)": 135.77,
      "step": 32350,
      "train_speed(iter/s)": 0.201771
    },
    {
      "acc": 0.77822075,
      "epoch": 0.7549833039268581,
      "grad_norm": 7.15625,
      "learning_rate": 7.15382000375035e-06,
      "loss": 0.79455347,
      "memory(GiB)": 135.77,
      "step": 32360,
      "train_speed(iter/s)": 0.201803
    },
    {
      "acc": 0.79522676,
      "epoch": 0.755216611499147,
      "grad_norm": 4.0625,
      "learning_rate": 7.152114993661394e-06,
      "loss": 0.71623573,
      "memory(GiB)": 135.77,
      "step": 32370,
      "train_speed(iter/s)": 0.201835
    },
    {
      "acc": 0.77095437,
      "epoch": 0.7554499190714359,
      "grad_norm": 4.03125,
      "learning_rate": 7.150409676358649e-06,
      "loss": 0.83390207,
      "memory(GiB)": 135.77,
      "step": 32380,
      "train_speed(iter/s)": 0.201866
    },
    {
      "acc": 0.76215563,
      "epoch": 0.7556832266437248,
      "grad_norm": 5.28125,
      "learning_rate": 7.148704052085547e-06,
      "loss": 0.87032471,
      "memory(GiB)": 135.77,
      "step": 32390,
      "train_speed(iter/s)": 0.201899
    },
    {
      "acc": 0.77448578,
      "epoch": 0.7559165342160137,
      "grad_norm": 15.5,
      "learning_rate": 7.146998121085566e-06,
      "loss": 0.80818844,
      "memory(GiB)": 135.77,
      "step": 32400,
      "train_speed(iter/s)": 0.20193
    },
    {
      "acc": 0.78134699,
      "epoch": 0.7561498417883026,
      "grad_norm": 4.5,
      "learning_rate": 7.145291883602226e-06,
      "loss": 0.76230822,
      "memory(GiB)": 135.77,
      "step": 32410,
      "train_speed(iter/s)": 0.201961
    },
    {
      "acc": 0.7758811,
      "epoch": 0.7563831493605915,
      "grad_norm": 5.65625,
      "learning_rate": 7.143585339879093e-06,
      "loss": 0.81609783,
      "memory(GiB)": 135.77,
      "step": 32420,
      "train_speed(iter/s)": 0.201993
    },
    {
      "acc": 0.76103873,
      "epoch": 0.7566164569328804,
      "grad_norm": 7.3125,
      "learning_rate": 7.141878490159777e-06,
      "loss": 0.88947668,
      "memory(GiB)": 135.77,
      "step": 32430,
      "train_speed(iter/s)": 0.202027
    },
    {
      "acc": 0.79736533,
      "epoch": 0.7568497645051693,
      "grad_norm": 6.03125,
      "learning_rate": 7.140171334687927e-06,
      "loss": 0.73429241,
      "memory(GiB)": 135.77,
      "step": 32440,
      "train_speed(iter/s)": 0.20206
    },
    {
      "acc": 0.76088171,
      "epoch": 0.7570830720774581,
      "grad_norm": 7.0,
      "learning_rate": 7.138463873707242e-06,
      "loss": 0.87091007,
      "memory(GiB)": 135.77,
      "step": 32450,
      "train_speed(iter/s)": 0.202093
    },
    {
      "acc": 0.77715158,
      "epoch": 0.757316379649747,
      "grad_norm": 4.21875,
      "learning_rate": 7.13675610746146e-06,
      "loss": 0.80676775,
      "memory(GiB)": 135.77,
      "step": 32460,
      "train_speed(iter/s)": 0.202125
    },
    {
      "acc": 0.77457027,
      "epoch": 0.7575496872220359,
      "grad_norm": 4.65625,
      "learning_rate": 7.135048036194364e-06,
      "loss": 0.81614847,
      "memory(GiB)": 135.77,
      "step": 32470,
      "train_speed(iter/s)": 0.202157
    },
    {
      "acc": 0.76460385,
      "epoch": 0.7577829947943248,
      "grad_norm": 6.46875,
      "learning_rate": 7.13333966014978e-06,
      "loss": 0.84198284,
      "memory(GiB)": 135.77,
      "step": 32480,
      "train_speed(iter/s)": 0.20219
    },
    {
      "acc": 0.77977839,
      "epoch": 0.7580163023666137,
      "grad_norm": 6.625,
      "learning_rate": 7.131630979571581e-06,
      "loss": 0.79072237,
      "memory(GiB)": 135.77,
      "step": 32490,
      "train_speed(iter/s)": 0.20222
    },
    {
      "acc": 0.77567453,
      "epoch": 0.7582496099389026,
      "grad_norm": 4.96875,
      "learning_rate": 7.1299219947036795e-06,
      "loss": 0.81002426,
      "memory(GiB)": 135.77,
      "step": 32500,
      "train_speed(iter/s)": 0.202251
    },
    {
      "epoch": 0.7582496099389026,
      "eval_acc": 0.742288214457469,
      "eval_loss": 0.8132081031799316,
      "eval_runtime": 1270.5334,
      "eval_samples_per_second": 28.327,
      "eval_steps_per_second": 14.164,
      "step": 32500
    },
    {
      "acc": 0.77000389,
      "epoch": 0.7584829175111915,
      "grad_norm": 5.34375,
      "learning_rate": 7.12821270579003e-06,
      "loss": 0.80007992,
      "memory(GiB)": 135.77,
      "step": 32510,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.78935204,
      "epoch": 0.7587162250834804,
      "grad_norm": 5.75,
      "learning_rate": 7.126503113074636e-06,
      "loss": 0.74440832,
      "memory(GiB)": 135.77,
      "step": 32520,
      "train_speed(iter/s)": 0.200709
    },
    {
      "acc": 0.77540474,
      "epoch": 0.7589495326557693,
      "grad_norm": 5.15625,
      "learning_rate": 7.1247932168015396e-06,
      "loss": 0.81684599,
      "memory(GiB)": 135.77,
      "step": 32530,
      "train_speed(iter/s)": 0.20074
    },
    {
      "acc": 0.77324686,
      "epoch": 0.7591828402280582,
      "grad_norm": 5.8125,
      "learning_rate": 7.123083017214829e-06,
      "loss": 0.79553862,
      "memory(GiB)": 135.77,
      "step": 32540,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.77071304,
      "epoch": 0.759416147800347,
      "grad_norm": 8.5,
      "learning_rate": 7.121372514558635e-06,
      "loss": 0.83940582,
      "memory(GiB)": 135.77,
      "step": 32550,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.76741009,
      "epoch": 0.759649455372636,
      "grad_norm": 5.28125,
      "learning_rate": 7.1196617090771305e-06,
      "loss": 0.83808603,
      "memory(GiB)": 135.77,
      "step": 32560,
      "train_speed(iter/s)": 0.200835
    },
    {
      "acc": 0.79151773,
      "epoch": 0.7598827629449248,
      "grad_norm": 4.40625,
      "learning_rate": 7.1179506010145335e-06,
      "loss": 0.75109611,
      "memory(GiB)": 135.77,
      "step": 32570,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.76236095,
      "epoch": 0.7601160705172137,
      "grad_norm": 3.59375,
      "learning_rate": 7.116239190615104e-06,
      "loss": 0.87826719,
      "memory(GiB)": 135.77,
      "step": 32580,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.76941271,
      "epoch": 0.7603493780895026,
      "grad_norm": 5.15625,
      "learning_rate": 7.1145274781231435e-06,
      "loss": 0.84898701,
      "memory(GiB)": 135.77,
      "step": 32590,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.7833683,
      "epoch": 0.7605826856617915,
      "grad_norm": 7.25,
      "learning_rate": 7.112815463782998e-06,
      "loss": 0.79329114,
      "memory(GiB)": 135.77,
      "step": 32600,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.77949257,
      "epoch": 0.7608159932340804,
      "grad_norm": 5.6875,
      "learning_rate": 7.111103147839062e-06,
      "loss": 0.78569312,
      "memory(GiB)": 135.77,
      "step": 32610,
      "train_speed(iter/s)": 0.200995
    },
    {
      "acc": 0.79783745,
      "epoch": 0.7610493008063693,
      "grad_norm": 5.46875,
      "learning_rate": 7.109390530535762e-06,
      "loss": 0.698351,
      "memory(GiB)": 135.77,
      "step": 32620,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.77755065,
      "epoch": 0.7612826083786582,
      "grad_norm": 6.71875,
      "learning_rate": 7.1076776121175794e-06,
      "loss": 0.82539043,
      "memory(GiB)": 135.77,
      "step": 32630,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.78371029,
      "epoch": 0.761515915950947,
      "grad_norm": 5.25,
      "learning_rate": 7.105964392829029e-06,
      "loss": 0.77058702,
      "memory(GiB)": 135.77,
      "step": 32640,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.76712589,
      "epoch": 0.7617492235232359,
      "grad_norm": 5.125,
      "learning_rate": 7.104250872914673e-06,
      "loss": 0.83793221,
      "memory(GiB)": 135.77,
      "step": 32650,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.78511324,
      "epoch": 0.7619825310955248,
      "grad_norm": 17.625,
      "learning_rate": 7.102537052619116e-06,
      "loss": 0.7639678,
      "memory(GiB)": 135.77,
      "step": 32660,
      "train_speed(iter/s)": 0.201156
    },
    {
      "acc": 0.78138227,
      "epoch": 0.7622158386678137,
      "grad_norm": 5.1875,
      "learning_rate": 7.100822932187006e-06,
      "loss": 0.78408322,
      "memory(GiB)": 135.77,
      "step": 32670,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.78767548,
      "epoch": 0.7624491462401026,
      "grad_norm": 5.5,
      "learning_rate": 7.099108511863032e-06,
      "loss": 0.75953522,
      "memory(GiB)": 135.77,
      "step": 32680,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.77689095,
      "epoch": 0.7626824538123915,
      "grad_norm": 4.9375,
      "learning_rate": 7.097393791891929e-06,
      "loss": 0.8115571,
      "memory(GiB)": 135.77,
      "step": 32690,
      "train_speed(iter/s)": 0.201248
    },
    {
      "acc": 0.79620209,
      "epoch": 0.7629157613846804,
      "grad_norm": 6.0,
      "learning_rate": 7.095678772518471e-06,
      "loss": 0.70747728,
      "memory(GiB)": 135.77,
      "step": 32700,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.77352734,
      "epoch": 0.7631490689569693,
      "grad_norm": 10.5625,
      "learning_rate": 7.093963453987476e-06,
      "loss": 0.81851959,
      "memory(GiB)": 135.77,
      "step": 32710,
      "train_speed(iter/s)": 0.201315
    },
    {
      "acc": 0.7703311,
      "epoch": 0.7633823765292582,
      "grad_norm": 6.03125,
      "learning_rate": 7.092247836543808e-06,
      "loss": 0.84133596,
      "memory(GiB)": 135.77,
      "step": 32720,
      "train_speed(iter/s)": 0.201343
    },
    {
      "acc": 0.79012804,
      "epoch": 0.7636156841015471,
      "grad_norm": 5.46875,
      "learning_rate": 7.090531920432368e-06,
      "loss": 0.7244441,
      "memory(GiB)": 135.77,
      "step": 32730,
      "train_speed(iter/s)": 0.201373
    },
    {
      "acc": 0.78044481,
      "epoch": 0.763848991673836,
      "grad_norm": 7.59375,
      "learning_rate": 7.088815705898103e-06,
      "loss": 0.80008049,
      "memory(GiB)": 135.77,
      "step": 32740,
      "train_speed(iter/s)": 0.201404
    },
    {
      "acc": 0.75984888,
      "epoch": 0.7640822992461249,
      "grad_norm": 4.46875,
      "learning_rate": 7.0870991931860044e-06,
      "loss": 0.85378962,
      "memory(GiB)": 135.77,
      "step": 32750,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.80223875,
      "epoch": 0.7643156068184138,
      "grad_norm": 5.40625,
      "learning_rate": 7.0853823825411005e-06,
      "loss": 0.69828153,
      "memory(GiB)": 135.77,
      "step": 32760,
      "train_speed(iter/s)": 0.201466
    },
    {
      "acc": 0.78329859,
      "epoch": 0.7645489143907027,
      "grad_norm": 8.0,
      "learning_rate": 7.083665274208469e-06,
      "loss": 0.75291567,
      "memory(GiB)": 135.77,
      "step": 32770,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.79095221,
      "epoch": 0.7647822219629916,
      "grad_norm": 5.1875,
      "learning_rate": 7.081947868433223e-06,
      "loss": 0.75146961,
      "memory(GiB)": 135.77,
      "step": 32780,
      "train_speed(iter/s)": 0.201528
    },
    {
      "acc": 0.75557308,
      "epoch": 0.7650155295352805,
      "grad_norm": 6.875,
      "learning_rate": 7.0802301654605255e-06,
      "loss": 0.88731985,
      "memory(GiB)": 135.77,
      "step": 32790,
      "train_speed(iter/s)": 0.201561
    },
    {
      "acc": 0.7902915,
      "epoch": 0.7652488371075694,
      "grad_norm": 5.375,
      "learning_rate": 7.078512165535576e-06,
      "loss": 0.74065533,
      "memory(GiB)": 135.77,
      "step": 32800,
      "train_speed(iter/s)": 0.201593
    },
    {
      "acc": 0.78259544,
      "epoch": 0.7654821446798583,
      "grad_norm": 5.75,
      "learning_rate": 7.076793868903617e-06,
      "loss": 0.77633686,
      "memory(GiB)": 135.77,
      "step": 32810,
      "train_speed(iter/s)": 0.201624
    },
    {
      "acc": 0.77786484,
      "epoch": 0.7657154522521472,
      "grad_norm": 3.9375,
      "learning_rate": 7.0750752758099384e-06,
      "loss": 0.81483288,
      "memory(GiB)": 135.77,
      "step": 32820,
      "train_speed(iter/s)": 0.201655
    },
    {
      "acc": 0.78939443,
      "epoch": 0.765948759824436,
      "grad_norm": 3.96875,
      "learning_rate": 7.073356386499865e-06,
      "loss": 0.76333141,
      "memory(GiB)": 135.77,
      "step": 32830,
      "train_speed(iter/s)": 0.201684
    },
    {
      "acc": 0.76923409,
      "epoch": 0.7661820673967249,
      "grad_norm": 5.21875,
      "learning_rate": 7.071637201218772e-06,
      "loss": 0.83631258,
      "memory(GiB)": 135.77,
      "step": 32840,
      "train_speed(iter/s)": 0.201716
    },
    {
      "acc": 0.78508043,
      "epoch": 0.7664153749690138,
      "grad_norm": 5.5,
      "learning_rate": 7.06991772021207e-06,
      "loss": 0.77355008,
      "memory(GiB)": 135.77,
      "step": 32850,
      "train_speed(iter/s)": 0.201748
    },
    {
      "acc": 0.78850126,
      "epoch": 0.7666486825413027,
      "grad_norm": 4.875,
      "learning_rate": 7.068197943725214e-06,
      "loss": 0.77419615,
      "memory(GiB)": 135.77,
      "step": 32860,
      "train_speed(iter/s)": 0.201779
    },
    {
      "acc": 0.77871504,
      "epoch": 0.7668819901135916,
      "grad_norm": 11.375,
      "learning_rate": 7.0664778720037034e-06,
      "loss": 0.78229799,
      "memory(GiB)": 135.77,
      "step": 32870,
      "train_speed(iter/s)": 0.201812
    },
    {
      "acc": 0.78440037,
      "epoch": 0.7671152976858805,
      "grad_norm": 7.875,
      "learning_rate": 7.064757505293075e-06,
      "loss": 0.77941685,
      "memory(GiB)": 135.77,
      "step": 32880,
      "train_speed(iter/s)": 0.201844
    },
    {
      "acc": 0.75921545,
      "epoch": 0.7673486052581694,
      "grad_norm": 6.0625,
      "learning_rate": 7.063036843838913e-06,
      "loss": 0.85018616,
      "memory(GiB)": 135.77,
      "step": 32890,
      "train_speed(iter/s)": 0.201875
    },
    {
      "acc": 0.77700405,
      "epoch": 0.7675819128304583,
      "grad_norm": 5.21875,
      "learning_rate": 7.061315887886841e-06,
      "loss": 0.80813074,
      "memory(GiB)": 135.77,
      "step": 32900,
      "train_speed(iter/s)": 0.201908
    },
    {
      "acc": 0.78043146,
      "epoch": 0.7678152204027472,
      "grad_norm": 6.46875,
      "learning_rate": 7.059594637682526e-06,
      "loss": 0.7803668,
      "memory(GiB)": 135.77,
      "step": 32910,
      "train_speed(iter/s)": 0.201936
    },
    {
      "acc": 0.76210136,
      "epoch": 0.7680485279750361,
      "grad_norm": 6.25,
      "learning_rate": 7.057873093471673e-06,
      "loss": 0.85728846,
      "memory(GiB)": 135.77,
      "step": 32920,
      "train_speed(iter/s)": 0.201968
    },
    {
      "acc": 0.78007121,
      "epoch": 0.768281835547325,
      "grad_norm": 4.8125,
      "learning_rate": 7.056151255500036e-06,
      "loss": 0.77832518,
      "memory(GiB)": 135.77,
      "step": 32930,
      "train_speed(iter/s)": 0.202
    },
    {
      "acc": 0.75704546,
      "epoch": 0.7685151431196139,
      "grad_norm": 5.59375,
      "learning_rate": 7.0544291240134025e-06,
      "loss": 0.88629284,
      "memory(GiB)": 135.77,
      "step": 32940,
      "train_speed(iter/s)": 0.202032
    },
    {
      "acc": 0.7763031,
      "epoch": 0.7687484506919028,
      "grad_norm": 16.625,
      "learning_rate": 7.052706699257609e-06,
      "loss": 0.79555397,
      "memory(GiB)": 135.77,
      "step": 32950,
      "train_speed(iter/s)": 0.202064
    },
    {
      "acc": 0.77789965,
      "epoch": 0.7689817582641917,
      "grad_norm": 7.25,
      "learning_rate": 7.05098398147853e-06,
      "loss": 0.80233002,
      "memory(GiB)": 135.77,
      "step": 32960,
      "train_speed(iter/s)": 0.202096
    },
    {
      "acc": 0.77640162,
      "epoch": 0.7692150658364806,
      "grad_norm": 6.3125,
      "learning_rate": 7.0492609709220835e-06,
      "loss": 0.81024342,
      "memory(GiB)": 135.77,
      "step": 32970,
      "train_speed(iter/s)": 0.202129
    },
    {
      "acc": 0.79174838,
      "epoch": 0.7694483734087695,
      "grad_norm": 4.5625,
      "learning_rate": 7.04753766783423e-06,
      "loss": 0.75231781,
      "memory(GiB)": 135.77,
      "step": 32980,
      "train_speed(iter/s)": 0.20216
    },
    {
      "acc": 0.77766156,
      "epoch": 0.7696816809810584,
      "grad_norm": 6.46875,
      "learning_rate": 7.045814072460968e-06,
      "loss": 0.78604116,
      "memory(GiB)": 135.77,
      "step": 32990,
      "train_speed(iter/s)": 0.202192
    },
    {
      "acc": 0.76858015,
      "epoch": 0.7699149885533473,
      "grad_norm": 4.96875,
      "learning_rate": 7.044090185048343e-06,
      "loss": 0.86624451,
      "memory(GiB)": 135.77,
      "step": 33000,
      "train_speed(iter/s)": 0.202225
    },
    {
      "epoch": 0.7699149885533473,
      "eval_acc": 0.742469978161131,
      "eval_loss": 0.8125158548355103,
      "eval_runtime": 1269.1292,
      "eval_samples_per_second": 28.359,
      "eval_steps_per_second": 14.18,
      "step": 33000
    },
    {
      "acc": 0.76679192,
      "epoch": 0.7701482961256362,
      "grad_norm": 4.75,
      "learning_rate": 7.042366005842437e-06,
      "loss": 0.83562012,
      "memory(GiB)": 135.77,
      "step": 33010,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.78599691,
      "epoch": 0.770381603697925,
      "grad_norm": 6.6875,
      "learning_rate": 7.040641535089377e-06,
      "loss": 0.82067356,
      "memory(GiB)": 135.77,
      "step": 33020,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.77557378,
      "epoch": 0.7706149112702139,
      "grad_norm": 6.3125,
      "learning_rate": 7.038916773035332e-06,
      "loss": 0.80831299,
      "memory(GiB)": 135.77,
      "step": 33030,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.78095722,
      "epoch": 0.7708482188425028,
      "grad_norm": 6.09375,
      "learning_rate": 7.037191719926507e-06,
      "loss": 0.77209196,
      "memory(GiB)": 135.77,
      "step": 33040,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.78821402,
      "epoch": 0.7710815264147917,
      "grad_norm": 8.0,
      "learning_rate": 7.035466376009157e-06,
      "loss": 0.75552502,
      "memory(GiB)": 135.77,
      "step": 33050,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.7637373,
      "epoch": 0.7713148339870806,
      "grad_norm": 4.25,
      "learning_rate": 7.033740741529573e-06,
      "loss": 0.84696598,
      "memory(GiB)": 135.77,
      "step": 33060,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.79854646,
      "epoch": 0.7715481415593695,
      "grad_norm": 4.65625,
      "learning_rate": 7.03201481673409e-06,
      "loss": 0.69255486,
      "memory(GiB)": 135.77,
      "step": 33070,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.79627194,
      "epoch": 0.7717814491316584,
      "grad_norm": 5.09375,
      "learning_rate": 7.030288601869082e-06,
      "loss": 0.74422112,
      "memory(GiB)": 135.77,
      "step": 33080,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.75897083,
      "epoch": 0.7720147567039473,
      "grad_norm": 5.0625,
      "learning_rate": 7.028562097180965e-06,
      "loss": 0.87837725,
      "memory(GiB)": 135.77,
      "step": 33090,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.7633698,
      "epoch": 0.7722480642762362,
      "grad_norm": 6.46875,
      "learning_rate": 7.026835302916198e-06,
      "loss": 0.87001724,
      "memory(GiB)": 135.77,
      "step": 33100,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.77343187,
      "epoch": 0.7724813718485251,
      "grad_norm": 5.8125,
      "learning_rate": 7.025108219321281e-06,
      "loss": 0.82100897,
      "memory(GiB)": 135.77,
      "step": 33110,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.78877459,
      "epoch": 0.772714679420814,
      "grad_norm": 4.28125,
      "learning_rate": 7.023380846642754e-06,
      "loss": 0.76161418,
      "memory(GiB)": 135.77,
      "step": 33120,
      "train_speed(iter/s)": 0.201023
    },
    {
      "acc": 0.77717133,
      "epoch": 0.7729479869931029,
      "grad_norm": 5.15625,
      "learning_rate": 7.021653185127197e-06,
      "loss": 0.79486446,
      "memory(GiB)": 135.77,
      "step": 33130,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.77787819,
      "epoch": 0.7731812945653918,
      "grad_norm": 7.53125,
      "learning_rate": 7.019925235021237e-06,
      "loss": 0.8035284,
      "memory(GiB)": 135.77,
      "step": 33140,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.77615976,
      "epoch": 0.7734146021376807,
      "grad_norm": 5.125,
      "learning_rate": 7.018196996571538e-06,
      "loss": 0.81237202,
      "memory(GiB)": 135.77,
      "step": 33150,
      "train_speed(iter/s)": 0.201114
    },
    {
      "acc": 0.78682756,
      "epoch": 0.7736479097099696,
      "grad_norm": 5.4375,
      "learning_rate": 7.016468470024802e-06,
      "loss": 0.74849882,
      "memory(GiB)": 135.77,
      "step": 33160,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.77469473,
      "epoch": 0.7738812172822584,
      "grad_norm": 5.28125,
      "learning_rate": 7.014739655627778e-06,
      "loss": 0.79529195,
      "memory(GiB)": 135.77,
      "step": 33170,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.7825881,
      "epoch": 0.7741145248545473,
      "grad_norm": 4.5625,
      "learning_rate": 7.013010553627253e-06,
      "loss": 0.76973028,
      "memory(GiB)": 135.77,
      "step": 33180,
      "train_speed(iter/s)": 0.201207
    },
    {
      "acc": 0.79780211,
      "epoch": 0.7743478324268362,
      "grad_norm": 5.59375,
      "learning_rate": 7.011281164270056e-06,
      "loss": 0.72503815,
      "memory(GiB)": 135.77,
      "step": 33190,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.78128815,
      "epoch": 0.7745811399991251,
      "grad_norm": 5.71875,
      "learning_rate": 7.009551487803058e-06,
      "loss": 0.7804842,
      "memory(GiB)": 135.77,
      "step": 33200,
      "train_speed(iter/s)": 0.201271
    },
    {
      "acc": 0.78946109,
      "epoch": 0.774814447571414,
      "grad_norm": 6.375,
      "learning_rate": 7.0078215244731685e-06,
      "loss": 0.74558954,
      "memory(GiB)": 135.77,
      "step": 33210,
      "train_speed(iter/s)": 0.201302
    },
    {
      "acc": 0.77743354,
      "epoch": 0.7750477551437028,
      "grad_norm": 5.875,
      "learning_rate": 7.00609127452734e-06,
      "loss": 0.80112734,
      "memory(GiB)": 135.77,
      "step": 33220,
      "train_speed(iter/s)": 0.201333
    },
    {
      "acc": 0.78358135,
      "epoch": 0.7752810627159917,
      "grad_norm": 5.625,
      "learning_rate": 7.0043607382125645e-06,
      "loss": 0.77119946,
      "memory(GiB)": 135.77,
      "step": 33230,
      "train_speed(iter/s)": 0.201365
    },
    {
      "acc": 0.78574867,
      "epoch": 0.7755143702882806,
      "grad_norm": 4.78125,
      "learning_rate": 7.002629915775876e-06,
      "loss": 0.77710333,
      "memory(GiB)": 135.77,
      "step": 33240,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.77628012,
      "epoch": 0.7757476778605695,
      "grad_norm": 6.6875,
      "learning_rate": 7.000898807464349e-06,
      "loss": 0.84035797,
      "memory(GiB)": 135.77,
      "step": 33250,
      "train_speed(iter/s)": 0.201424
    },
    {
      "acc": 0.78520012,
      "epoch": 0.7759809854328584,
      "grad_norm": 4.40625,
      "learning_rate": 6.999167413525099e-06,
      "loss": 0.76161404,
      "memory(GiB)": 135.77,
      "step": 33260,
      "train_speed(iter/s)": 0.201455
    },
    {
      "acc": 0.78030691,
      "epoch": 0.7762142930051473,
      "grad_norm": 6.15625,
      "learning_rate": 6.9974357342052805e-06,
      "loss": 0.79453855,
      "memory(GiB)": 135.77,
      "step": 33270,
      "train_speed(iter/s)": 0.201487
    },
    {
      "acc": 0.79090891,
      "epoch": 0.7764476005774362,
      "grad_norm": 4.65625,
      "learning_rate": 6.995703769752091e-06,
      "loss": 0.73884726,
      "memory(GiB)": 135.77,
      "step": 33280,
      "train_speed(iter/s)": 0.201521
    },
    {
      "acc": 0.804249,
      "epoch": 0.7766809081497251,
      "grad_norm": 4.875,
      "learning_rate": 6.993971520412769e-06,
      "loss": 0.70278206,
      "memory(GiB)": 135.77,
      "step": 33290,
      "train_speed(iter/s)": 0.201554
    },
    {
      "acc": 0.78132133,
      "epoch": 0.776914215722014,
      "grad_norm": 6.875,
      "learning_rate": 6.992238986434591e-06,
      "loss": 0.79448328,
      "memory(GiB)": 135.77,
      "step": 33300,
      "train_speed(iter/s)": 0.201587
    },
    {
      "acc": 0.77671242,
      "epoch": 0.7771475232943029,
      "grad_norm": 4.75,
      "learning_rate": 6.9905061680648765e-06,
      "loss": 0.80529804,
      "memory(GiB)": 135.77,
      "step": 33310,
      "train_speed(iter/s)": 0.201617
    },
    {
      "acc": 0.78557901,
      "epoch": 0.7773808308665918,
      "grad_norm": 5.03125,
      "learning_rate": 6.9887730655509855e-06,
      "loss": 0.76596785,
      "memory(GiB)": 135.77,
      "step": 33320,
      "train_speed(iter/s)": 0.201648
    },
    {
      "acc": 0.7802887,
      "epoch": 0.7776141384388807,
      "grad_norm": 8.375,
      "learning_rate": 6.987039679140316e-06,
      "loss": 0.79592075,
      "memory(GiB)": 135.77,
      "step": 33330,
      "train_speed(iter/s)": 0.20168
    },
    {
      "acc": 0.77079372,
      "epoch": 0.7778474460111696,
      "grad_norm": 6.1875,
      "learning_rate": 6.9853060090803105e-06,
      "loss": 0.8273881,
      "memory(GiB)": 135.77,
      "step": 33340,
      "train_speed(iter/s)": 0.201712
    },
    {
      "acc": 0.76755457,
      "epoch": 0.7780807535834585,
      "grad_norm": 6.09375,
      "learning_rate": 6.983572055618449e-06,
      "loss": 0.84588947,
      "memory(GiB)": 135.77,
      "step": 33350,
      "train_speed(iter/s)": 0.201741
    },
    {
      "acc": 0.7820858,
      "epoch": 0.7783140611557474,
      "grad_norm": 3.640625,
      "learning_rate": 6.981837819002252e-06,
      "loss": 0.78927259,
      "memory(GiB)": 135.77,
      "step": 33360,
      "train_speed(iter/s)": 0.201773
    },
    {
      "acc": 0.78572111,
      "epoch": 0.7785473687280363,
      "grad_norm": 4.625,
      "learning_rate": 6.980103299479281e-06,
      "loss": 0.76833,
      "memory(GiB)": 135.77,
      "step": 33370,
      "train_speed(iter/s)": 0.201804
    },
    {
      "acc": 0.78051925,
      "epoch": 0.7787806763003252,
      "grad_norm": 5.65625,
      "learning_rate": 6.978368497297143e-06,
      "loss": 0.79444427,
      "memory(GiB)": 135.77,
      "step": 33380,
      "train_speed(iter/s)": 0.201834
    },
    {
      "acc": 0.76521983,
      "epoch": 0.7790139838726141,
      "grad_norm": 4.84375,
      "learning_rate": 6.976633412703474e-06,
      "loss": 0.85772009,
      "memory(GiB)": 135.77,
      "step": 33390,
      "train_speed(iter/s)": 0.201864
    },
    {
      "acc": 0.79295754,
      "epoch": 0.779247291444903,
      "grad_norm": 5.28125,
      "learning_rate": 6.974898045945959e-06,
      "loss": 0.74175,
      "memory(GiB)": 135.77,
      "step": 33400,
      "train_speed(iter/s)": 0.201895
    },
    {
      "acc": 0.78126574,
      "epoch": 0.7794805990171918,
      "grad_norm": 6.03125,
      "learning_rate": 6.973162397272323e-06,
      "loss": 0.79368715,
      "memory(GiB)": 135.77,
      "step": 33410,
      "train_speed(iter/s)": 0.201925
    },
    {
      "acc": 0.79608712,
      "epoch": 0.7797139065894807,
      "grad_norm": 6.59375,
      "learning_rate": 6.971426466930327e-06,
      "loss": 0.74015937,
      "memory(GiB)": 135.77,
      "step": 33420,
      "train_speed(iter/s)": 0.201954
    },
    {
      "acc": 0.78107915,
      "epoch": 0.7799472141617696,
      "grad_norm": 4.8125,
      "learning_rate": 6.969690255167777e-06,
      "loss": 0.77917404,
      "memory(GiB)": 135.77,
      "step": 33430,
      "train_speed(iter/s)": 0.201986
    },
    {
      "acc": 0.79521189,
      "epoch": 0.7801805217340585,
      "grad_norm": 6.90625,
      "learning_rate": 6.9679537622325154e-06,
      "loss": 0.73068018,
      "memory(GiB)": 135.77,
      "step": 33440,
      "train_speed(iter/s)": 0.202018
    },
    {
      "acc": 0.80074768,
      "epoch": 0.7804138293063474,
      "grad_norm": 4.8125,
      "learning_rate": 6.966216988372424e-06,
      "loss": 0.72729893,
      "memory(GiB)": 135.77,
      "step": 33450,
      "train_speed(iter/s)": 0.202049
    },
    {
      "acc": 0.78108997,
      "epoch": 0.7806471368786363,
      "grad_norm": 4.21875,
      "learning_rate": 6.964479933835429e-06,
      "loss": 0.78009777,
      "memory(GiB)": 135.77,
      "step": 33460,
      "train_speed(iter/s)": 0.202079
    },
    {
      "acc": 0.76426306,
      "epoch": 0.7808804444509252,
      "grad_norm": 6.15625,
      "learning_rate": 6.962742598869495e-06,
      "loss": 0.87753067,
      "memory(GiB)": 135.77,
      "step": 33470,
      "train_speed(iter/s)": 0.202111
    },
    {
      "acc": 0.77631526,
      "epoch": 0.7811137520232141,
      "grad_norm": 5.46875,
      "learning_rate": 6.961004983722625e-06,
      "loss": 0.80698681,
      "memory(GiB)": 135.77,
      "step": 33480,
      "train_speed(iter/s)": 0.20214
    },
    {
      "acc": 0.75766716,
      "epoch": 0.781347059595503,
      "grad_norm": 5.5,
      "learning_rate": 6.959267088642864e-06,
      "loss": 0.89244823,
      "memory(GiB)": 135.77,
      "step": 33490,
      "train_speed(iter/s)": 0.202169
    },
    {
      "acc": 0.77791281,
      "epoch": 0.7815803671677919,
      "grad_norm": 6.0625,
      "learning_rate": 6.9575289138782944e-06,
      "loss": 0.78410869,
      "memory(GiB)": 135.77,
      "step": 33500,
      "train_speed(iter/s)": 0.202201
    },
    {
      "epoch": 0.7815803671677919,
      "eval_acc": 0.7424501202763256,
      "eval_loss": 0.8124200105667114,
      "eval_runtime": 1269.5877,
      "eval_samples_per_second": 28.349,
      "eval_steps_per_second": 14.175,
      "step": 33500
    },
    {
      "acc": 0.79002161,
      "epoch": 0.7818136747400808,
      "grad_norm": 5.15625,
      "learning_rate": 6.955790459677041e-06,
      "loss": 0.74639416,
      "memory(GiB)": 135.77,
      "step": 33510,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.7705555,
      "epoch": 0.7820469823123697,
      "grad_norm": 4.90625,
      "learning_rate": 6.9540517262872675e-06,
      "loss": 0.82598076,
      "memory(GiB)": 135.77,
      "step": 33520,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.76916981,
      "epoch": 0.7822802898846586,
      "grad_norm": 6.28125,
      "learning_rate": 6.952312713957179e-06,
      "loss": 0.82480154,
      "memory(GiB)": 135.77,
      "step": 33530,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.77232513,
      "epoch": 0.7825135974569475,
      "grad_norm": 4.875,
      "learning_rate": 6.9505734229350155e-06,
      "loss": 0.8362215,
      "memory(GiB)": 135.77,
      "step": 33540,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.78856249,
      "epoch": 0.7827469050292364,
      "grad_norm": 5.0625,
      "learning_rate": 6.948833853469065e-06,
      "loss": 0.76099873,
      "memory(GiB)": 135.77,
      "step": 33550,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.78895731,
      "epoch": 0.7829802126015253,
      "grad_norm": 4.25,
      "learning_rate": 6.947094005807646e-06,
      "loss": 0.76939201,
      "memory(GiB)": 135.77,
      "step": 33560,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.76670189,
      "epoch": 0.7832135201738142,
      "grad_norm": 4.59375,
      "learning_rate": 6.945353880199124e-06,
      "loss": 0.84993162,
      "memory(GiB)": 135.77,
      "step": 33570,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.76875811,
      "epoch": 0.7834468277461031,
      "grad_norm": 5.25,
      "learning_rate": 6.943613476891902e-06,
      "loss": 0.84574566,
      "memory(GiB)": 135.77,
      "step": 33580,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.77109642,
      "epoch": 0.783680135318392,
      "grad_norm": 4.3125,
      "learning_rate": 6.941872796134419e-06,
      "loss": 0.82063408,
      "memory(GiB)": 135.77,
      "step": 33590,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.76629744,
      "epoch": 0.7839134428906808,
      "grad_norm": 4.1875,
      "learning_rate": 6.940131838175159e-06,
      "loss": 0.85809727,
      "memory(GiB)": 135.77,
      "step": 33600,
      "train_speed(iter/s)": 0.200955
    },
    {
      "acc": 0.77465167,
      "epoch": 0.7841467504629697,
      "grad_norm": 5.3125,
      "learning_rate": 6.938390603262644e-06,
      "loss": 0.80659037,
      "memory(GiB)": 135.77,
      "step": 33610,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.77369032,
      "epoch": 0.7843800580352586,
      "grad_norm": 5.09375,
      "learning_rate": 6.936649091645431e-06,
      "loss": 0.82941837,
      "memory(GiB)": 135.77,
      "step": 33620,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.79202423,
      "epoch": 0.7846133656075475,
      "grad_norm": 4.5,
      "learning_rate": 6.9349073035721235e-06,
      "loss": 0.7462399,
      "memory(GiB)": 135.77,
      "step": 33630,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.770962,
      "epoch": 0.7848466731798364,
      "grad_norm": 5.6875,
      "learning_rate": 6.933165239291362e-06,
      "loss": 0.82215929,
      "memory(GiB)": 135.77,
      "step": 33640,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78424397,
      "epoch": 0.7850799807521253,
      "grad_norm": 4.90625,
      "learning_rate": 6.931422899051823e-06,
      "loss": 0.76600466,
      "memory(GiB)": 135.77,
      "step": 33650,
      "train_speed(iter/s)": 0.201109
    },
    {
      "acc": 0.76943855,
      "epoch": 0.7853132883244142,
      "grad_norm": 5.09375,
      "learning_rate": 6.929680283102227e-06,
      "loss": 0.83709068,
      "memory(GiB)": 135.77,
      "step": 33660,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.78439827,
      "epoch": 0.7855465958967031,
      "grad_norm": 5.1875,
      "learning_rate": 6.9279373916913305e-06,
      "loss": 0.78921213,
      "memory(GiB)": 135.77,
      "step": 33670,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.78686228,
      "epoch": 0.785779903468992,
      "grad_norm": 4.53125,
      "learning_rate": 6.926194225067932e-06,
      "loss": 0.77492542,
      "memory(GiB)": 135.77,
      "step": 33680,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.79153872,
      "epoch": 0.7860132110412809,
      "grad_norm": 5.78125,
      "learning_rate": 6.924450783480866e-06,
      "loss": 0.74138288,
      "memory(GiB)": 135.77,
      "step": 33690,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.78364296,
      "epoch": 0.7862465186135698,
      "grad_norm": 4.46875,
      "learning_rate": 6.922707067179011e-06,
      "loss": 0.78220735,
      "memory(GiB)": 135.77,
      "step": 33700,
      "train_speed(iter/s)": 0.201263
    },
    {
      "acc": 0.76618633,
      "epoch": 0.7864798261858587,
      "grad_norm": 5.34375,
      "learning_rate": 6.92096307641128e-06,
      "loss": 0.84831276,
      "memory(GiB)": 135.77,
      "step": 33710,
      "train_speed(iter/s)": 0.201293
    },
    {
      "acc": 0.78253098,
      "epoch": 0.7867131337581476,
      "grad_norm": 4.84375,
      "learning_rate": 6.919218811426629e-06,
      "loss": 0.78605328,
      "memory(GiB)": 135.77,
      "step": 33720,
      "train_speed(iter/s)": 0.201324
    },
    {
      "acc": 0.79246273,
      "epoch": 0.7869464413304365,
      "grad_norm": 5.5625,
      "learning_rate": 6.91747427247405e-06,
      "loss": 0.72274895,
      "memory(GiB)": 135.77,
      "step": 33730,
      "train_speed(iter/s)": 0.201356
    },
    {
      "acc": 0.7848577,
      "epoch": 0.7871797489027254,
      "grad_norm": 11.9375,
      "learning_rate": 6.915729459802575e-06,
      "loss": 0.76760449,
      "memory(GiB)": 135.77,
      "step": 33740,
      "train_speed(iter/s)": 0.201388
    },
    {
      "acc": 0.80589046,
      "epoch": 0.7874130564750143,
      "grad_norm": 5.25,
      "learning_rate": 6.913984373661275e-06,
      "loss": 0.68049059,
      "memory(GiB)": 135.77,
      "step": 33750,
      "train_speed(iter/s)": 0.20142
    },
    {
      "acc": 0.7855567,
      "epoch": 0.7876463640473031,
      "grad_norm": 5.09375,
      "learning_rate": 6.9122390142992634e-06,
      "loss": 0.77877674,
      "memory(GiB)": 135.77,
      "step": 33760,
      "train_speed(iter/s)": 0.201453
    },
    {
      "acc": 0.76452026,
      "epoch": 0.787879671619592,
      "grad_norm": 6.875,
      "learning_rate": 6.910493381965687e-06,
      "loss": 0.85262928,
      "memory(GiB)": 135.77,
      "step": 33770,
      "train_speed(iter/s)": 0.201485
    },
    {
      "acc": 0.77853546,
      "epoch": 0.788112979191881,
      "grad_norm": 6.59375,
      "learning_rate": 6.9087474769097366e-06,
      "loss": 0.7785665,
      "memory(GiB)": 135.77,
      "step": 33780,
      "train_speed(iter/s)": 0.201516
    },
    {
      "acc": 0.78647032,
      "epoch": 0.7883462867641697,
      "grad_norm": 6.65625,
      "learning_rate": 6.907001299380639e-06,
      "loss": 0.77372932,
      "memory(GiB)": 135.77,
      "step": 33790,
      "train_speed(iter/s)": 0.201546
    },
    {
      "acc": 0.78218312,
      "epoch": 0.7885795943364586,
      "grad_norm": 5.375,
      "learning_rate": 6.905254849627658e-06,
      "loss": 0.78796425,
      "memory(GiB)": 135.77,
      "step": 33800,
      "train_speed(iter/s)": 0.201576
    },
    {
      "acc": 0.76979618,
      "epoch": 0.7888129019087475,
      "grad_norm": 6.5,
      "learning_rate": 6.9035081279001e-06,
      "loss": 0.82921257,
      "memory(GiB)": 135.77,
      "step": 33810,
      "train_speed(iter/s)": 0.201608
    },
    {
      "acc": 0.79161835,
      "epoch": 0.7890462094810364,
      "grad_norm": 4.09375,
      "learning_rate": 6.901761134447311e-06,
      "loss": 0.76767821,
      "memory(GiB)": 135.77,
      "step": 33820,
      "train_speed(iter/s)": 0.201636
    },
    {
      "acc": 0.79137611,
      "epoch": 0.7892795170533253,
      "grad_norm": 6.375,
      "learning_rate": 6.900013869518673e-06,
      "loss": 0.74047461,
      "memory(GiB)": 135.77,
      "step": 33830,
      "train_speed(iter/s)": 0.201666
    },
    {
      "acc": 0.77348399,
      "epoch": 0.7895128246256142,
      "grad_norm": 7.4375,
      "learning_rate": 6.898266333363607e-06,
      "loss": 0.81377163,
      "memory(GiB)": 135.77,
      "step": 33840,
      "train_speed(iter/s)": 0.201698
    },
    {
      "acc": 0.77934666,
      "epoch": 0.7897461321979031,
      "grad_norm": 7.25,
      "learning_rate": 6.8965185262315725e-06,
      "loss": 0.79142418,
      "memory(GiB)": 135.77,
      "step": 33850,
      "train_speed(iter/s)": 0.201728
    },
    {
      "acc": 0.78089814,
      "epoch": 0.789979439770192,
      "grad_norm": 4.875,
      "learning_rate": 6.89477044837207e-06,
      "loss": 0.77429733,
      "memory(GiB)": 135.77,
      "step": 33860,
      "train_speed(iter/s)": 0.201757
    },
    {
      "acc": 0.7920969,
      "epoch": 0.7902127473424809,
      "grad_norm": 4.8125,
      "learning_rate": 6.893022100034636e-06,
      "loss": 0.74223742,
      "memory(GiB)": 135.77,
      "step": 33870,
      "train_speed(iter/s)": 0.201785
    },
    {
      "acc": 0.78182831,
      "epoch": 0.7904460549147698,
      "grad_norm": 5.28125,
      "learning_rate": 6.891273481468847e-06,
      "loss": 0.79486198,
      "memory(GiB)": 135.77,
      "step": 33880,
      "train_speed(iter/s)": 0.201814
    },
    {
      "acc": 0.78645382,
      "epoch": 0.7906793624870587,
      "grad_norm": 4.40625,
      "learning_rate": 6.889524592924319e-06,
      "loss": 0.75907497,
      "memory(GiB)": 135.77,
      "step": 33890,
      "train_speed(iter/s)": 0.201846
    },
    {
      "acc": 0.78487177,
      "epoch": 0.7909126700593476,
      "grad_norm": 8.5,
      "learning_rate": 6.887775434650704e-06,
      "loss": 0.76534662,
      "memory(GiB)": 135.77,
      "step": 33900,
      "train_speed(iter/s)": 0.201877
    },
    {
      "acc": 0.81099415,
      "epoch": 0.7911459776316365,
      "grad_norm": 5.65625,
      "learning_rate": 6.8860260068976935e-06,
      "loss": 0.66781607,
      "memory(GiB)": 135.77,
      "step": 33910,
      "train_speed(iter/s)": 0.201908
    },
    {
      "acc": 0.7786212,
      "epoch": 0.7913792852039254,
      "grad_norm": 10.1875,
      "learning_rate": 6.884276309915018e-06,
      "loss": 0.77360516,
      "memory(GiB)": 135.77,
      "step": 33920,
      "train_speed(iter/s)": 0.20194
    },
    {
      "acc": 0.77263908,
      "epoch": 0.7916125927762143,
      "grad_norm": 8.3125,
      "learning_rate": 6.882526343952448e-06,
      "loss": 0.81154432,
      "memory(GiB)": 135.77,
      "step": 33930,
      "train_speed(iter/s)": 0.201971
    },
    {
      "acc": 0.7702785,
      "epoch": 0.7918459003485032,
      "grad_norm": 5.375,
      "learning_rate": 6.880776109259788e-06,
      "loss": 0.81274462,
      "memory(GiB)": 135.77,
      "step": 33940,
      "train_speed(iter/s)": 0.202002
    },
    {
      "acc": 0.77342148,
      "epoch": 0.7920792079207921,
      "grad_norm": 4.6875,
      "learning_rate": 6.8790256060868866e-06,
      "loss": 0.80101671,
      "memory(GiB)": 135.77,
      "step": 33950,
      "train_speed(iter/s)": 0.202032
    },
    {
      "acc": 0.78165035,
      "epoch": 0.792312515493081,
      "grad_norm": 6.03125,
      "learning_rate": 6.8772748346836235e-06,
      "loss": 0.80019588,
      "memory(GiB)": 135.77,
      "step": 33960,
      "train_speed(iter/s)": 0.202064
    },
    {
      "acc": 0.7831409,
      "epoch": 0.7925458230653699,
      "grad_norm": 5.75,
      "learning_rate": 6.875523795299925e-06,
      "loss": 0.77129364,
      "memory(GiB)": 135.77,
      "step": 33970,
      "train_speed(iter/s)": 0.202096
    },
    {
      "acc": 0.77243023,
      "epoch": 0.7927791306376588,
      "grad_norm": 5.46875,
      "learning_rate": 6.873772488185747e-06,
      "loss": 0.83966198,
      "memory(GiB)": 135.77,
      "step": 33980,
      "train_speed(iter/s)": 0.202126
    },
    {
      "acc": 0.79374552,
      "epoch": 0.7930124382099476,
      "grad_norm": 5.71875,
      "learning_rate": 6.872020913591092e-06,
      "loss": 0.73950348,
      "memory(GiB)": 135.77,
      "step": 33990,
      "train_speed(iter/s)": 0.202158
    },
    {
      "acc": 0.78279028,
      "epoch": 0.7932457457822365,
      "grad_norm": 5.78125,
      "learning_rate": 6.870269071765997e-06,
      "loss": 0.79019928,
      "memory(GiB)": 135.77,
      "step": 34000,
      "train_speed(iter/s)": 0.202189
    },
    {
      "epoch": 0.7932457457822365,
      "eval_acc": 0.7426201934510297,
      "eval_loss": 0.8120821714401245,
      "eval_runtime": 1267.8986,
      "eval_samples_per_second": 28.386,
      "eval_steps_per_second": 14.194,
      "step": 34000
    },
    {
      "acc": 0.78353124,
      "epoch": 0.7934790533545254,
      "grad_norm": 6.4375,
      "learning_rate": 6.868516962960534e-06,
      "loss": 0.75921488,
      "memory(GiB)": 135.77,
      "step": 34010,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.78991613,
      "epoch": 0.7937123609268143,
      "grad_norm": 4.875,
      "learning_rate": 6.866764587424818e-06,
      "loss": 0.75100489,
      "memory(GiB)": 135.77,
      "step": 34020,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.78309031,
      "epoch": 0.7939456684991032,
      "grad_norm": 5.625,
      "learning_rate": 6.865011945408998e-06,
      "loss": 0.78480835,
      "memory(GiB)": 135.77,
      "step": 34030,
      "train_speed(iter/s)": 0.200748
    },
    {
      "acc": 0.7720583,
      "epoch": 0.7941789760713921,
      "grad_norm": 6.78125,
      "learning_rate": 6.863259037163266e-06,
      "loss": 0.81258068,
      "memory(GiB)": 135.77,
      "step": 34040,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.77889328,
      "epoch": 0.794412283643681,
      "grad_norm": 5.625,
      "learning_rate": 6.8615058629378465e-06,
      "loss": 0.79917364,
      "memory(GiB)": 135.77,
      "step": 34050,
      "train_speed(iter/s)": 0.200807
    },
    {
      "acc": 0.79887199,
      "epoch": 0.7946455912159699,
      "grad_norm": 5.375,
      "learning_rate": 6.859752422983006e-06,
      "loss": 0.72141008,
      "memory(GiB)": 135.77,
      "step": 34060,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.78431706,
      "epoch": 0.7948788987882588,
      "grad_norm": 7.625,
      "learning_rate": 6.857998717549048e-06,
      "loss": 0.7462605,
      "memory(GiB)": 135.77,
      "step": 34070,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.78418608,
      "epoch": 0.7951122063605477,
      "grad_norm": 5.5,
      "learning_rate": 6.856244746886313e-06,
      "loss": 0.77865896,
      "memory(GiB)": 135.77,
      "step": 34080,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.75828471,
      "epoch": 0.7953455139328366,
      "grad_norm": 4.46875,
      "learning_rate": 6.85449051124518e-06,
      "loss": 0.85784569,
      "memory(GiB)": 135.77,
      "step": 34090,
      "train_speed(iter/s)": 0.200935
    },
    {
      "acc": 0.78311863,
      "epoch": 0.7955788215051255,
      "grad_norm": 7.40625,
      "learning_rate": 6.852736010876063e-06,
      "loss": 0.76896353,
      "memory(GiB)": 135.77,
      "step": 34100,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.78136911,
      "epoch": 0.7958121290774144,
      "grad_norm": 4.96875,
      "learning_rate": 6.85098124602942e-06,
      "loss": 0.78809237,
      "memory(GiB)": 135.77,
      "step": 34110,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.78460474,
      "epoch": 0.7960454366497033,
      "grad_norm": 6.125,
      "learning_rate": 6.8492262169557435e-06,
      "loss": 0.74483161,
      "memory(GiB)": 135.77,
      "step": 34120,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.80227928,
      "epoch": 0.7962787442219922,
      "grad_norm": 4.375,
      "learning_rate": 6.847470923905559e-06,
      "loss": 0.71104026,
      "memory(GiB)": 135.77,
      "step": 34130,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.78113799,
      "epoch": 0.7965120517942811,
      "grad_norm": 7.125,
      "learning_rate": 6.845715367129438e-06,
      "loss": 0.79373302,
      "memory(GiB)": 135.77,
      "step": 34140,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.78072958,
      "epoch": 0.79674535936657,
      "grad_norm": 5.5,
      "learning_rate": 6.843959546877985e-06,
      "loss": 0.77883039,
      "memory(GiB)": 135.77,
      "step": 34150,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.78991518,
      "epoch": 0.7969786669388589,
      "grad_norm": 4.78125,
      "learning_rate": 6.842203463401842e-06,
      "loss": 0.74006615,
      "memory(GiB)": 135.77,
      "step": 34160,
      "train_speed(iter/s)": 0.201145
    },
    {
      "acc": 0.76768169,
      "epoch": 0.7972119745111478,
      "grad_norm": 6.8125,
      "learning_rate": 6.84044711695169e-06,
      "loss": 0.8356452,
      "memory(GiB)": 135.77,
      "step": 34170,
      "train_speed(iter/s)": 0.201176
    },
    {
      "acc": 0.78966255,
      "epoch": 0.7974452820834366,
      "grad_norm": 5.21875,
      "learning_rate": 6.838690507778247e-06,
      "loss": 0.75411134,
      "memory(GiB)": 135.77,
      "step": 34180,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78948689,
      "epoch": 0.7976785896557255,
      "grad_norm": 7.0,
      "learning_rate": 6.836933636132267e-06,
      "loss": 0.75135803,
      "memory(GiB)": 135.77,
      "step": 34190,
      "train_speed(iter/s)": 0.201235
    },
    {
      "acc": 0.77856131,
      "epoch": 0.7979118972280144,
      "grad_norm": 8.9375,
      "learning_rate": 6.835176502264544e-06,
      "loss": 0.81702185,
      "memory(GiB)": 135.77,
      "step": 34200,
      "train_speed(iter/s)": 0.201263
    },
    {
      "acc": 0.76752882,
      "epoch": 0.7981452048003033,
      "grad_norm": 7.84375,
      "learning_rate": 6.8334191064259095e-06,
      "loss": 0.85097122,
      "memory(GiB)": 135.77,
      "step": 34210,
      "train_speed(iter/s)": 0.201294
    },
    {
      "acc": 0.76000738,
      "epoch": 0.7983785123725922,
      "grad_norm": 4.375,
      "learning_rate": 6.8316614488672305e-06,
      "loss": 0.88621998,
      "memory(GiB)": 135.77,
      "step": 34220,
      "train_speed(iter/s)": 0.201324
    },
    {
      "acc": 0.77623281,
      "epoch": 0.7986118199448811,
      "grad_norm": 5.875,
      "learning_rate": 6.829903529839411e-06,
      "loss": 0.79441414,
      "memory(GiB)": 135.77,
      "step": 34230,
      "train_speed(iter/s)": 0.201353
    },
    {
      "acc": 0.77455211,
      "epoch": 0.79884512751717,
      "grad_norm": 6.8125,
      "learning_rate": 6.828145349593395e-06,
      "loss": 0.80323877,
      "memory(GiB)": 135.77,
      "step": 34240,
      "train_speed(iter/s)": 0.201383
    },
    {
      "acc": 0.77257781,
      "epoch": 0.7990784350894589,
      "grad_norm": 5.0625,
      "learning_rate": 6.82638690838016e-06,
      "loss": 0.83061676,
      "memory(GiB)": 135.77,
      "step": 34250,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.76844459,
      "epoch": 0.7993117426617478,
      "grad_norm": 4.65625,
      "learning_rate": 6.824628206450724e-06,
      "loss": 0.84785528,
      "memory(GiB)": 135.77,
      "step": 34260,
      "train_speed(iter/s)": 0.201443
    },
    {
      "acc": 0.78014994,
      "epoch": 0.7995450502340367,
      "grad_norm": 6.21875,
      "learning_rate": 6.822869244056143e-06,
      "loss": 0.79277802,
      "memory(GiB)": 135.77,
      "step": 34270,
      "train_speed(iter/s)": 0.201472
    },
    {
      "acc": 0.78358684,
      "epoch": 0.7997783578063256,
      "grad_norm": 5.46875,
      "learning_rate": 6.821110021447506e-06,
      "loss": 0.79087338,
      "memory(GiB)": 135.77,
      "step": 34280,
      "train_speed(iter/s)": 0.201502
    },
    {
      "acc": 0.76477342,
      "epoch": 0.8000116653786145,
      "grad_norm": 5.9375,
      "learning_rate": 6.819350538875944e-06,
      "loss": 0.86270304,
      "memory(GiB)": 135.77,
      "step": 34290,
      "train_speed(iter/s)": 0.201534
    },
    {
      "acc": 0.77901411,
      "epoch": 0.8002449729509034,
      "grad_norm": 4.78125,
      "learning_rate": 6.817590796592621e-06,
      "loss": 0.7760778,
      "memory(GiB)": 135.77,
      "step": 34300,
      "train_speed(iter/s)": 0.201563
    },
    {
      "acc": 0.786831,
      "epoch": 0.8004782805231923,
      "grad_norm": 5.125,
      "learning_rate": 6.815830794848739e-06,
      "loss": 0.75008812,
      "memory(GiB)": 135.77,
      "step": 34310,
      "train_speed(iter/s)": 0.201594
    },
    {
      "acc": 0.77648001,
      "epoch": 0.8007115880954812,
      "grad_norm": 4.46875,
      "learning_rate": 6.8140705338955386e-06,
      "loss": 0.80113869,
      "memory(GiB)": 135.77,
      "step": 34320,
      "train_speed(iter/s)": 0.201625
    },
    {
      "acc": 0.76347504,
      "epoch": 0.80094489566777,
      "grad_norm": 7.34375,
      "learning_rate": 6.812310013984296e-06,
      "loss": 0.85464201,
      "memory(GiB)": 135.77,
      "step": 34330,
      "train_speed(iter/s)": 0.201655
    },
    {
      "acc": 0.77026582,
      "epoch": 0.801178203240059,
      "grad_norm": 5.34375,
      "learning_rate": 6.810549235366325e-06,
      "loss": 0.8082571,
      "memory(GiB)": 135.77,
      "step": 34340,
      "train_speed(iter/s)": 0.201686
    },
    {
      "acc": 0.77789001,
      "epoch": 0.8014115108123478,
      "grad_norm": 6.1875,
      "learning_rate": 6.808788198292977e-06,
      "loss": 0.78789454,
      "memory(GiB)": 135.77,
      "step": 34350,
      "train_speed(iter/s)": 0.201712
    },
    {
      "acc": 0.76755538,
      "epoch": 0.8016448183846367,
      "grad_norm": 6.25,
      "learning_rate": 6.80702690301564e-06,
      "loss": 0.83963947,
      "memory(GiB)": 135.77,
      "step": 34360,
      "train_speed(iter/s)": 0.201742
    },
    {
      "acc": 0.78171759,
      "epoch": 0.8018781259569255,
      "grad_norm": 6.5,
      "learning_rate": 6.805265349785738e-06,
      "loss": 0.78367205,
      "memory(GiB)": 135.77,
      "step": 34370,
      "train_speed(iter/s)": 0.201774
    },
    {
      "acc": 0.77745686,
      "epoch": 0.8021114335292144,
      "grad_norm": 5.90625,
      "learning_rate": 6.80350353885473e-06,
      "loss": 0.79816732,
      "memory(GiB)": 135.77,
      "step": 34380,
      "train_speed(iter/s)": 0.201804
    },
    {
      "acc": 0.76283531,
      "epoch": 0.8023447411015033,
      "grad_norm": 6.9375,
      "learning_rate": 6.801741470474117e-06,
      "loss": 0.86973667,
      "memory(GiB)": 135.77,
      "step": 34390,
      "train_speed(iter/s)": 0.201837
    },
    {
      "acc": 0.7806407,
      "epoch": 0.8025780486737922,
      "grad_norm": 5.0625,
      "learning_rate": 6.799979144895432e-06,
      "loss": 0.79195518,
      "memory(GiB)": 135.77,
      "step": 34400,
      "train_speed(iter/s)": 0.201868
    },
    {
      "acc": 0.75355606,
      "epoch": 0.8028113562460811,
      "grad_norm": 5.40625,
      "learning_rate": 6.798216562370247e-06,
      "loss": 0.90384073,
      "memory(GiB)": 135.77,
      "step": 34410,
      "train_speed(iter/s)": 0.201899
    },
    {
      "acc": 0.77500319,
      "epoch": 0.80304466381837,
      "grad_norm": 7.1875,
      "learning_rate": 6.79645372315017e-06,
      "loss": 0.82036419,
      "memory(GiB)": 135.77,
      "step": 34420,
      "train_speed(iter/s)": 0.201929
    },
    {
      "acc": 0.7839057,
      "epoch": 0.8032779713906589,
      "grad_norm": 4.09375,
      "learning_rate": 6.794690627486846e-06,
      "loss": 0.76988935,
      "memory(GiB)": 135.77,
      "step": 34430,
      "train_speed(iter/s)": 0.201958
    },
    {
      "acc": 0.78427267,
      "epoch": 0.8035112789629478,
      "grad_norm": 5.625,
      "learning_rate": 6.792927275631957e-06,
      "loss": 0.77875066,
      "memory(GiB)": 135.77,
      "step": 34440,
      "train_speed(iter/s)": 0.201985
    },
    {
      "acc": 0.78074837,
      "epoch": 0.8037445865352367,
      "grad_norm": 5.59375,
      "learning_rate": 6.791163667837219e-06,
      "loss": 0.77121277,
      "memory(GiB)": 135.77,
      "step": 34450,
      "train_speed(iter/s)": 0.202014
    },
    {
      "acc": 0.77221909,
      "epoch": 0.8039778941075256,
      "grad_norm": 6.21875,
      "learning_rate": 6.789399804354389e-06,
      "loss": 0.82327061,
      "memory(GiB)": 135.77,
      "step": 34460,
      "train_speed(iter/s)": 0.202044
    },
    {
      "acc": 0.78617983,
      "epoch": 0.8042112016798145,
      "grad_norm": 4.90625,
      "learning_rate": 6.787635685435255e-06,
      "loss": 0.78236666,
      "memory(GiB)": 135.77,
      "step": 34470,
      "train_speed(iter/s)": 0.202072
    },
    {
      "acc": 0.77123075,
      "epoch": 0.8044445092521034,
      "grad_norm": 6.90625,
      "learning_rate": 6.785871311331648e-06,
      "loss": 0.83095236,
      "memory(GiB)": 135.77,
      "step": 34480,
      "train_speed(iter/s)": 0.202104
    },
    {
      "acc": 0.7779706,
      "epoch": 0.8046778168243923,
      "grad_norm": 4.1875,
      "learning_rate": 6.7841066822954284e-06,
      "loss": 0.80731964,
      "memory(GiB)": 135.77,
      "step": 34490,
      "train_speed(iter/s)": 0.202133
    },
    {
      "acc": 0.78801842,
      "epoch": 0.8049111243966812,
      "grad_norm": 7.28125,
      "learning_rate": 6.7823417985784986e-06,
      "loss": 0.74819498,
      "memory(GiB)": 135.77,
      "step": 34500,
      "train_speed(iter/s)": 0.202161
    },
    {
      "epoch": 0.8049111243966812,
      "eval_acc": 0.7425651038351181,
      "eval_loss": 0.8118574619293213,
      "eval_runtime": 1269.6887,
      "eval_samples_per_second": 28.346,
      "eval_steps_per_second": 14.174,
      "step": 34500
    },
    {
      "acc": 0.77680259,
      "epoch": 0.8051444319689701,
      "grad_norm": 5.75,
      "learning_rate": 6.780576660432797e-06,
      "loss": 0.81476183,
      "memory(GiB)": 135.77,
      "step": 34510,
      "train_speed(iter/s)": 0.20068
    },
    {
      "acc": 0.75909719,
      "epoch": 0.805377739541259,
      "grad_norm": 5.90625,
      "learning_rate": 6.778811268110294e-06,
      "loss": 0.86525078,
      "memory(GiB)": 135.77,
      "step": 34520,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.76516881,
      "epoch": 0.8056110471135479,
      "grad_norm": 6.4375,
      "learning_rate": 6.777045621862997e-06,
      "loss": 0.83675842,
      "memory(GiB)": 135.77,
      "step": 34530,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.78065295,
      "epoch": 0.8058443546858368,
      "grad_norm": 7.59375,
      "learning_rate": 6.775279721942954e-06,
      "loss": 0.77787828,
      "memory(GiB)": 135.77,
      "step": 34540,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.78033791,
      "epoch": 0.8060776622581257,
      "grad_norm": 4.53125,
      "learning_rate": 6.773513568602248e-06,
      "loss": 0.78399673,
      "memory(GiB)": 135.77,
      "step": 34550,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.78510847,
      "epoch": 0.8063109698304145,
      "grad_norm": 4.90625,
      "learning_rate": 6.771747162092993e-06,
      "loss": 0.75295162,
      "memory(GiB)": 135.77,
      "step": 34560,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.75793881,
      "epoch": 0.8065442774027034,
      "grad_norm": 6.5625,
      "learning_rate": 6.769980502667348e-06,
      "loss": 0.85168362,
      "memory(GiB)": 135.77,
      "step": 34570,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.77159257,
      "epoch": 0.8067775849749923,
      "grad_norm": 5.8125,
      "learning_rate": 6.7682135905775e-06,
      "loss": 0.7980968,
      "memory(GiB)": 135.77,
      "step": 34580,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.77705431,
      "epoch": 0.8070108925472812,
      "grad_norm": 5.90625,
      "learning_rate": 6.7664464260756745e-06,
      "loss": 0.81086655,
      "memory(GiB)": 135.77,
      "step": 34590,
      "train_speed(iter/s)": 0.200921
    },
    {
      "acc": 0.77288313,
      "epoch": 0.8072442001195701,
      "grad_norm": 5.0,
      "learning_rate": 6.764679009414135e-06,
      "loss": 0.82601871,
      "memory(GiB)": 135.77,
      "step": 34600,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.77113543,
      "epoch": 0.807477507691859,
      "grad_norm": 6.3125,
      "learning_rate": 6.76291134084518e-06,
      "loss": 0.81248341,
      "memory(GiB)": 135.77,
      "step": 34610,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.77865152,
      "epoch": 0.8077108152641479,
      "grad_norm": 5.53125,
      "learning_rate": 6.761143420621141e-06,
      "loss": 0.81596804,
      "memory(GiB)": 135.77,
      "step": 34620,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.75378828,
      "epoch": 0.8079441228364368,
      "grad_norm": 7.0,
      "learning_rate": 6.759375248994393e-06,
      "loss": 0.90001888,
      "memory(GiB)": 135.77,
      "step": 34630,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.79571762,
      "epoch": 0.8081774304087257,
      "grad_norm": 3.953125,
      "learning_rate": 6.757606826217339e-06,
      "loss": 0.72643576,
      "memory(GiB)": 135.77,
      "step": 34640,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.77722073,
      "epoch": 0.8084107379810146,
      "grad_norm": 6.75,
      "learning_rate": 6.755838152542421e-06,
      "loss": 0.81900959,
      "memory(GiB)": 135.77,
      "step": 34650,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.76127481,
      "epoch": 0.8086440455533035,
      "grad_norm": 4.53125,
      "learning_rate": 6.754069228222117e-06,
      "loss": 0.85369015,
      "memory(GiB)": 135.77,
      "step": 34660,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.77648983,
      "epoch": 0.8088773531255924,
      "grad_norm": 4.8125,
      "learning_rate": 6.752300053508939e-06,
      "loss": 0.80511589,
      "memory(GiB)": 135.77,
      "step": 34670,
      "train_speed(iter/s)": 0.201162
    },
    {
      "acc": 0.7765605,
      "epoch": 0.8091106606978813,
      "grad_norm": 5.71875,
      "learning_rate": 6.750530628655437e-06,
      "loss": 0.80522938,
      "memory(GiB)": 135.77,
      "step": 34680,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.77244101,
      "epoch": 0.8093439682701702,
      "grad_norm": 5.0,
      "learning_rate": 6.748760953914198e-06,
      "loss": 0.81097126,
      "memory(GiB)": 135.77,
      "step": 34690,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.76321468,
      "epoch": 0.8095772758424591,
      "grad_norm": 24.375,
      "learning_rate": 6.746991029537841e-06,
      "loss": 0.84920597,
      "memory(GiB)": 135.77,
      "step": 34700,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.76385412,
      "epoch": 0.809810583414748,
      "grad_norm": 6.0,
      "learning_rate": 6.74522085577902e-06,
      "loss": 0.84149389,
      "memory(GiB)": 135.77,
      "step": 34710,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.77857213,
      "epoch": 0.8100438909870369,
      "grad_norm": 7.21875,
      "learning_rate": 6.743450432890431e-06,
      "loss": 0.77943163,
      "memory(GiB)": 135.77,
      "step": 34720,
      "train_speed(iter/s)": 0.201312
    },
    {
      "acc": 0.76652179,
      "epoch": 0.8102771985593258,
      "grad_norm": 5.84375,
      "learning_rate": 6.741679761124798e-06,
      "loss": 0.81627655,
      "memory(GiB)": 135.77,
      "step": 34730,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.78065634,
      "epoch": 0.8105105061316147,
      "grad_norm": 6.375,
      "learning_rate": 6.739908840734885e-06,
      "loss": 0.7564887,
      "memory(GiB)": 135.77,
      "step": 34740,
      "train_speed(iter/s)": 0.201371
    },
    {
      "acc": 0.78042436,
      "epoch": 0.8107438137039036,
      "grad_norm": 5.375,
      "learning_rate": 6.738137671973492e-06,
      "loss": 0.78124051,
      "memory(GiB)": 135.77,
      "step": 34750,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.78339186,
      "epoch": 0.8109771212761924,
      "grad_norm": 7.21875,
      "learning_rate": 6.736366255093449e-06,
      "loss": 0.75932827,
      "memory(GiB)": 135.77,
      "step": 34760,
      "train_speed(iter/s)": 0.201431
    },
    {
      "acc": 0.77975254,
      "epoch": 0.8112104288484813,
      "grad_norm": 5.4375,
      "learning_rate": 6.73459459034763e-06,
      "loss": 0.78994699,
      "memory(GiB)": 135.77,
      "step": 34770,
      "train_speed(iter/s)": 0.20146
    },
    {
      "acc": 0.77287989,
      "epoch": 0.8114437364207702,
      "grad_norm": 6.65625,
      "learning_rate": 6.732822677988935e-06,
      "loss": 0.81505728,
      "memory(GiB)": 135.77,
      "step": 34780,
      "train_speed(iter/s)": 0.20149
    },
    {
      "acc": 0.77970748,
      "epoch": 0.8116770439930591,
      "grad_norm": 7.3125,
      "learning_rate": 6.731050518270307e-06,
      "loss": 0.79578314,
      "memory(GiB)": 135.77,
      "step": 34790,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.77548347,
      "epoch": 0.811910351565348,
      "grad_norm": 5.75,
      "learning_rate": 6.729278111444721e-06,
      "loss": 0.80311823,
      "memory(GiB)": 135.77,
      "step": 34800,
      "train_speed(iter/s)": 0.201549
    },
    {
      "acc": 0.77532001,
      "epoch": 0.8121436591376369,
      "grad_norm": 4.6875,
      "learning_rate": 6.727505457765185e-06,
      "loss": 0.81310711,
      "memory(GiB)": 135.77,
      "step": 34810,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.77574654,
      "epoch": 0.8123769667099258,
      "grad_norm": 6.09375,
      "learning_rate": 6.725732557484748e-06,
      "loss": 0.79087381,
      "memory(GiB)": 135.77,
      "step": 34820,
      "train_speed(iter/s)": 0.201612
    },
    {
      "acc": 0.79127479,
      "epoch": 0.8126102742822147,
      "grad_norm": 5.96875,
      "learning_rate": 6.723959410856489e-06,
      "loss": 0.74439774,
      "memory(GiB)": 135.77,
      "step": 34830,
      "train_speed(iter/s)": 0.201639
    },
    {
      "acc": 0.78656416,
      "epoch": 0.8128435818545036,
      "grad_norm": 6.65625,
      "learning_rate": 6.722186018133525e-06,
      "loss": 0.76410608,
      "memory(GiB)": 135.77,
      "step": 34840,
      "train_speed(iter/s)": 0.20167
    },
    {
      "acc": 0.7781281,
      "epoch": 0.8130768894267925,
      "grad_norm": 7.25,
      "learning_rate": 6.720412379569008e-06,
      "loss": 0.79291306,
      "memory(GiB)": 135.77,
      "step": 34850,
      "train_speed(iter/s)": 0.201701
    },
    {
      "acc": 0.80160904,
      "epoch": 0.8133101969990814,
      "grad_norm": 4.8125,
      "learning_rate": 6.718638495416124e-06,
      "loss": 0.7177484,
      "memory(GiB)": 135.77,
      "step": 34860,
      "train_speed(iter/s)": 0.20173
    },
    {
      "acc": 0.77137651,
      "epoch": 0.8135435045713703,
      "grad_norm": 6.21875,
      "learning_rate": 6.716864365928094e-06,
      "loss": 0.82935047,
      "memory(GiB)": 135.77,
      "step": 34870,
      "train_speed(iter/s)": 0.201759
    },
    {
      "acc": 0.76962833,
      "epoch": 0.8137768121436592,
      "grad_norm": 5.34375,
      "learning_rate": 6.715089991358174e-06,
      "loss": 0.83387051,
      "memory(GiB)": 135.77,
      "step": 34880,
      "train_speed(iter/s)": 0.201789
    },
    {
      "acc": 0.74992967,
      "epoch": 0.814010119715948,
      "grad_norm": 5.625,
      "learning_rate": 6.713315371959656e-06,
      "loss": 0.89630518,
      "memory(GiB)": 135.77,
      "step": 34890,
      "train_speed(iter/s)": 0.201819
    },
    {
      "acc": 0.78066454,
      "epoch": 0.814243427288237,
      "grad_norm": 6.53125,
      "learning_rate": 6.7115405079858656e-06,
      "loss": 0.77816544,
      "memory(GiB)": 135.77,
      "step": 34900,
      "train_speed(iter/s)": 0.201848
    },
    {
      "acc": 0.77989941,
      "epoch": 0.8144767348605259,
      "grad_norm": 5.40625,
      "learning_rate": 6.709765399690164e-06,
      "loss": 0.77061434,
      "memory(GiB)": 135.77,
      "step": 34910,
      "train_speed(iter/s)": 0.20188
    },
    {
      "acc": 0.77812452,
      "epoch": 0.8147100424328148,
      "grad_norm": 4.375,
      "learning_rate": 6.707990047325952e-06,
      "loss": 0.79744072,
      "memory(GiB)": 135.77,
      "step": 34920,
      "train_speed(iter/s)": 0.201909
    },
    {
      "acc": 0.7687603,
      "epoch": 0.8149433500051037,
      "grad_norm": 5.875,
      "learning_rate": 6.706214451146654e-06,
      "loss": 0.82320156,
      "memory(GiB)": 135.77,
      "step": 34930,
      "train_speed(iter/s)": 0.201939
    },
    {
      "acc": 0.7765502,
      "epoch": 0.8151766575773925,
      "grad_norm": 4.96875,
      "learning_rate": 6.70443861140574e-06,
      "loss": 0.78917093,
      "memory(GiB)": 135.77,
      "step": 34940,
      "train_speed(iter/s)": 0.201968
    },
    {
      "acc": 0.7818922,
      "epoch": 0.8154099651496813,
      "grad_norm": 5.875,
      "learning_rate": 6.702662528356709e-06,
      "loss": 0.77447877,
      "memory(GiB)": 135.77,
      "step": 34950,
      "train_speed(iter/s)": 0.201998
    },
    {
      "acc": 0.77006559,
      "epoch": 0.8156432727219702,
      "grad_norm": 5.53125,
      "learning_rate": 6.700886202253096e-06,
      "loss": 0.83356829,
      "memory(GiB)": 135.77,
      "step": 34960,
      "train_speed(iter/s)": 0.202028
    },
    {
      "acc": 0.76896172,
      "epoch": 0.8158765802942591,
      "grad_norm": 6.0625,
      "learning_rate": 6.699109633348473e-06,
      "loss": 0.853722,
      "memory(GiB)": 135.77,
      "step": 34970,
      "train_speed(iter/s)": 0.202058
    },
    {
      "acc": 0.78813248,
      "epoch": 0.816109887866548,
      "grad_norm": 8.75,
      "learning_rate": 6.697332821896443e-06,
      "loss": 0.78391209,
      "memory(GiB)": 135.77,
      "step": 34980,
      "train_speed(iter/s)": 0.202086
    },
    {
      "acc": 0.77606354,
      "epoch": 0.8163431954388369,
      "grad_norm": 4.4375,
      "learning_rate": 6.695555768150644e-06,
      "loss": 0.78519859,
      "memory(GiB)": 135.77,
      "step": 34990,
      "train_speed(iter/s)": 0.202116
    },
    {
      "acc": 0.75625358,
      "epoch": 0.8165765030111258,
      "grad_norm": 5.71875,
      "learning_rate": 6.693778472364754e-06,
      "loss": 0.89699631,
      "memory(GiB)": 135.77,
      "step": 35000,
      "train_speed(iter/s)": 0.202146
    },
    {
      "epoch": 0.8165765030111258,
      "eval_acc": 0.7427814586929572,
      "eval_loss": 0.8112027049064636,
      "eval_runtime": 1268.9891,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "step": 35000
    },
    {
      "acc": 0.76805382,
      "epoch": 0.8168098105834147,
      "grad_norm": 5.4375,
      "learning_rate": 6.692000934792479e-06,
      "loss": 0.81355896,
      "memory(GiB)": 135.77,
      "step": 35010,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.78000903,
      "epoch": 0.8170431181557036,
      "grad_norm": 6.96875,
      "learning_rate": 6.6902231556875605e-06,
      "loss": 0.79382019,
      "memory(GiB)": 135.77,
      "step": 35020,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.77730827,
      "epoch": 0.8172764257279925,
      "grad_norm": 4.59375,
      "learning_rate": 6.688445135303779e-06,
      "loss": 0.78697395,
      "memory(GiB)": 135.77,
      "step": 35030,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.80807705,
      "epoch": 0.8175097333002814,
      "grad_norm": 5.84375,
      "learning_rate": 6.686666873894945e-06,
      "loss": 0.6698247,
      "memory(GiB)": 135.77,
      "step": 35040,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.78960042,
      "epoch": 0.8177430408725703,
      "grad_norm": 5.3125,
      "learning_rate": 6.684888371714903e-06,
      "loss": 0.74742336,
      "memory(GiB)": 135.77,
      "step": 35050,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.77717094,
      "epoch": 0.8179763484448592,
      "grad_norm": 5.0625,
      "learning_rate": 6.683109629017536e-06,
      "loss": 0.79351301,
      "memory(GiB)": 135.77,
      "step": 35060,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.77711973,
      "epoch": 0.8182096560171481,
      "grad_norm": 10.8125,
      "learning_rate": 6.681330646056758e-06,
      "loss": 0.80530987,
      "memory(GiB)": 135.77,
      "step": 35070,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.78023176,
      "epoch": 0.818442963589437,
      "grad_norm": 5.53125,
      "learning_rate": 6.679551423086521e-06,
      "loss": 0.80717087,
      "memory(GiB)": 135.77,
      "step": 35080,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.76191292,
      "epoch": 0.8186762711617259,
      "grad_norm": 4.9375,
      "learning_rate": 6.677771960360806e-06,
      "loss": 0.8590991,
      "memory(GiB)": 135.77,
      "step": 35090,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.76519403,
      "epoch": 0.8189095787340148,
      "grad_norm": 4.875,
      "learning_rate": 6.6759922581336285e-06,
      "loss": 0.83596706,
      "memory(GiB)": 135.77,
      "step": 35100,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.77768407,
      "epoch": 0.8191428863063037,
      "grad_norm": 7.1875,
      "learning_rate": 6.674212316659045e-06,
      "loss": 0.79807773,
      "memory(GiB)": 135.77,
      "step": 35110,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.76411262,
      "epoch": 0.8193761938785926,
      "grad_norm": 5.09375,
      "learning_rate": 6.6724321361911384e-06,
      "loss": 0.84451618,
      "memory(GiB)": 135.77,
      "step": 35120,
      "train_speed(iter/s)": 0.201002
    },
    {
      "acc": 0.76351461,
      "epoch": 0.8196095014508815,
      "grad_norm": 4.875,
      "learning_rate": 6.6706517169840305e-06,
      "loss": 0.85793591,
      "memory(GiB)": 135.77,
      "step": 35130,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.77693996,
      "epoch": 0.8198428090231703,
      "grad_norm": 8.75,
      "learning_rate": 6.668871059291875e-06,
      "loss": 0.78128543,
      "memory(GiB)": 135.77,
      "step": 35140,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.78849602,
      "epoch": 0.8200761165954592,
      "grad_norm": 4.5,
      "learning_rate": 6.667090163368863e-06,
      "loss": 0.7485117,
      "memory(GiB)": 135.77,
      "step": 35150,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.77250385,
      "epoch": 0.8203094241677481,
      "grad_norm": 4.5625,
      "learning_rate": 6.665309029469214e-06,
      "loss": 0.81979828,
      "memory(GiB)": 135.77,
      "step": 35160,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.77600441,
      "epoch": 0.820542731740037,
      "grad_norm": 4.71875,
      "learning_rate": 6.663527657847182e-06,
      "loss": 0.81544447,
      "memory(GiB)": 135.77,
      "step": 35170,
      "train_speed(iter/s)": 0.201153
    },
    {
      "acc": 0.789469,
      "epoch": 0.8207760393123259,
      "grad_norm": 4.40625,
      "learning_rate": 6.661746048757061e-06,
      "loss": 0.75934238,
      "memory(GiB)": 135.77,
      "step": 35180,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.75972357,
      "epoch": 0.8210093468846148,
      "grad_norm": 4.21875,
      "learning_rate": 6.6599642024531755e-06,
      "loss": 0.87705612,
      "memory(GiB)": 135.77,
      "step": 35190,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.80127926,
      "epoch": 0.8212426544569037,
      "grad_norm": 5.125,
      "learning_rate": 6.658182119189882e-06,
      "loss": 0.71490326,
      "memory(GiB)": 135.77,
      "step": 35200,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.77492847,
      "epoch": 0.8214759620291926,
      "grad_norm": 5.125,
      "learning_rate": 6.656399799221572e-06,
      "loss": 0.79777231,
      "memory(GiB)": 135.77,
      "step": 35210,
      "train_speed(iter/s)": 0.201272
    },
    {
      "acc": 0.78476725,
      "epoch": 0.8217092696014815,
      "grad_norm": 6.0625,
      "learning_rate": 6.654617242802672e-06,
      "loss": 0.78095455,
      "memory(GiB)": 135.77,
      "step": 35220,
      "train_speed(iter/s)": 0.201303
    },
    {
      "acc": 0.78755674,
      "epoch": 0.8219425771737704,
      "grad_norm": 5.5625,
      "learning_rate": 6.652834450187643e-06,
      "loss": 0.77176509,
      "memory(GiB)": 135.77,
      "step": 35230,
      "train_speed(iter/s)": 0.201333
    },
    {
      "acc": 0.78302078,
      "epoch": 0.8221758847460593,
      "grad_norm": 4.84375,
      "learning_rate": 6.651051421630974e-06,
      "loss": 0.77143784,
      "memory(GiB)": 135.77,
      "step": 35240,
      "train_speed(iter/s)": 0.201363
    },
    {
      "acc": 0.80360794,
      "epoch": 0.8224091923183482,
      "grad_norm": 5.5,
      "learning_rate": 6.649268157387195e-06,
      "loss": 0.69977317,
      "memory(GiB)": 135.77,
      "step": 35250,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.7762537,
      "epoch": 0.8226424998906371,
      "grad_norm": 5.65625,
      "learning_rate": 6.647484657710867e-06,
      "loss": 0.80590124,
      "memory(GiB)": 135.77,
      "step": 35260,
      "train_speed(iter/s)": 0.201424
    },
    {
      "acc": 0.79981756,
      "epoch": 0.822875807462926,
      "grad_norm": 4.875,
      "learning_rate": 6.645700922856582e-06,
      "loss": 0.71729341,
      "memory(GiB)": 135.77,
      "step": 35270,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.77900333,
      "epoch": 0.8231091150352149,
      "grad_norm": 4.25,
      "learning_rate": 6.643916953078966e-06,
      "loss": 0.79960418,
      "memory(GiB)": 135.77,
      "step": 35280,
      "train_speed(iter/s)": 0.201482
    },
    {
      "acc": 0.783249,
      "epoch": 0.8233424226075038,
      "grad_norm": 4.875,
      "learning_rate": 6.642132748632685e-06,
      "loss": 0.79931679,
      "memory(GiB)": 135.77,
      "step": 35290,
      "train_speed(iter/s)": 0.201512
    },
    {
      "acc": 0.77131052,
      "epoch": 0.8235757301797927,
      "grad_norm": 7.09375,
      "learning_rate": 6.640348309772431e-06,
      "loss": 0.82600031,
      "memory(GiB)": 135.77,
      "step": 35300,
      "train_speed(iter/s)": 0.201543
    },
    {
      "acc": 0.77425413,
      "epoch": 0.8238090377520816,
      "grad_norm": 5.8125,
      "learning_rate": 6.638563636752932e-06,
      "loss": 0.81250858,
      "memory(GiB)": 135.77,
      "step": 35310,
      "train_speed(iter/s)": 0.201574
    },
    {
      "acc": 0.7789402,
      "epoch": 0.8240423453243705,
      "grad_norm": 6.59375,
      "learning_rate": 6.63677872982895e-06,
      "loss": 0.77923145,
      "memory(GiB)": 135.77,
      "step": 35320,
      "train_speed(iter/s)": 0.201602
    },
    {
      "acc": 0.79925971,
      "epoch": 0.8242756528966593,
      "grad_norm": 5.125,
      "learning_rate": 6.634993589255278e-06,
      "loss": 0.71169586,
      "memory(GiB)": 135.77,
      "step": 35330,
      "train_speed(iter/s)": 0.201631
    },
    {
      "acc": 0.79988279,
      "epoch": 0.8245089604689482,
      "grad_norm": 6.03125,
      "learning_rate": 6.633208215286748e-06,
      "loss": 0.70508504,
      "memory(GiB)": 135.77,
      "step": 35340,
      "train_speed(iter/s)": 0.201658
    },
    {
      "acc": 0.77253027,
      "epoch": 0.8247422680412371,
      "grad_norm": 5.15625,
      "learning_rate": 6.6314226081782195e-06,
      "loss": 0.80938549,
      "memory(GiB)": 135.77,
      "step": 35350,
      "train_speed(iter/s)": 0.201685
    },
    {
      "acc": 0.78187051,
      "epoch": 0.824975575613526,
      "grad_norm": 5.3125,
      "learning_rate": 6.6296367681845875e-06,
      "loss": 0.76466885,
      "memory(GiB)": 135.77,
      "step": 35360,
      "train_speed(iter/s)": 0.201716
    },
    {
      "acc": 0.76568704,
      "epoch": 0.8252088831858149,
      "grad_norm": 5.5,
      "learning_rate": 6.62785069556078e-06,
      "loss": 0.86667337,
      "memory(GiB)": 135.77,
      "step": 35370,
      "train_speed(iter/s)": 0.201747
    },
    {
      "acc": 0.78236561,
      "epoch": 0.8254421907581038,
      "grad_norm": 5.21875,
      "learning_rate": 6.6260643905617605e-06,
      "loss": 0.76878605,
      "memory(GiB)": 135.77,
      "step": 35380,
      "train_speed(iter/s)": 0.201776
    },
    {
      "acc": 0.80533562,
      "epoch": 0.8256754983303927,
      "grad_norm": 7.34375,
      "learning_rate": 6.624277853442519e-06,
      "loss": 0.69753013,
      "memory(GiB)": 135.77,
      "step": 35390,
      "train_speed(iter/s)": 0.201806
    },
    {
      "acc": 0.78438272,
      "epoch": 0.8259088059026816,
      "grad_norm": 7.625,
      "learning_rate": 6.622491084458087e-06,
      "loss": 0.75821857,
      "memory(GiB)": 135.77,
      "step": 35400,
      "train_speed(iter/s)": 0.201837
    },
    {
      "acc": 0.77371559,
      "epoch": 0.8261421134749705,
      "grad_norm": 5.3125,
      "learning_rate": 6.620704083863523e-06,
      "loss": 0.82229176,
      "memory(GiB)": 135.77,
      "step": 35410,
      "train_speed(iter/s)": 0.201866
    },
    {
      "acc": 0.77689691,
      "epoch": 0.8263754210472594,
      "grad_norm": 5.15625,
      "learning_rate": 6.618916851913923e-06,
      "loss": 0.79822955,
      "memory(GiB)": 135.77,
      "step": 35420,
      "train_speed(iter/s)": 0.201896
    },
    {
      "acc": 0.76417141,
      "epoch": 0.8266087286195483,
      "grad_norm": 10.0,
      "learning_rate": 6.617129388864412e-06,
      "loss": 0.86350403,
      "memory(GiB)": 135.77,
      "step": 35430,
      "train_speed(iter/s)": 0.201926
    },
    {
      "acc": 0.7825284,
      "epoch": 0.8268420361918372,
      "grad_norm": 11.75,
      "learning_rate": 6.615341694970151e-06,
      "loss": 0.76954165,
      "memory(GiB)": 135.77,
      "step": 35440,
      "train_speed(iter/s)": 0.201953
    },
    {
      "acc": 0.78282237,
      "epoch": 0.8270753437641261,
      "grad_norm": 7.4375,
      "learning_rate": 6.613553770486333e-06,
      "loss": 0.77155132,
      "memory(GiB)": 135.77,
      "step": 35450,
      "train_speed(iter/s)": 0.201981
    },
    {
      "acc": 0.76642599,
      "epoch": 0.827308651336415,
      "grad_norm": 4.28125,
      "learning_rate": 6.611765615668182e-06,
      "loss": 0.8688242,
      "memory(GiB)": 135.77,
      "step": 35460,
      "train_speed(iter/s)": 0.202011
    },
    {
      "acc": 0.79817481,
      "epoch": 0.8275419589087039,
      "grad_norm": 4.9375,
      "learning_rate": 6.609977230770957e-06,
      "loss": 0.71652417,
      "memory(GiB)": 135.77,
      "step": 35470,
      "train_speed(iter/s)": 0.20204
    },
    {
      "acc": 0.76829128,
      "epoch": 0.8277752664809928,
      "grad_norm": 5.5625,
      "learning_rate": 6.608188616049951e-06,
      "loss": 0.83743439,
      "memory(GiB)": 135.77,
      "step": 35480,
      "train_speed(iter/s)": 0.20207
    },
    {
      "acc": 0.78781743,
      "epoch": 0.8280085740532817,
      "grad_norm": 7.53125,
      "learning_rate": 6.606399771760487e-06,
      "loss": 0.76809006,
      "memory(GiB)": 135.77,
      "step": 35490,
      "train_speed(iter/s)": 0.2021
    },
    {
      "acc": 0.78514261,
      "epoch": 0.8282418816255706,
      "grad_norm": 6.03125,
      "learning_rate": 6.6046106981579216e-06,
      "loss": 0.76895576,
      "memory(GiB)": 135.77,
      "step": 35500,
      "train_speed(iter/s)": 0.202131
    },
    {
      "epoch": 0.8282418816255706,
      "eval_acc": 0.7427417429233465,
      "eval_loss": 0.8110933899879456,
      "eval_runtime": 1270.2164,
      "eval_samples_per_second": 28.335,
      "eval_steps_per_second": 14.168,
      "step": 35500
    },
    {
      "acc": 0.7827517,
      "epoch": 0.8284751891978595,
      "grad_norm": 5.8125,
      "learning_rate": 6.6028213954976474e-06,
      "loss": 0.79220624,
      "memory(GiB)": 135.77,
      "step": 35510,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.77409925,
      "epoch": 0.8287084967701484,
      "grad_norm": 6.15625,
      "learning_rate": 6.601031864035082e-06,
      "loss": 0.80091858,
      "memory(GiB)": 135.77,
      "step": 35520,
      "train_speed(iter/s)": 0.200717
    },
    {
      "acc": 0.78355837,
      "epoch": 0.8289418043424371,
      "grad_norm": 5.25,
      "learning_rate": 6.5992421040256834e-06,
      "loss": 0.75921507,
      "memory(GiB)": 135.77,
      "step": 35530,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76949062,
      "epoch": 0.829175111914726,
      "grad_norm": 5.375,
      "learning_rate": 6.597452115724939e-06,
      "loss": 0.81914463,
      "memory(GiB)": 135.77,
      "step": 35540,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.77408943,
      "epoch": 0.8294084194870149,
      "grad_norm": 5.59375,
      "learning_rate": 6.5956618993883716e-06,
      "loss": 0.80963411,
      "memory(GiB)": 135.77,
      "step": 35550,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.80136003,
      "epoch": 0.8296417270593038,
      "grad_norm": 4.53125,
      "learning_rate": 6.59387145527153e-06,
      "loss": 0.71679068,
      "memory(GiB)": 135.77,
      "step": 35560,
      "train_speed(iter/s)": 0.200835
    },
    {
      "acc": 0.76815443,
      "epoch": 0.8298750346315927,
      "grad_norm": 5.4375,
      "learning_rate": 6.59208078363e-06,
      "loss": 0.8367054,
      "memory(GiB)": 135.77,
      "step": 35570,
      "train_speed(iter/s)": 0.200866
    },
    {
      "acc": 0.7694231,
      "epoch": 0.8301083422038816,
      "grad_norm": 5.34375,
      "learning_rate": 6.590289884719403e-06,
      "loss": 0.85313902,
      "memory(GiB)": 135.77,
      "step": 35580,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.79432936,
      "epoch": 0.8303416497761705,
      "grad_norm": 5.09375,
      "learning_rate": 6.588498758795386e-06,
      "loss": 0.75772634,
      "memory(GiB)": 135.77,
      "step": 35590,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.77298503,
      "epoch": 0.8305749573484594,
      "grad_norm": 5.625,
      "learning_rate": 6.586707406113632e-06,
      "loss": 0.81344709,
      "memory(GiB)": 135.77,
      "step": 35600,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.78691454,
      "epoch": 0.8308082649207483,
      "grad_norm": 5.5,
      "learning_rate": 6.5849158269298565e-06,
      "loss": 0.76828012,
      "memory(GiB)": 135.77,
      "step": 35610,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.76740513,
      "epoch": 0.8310415724930372,
      "grad_norm": 6.34375,
      "learning_rate": 6.583124021499807e-06,
      "loss": 0.83374586,
      "memory(GiB)": 135.77,
      "step": 35620,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.76556411,
      "epoch": 0.8312748800653261,
      "grad_norm": 6.0625,
      "learning_rate": 6.581331990079264e-06,
      "loss": 0.84218559,
      "memory(GiB)": 135.77,
      "step": 35630,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.77049723,
      "epoch": 0.831508187637615,
      "grad_norm": 5.75,
      "learning_rate": 6.579539732924038e-06,
      "loss": 0.83664169,
      "memory(GiB)": 135.77,
      "step": 35640,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.76034393,
      "epoch": 0.8317414952099039,
      "grad_norm": 7.5625,
      "learning_rate": 6.5777472502899765e-06,
      "loss": 0.87293215,
      "memory(GiB)": 135.77,
      "step": 35650,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.78872604,
      "epoch": 0.8319748027821928,
      "grad_norm": 5.03125,
      "learning_rate": 6.5759545424329514e-06,
      "loss": 0.7763586,
      "memory(GiB)": 135.77,
      "step": 35660,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.76089096,
      "epoch": 0.8322081103544817,
      "grad_norm": 5.90625,
      "learning_rate": 6.574161609608873e-06,
      "loss": 0.84671097,
      "memory(GiB)": 135.77,
      "step": 35670,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.78287115,
      "epoch": 0.8324414179267706,
      "grad_norm": 7.78125,
      "learning_rate": 6.572368452073683e-06,
      "loss": 0.79170933,
      "memory(GiB)": 135.77,
      "step": 35680,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.79201784,
      "epoch": 0.8326747254990595,
      "grad_norm": 5.46875,
      "learning_rate": 6.570575070083351e-06,
      "loss": 0.72850819,
      "memory(GiB)": 135.77,
      "step": 35690,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.76989765,
      "epoch": 0.8329080330713484,
      "grad_norm": 4.3125,
      "learning_rate": 6.5687814638938865e-06,
      "loss": 0.80628548,
      "memory(GiB)": 135.77,
      "step": 35700,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.77745337,
      "epoch": 0.8331413406436373,
      "grad_norm": 4.78125,
      "learning_rate": 6.566987633761323e-06,
      "loss": 0.80288849,
      "memory(GiB)": 135.77,
      "step": 35710,
      "train_speed(iter/s)": 0.201276
    },
    {
      "acc": 0.77031932,
      "epoch": 0.8333746482159261,
      "grad_norm": 5.25,
      "learning_rate": 6.5651935799417295e-06,
      "loss": 0.84479132,
      "memory(GiB)": 135.77,
      "step": 35720,
      "train_speed(iter/s)": 0.201306
    },
    {
      "acc": 0.7754509,
      "epoch": 0.833607955788215,
      "grad_norm": 5.53125,
      "learning_rate": 6.563399302691209e-06,
      "loss": 0.78503389,
      "memory(GiB)": 135.77,
      "step": 35730,
      "train_speed(iter/s)": 0.201336
    },
    {
      "acc": 0.77183795,
      "epoch": 0.8338412633605039,
      "grad_norm": 5.65625,
      "learning_rate": 6.561604802265891e-06,
      "loss": 0.82641964,
      "memory(GiB)": 135.77,
      "step": 35740,
      "train_speed(iter/s)": 0.201364
    },
    {
      "acc": 0.78574123,
      "epoch": 0.8340745709327928,
      "grad_norm": 4.21875,
      "learning_rate": 6.55981007892194e-06,
      "loss": 0.76580262,
      "memory(GiB)": 135.77,
      "step": 35750,
      "train_speed(iter/s)": 0.201393
    },
    {
      "acc": 0.79073353,
      "epoch": 0.8343078785050817,
      "grad_norm": 7.90625,
      "learning_rate": 6.558015132915554e-06,
      "loss": 0.7411293,
      "memory(GiB)": 135.77,
      "step": 35760,
      "train_speed(iter/s)": 0.201422
    },
    {
      "acc": 0.77999744,
      "epoch": 0.8345411860773706,
      "grad_norm": 5.5,
      "learning_rate": 6.556219964502961e-06,
      "loss": 0.77981787,
      "memory(GiB)": 135.77,
      "step": 35770,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.77890453,
      "epoch": 0.8347744936496595,
      "grad_norm": 6.5,
      "learning_rate": 6.5544245739404196e-06,
      "loss": 0.76951075,
      "memory(GiB)": 135.77,
      "step": 35780,
      "train_speed(iter/s)": 0.201478
    },
    {
      "acc": 0.78939867,
      "epoch": 0.8350078012219484,
      "grad_norm": 5.4375,
      "learning_rate": 6.552628961484222e-06,
      "loss": 0.76422501,
      "memory(GiB)": 135.77,
      "step": 35790,
      "train_speed(iter/s)": 0.201507
    },
    {
      "acc": 0.78203793,
      "epoch": 0.8352411087942373,
      "grad_norm": 8.1875,
      "learning_rate": 6.550833127390692e-06,
      "loss": 0.7660037,
      "memory(GiB)": 135.77,
      "step": 35800,
      "train_speed(iter/s)": 0.201535
    },
    {
      "acc": 0.7650146,
      "epoch": 0.8354744163665262,
      "grad_norm": 4.96875,
      "learning_rate": 6.549037071916184e-06,
      "loss": 0.87431145,
      "memory(GiB)": 135.77,
      "step": 35810,
      "train_speed(iter/s)": 0.201565
    },
    {
      "acc": 0.75421095,
      "epoch": 0.8357077239388151,
      "grad_norm": 4.21875,
      "learning_rate": 6.547240795317081e-06,
      "loss": 0.86133585,
      "memory(GiB)": 135.77,
      "step": 35820,
      "train_speed(iter/s)": 0.201593
    },
    {
      "acc": 0.76499453,
      "epoch": 0.835941031511104,
      "grad_norm": 8.5625,
      "learning_rate": 6.545444297849808e-06,
      "loss": 0.85870771,
      "memory(GiB)": 135.77,
      "step": 35830,
      "train_speed(iter/s)": 0.201622
    },
    {
      "acc": 0.77958255,
      "epoch": 0.8361743390833929,
      "grad_norm": 6.0625,
      "learning_rate": 6.543647579770806e-06,
      "loss": 0.78675489,
      "memory(GiB)": 135.77,
      "step": 35840,
      "train_speed(iter/s)": 0.20165
    },
    {
      "acc": 0.77734413,
      "epoch": 0.8364076466556818,
      "grad_norm": 5.65625,
      "learning_rate": 6.5418506413365634e-06,
      "loss": 0.78364563,
      "memory(GiB)": 135.77,
      "step": 35850,
      "train_speed(iter/s)": 0.201679
    },
    {
      "acc": 0.78868957,
      "epoch": 0.8366409542279707,
      "grad_norm": 5.59375,
      "learning_rate": 6.5400534828035885e-06,
      "loss": 0.74761543,
      "memory(GiB)": 135.77,
      "step": 35860,
      "train_speed(iter/s)": 0.201708
    },
    {
      "acc": 0.76377254,
      "epoch": 0.8368742618002596,
      "grad_norm": 5.90625,
      "learning_rate": 6.538256104428427e-06,
      "loss": 0.85419159,
      "memory(GiB)": 135.77,
      "step": 35870,
      "train_speed(iter/s)": 0.201738
    },
    {
      "acc": 0.77922173,
      "epoch": 0.8371075693725485,
      "grad_norm": 6.5,
      "learning_rate": 6.536458506467654e-06,
      "loss": 0.78688426,
      "memory(GiB)": 135.77,
      "step": 35880,
      "train_speed(iter/s)": 0.201765
    },
    {
      "acc": 0.76489186,
      "epoch": 0.8373408769448374,
      "grad_norm": 4.875,
      "learning_rate": 6.5346606891778755e-06,
      "loss": 0.84113293,
      "memory(GiB)": 135.77,
      "step": 35890,
      "train_speed(iter/s)": 0.201795
    },
    {
      "acc": 0.77273259,
      "epoch": 0.8375741845171263,
      "grad_norm": 7.375,
      "learning_rate": 6.532862652815728e-06,
      "loss": 0.8084506,
      "memory(GiB)": 135.77,
      "step": 35900,
      "train_speed(iter/s)": 0.201825
    },
    {
      "acc": 0.74848862,
      "epoch": 0.8378074920894151,
      "grad_norm": 7.0625,
      "learning_rate": 6.531064397637883e-06,
      "loss": 0.90964317,
      "memory(GiB)": 135.77,
      "step": 35910,
      "train_speed(iter/s)": 0.201853
    },
    {
      "acc": 0.78161101,
      "epoch": 0.838040799661704,
      "grad_norm": 5.8125,
      "learning_rate": 6.529265923901039e-06,
      "loss": 0.78454123,
      "memory(GiB)": 135.77,
      "step": 35920,
      "train_speed(iter/s)": 0.201882
    },
    {
      "acc": 0.78447218,
      "epoch": 0.8382741072339929,
      "grad_norm": 5.25,
      "learning_rate": 6.527467231861929e-06,
      "loss": 0.76308413,
      "memory(GiB)": 135.77,
      "step": 35930,
      "train_speed(iter/s)": 0.201912
    },
    {
      "acc": 0.78148451,
      "epoch": 0.8385074148062818,
      "grad_norm": 97.5,
      "learning_rate": 6.525668321777317e-06,
      "loss": 0.78811293,
      "memory(GiB)": 135.77,
      "step": 35940,
      "train_speed(iter/s)": 0.201942
    },
    {
      "acc": 0.76809978,
      "epoch": 0.8387407223785707,
      "grad_norm": 5.09375,
      "learning_rate": 6.523869193903994e-06,
      "loss": 0.84421577,
      "memory(GiB)": 135.77,
      "step": 35950,
      "train_speed(iter/s)": 0.201973
    },
    {
      "acc": 0.76690893,
      "epoch": 0.8389740299508596,
      "grad_norm": 4.28125,
      "learning_rate": 6.522069848498787e-06,
      "loss": 0.8514679,
      "memory(GiB)": 135.77,
      "step": 35960,
      "train_speed(iter/s)": 0.202001
    },
    {
      "acc": 0.78587494,
      "epoch": 0.8392073375231485,
      "grad_norm": 7.875,
      "learning_rate": 6.5202702858185495e-06,
      "loss": 0.75838385,
      "memory(GiB)": 135.77,
      "step": 35970,
      "train_speed(iter/s)": 0.20203
    },
    {
      "acc": 0.77815952,
      "epoch": 0.8394406450954374,
      "grad_norm": 5.15625,
      "learning_rate": 6.518470506120171e-06,
      "loss": 0.7992382,
      "memory(GiB)": 135.77,
      "step": 35980,
      "train_speed(iter/s)": 0.202058
    },
    {
      "acc": 0.77503004,
      "epoch": 0.8396739526677263,
      "grad_norm": 6.75,
      "learning_rate": 6.51667050966057e-06,
      "loss": 0.79922199,
      "memory(GiB)": 135.77,
      "step": 35990,
      "train_speed(iter/s)": 0.202088
    },
    {
      "acc": 0.76703901,
      "epoch": 0.8399072602400152,
      "grad_norm": 5.875,
      "learning_rate": 6.514870296696694e-06,
      "loss": 0.84329872,
      "memory(GiB)": 135.77,
      "step": 36000,
      "train_speed(iter/s)": 0.202117
    },
    {
      "epoch": 0.8399072602400152,
      "eval_acc": 0.7429556956177011,
      "eval_loss": 0.8103926777839661,
      "eval_runtime": 1268.5973,
      "eval_samples_per_second": 28.371,
      "eval_steps_per_second": 14.186,
      "step": 36000
    },
    {
      "acc": 0.76743212,
      "epoch": 0.8401405678123041,
      "grad_norm": 5.5,
      "learning_rate": 6.513069867485523e-06,
      "loss": 0.8609807,
      "memory(GiB)": 135.77,
      "step": 36010,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.75457373,
      "epoch": 0.840373875384593,
      "grad_norm": 6.21875,
      "learning_rate": 6.511269222284069e-06,
      "loss": 0.86449337,
      "memory(GiB)": 135.77,
      "step": 36020,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.76079054,
      "epoch": 0.8406071829568819,
      "grad_norm": 4.71875,
      "learning_rate": 6.509468361349371e-06,
      "loss": 0.87414017,
      "memory(GiB)": 135.77,
      "step": 36030,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.79017248,
      "epoch": 0.8408404905291708,
      "grad_norm": 4.625,
      "learning_rate": 6.507667284938502e-06,
      "loss": 0.75717969,
      "memory(GiB)": 135.77,
      "step": 36040,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.77153301,
      "epoch": 0.8410737981014597,
      "grad_norm": 5.5,
      "learning_rate": 6.505865993308568e-06,
      "loss": 0.84561863,
      "memory(GiB)": 135.77,
      "step": 36050,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.7702539,
      "epoch": 0.8413071056737486,
      "grad_norm": 10.75,
      "learning_rate": 6.5040644867167e-06,
      "loss": 0.83923016,
      "memory(GiB)": 135.77,
      "step": 36060,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.78410187,
      "epoch": 0.8415404132460375,
      "grad_norm": 7.4375,
      "learning_rate": 6.502262765420064e-06,
      "loss": 0.78843861,
      "memory(GiB)": 135.77,
      "step": 36070,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.76989207,
      "epoch": 0.8417737208183264,
      "grad_norm": 5.03125,
      "learning_rate": 6.500460829675854e-06,
      "loss": 0.81716394,
      "memory(GiB)": 135.77,
      "step": 36080,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.75923405,
      "epoch": 0.8420070283906153,
      "grad_norm": 6.4375,
      "learning_rate": 6.498658679741298e-06,
      "loss": 0.86520176,
      "memory(GiB)": 135.77,
      "step": 36090,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.79458346,
      "epoch": 0.842240335962904,
      "grad_norm": 6.75,
      "learning_rate": 6.49685631587365e-06,
      "loss": 0.74665785,
      "memory(GiB)": 135.77,
      "step": 36100,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.7725565,
      "epoch": 0.8424736435351929,
      "grad_norm": 6.46875,
      "learning_rate": 6.495053738330196e-06,
      "loss": 0.80447502,
      "memory(GiB)": 135.77,
      "step": 36110,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.78278484,
      "epoch": 0.8427069511074818,
      "grad_norm": 5.5625,
      "learning_rate": 6.493250947368257e-06,
      "loss": 0.77641745,
      "memory(GiB)": 135.77,
      "step": 36120,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.79837875,
      "epoch": 0.8429402586797707,
      "grad_norm": 5.0625,
      "learning_rate": 6.491447943245179e-06,
      "loss": 0.71321259,
      "memory(GiB)": 135.77,
      "step": 36130,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.79435396,
      "epoch": 0.8431735662520596,
      "grad_norm": 5.125,
      "learning_rate": 6.489644726218339e-06,
      "loss": 0.7343874,
      "memory(GiB)": 135.77,
      "step": 36140,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.78385534,
      "epoch": 0.8434068738243485,
      "grad_norm": 6.59375,
      "learning_rate": 6.4878412965451485e-06,
      "loss": 0.79530344,
      "memory(GiB)": 135.77,
      "step": 36150,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.78743668,
      "epoch": 0.8436401813966374,
      "grad_norm": 7.125,
      "learning_rate": 6.486037654483046e-06,
      "loss": 0.74157157,
      "memory(GiB)": 135.77,
      "step": 36160,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.80141068,
      "epoch": 0.8438734889689263,
      "grad_norm": 4.8125,
      "learning_rate": 6.484233800289499e-06,
      "loss": 0.70053635,
      "memory(GiB)": 135.77,
      "step": 36170,
      "train_speed(iter/s)": 0.201156
    },
    {
      "acc": 0.76834717,
      "epoch": 0.8441067965412152,
      "grad_norm": 6.09375,
      "learning_rate": 6.482429734222008e-06,
      "loss": 0.85586014,
      "memory(GiB)": 135.77,
      "step": 36180,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.77482719,
      "epoch": 0.8443401041135041,
      "grad_norm": 6.0625,
      "learning_rate": 6.4806254565381025e-06,
      "loss": 0.81930189,
      "memory(GiB)": 135.77,
      "step": 36190,
      "train_speed(iter/s)": 0.20121
    },
    {
      "acc": 0.81716328,
      "epoch": 0.844573411685793,
      "grad_norm": 5.375,
      "learning_rate": 6.478820967495343e-06,
      "loss": 0.6350563,
      "memory(GiB)": 135.77,
      "step": 36200,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.77658648,
      "epoch": 0.8448067192580819,
      "grad_norm": 4.34375,
      "learning_rate": 6.47701626735132e-06,
      "loss": 0.78943858,
      "memory(GiB)": 135.77,
      "step": 36210,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.78699703,
      "epoch": 0.8450400268303708,
      "grad_norm": 5.96875,
      "learning_rate": 6.475211356363655e-06,
      "loss": 0.75699453,
      "memory(GiB)": 135.77,
      "step": 36220,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.80862513,
      "epoch": 0.8452733344026597,
      "grad_norm": 4.625,
      "learning_rate": 6.473406234789998e-06,
      "loss": 0.67446146,
      "memory(GiB)": 135.77,
      "step": 36230,
      "train_speed(iter/s)": 0.20132
    },
    {
      "acc": 0.77075014,
      "epoch": 0.8455066419749486,
      "grad_norm": 6.25,
      "learning_rate": 6.471600902888029e-06,
      "loss": 0.81247129,
      "memory(GiB)": 135.77,
      "step": 36240,
      "train_speed(iter/s)": 0.20135
    },
    {
      "acc": 0.77357235,
      "epoch": 0.8457399495472375,
      "grad_norm": 6.21875,
      "learning_rate": 6.4697953609154575e-06,
      "loss": 0.81807146,
      "memory(GiB)": 135.77,
      "step": 36250,
      "train_speed(iter/s)": 0.201378
    },
    {
      "acc": 0.78533292,
      "epoch": 0.8459732571195264,
      "grad_norm": 5.34375,
      "learning_rate": 6.467989609130024e-06,
      "loss": 0.78466349,
      "memory(GiB)": 135.77,
      "step": 36260,
      "train_speed(iter/s)": 0.201406
    },
    {
      "acc": 0.77039223,
      "epoch": 0.8462065646918153,
      "grad_norm": 5.34375,
      "learning_rate": 6.466183647789502e-06,
      "loss": 0.83654995,
      "memory(GiB)": 135.77,
      "step": 36270,
      "train_speed(iter/s)": 0.201434
    },
    {
      "acc": 0.7773541,
      "epoch": 0.8464398722641042,
      "grad_norm": 4.40625,
      "learning_rate": 6.46437747715169e-06,
      "loss": 0.79755011,
      "memory(GiB)": 135.77,
      "step": 36280,
      "train_speed(iter/s)": 0.201462
    },
    {
      "acc": 0.77812119,
      "epoch": 0.8466731798363931,
      "grad_norm": 5.46875,
      "learning_rate": 6.462571097474419e-06,
      "loss": 0.78683367,
      "memory(GiB)": 135.77,
      "step": 36290,
      "train_speed(iter/s)": 0.201491
    },
    {
      "acc": 0.78762054,
      "epoch": 0.8469064874086819,
      "grad_norm": 4.28125,
      "learning_rate": 6.460764509015547e-06,
      "loss": 0.76013975,
      "memory(GiB)": 135.77,
      "step": 36300,
      "train_speed(iter/s)": 0.201516
    },
    {
      "acc": 0.78863735,
      "epoch": 0.8471397949809708,
      "grad_norm": 6.125,
      "learning_rate": 6.4589577120329685e-06,
      "loss": 0.7554966,
      "memory(GiB)": 135.77,
      "step": 36310,
      "train_speed(iter/s)": 0.201544
    },
    {
      "acc": 0.78832941,
      "epoch": 0.8473731025532597,
      "grad_norm": 6.375,
      "learning_rate": 6.4571507067845985e-06,
      "loss": 0.78529215,
      "memory(GiB)": 135.77,
      "step": 36320,
      "train_speed(iter/s)": 0.201569
    },
    {
      "acc": 0.78547802,
      "epoch": 0.8476064101255486,
      "grad_norm": 4.8125,
      "learning_rate": 6.455343493528388e-06,
      "loss": 0.77054935,
      "memory(GiB)": 135.77,
      "step": 36330,
      "train_speed(iter/s)": 0.201598
    },
    {
      "acc": 0.7642107,
      "epoch": 0.8478397176978375,
      "grad_norm": 5.96875,
      "learning_rate": 6.4535360725223175e-06,
      "loss": 0.85638161,
      "memory(GiB)": 135.77,
      "step": 36340,
      "train_speed(iter/s)": 0.201628
    },
    {
      "acc": 0.77951527,
      "epoch": 0.8480730252701264,
      "grad_norm": 4.15625,
      "learning_rate": 6.451728444024394e-06,
      "loss": 0.77187543,
      "memory(GiB)": 135.77,
      "step": 36350,
      "train_speed(iter/s)": 0.201656
    },
    {
      "acc": 0.78816013,
      "epoch": 0.8483063328424153,
      "grad_norm": 4.84375,
      "learning_rate": 6.449920608292658e-06,
      "loss": 0.75378218,
      "memory(GiB)": 135.77,
      "step": 36360,
      "train_speed(iter/s)": 0.201683
    },
    {
      "acc": 0.7518692,
      "epoch": 0.8485396404147042,
      "grad_norm": 5.65625,
      "learning_rate": 6.448112565585176e-06,
      "loss": 0.89817324,
      "memory(GiB)": 135.77,
      "step": 36370,
      "train_speed(iter/s)": 0.201712
    },
    {
      "acc": 0.77250676,
      "epoch": 0.8487729479869931,
      "grad_norm": 4.34375,
      "learning_rate": 6.446304316160046e-06,
      "loss": 0.80824423,
      "memory(GiB)": 135.77,
      "step": 36380,
      "train_speed(iter/s)": 0.20174
    },
    {
      "acc": 0.78365912,
      "epoch": 0.849006255559282,
      "grad_norm": 4.40625,
      "learning_rate": 6.444495860275395e-06,
      "loss": 0.76890135,
      "memory(GiB)": 135.77,
      "step": 36390,
      "train_speed(iter/s)": 0.201768
    },
    {
      "acc": 0.75655127,
      "epoch": 0.8492395631315709,
      "grad_norm": 5.625,
      "learning_rate": 6.442687198189379e-06,
      "loss": 0.86654816,
      "memory(GiB)": 135.77,
      "step": 36400,
      "train_speed(iter/s)": 0.201797
    },
    {
      "acc": 0.77825761,
      "epoch": 0.8494728707038598,
      "grad_norm": 5.3125,
      "learning_rate": 6.440878330160185e-06,
      "loss": 0.79696665,
      "memory(GiB)": 135.77,
      "step": 36410,
      "train_speed(iter/s)": 0.201829
    },
    {
      "acc": 0.78085461,
      "epoch": 0.8497061782761487,
      "grad_norm": 4.46875,
      "learning_rate": 6.439069256446027e-06,
      "loss": 0.78490944,
      "memory(GiB)": 135.77,
      "step": 36420,
      "train_speed(iter/s)": 0.201857
    },
    {
      "acc": 0.78769636,
      "epoch": 0.8499394858484376,
      "grad_norm": 7.34375,
      "learning_rate": 6.437259977305152e-06,
      "loss": 0.77664795,
      "memory(GiB)": 135.77,
      "step": 36430,
      "train_speed(iter/s)": 0.201887
    },
    {
      "acc": 0.77704792,
      "epoch": 0.8501727934207265,
      "grad_norm": 4.875,
      "learning_rate": 6.435450492995833e-06,
      "loss": 0.80361004,
      "memory(GiB)": 135.77,
      "step": 36440,
      "train_speed(iter/s)": 0.201916
    },
    {
      "acc": 0.79341078,
      "epoch": 0.8504061009930154,
      "grad_norm": 5.21875,
      "learning_rate": 6.433640803776372e-06,
      "loss": 0.7458343,
      "memory(GiB)": 135.77,
      "step": 36450,
      "train_speed(iter/s)": 0.201944
    },
    {
      "acc": 0.78444099,
      "epoch": 0.8506394085653043,
      "grad_norm": 6.5,
      "learning_rate": 6.431830909905105e-06,
      "loss": 0.77440691,
      "memory(GiB)": 135.77,
      "step": 36460,
      "train_speed(iter/s)": 0.20197
    },
    {
      "acc": 0.78535118,
      "epoch": 0.8508727161375932,
      "grad_norm": 4.28125,
      "learning_rate": 6.43002081164039e-06,
      "loss": 0.76964531,
      "memory(GiB)": 135.77,
      "step": 36470,
      "train_speed(iter/s)": 0.201997
    },
    {
      "acc": 0.7824791,
      "epoch": 0.8511060237098821,
      "grad_norm": 6.15625,
      "learning_rate": 6.428210509240618e-06,
      "loss": 0.80835476,
      "memory(GiB)": 135.77,
      "step": 36480,
      "train_speed(iter/s)": 0.202024
    },
    {
      "acc": 0.79336419,
      "epoch": 0.8513393312821709,
      "grad_norm": 4.90625,
      "learning_rate": 6.426400002964211e-06,
      "loss": 0.73920593,
      "memory(GiB)": 135.77,
      "step": 36490,
      "train_speed(iter/s)": 0.202054
    },
    {
      "acc": 0.77750196,
      "epoch": 0.8515726388544598,
      "grad_norm": 4.84375,
      "learning_rate": 6.42458929306962e-06,
      "loss": 0.81381168,
      "memory(GiB)": 135.77,
      "step": 36500,
      "train_speed(iter/s)": 0.202082
    },
    {
      "epoch": 0.8515726388544598,
      "eval_acc": 0.7429995751373517,
      "eval_loss": 0.8101937174797058,
      "eval_runtime": 1268.6867,
      "eval_samples_per_second": 28.369,
      "eval_steps_per_second": 14.185,
      "step": 36500
    },
    {
      "acc": 0.78311806,
      "epoch": 0.8518059464267487,
      "grad_norm": 5.34375,
      "learning_rate": 6.42277837981532e-06,
      "loss": 0.75560861,
      "memory(GiB)": 135.77,
      "step": 36510,
      "train_speed(iter/s)": 0.200682
    },
    {
      "acc": 0.76936331,
      "epoch": 0.8520392539990376,
      "grad_norm": 7.1875,
      "learning_rate": 6.420967263459821e-06,
      "loss": 0.82802315,
      "memory(GiB)": 135.77,
      "step": 36520,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.76965437,
      "epoch": 0.8522725615713265,
      "grad_norm": 5.3125,
      "learning_rate": 6.419155944261657e-06,
      "loss": 0.82444363,
      "memory(GiB)": 135.77,
      "step": 36530,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.77062273,
      "epoch": 0.8525058691436154,
      "grad_norm": 13.5,
      "learning_rate": 6.4173444224793935e-06,
      "loss": 0.82105751,
      "memory(GiB)": 135.77,
      "step": 36540,
      "train_speed(iter/s)": 0.200772
    },
    {
      "acc": 0.77684073,
      "epoch": 0.8527391767159043,
      "grad_norm": 5.96875,
      "learning_rate": 6.415532698371625e-06,
      "loss": 0.80583134,
      "memory(GiB)": 135.77,
      "step": 36550,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.78810062,
      "epoch": 0.8529724842881932,
      "grad_norm": 4.25,
      "learning_rate": 6.413720772196976e-06,
      "loss": 0.76445894,
      "memory(GiB)": 135.77,
      "step": 36560,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.7537466,
      "epoch": 0.8532057918604821,
      "grad_norm": 6.34375,
      "learning_rate": 6.411908644214098e-06,
      "loss": 0.89372339,
      "memory(GiB)": 135.77,
      "step": 36570,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.77111721,
      "epoch": 0.853439099432771,
      "grad_norm": 5.53125,
      "learning_rate": 6.410096314681671e-06,
      "loss": 0.82660351,
      "memory(GiB)": 135.77,
      "step": 36580,
      "train_speed(iter/s)": 0.200892
    },
    {
      "acc": 0.76293507,
      "epoch": 0.8536724070050599,
      "grad_norm": 5.34375,
      "learning_rate": 6.408283783858405e-06,
      "loss": 0.87582846,
      "memory(GiB)": 135.77,
      "step": 36590,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.77199388,
      "epoch": 0.8539057145773488,
      "grad_norm": 5.84375,
      "learning_rate": 6.406471052003036e-06,
      "loss": 0.82917595,
      "memory(GiB)": 135.77,
      "step": 36600,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78203735,
      "epoch": 0.8541390221496377,
      "grad_norm": 5.28125,
      "learning_rate": 6.4046581193743344e-06,
      "loss": 0.79196444,
      "memory(GiB)": 135.77,
      "step": 36610,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.7757906,
      "epoch": 0.8543723297219266,
      "grad_norm": 5.6875,
      "learning_rate": 6.402844986231094e-06,
      "loss": 0.81577177,
      "memory(GiB)": 135.77,
      "step": 36620,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.77216077,
      "epoch": 0.8546056372942155,
      "grad_norm": 5.1875,
      "learning_rate": 6.401031652832141e-06,
      "loss": 0.79967194,
      "memory(GiB)": 135.77,
      "step": 36630,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.78772125,
      "epoch": 0.8548389448665044,
      "grad_norm": 4.875,
      "learning_rate": 6.3992181194363234e-06,
      "loss": 0.75887017,
      "memory(GiB)": 135.77,
      "step": 36640,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.76235476,
      "epoch": 0.8550722524387933,
      "grad_norm": 6.28125,
      "learning_rate": 6.397404386302528e-06,
      "loss": 0.84511347,
      "memory(GiB)": 135.77,
      "step": 36650,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.7826313,
      "epoch": 0.8553055600110822,
      "grad_norm": 4.21875,
      "learning_rate": 6.395590453689662e-06,
      "loss": 0.8007947,
      "memory(GiB)": 135.77,
      "step": 36660,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.77450428,
      "epoch": 0.855538867583371,
      "grad_norm": 6.53125,
      "learning_rate": 6.393776321856664e-06,
      "loss": 0.80132704,
      "memory(GiB)": 135.77,
      "step": 36670,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.75675406,
      "epoch": 0.8557721751556598,
      "grad_norm": 6.75,
      "learning_rate": 6.391961991062501e-06,
      "loss": 0.86429386,
      "memory(GiB)": 135.77,
      "step": 36680,
      "train_speed(iter/s)": 0.201179
    },
    {
      "acc": 0.75716515,
      "epoch": 0.8560054827279487,
      "grad_norm": 11.5,
      "learning_rate": 6.390147461566167e-06,
      "loss": 0.86847095,
      "memory(GiB)": 135.77,
      "step": 36690,
      "train_speed(iter/s)": 0.201205
    },
    {
      "acc": 0.78775959,
      "epoch": 0.8562387903002376,
      "grad_norm": 4.875,
      "learning_rate": 6.388332733626689e-06,
      "loss": 0.75113401,
      "memory(GiB)": 147.13,
      "step": 36700,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.76984215,
      "epoch": 0.8564720978725265,
      "grad_norm": 6.0625,
      "learning_rate": 6.386517807503114e-06,
      "loss": 0.83430986,
      "memory(GiB)": 147.13,
      "step": 36710,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.77020178,
      "epoch": 0.8567054054448154,
      "grad_norm": 6.28125,
      "learning_rate": 6.384702683454527e-06,
      "loss": 0.82754822,
      "memory(GiB)": 147.13,
      "step": 36720,
      "train_speed(iter/s)": 0.20129
    },
    {
      "acc": 0.7816741,
      "epoch": 0.8569387130171043,
      "grad_norm": 6.125,
      "learning_rate": 6.382887361740033e-06,
      "loss": 0.78474622,
      "memory(GiB)": 147.13,
      "step": 36730,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.79666619,
      "epoch": 0.8571720205893932,
      "grad_norm": 5.0,
      "learning_rate": 6.38107184261877e-06,
      "loss": 0.72450123,
      "memory(GiB)": 147.13,
      "step": 36740,
      "train_speed(iter/s)": 0.201344
    },
    {
      "acc": 0.78614969,
      "epoch": 0.8574053281616821,
      "grad_norm": 9.125,
      "learning_rate": 6.379256126349903e-06,
      "loss": 0.76010709,
      "memory(GiB)": 147.13,
      "step": 36750,
      "train_speed(iter/s)": 0.201372
    },
    {
      "acc": 0.77783051,
      "epoch": 0.857638635733971,
      "grad_norm": 5.53125,
      "learning_rate": 6.377440213192625e-06,
      "loss": 0.80419178,
      "memory(GiB)": 147.13,
      "step": 36760,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.78635902,
      "epoch": 0.8578719433062599,
      "grad_norm": 5.03125,
      "learning_rate": 6.375624103406155e-06,
      "loss": 0.77373714,
      "memory(GiB)": 147.13,
      "step": 36770,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.79163675,
      "epoch": 0.8581052508785488,
      "grad_norm": 5.71875,
      "learning_rate": 6.373807797249744e-06,
      "loss": 0.74694929,
      "memory(GiB)": 147.13,
      "step": 36780,
      "train_speed(iter/s)": 0.201456
    },
    {
      "acc": 0.77907381,
      "epoch": 0.8583385584508377,
      "grad_norm": 5.875,
      "learning_rate": 6.371991294982671e-06,
      "loss": 0.79713793,
      "memory(GiB)": 147.13,
      "step": 36790,
      "train_speed(iter/s)": 0.201484
    },
    {
      "acc": 0.75770316,
      "epoch": 0.8585718660231266,
      "grad_norm": 6.75,
      "learning_rate": 6.370174596864238e-06,
      "loss": 0.88794937,
      "memory(GiB)": 147.13,
      "step": 36800,
      "train_speed(iter/s)": 0.201513
    },
    {
      "acc": 0.77909842,
      "epoch": 0.8588051735954155,
      "grad_norm": 4.46875,
      "learning_rate": 6.368357703153782e-06,
      "loss": 0.80933161,
      "memory(GiB)": 147.13,
      "step": 36810,
      "train_speed(iter/s)": 0.201542
    },
    {
      "acc": 0.78378386,
      "epoch": 0.8590384811677044,
      "grad_norm": 4.3125,
      "learning_rate": 6.366540614110658e-06,
      "loss": 0.78677444,
      "memory(GiB)": 147.13,
      "step": 36820,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.77209501,
      "epoch": 0.8592717887399933,
      "grad_norm": 5.9375,
      "learning_rate": 6.364723329994259e-06,
      "loss": 0.81071463,
      "memory(GiB)": 147.13,
      "step": 36830,
      "train_speed(iter/s)": 0.201599
    },
    {
      "acc": 0.77325363,
      "epoch": 0.8595050963122822,
      "grad_norm": 6.4375,
      "learning_rate": 6.362905851064001e-06,
      "loss": 0.81591139,
      "memory(GiB)": 147.13,
      "step": 36840,
      "train_speed(iter/s)": 0.201627
    },
    {
      "acc": 0.76463056,
      "epoch": 0.8597384038845711,
      "grad_norm": 4.6875,
      "learning_rate": 6.361088177579329e-06,
      "loss": 0.83884716,
      "memory(GiB)": 147.13,
      "step": 36850,
      "train_speed(iter/s)": 0.201656
    },
    {
      "acc": 0.77708492,
      "epoch": 0.85997171145686,
      "grad_norm": 6.1875,
      "learning_rate": 6.359270309799715e-06,
      "loss": 0.80096121,
      "memory(GiB)": 147.13,
      "step": 36860,
      "train_speed(iter/s)": 0.201683
    },
    {
      "acc": 0.78316288,
      "epoch": 0.8602050190291488,
      "grad_norm": 3.859375,
      "learning_rate": 6.357452247984659e-06,
      "loss": 0.79706192,
      "memory(GiB)": 147.13,
      "step": 36870,
      "train_speed(iter/s)": 0.201711
    },
    {
      "acc": 0.77961206,
      "epoch": 0.8604383266014377,
      "grad_norm": 5.40625,
      "learning_rate": 6.35563399239369e-06,
      "loss": 0.80441303,
      "memory(GiB)": 147.13,
      "step": 36880,
      "train_speed(iter/s)": 0.201737
    },
    {
      "acc": 0.78116093,
      "epoch": 0.8606716341737266,
      "grad_norm": 4.59375,
      "learning_rate": 6.353815543286361e-06,
      "loss": 0.78395686,
      "memory(GiB)": 147.13,
      "step": 36890,
      "train_speed(iter/s)": 0.201765
    },
    {
      "acc": 0.78719234,
      "epoch": 0.8609049417460155,
      "grad_norm": 4.6875,
      "learning_rate": 6.351996900922257e-06,
      "loss": 0.75968537,
      "memory(GiB)": 147.13,
      "step": 36900,
      "train_speed(iter/s)": 0.201793
    },
    {
      "acc": 0.78203468,
      "epoch": 0.8611382493183044,
      "grad_norm": 5.1875,
      "learning_rate": 6.3501780655609875e-06,
      "loss": 0.78212891,
      "memory(GiB)": 147.13,
      "step": 36910,
      "train_speed(iter/s)": 0.201819
    },
    {
      "acc": 0.77617979,
      "epoch": 0.8613715568905933,
      "grad_norm": 5.75,
      "learning_rate": 6.348359037462194e-06,
      "loss": 0.79318113,
      "memory(GiB)": 147.13,
      "step": 36920,
      "train_speed(iter/s)": 0.201847
    },
    {
      "acc": 0.79143863,
      "epoch": 0.8616048644628822,
      "grad_norm": 5.5,
      "learning_rate": 6.346539816885537e-06,
      "loss": 0.73529739,
      "memory(GiB)": 147.13,
      "step": 36930,
      "train_speed(iter/s)": 0.201875
    },
    {
      "acc": 0.76003532,
      "epoch": 0.8618381720351711,
      "grad_norm": 5.1875,
      "learning_rate": 6.3447204040907125e-06,
      "loss": 0.87924767,
      "memory(GiB)": 147.13,
      "step": 36940,
      "train_speed(iter/s)": 0.201904
    },
    {
      "acc": 0.78279285,
      "epoch": 0.86207147960746,
      "grad_norm": 4.75,
      "learning_rate": 6.342900799337443e-06,
      "loss": 0.77494459,
      "memory(GiB)": 147.13,
      "step": 36950,
      "train_speed(iter/s)": 0.201933
    },
    {
      "acc": 0.78726845,
      "epoch": 0.8623047871797489,
      "grad_norm": 6.28125,
      "learning_rate": 6.341081002885472e-06,
      "loss": 0.7636848,
      "memory(GiB)": 147.13,
      "step": 36960,
      "train_speed(iter/s)": 0.201962
    },
    {
      "acc": 0.78955112,
      "epoch": 0.8625380947520378,
      "grad_norm": 22.25,
      "learning_rate": 6.33926101499458e-06,
      "loss": 0.75000238,
      "memory(GiB)": 147.13,
      "step": 36970,
      "train_speed(iter/s)": 0.201987
    },
    {
      "acc": 0.7933239,
      "epoch": 0.8627714023243267,
      "grad_norm": 5.8125,
      "learning_rate": 6.337440835924564e-06,
      "loss": 0.75319366,
      "memory(GiB)": 147.13,
      "step": 36980,
      "train_speed(iter/s)": 0.202016
    },
    {
      "acc": 0.7852212,
      "epoch": 0.8630047098966156,
      "grad_norm": 5.53125,
      "learning_rate": 6.335620465935259e-06,
      "loss": 0.76989651,
      "memory(GiB)": 147.13,
      "step": 36990,
      "train_speed(iter/s)": 0.202045
    },
    {
      "acc": 0.78815894,
      "epoch": 0.8632380174689045,
      "grad_norm": 5.4375,
      "learning_rate": 6.333799905286519e-06,
      "loss": 0.76841455,
      "memory(GiB)": 147.13,
      "step": 37000,
      "train_speed(iter/s)": 0.202073
    },
    {
      "epoch": 0.8632380174689045,
      "eval_acc": 0.7431547548984517,
      "eval_loss": 0.8100025653839111,
      "eval_runtime": 1269.5429,
      "eval_samples_per_second": 28.35,
      "eval_steps_per_second": 14.175,
      "step": 37000
    },
    {
      "acc": 0.76950207,
      "epoch": 0.8634713250411934,
      "grad_norm": 5.25,
      "learning_rate": 6.331979154238232e-06,
      "loss": 0.82681618,
      "memory(GiB)": 147.13,
      "step": 37010,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.77774849,
      "epoch": 0.8637046326134823,
      "grad_norm": 4.71875,
      "learning_rate": 6.330158213050308e-06,
      "loss": 0.81203289,
      "memory(GiB)": 147.13,
      "step": 37020,
      "train_speed(iter/s)": 0.20072
    },
    {
      "acc": 0.77672329,
      "epoch": 0.8639379401857712,
      "grad_norm": 9.375,
      "learning_rate": 6.328337081982685e-06,
      "loss": 0.78580074,
      "memory(GiB)": 147.13,
      "step": 37030,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.79651041,
      "epoch": 0.8641712477580601,
      "grad_norm": 5.125,
      "learning_rate": 6.326515761295328e-06,
      "loss": 0.72057328,
      "memory(GiB)": 147.13,
      "step": 37040,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.767904,
      "epoch": 0.864404555330349,
      "grad_norm": 6.1875,
      "learning_rate": 6.3246942512482325e-06,
      "loss": 0.84308643,
      "memory(GiB)": 147.13,
      "step": 37050,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.77300673,
      "epoch": 0.8646378629026379,
      "grad_norm": 4.9375,
      "learning_rate": 6.3228725521014165e-06,
      "loss": 0.8220705,
      "memory(GiB)": 147.13,
      "step": 37060,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.81248608,
      "epoch": 0.8648711704749267,
      "grad_norm": 8.0625,
      "learning_rate": 6.32105066411493e-06,
      "loss": 0.68309054,
      "memory(GiB)": 147.13,
      "step": 37070,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.76430616,
      "epoch": 0.8651044780472156,
      "grad_norm": 5.09375,
      "learning_rate": 6.319228587548843e-06,
      "loss": 0.85580292,
      "memory(GiB)": 147.13,
      "step": 37080,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.76082668,
      "epoch": 0.8653377856195045,
      "grad_norm": 5.0,
      "learning_rate": 6.317406322663259e-06,
      "loss": 0.86531963,
      "memory(GiB)": 147.13,
      "step": 37090,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.77870302,
      "epoch": 0.8655710931917934,
      "grad_norm": 4.40625,
      "learning_rate": 6.315583869718306e-06,
      "loss": 0.79484749,
      "memory(GiB)": 147.13,
      "step": 37100,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.7822587,
      "epoch": 0.8658044007640823,
      "grad_norm": 4.71875,
      "learning_rate": 6.313761228974137e-06,
      "loss": 0.80919828,
      "memory(GiB)": 147.13,
      "step": 37110,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.76600981,
      "epoch": 0.8660377083363712,
      "grad_norm": 6.15625,
      "learning_rate": 6.311938400690933e-06,
      "loss": 0.83051043,
      "memory(GiB)": 147.13,
      "step": 37120,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.78756142,
      "epoch": 0.8662710159086601,
      "grad_norm": 4.21875,
      "learning_rate": 6.310115385128905e-06,
      "loss": 0.73773327,
      "memory(GiB)": 147.13,
      "step": 37130,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.77740078,
      "epoch": 0.866504323480949,
      "grad_norm": 5.6875,
      "learning_rate": 6.308292182548287e-06,
      "loss": 0.79663916,
      "memory(GiB)": 147.13,
      "step": 37140,
      "train_speed(iter/s)": 0.20105
    },
    {
      "acc": 0.7775095,
      "epoch": 0.8667376310532379,
      "grad_norm": 4.3125,
      "learning_rate": 6.3064687932093386e-06,
      "loss": 0.79716334,
      "memory(GiB)": 147.13,
      "step": 37150,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78595037,
      "epoch": 0.8669709386255268,
      "grad_norm": 4.0,
      "learning_rate": 6.3046452173723495e-06,
      "loss": 0.76871729,
      "memory(GiB)": 147.13,
      "step": 37160,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.76833076,
      "epoch": 0.8672042461978157,
      "grad_norm": 4.90625,
      "learning_rate": 6.302821455297635e-06,
      "loss": 0.83416252,
      "memory(GiB)": 147.13,
      "step": 37170,
      "train_speed(iter/s)": 0.201131
    },
    {
      "acc": 0.77176652,
      "epoch": 0.8674375537701046,
      "grad_norm": 4.84375,
      "learning_rate": 6.300997507245537e-06,
      "loss": 0.82635098,
      "memory(GiB)": 147.13,
      "step": 37180,
      "train_speed(iter/s)": 0.201159
    },
    {
      "acc": 0.78389416,
      "epoch": 0.8676708613423935,
      "grad_norm": 5.4375,
      "learning_rate": 6.299173373476422e-06,
      "loss": 0.77835617,
      "memory(GiB)": 147.13,
      "step": 37190,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.7786643,
      "epoch": 0.8679041689146824,
      "grad_norm": 4.65625,
      "learning_rate": 6.2973490542506854e-06,
      "loss": 0.81813984,
      "memory(GiB)": 147.13,
      "step": 37200,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.78079252,
      "epoch": 0.8681374764869713,
      "grad_norm": 5.625,
      "learning_rate": 6.295524549828747e-06,
      "loss": 0.76534152,
      "memory(GiB)": 147.13,
      "step": 37210,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.77490101,
      "epoch": 0.8683707840592602,
      "grad_norm": 5.5,
      "learning_rate": 6.293699860471057e-06,
      "loss": 0.77603149,
      "memory(GiB)": 147.13,
      "step": 37220,
      "train_speed(iter/s)": 0.201269
    },
    {
      "acc": 0.75890179,
      "epoch": 0.8686040916315491,
      "grad_norm": 6.375,
      "learning_rate": 6.2918749864380875e-06,
      "loss": 0.8679534,
      "memory(GiB)": 147.13,
      "step": 37230,
      "train_speed(iter/s)": 0.201298
    },
    {
      "acc": 0.7783083,
      "epoch": 0.868837399203838,
      "grad_norm": 4.3125,
      "learning_rate": 6.290049927990339e-06,
      "loss": 0.79737039,
      "memory(GiB)": 147.13,
      "step": 37240,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.77292824,
      "epoch": 0.8690707067761269,
      "grad_norm": 5.25,
      "learning_rate": 6.288224685388337e-06,
      "loss": 0.82085094,
      "memory(GiB)": 147.13,
      "step": 37250,
      "train_speed(iter/s)": 0.201351
    },
    {
      "acc": 0.78002787,
      "epoch": 0.8693040143484156,
      "grad_norm": 4.5625,
      "learning_rate": 6.286399258892638e-06,
      "loss": 0.80144997,
      "memory(GiB)": 147.13,
      "step": 37260,
      "train_speed(iter/s)": 0.201378
    },
    {
      "acc": 0.78442354,
      "epoch": 0.8695373219207045,
      "grad_norm": 3.90625,
      "learning_rate": 6.284573648763816e-06,
      "loss": 0.77326307,
      "memory(GiB)": 147.13,
      "step": 37270,
      "train_speed(iter/s)": 0.201406
    },
    {
      "acc": 0.78011799,
      "epoch": 0.8697706294929934,
      "grad_norm": 4.96875,
      "learning_rate": 6.28274785526248e-06,
      "loss": 0.82026291,
      "memory(GiB)": 147.13,
      "step": 37280,
      "train_speed(iter/s)": 0.201434
    },
    {
      "acc": 0.76693525,
      "epoch": 0.8700039370652823,
      "grad_norm": 12.875,
      "learning_rate": 6.2809218786492595e-06,
      "loss": 0.84187355,
      "memory(GiB)": 147.13,
      "step": 37290,
      "train_speed(iter/s)": 0.201462
    },
    {
      "acc": 0.79030437,
      "epoch": 0.8702372446375712,
      "grad_norm": 6.53125,
      "learning_rate": 6.279095719184813e-06,
      "loss": 0.7504972,
      "memory(GiB)": 147.13,
      "step": 37300,
      "train_speed(iter/s)": 0.201489
    },
    {
      "acc": 0.77714148,
      "epoch": 0.8704705522098601,
      "grad_norm": 4.71875,
      "learning_rate": 6.277269377129826e-06,
      "loss": 0.79407024,
      "memory(GiB)": 147.13,
      "step": 37310,
      "train_speed(iter/s)": 0.201517
    },
    {
      "acc": 0.77218485,
      "epoch": 0.870703859782149,
      "grad_norm": 6.09375,
      "learning_rate": 6.275442852745005e-06,
      "loss": 0.82651482,
      "memory(GiB)": 147.13,
      "step": 37320,
      "train_speed(iter/s)": 0.201546
    },
    {
      "acc": 0.77186918,
      "epoch": 0.8709371673544379,
      "grad_norm": 9.3125,
      "learning_rate": 6.273616146291086e-06,
      "loss": 0.81594334,
      "memory(GiB)": 147.13,
      "step": 37330,
      "train_speed(iter/s)": 0.201575
    },
    {
      "acc": 0.7657239,
      "epoch": 0.8711704749267268,
      "grad_norm": 5.59375,
      "learning_rate": 6.2717892580288335e-06,
      "loss": 0.83614559,
      "memory(GiB)": 147.13,
      "step": 37340,
      "train_speed(iter/s)": 0.201603
    },
    {
      "acc": 0.77476711,
      "epoch": 0.8714037824990157,
      "grad_norm": 4.75,
      "learning_rate": 6.269962188219034e-06,
      "loss": 0.80064745,
      "memory(GiB)": 147.13,
      "step": 37350,
      "train_speed(iter/s)": 0.20163
    },
    {
      "acc": 0.78318167,
      "epoch": 0.8716370900713046,
      "grad_norm": 5.9375,
      "learning_rate": 6.2681349371225e-06,
      "loss": 0.78883448,
      "memory(GiB)": 147.13,
      "step": 37360,
      "train_speed(iter/s)": 0.201654
    },
    {
      "acc": 0.79107747,
      "epoch": 0.8718703976435935,
      "grad_norm": 3.90625,
      "learning_rate": 6.266307505000073e-06,
      "loss": 0.72986178,
      "memory(GiB)": 147.13,
      "step": 37370,
      "train_speed(iter/s)": 0.201681
    },
    {
      "acc": 0.77610064,
      "epoch": 0.8721037052158824,
      "grad_norm": 6.8125,
      "learning_rate": 6.264479892112619e-06,
      "loss": 0.79103355,
      "memory(GiB)": 147.13,
      "step": 37380,
      "train_speed(iter/s)": 0.201711
    },
    {
      "acc": 0.77150335,
      "epoch": 0.8723370127881713,
      "grad_norm": 6.0625,
      "learning_rate": 6.262652098721026e-06,
      "loss": 0.82169552,
      "memory(GiB)": 147.13,
      "step": 37390,
      "train_speed(iter/s)": 0.201738
    },
    {
      "acc": 0.77216101,
      "epoch": 0.8725703203604602,
      "grad_norm": 5.65625,
      "learning_rate": 6.260824125086212e-06,
      "loss": 0.8190197,
      "memory(GiB)": 147.13,
      "step": 37400,
      "train_speed(iter/s)": 0.201767
    },
    {
      "acc": 0.77426405,
      "epoch": 0.8728036279327491,
      "grad_norm": 5.375,
      "learning_rate": 6.258995971469122e-06,
      "loss": 0.83238316,
      "memory(GiB)": 147.13,
      "step": 37410,
      "train_speed(iter/s)": 0.201795
    },
    {
      "acc": 0.78717422,
      "epoch": 0.873036935505038,
      "grad_norm": 6.0625,
      "learning_rate": 6.2571676381307215e-06,
      "loss": 0.79043889,
      "memory(GiB)": 147.13,
      "step": 37420,
      "train_speed(iter/s)": 0.201823
    },
    {
      "acc": 0.78104959,
      "epoch": 0.8732702430773269,
      "grad_norm": 5.0625,
      "learning_rate": 6.255339125332007e-06,
      "loss": 0.79519758,
      "memory(GiB)": 147.13,
      "step": 37430,
      "train_speed(iter/s)": 0.201849
    },
    {
      "acc": 0.7618619,
      "epoch": 0.8735035506496158,
      "grad_norm": 4.90625,
      "learning_rate": 6.253510433333996e-06,
      "loss": 0.86247349,
      "memory(GiB)": 147.13,
      "step": 37440,
      "train_speed(iter/s)": 0.201877
    },
    {
      "acc": 0.77435389,
      "epoch": 0.8737368582219046,
      "grad_norm": 5.84375,
      "learning_rate": 6.251681562397736e-06,
      "loss": 0.79898653,
      "memory(GiB)": 147.13,
      "step": 37450,
      "train_speed(iter/s)": 0.201905
    },
    {
      "acc": 0.78359423,
      "epoch": 0.8739701657941935,
      "grad_norm": 5.4375,
      "learning_rate": 6.2498525127842955e-06,
      "loss": 0.76345568,
      "memory(GiB)": 147.13,
      "step": 37460,
      "train_speed(iter/s)": 0.201931
    },
    {
      "acc": 0.7648387,
      "epoch": 0.8742034733664824,
      "grad_norm": 5.25,
      "learning_rate": 6.248023284754772e-06,
      "loss": 0.83345156,
      "memory(GiB)": 147.13,
      "step": 37470,
      "train_speed(iter/s)": 0.20196
    },
    {
      "acc": 0.78857408,
      "epoch": 0.8744367809387713,
      "grad_norm": 5.0625,
      "learning_rate": 6.2461938785702866e-06,
      "loss": 0.75806623,
      "memory(GiB)": 147.13,
      "step": 37480,
      "train_speed(iter/s)": 0.201987
    },
    {
      "acc": 0.76943989,
      "epoch": 0.8746700885110602,
      "grad_norm": 3.984375,
      "learning_rate": 6.244364294491989e-06,
      "loss": 0.82730265,
      "memory(GiB)": 147.13,
      "step": 37490,
      "train_speed(iter/s)": 0.202015
    },
    {
      "acc": 0.77271547,
      "epoch": 0.8749033960833491,
      "grad_norm": 7.25,
      "learning_rate": 6.2425345327810485e-06,
      "loss": 0.81668425,
      "memory(GiB)": 147.13,
      "step": 37500,
      "train_speed(iter/s)": 0.202043
    },
    {
      "epoch": 0.8749033960833491,
      "eval_acc": 0.7432971231209676,
      "eval_loss": 0.8092904686927795,
      "eval_runtime": 1270.0898,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 37500
    },
    {
      "acc": 0.77455845,
      "epoch": 0.875136703655638,
      "grad_norm": 5.8125,
      "learning_rate": 6.240704593698664e-06,
      "loss": 0.79787617,
      "memory(GiB)": 147.13,
      "step": 37510,
      "train_speed(iter/s)": 0.20068
    },
    {
      "acc": 0.77178693,
      "epoch": 0.8753700112279269,
      "grad_norm": 6.84375,
      "learning_rate": 6.238874477506061e-06,
      "loss": 0.81941929,
      "memory(GiB)": 147.13,
      "step": 37520,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.75730181,
      "epoch": 0.8756033188002158,
      "grad_norm": 6.46875,
      "learning_rate": 6.237044184464485e-06,
      "loss": 0.89033489,
      "memory(GiB)": 147.13,
      "step": 37530,
      "train_speed(iter/s)": 0.20073
    },
    {
      "acc": 0.77331123,
      "epoch": 0.8758366263725047,
      "grad_norm": 4.3125,
      "learning_rate": 6.235213714835211e-06,
      "loss": 0.82823706,
      "memory(GiB)": 147.13,
      "step": 37540,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.78762941,
      "epoch": 0.8760699339447936,
      "grad_norm": 4.125,
      "learning_rate": 6.233383068879538e-06,
      "loss": 0.75171795,
      "memory(GiB)": 147.13,
      "step": 37550,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.7689115,
      "epoch": 0.8763032415170825,
      "grad_norm": 5.28125,
      "learning_rate": 6.231552246858791e-06,
      "loss": 0.83698616,
      "memory(GiB)": 147.13,
      "step": 37560,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.78390293,
      "epoch": 0.8765365490893714,
      "grad_norm": 5.78125,
      "learning_rate": 6.229721249034318e-06,
      "loss": 0.79848022,
      "memory(GiB)": 147.13,
      "step": 37570,
      "train_speed(iter/s)": 0.200839
    },
    {
      "acc": 0.78724899,
      "epoch": 0.8767698566616603,
      "grad_norm": 5.59375,
      "learning_rate": 6.227890075667492e-06,
      "loss": 0.77459116,
      "memory(GiB)": 147.13,
      "step": 37580,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.79053335,
      "epoch": 0.8770031642339492,
      "grad_norm": 5.65625,
      "learning_rate": 6.226058727019717e-06,
      "loss": 0.75305357,
      "memory(GiB)": 147.13,
      "step": 37590,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.75518756,
      "epoch": 0.8772364718062381,
      "grad_norm": 5.8125,
      "learning_rate": 6.224227203352415e-06,
      "loss": 0.87923584,
      "memory(GiB)": 147.13,
      "step": 37600,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.78576441,
      "epoch": 0.877469779378527,
      "grad_norm": 4.875,
      "learning_rate": 6.222395504927035e-06,
      "loss": 0.77100844,
      "memory(GiB)": 147.13,
      "step": 37610,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.7848784,
      "epoch": 0.8777030869508159,
      "grad_norm": 6.3125,
      "learning_rate": 6.22056363200505e-06,
      "loss": 0.7881711,
      "memory(GiB)": 147.13,
      "step": 37620,
      "train_speed(iter/s)": 0.200976
    },
    {
      "acc": 0.77304902,
      "epoch": 0.8779363945231048,
      "grad_norm": 10.875,
      "learning_rate": 6.218731584847963e-06,
      "loss": 0.81832428,
      "memory(GiB)": 147.13,
      "step": 37630,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.78921366,
      "epoch": 0.8781697020953936,
      "grad_norm": 6.65625,
      "learning_rate": 6.216899363717295e-06,
      "loss": 0.76330223,
      "memory(GiB)": 147.13,
      "step": 37640,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.77204285,
      "epoch": 0.8784030096676825,
      "grad_norm": 7.15625,
      "learning_rate": 6.215066968874596e-06,
      "loss": 0.83313923,
      "memory(GiB)": 147.13,
      "step": 37650,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.78158464,
      "epoch": 0.8786363172399714,
      "grad_norm": 5.75,
      "learning_rate": 6.213234400581442e-06,
      "loss": 0.7798768,
      "memory(GiB)": 147.13,
      "step": 37660,
      "train_speed(iter/s)": 0.201087
    },
    {
      "acc": 0.77431865,
      "epoch": 0.8788696248122603,
      "grad_norm": 13.3125,
      "learning_rate": 6.2114016590994295e-06,
      "loss": 0.82120934,
      "memory(GiB)": 147.13,
      "step": 37670,
      "train_speed(iter/s)": 0.201113
    },
    {
      "acc": 0.76597242,
      "epoch": 0.8791029323845492,
      "grad_norm": 6.1875,
      "learning_rate": 6.209568744690181e-06,
      "loss": 0.82842293,
      "memory(GiB)": 147.13,
      "step": 37680,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.7780242,
      "epoch": 0.8793362399568381,
      "grad_norm": 7.125,
      "learning_rate": 6.207735657615346e-06,
      "loss": 0.7849369,
      "memory(GiB)": 147.13,
      "step": 37690,
      "train_speed(iter/s)": 0.201167
    },
    {
      "acc": 0.79243302,
      "epoch": 0.879569547529127,
      "grad_norm": 4.15625,
      "learning_rate": 6.2059023981365965e-06,
      "loss": 0.75370507,
      "memory(GiB)": 147.13,
      "step": 37700,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.78596573,
      "epoch": 0.8798028551014159,
      "grad_norm": 5.46875,
      "learning_rate": 6.20406896651563e-06,
      "loss": 0.76070099,
      "memory(GiB)": 147.13,
      "step": 37710,
      "train_speed(iter/s)": 0.201224
    },
    {
      "acc": 0.77707858,
      "epoch": 0.8800361626737048,
      "grad_norm": 6.59375,
      "learning_rate": 6.202235363014169e-06,
      "loss": 0.82894115,
      "memory(GiB)": 147.13,
      "step": 37720,
      "train_speed(iter/s)": 0.201248
    },
    {
      "acc": 0.77520704,
      "epoch": 0.8802694702459937,
      "grad_norm": 6.8125,
      "learning_rate": 6.2004015878939585e-06,
      "loss": 0.81481781,
      "memory(GiB)": 147.13,
      "step": 37730,
      "train_speed(iter/s)": 0.201275
    },
    {
      "acc": 0.78262343,
      "epoch": 0.8805027778182826,
      "grad_norm": 6.28125,
      "learning_rate": 6.198567641416772e-06,
      "loss": 0.78242087,
      "memory(GiB)": 147.13,
      "step": 37740,
      "train_speed(iter/s)": 0.201304
    },
    {
      "acc": 0.77590933,
      "epoch": 0.8807360853905715,
      "grad_norm": 5.09375,
      "learning_rate": 6.1967335238444004e-06,
      "loss": 0.81266613,
      "memory(GiB)": 147.13,
      "step": 37750,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.7834343,
      "epoch": 0.8809693929628604,
      "grad_norm": 6.9375,
      "learning_rate": 6.194899235438666e-06,
      "loss": 0.75194969,
      "memory(GiB)": 147.13,
      "step": 37760,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.75540552,
      "epoch": 0.8812027005351493,
      "grad_norm": 4.34375,
      "learning_rate": 6.193064776461415e-06,
      "loss": 0.88943443,
      "memory(GiB)": 147.13,
      "step": 37770,
      "train_speed(iter/s)": 0.201387
    },
    {
      "acc": 0.78063102,
      "epoch": 0.8814360081074382,
      "grad_norm": 4.25,
      "learning_rate": 6.191230147174512e-06,
      "loss": 0.80147295,
      "memory(GiB)": 147.13,
      "step": 37780,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.77797747,
      "epoch": 0.8816693156797271,
      "grad_norm": 5.5625,
      "learning_rate": 6.1893953478398515e-06,
      "loss": 0.81042271,
      "memory(GiB)": 147.13,
      "step": 37790,
      "train_speed(iter/s)": 0.201438
    },
    {
      "acc": 0.78152409,
      "epoch": 0.881902623252016,
      "grad_norm": 5.96875,
      "learning_rate": 6.18756037871935e-06,
      "loss": 0.77627401,
      "memory(GiB)": 147.13,
      "step": 37800,
      "train_speed(iter/s)": 0.201464
    },
    {
      "acc": 0.76364655,
      "epoch": 0.8821359308243049,
      "grad_norm": 4.125,
      "learning_rate": 6.185725240074951e-06,
      "loss": 0.85897408,
      "memory(GiB)": 147.13,
      "step": 37810,
      "train_speed(iter/s)": 0.201494
    },
    {
      "acc": 0.78259721,
      "epoch": 0.8823692383965938,
      "grad_norm": 5.28125,
      "learning_rate": 6.1838899321686185e-06,
      "loss": 0.78828173,
      "memory(GiB)": 147.13,
      "step": 37820,
      "train_speed(iter/s)": 0.201523
    },
    {
      "acc": 0.77779245,
      "epoch": 0.8826025459688827,
      "grad_norm": 4.625,
      "learning_rate": 6.1820544552623415e-06,
      "loss": 0.80315275,
      "memory(GiB)": 147.13,
      "step": 37830,
      "train_speed(iter/s)": 0.201551
    },
    {
      "acc": 0.77334061,
      "epoch": 0.8828358535411714,
      "grad_norm": 5.78125,
      "learning_rate": 6.180218809618135e-06,
      "loss": 0.81666965,
      "memory(GiB)": 147.13,
      "step": 37840,
      "train_speed(iter/s)": 0.201578
    },
    {
      "acc": 0.76984267,
      "epoch": 0.8830691611134603,
      "grad_norm": 6.96875,
      "learning_rate": 6.1783829954980345e-06,
      "loss": 0.83126059,
      "memory(GiB)": 147.13,
      "step": 37850,
      "train_speed(iter/s)": 0.201607
    },
    {
      "acc": 0.77715025,
      "epoch": 0.8833024686857492,
      "grad_norm": 6.84375,
      "learning_rate": 6.176547013164104e-06,
      "loss": 0.79118605,
      "memory(GiB)": 147.13,
      "step": 37860,
      "train_speed(iter/s)": 0.201634
    },
    {
      "acc": 0.76471095,
      "epoch": 0.8835357762580381,
      "grad_norm": 5.34375,
      "learning_rate": 6.17471086287843e-06,
      "loss": 0.83969116,
      "memory(GiB)": 147.13,
      "step": 37870,
      "train_speed(iter/s)": 0.201663
    },
    {
      "acc": 0.78056355,
      "epoch": 0.883769083830327,
      "grad_norm": 7.3125,
      "learning_rate": 6.172874544903122e-06,
      "loss": 0.76982565,
      "memory(GiB)": 147.13,
      "step": 37880,
      "train_speed(iter/s)": 0.201689
    },
    {
      "acc": 0.78541079,
      "epoch": 0.8840023914026159,
      "grad_norm": 6.625,
      "learning_rate": 6.171038059500315e-06,
      "loss": 0.77315903,
      "memory(GiB)": 147.13,
      "step": 37890,
      "train_speed(iter/s)": 0.201715
    },
    {
      "acc": 0.78337684,
      "epoch": 0.8842356989749048,
      "grad_norm": 4.15625,
      "learning_rate": 6.169201406932163e-06,
      "loss": 0.78238239,
      "memory(GiB)": 147.13,
      "step": 37900,
      "train_speed(iter/s)": 0.201744
    },
    {
      "acc": 0.7873363,
      "epoch": 0.8844690065471937,
      "grad_norm": 6.03125,
      "learning_rate": 6.167364587460849e-06,
      "loss": 0.75477743,
      "memory(GiB)": 147.13,
      "step": 37910,
      "train_speed(iter/s)": 0.201772
    },
    {
      "acc": 0.7772418,
      "epoch": 0.8847023141194826,
      "grad_norm": 5.71875,
      "learning_rate": 6.16552760134858e-06,
      "loss": 0.84283714,
      "memory(GiB)": 147.13,
      "step": 37920,
      "train_speed(iter/s)": 0.201801
    },
    {
      "acc": 0.78096313,
      "epoch": 0.8849356216917715,
      "grad_norm": 5.34375,
      "learning_rate": 6.1636904488575845e-06,
      "loss": 0.79128194,
      "memory(GiB)": 147.13,
      "step": 37930,
      "train_speed(iter/s)": 0.201826
    },
    {
      "acc": 0.79532461,
      "epoch": 0.8851689292640604,
      "grad_norm": 5.5625,
      "learning_rate": 6.161853130250117e-06,
      "loss": 0.74347682,
      "memory(GiB)": 147.13,
      "step": 37940,
      "train_speed(iter/s)": 0.201852
    },
    {
      "acc": 0.77315941,
      "epoch": 0.8854022368363493,
      "grad_norm": 4.75,
      "learning_rate": 6.160015645788451e-06,
      "loss": 0.8172802,
      "memory(GiB)": 147.13,
      "step": 37950,
      "train_speed(iter/s)": 0.201879
    },
    {
      "acc": 0.77103496,
      "epoch": 0.8856355444086382,
      "grad_norm": 5.1875,
      "learning_rate": 6.15817799573489e-06,
      "loss": 0.83767815,
      "memory(GiB)": 147.13,
      "step": 37960,
      "train_speed(iter/s)": 0.201905
    },
    {
      "acc": 0.80051785,
      "epoch": 0.8858688519809271,
      "grad_norm": 5.53125,
      "learning_rate": 6.1563401803517545e-06,
      "loss": 0.70509105,
      "memory(GiB)": 147.13,
      "step": 37970,
      "train_speed(iter/s)": 0.201931
    },
    {
      "acc": 0.77665696,
      "epoch": 0.886102159553216,
      "grad_norm": 5.375,
      "learning_rate": 6.154502199901396e-06,
      "loss": 0.81156635,
      "memory(GiB)": 147.13,
      "step": 37980,
      "train_speed(iter/s)": 0.201956
    },
    {
      "acc": 0.76781158,
      "epoch": 0.8863354671255049,
      "grad_norm": 4.46875,
      "learning_rate": 6.152664054646183e-06,
      "loss": 0.83875208,
      "memory(GiB)": 147.13,
      "step": 37990,
      "train_speed(iter/s)": 0.201984
    },
    {
      "acc": 0.75077744,
      "epoch": 0.8865687746977938,
      "grad_norm": 5.875,
      "learning_rate": 6.150825744848511e-06,
      "loss": 0.90200367,
      "memory(GiB)": 147.13,
      "step": 38000,
      "train_speed(iter/s)": 0.202011
    },
    {
      "epoch": 0.8865687746977938,
      "eval_acc": 0.74343228485303,
      "eval_loss": 0.8092561960220337,
      "eval_runtime": 1268.5776,
      "eval_samples_per_second": 28.371,
      "eval_steps_per_second": 14.186,
      "step": 38000
    },
    {
      "acc": 0.7838891,
      "epoch": 0.8868020822700827,
      "grad_norm": 5.5625,
      "learning_rate": 6.148987270770798e-06,
      "loss": 0.78417902,
      "memory(GiB)": 147.13,
      "step": 38010,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.79363413,
      "epoch": 0.8870353898423716,
      "grad_norm": 4.78125,
      "learning_rate": 6.147148632675486e-06,
      "loss": 0.73398776,
      "memory(GiB)": 147.13,
      "step": 38020,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.77010832,
      "epoch": 0.8872686974146604,
      "grad_norm": 12.5,
      "learning_rate": 6.145309830825041e-06,
      "loss": 0.87349873,
      "memory(GiB)": 147.13,
      "step": 38030,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.76912684,
      "epoch": 0.8875020049869493,
      "grad_norm": 5.15625,
      "learning_rate": 6.143470865481948e-06,
      "loss": 0.81946507,
      "memory(GiB)": 147.13,
      "step": 38040,
      "train_speed(iter/s)": 0.200748
    },
    {
      "acc": 0.76783762,
      "epoch": 0.8877353125592382,
      "grad_norm": 5.125,
      "learning_rate": 6.141631736908723e-06,
      "loss": 0.82789316,
      "memory(GiB)": 147.13,
      "step": 38050,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.78408384,
      "epoch": 0.8879686201315271,
      "grad_norm": 7.0625,
      "learning_rate": 6.1397924453679e-06,
      "loss": 0.7791399,
      "memory(GiB)": 147.13,
      "step": 38060,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.77008028,
      "epoch": 0.888201927703816,
      "grad_norm": 7.625,
      "learning_rate": 6.137952991122035e-06,
      "loss": 0.83867502,
      "memory(GiB)": 147.13,
      "step": 38070,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.79044018,
      "epoch": 0.8884352352761049,
      "grad_norm": 3.90625,
      "learning_rate": 6.136113374433712e-06,
      "loss": 0.74994631,
      "memory(GiB)": 147.13,
      "step": 38080,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.76272373,
      "epoch": 0.8886685428483938,
      "grad_norm": 4.84375,
      "learning_rate": 6.134273595565534e-06,
      "loss": 0.84841614,
      "memory(GiB)": 147.13,
      "step": 38090,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.76892252,
      "epoch": 0.8889018504206827,
      "grad_norm": 6.8125,
      "learning_rate": 6.13243365478013e-06,
      "loss": 0.81232853,
      "memory(GiB)": 147.13,
      "step": 38100,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.78824024,
      "epoch": 0.8891351579929716,
      "grad_norm": 5.59375,
      "learning_rate": 6.13059355234015e-06,
      "loss": 0.75986581,
      "memory(GiB)": 147.13,
      "step": 38110,
      "train_speed(iter/s)": 0.200939
    },
    {
      "acc": 0.77105479,
      "epoch": 0.8893684655652605,
      "grad_norm": 6.46875,
      "learning_rate": 6.128753288508271e-06,
      "loss": 0.83737135,
      "memory(GiB)": 147.13,
      "step": 38120,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.79165573,
      "epoch": 0.8896017731375494,
      "grad_norm": 7.875,
      "learning_rate": 6.126912863547186e-06,
      "loss": 0.74300852,
      "memory(GiB)": 147.13,
      "step": 38130,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.79510765,
      "epoch": 0.8898350807098383,
      "grad_norm": 5.0,
      "learning_rate": 6.125072277719618e-06,
      "loss": 0.7389545,
      "memory(GiB)": 147.13,
      "step": 38140,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.79052238,
      "epoch": 0.8900683882821272,
      "grad_norm": 6.25,
      "learning_rate": 6.123231531288308e-06,
      "loss": 0.75054169,
      "memory(GiB)": 147.13,
      "step": 38150,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.76992264,
      "epoch": 0.8903016958544161,
      "grad_norm": 7.71875,
      "learning_rate": 6.121390624516026e-06,
      "loss": 0.83893986,
      "memory(GiB)": 147.13,
      "step": 38160,
      "train_speed(iter/s)": 0.201076
    },
    {
      "acc": 0.78948836,
      "epoch": 0.890535003426705,
      "grad_norm": 8.0625,
      "learning_rate": 6.119549557665556e-06,
      "loss": 0.74797955,
      "memory(GiB)": 147.13,
      "step": 38170,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.77837534,
      "epoch": 0.8907683109989939,
      "grad_norm": 6.4375,
      "learning_rate": 6.117708330999712e-06,
      "loss": 0.78834462,
      "memory(GiB)": 147.13,
      "step": 38180,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.75297174,
      "epoch": 0.8910016185712828,
      "grad_norm": 4.59375,
      "learning_rate": 6.115866944781329e-06,
      "loss": 0.86872931,
      "memory(GiB)": 147.13,
      "step": 38190,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.79166384,
      "epoch": 0.8912349261435717,
      "grad_norm": 6.21875,
      "learning_rate": 6.114025399273264e-06,
      "loss": 0.72422724,
      "memory(GiB)": 147.13,
      "step": 38200,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.788766,
      "epoch": 0.8914682337158606,
      "grad_norm": 5.21875,
      "learning_rate": 6.112183694738395e-06,
      "loss": 0.74276934,
      "memory(GiB)": 147.13,
      "step": 38210,
      "train_speed(iter/s)": 0.20121
    },
    {
      "acc": 0.78757305,
      "epoch": 0.8917015412881494,
      "grad_norm": 5.75,
      "learning_rate": 6.110341831439628e-06,
      "loss": 0.76513844,
      "memory(GiB)": 147.13,
      "step": 38220,
      "train_speed(iter/s)": 0.201236
    },
    {
      "acc": 0.77797546,
      "epoch": 0.8919348488604383,
      "grad_norm": 5.1875,
      "learning_rate": 6.108499809639887e-06,
      "loss": 0.80226593,
      "memory(GiB)": 147.13,
      "step": 38230,
      "train_speed(iter/s)": 0.201263
    },
    {
      "acc": 0.78466263,
      "epoch": 0.8921681564327272,
      "grad_norm": 6.09375,
      "learning_rate": 6.106657629602122e-06,
      "loss": 0.76339788,
      "memory(GiB)": 147.13,
      "step": 38240,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.78802223,
      "epoch": 0.8924014640050161,
      "grad_norm": 4.5,
      "learning_rate": 6.104815291589299e-06,
      "loss": 0.75786328,
      "memory(GiB)": 147.13,
      "step": 38250,
      "train_speed(iter/s)": 0.20132
    },
    {
      "acc": 0.77927275,
      "epoch": 0.892634771577305,
      "grad_norm": 7.09375,
      "learning_rate": 6.1029727958644144e-06,
      "loss": 0.80576792,
      "memory(GiB)": 147.13,
      "step": 38260,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.78089199,
      "epoch": 0.8928680791495939,
      "grad_norm": 5.84375,
      "learning_rate": 6.1011301426904845e-06,
      "loss": 0.80531197,
      "memory(GiB)": 147.13,
      "step": 38270,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.79331675,
      "epoch": 0.8931013867218828,
      "grad_norm": 5.59375,
      "learning_rate": 6.0992873323305465e-06,
      "loss": 0.74135227,
      "memory(GiB)": 147.13,
      "step": 38280,
      "train_speed(iter/s)": 0.201404
    },
    {
      "acc": 0.78814082,
      "epoch": 0.8933346942941717,
      "grad_norm": 4.9375,
      "learning_rate": 6.097444365047662e-06,
      "loss": 0.76247959,
      "memory(GiB)": 147.13,
      "step": 38290,
      "train_speed(iter/s)": 0.201432
    },
    {
      "acc": 0.76565762,
      "epoch": 0.8935680018664606,
      "grad_norm": 6.375,
      "learning_rate": 6.095601241104911e-06,
      "loss": 0.85332813,
      "memory(GiB)": 147.13,
      "step": 38300,
      "train_speed(iter/s)": 0.201459
    },
    {
      "acc": 0.79775143,
      "epoch": 0.8938013094387495,
      "grad_norm": 4.78125,
      "learning_rate": 6.093757960765404e-06,
      "loss": 0.72894554,
      "memory(GiB)": 147.13,
      "step": 38310,
      "train_speed(iter/s)": 0.201486
    },
    {
      "acc": 0.76751165,
      "epoch": 0.8940346170110384,
      "grad_norm": 4.90625,
      "learning_rate": 6.091914524292264e-06,
      "loss": 0.86291084,
      "memory(GiB)": 147.13,
      "step": 38320,
      "train_speed(iter/s)": 0.201515
    },
    {
      "acc": 0.76315336,
      "epoch": 0.8942679245833273,
      "grad_norm": 6.5,
      "learning_rate": 6.090070931948643e-06,
      "loss": 0.85644455,
      "memory(GiB)": 147.13,
      "step": 38330,
      "train_speed(iter/s)": 0.201542
    },
    {
      "acc": 0.79633956,
      "epoch": 0.8945012321556162,
      "grad_norm": 6.65625,
      "learning_rate": 6.088227183997715e-06,
      "loss": 0.72798471,
      "memory(GiB)": 147.13,
      "step": 38340,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.78646612,
      "epoch": 0.8947345397279051,
      "grad_norm": 5.65625,
      "learning_rate": 6.08638328070267e-06,
      "loss": 0.75003166,
      "memory(GiB)": 147.13,
      "step": 38350,
      "train_speed(iter/s)": 0.201596
    },
    {
      "acc": 0.77957439,
      "epoch": 0.894967847300194,
      "grad_norm": 5.875,
      "learning_rate": 6.084539222326728e-06,
      "loss": 0.78398819,
      "memory(GiB)": 147.13,
      "step": 38360,
      "train_speed(iter/s)": 0.201624
    },
    {
      "acc": 0.78460279,
      "epoch": 0.8952011548724829,
      "grad_norm": 6.96875,
      "learning_rate": 6.082695009133126e-06,
      "loss": 0.77256527,
      "memory(GiB)": 147.13,
      "step": 38370,
      "train_speed(iter/s)": 0.201652
    },
    {
      "acc": 0.78010826,
      "epoch": 0.8954344624447718,
      "grad_norm": 6.8125,
      "learning_rate": 6.080850641385129e-06,
      "loss": 0.79071956,
      "memory(GiB)": 147.13,
      "step": 38380,
      "train_speed(iter/s)": 0.201679
    },
    {
      "acc": 0.76988091,
      "epoch": 0.8956677700170607,
      "grad_norm": 4.6875,
      "learning_rate": 6.079006119346015e-06,
      "loss": 0.83236942,
      "memory(GiB)": 147.13,
      "step": 38390,
      "train_speed(iter/s)": 0.201707
    },
    {
      "acc": 0.78275881,
      "epoch": 0.8959010775893496,
      "grad_norm": 4.90625,
      "learning_rate": 6.0771614432790915e-06,
      "loss": 0.76272345,
      "memory(GiB)": 147.13,
      "step": 38400,
      "train_speed(iter/s)": 0.201736
    },
    {
      "acc": 0.79821758,
      "epoch": 0.8961343851616383,
      "grad_norm": 3.765625,
      "learning_rate": 6.075316613447684e-06,
      "loss": 0.7420167,
      "memory(GiB)": 147.13,
      "step": 38410,
      "train_speed(iter/s)": 0.20176
    },
    {
      "acc": 0.78013487,
      "epoch": 0.8963676927339272,
      "grad_norm": 5.03125,
      "learning_rate": 6.073471630115142e-06,
      "loss": 0.77953615,
      "memory(GiB)": 147.13,
      "step": 38420,
      "train_speed(iter/s)": 0.201788
    },
    {
      "acc": 0.76554446,
      "epoch": 0.8966010003062161,
      "grad_norm": 5.4375,
      "learning_rate": 6.071626493544838e-06,
      "loss": 0.85112772,
      "memory(GiB)": 147.13,
      "step": 38430,
      "train_speed(iter/s)": 0.201815
    },
    {
      "acc": 0.78542604,
      "epoch": 0.896834307878505,
      "grad_norm": 4.34375,
      "learning_rate": 6.0697812040001625e-06,
      "loss": 0.7776813,
      "memory(GiB)": 147.13,
      "step": 38440,
      "train_speed(iter/s)": 0.201842
    },
    {
      "acc": 0.79344821,
      "epoch": 0.8970676154507939,
      "grad_norm": 8.5,
      "learning_rate": 6.067935761744531e-06,
      "loss": 0.74864655,
      "memory(GiB)": 147.13,
      "step": 38450,
      "train_speed(iter/s)": 0.201867
    },
    {
      "acc": 0.77485456,
      "epoch": 0.8973009230230828,
      "grad_norm": 5.4375,
      "learning_rate": 6.066090167041381e-06,
      "loss": 0.78871832,
      "memory(GiB)": 147.13,
      "step": 38460,
      "train_speed(iter/s)": 0.201894
    },
    {
      "acc": 0.7692565,
      "epoch": 0.8975342305953717,
      "grad_norm": 6.0625,
      "learning_rate": 6.0642444201541686e-06,
      "loss": 0.83406401,
      "memory(GiB)": 147.13,
      "step": 38470,
      "train_speed(iter/s)": 0.201921
    },
    {
      "acc": 0.78231134,
      "epoch": 0.8977675381676606,
      "grad_norm": 4.28125,
      "learning_rate": 6.062398521346374e-06,
      "loss": 0.77111473,
      "memory(GiB)": 147.13,
      "step": 38480,
      "train_speed(iter/s)": 0.201948
    },
    {
      "acc": 0.78063517,
      "epoch": 0.8980008457399495,
      "grad_norm": 5.96875,
      "learning_rate": 6.060552470881498e-06,
      "loss": 0.77957792,
      "memory(GiB)": 147.13,
      "step": 38490,
      "train_speed(iter/s)": 0.201974
    },
    {
      "acc": 0.77691154,
      "epoch": 0.8982341533122384,
      "grad_norm": 4.65625,
      "learning_rate": 6.0587062690230654e-06,
      "loss": 0.79885607,
      "memory(GiB)": 147.13,
      "step": 38500,
      "train_speed(iter/s)": 0.202001
    },
    {
      "epoch": 0.8982341533122384,
      "eval_acc": 0.7434691180264593,
      "eval_loss": 0.8091056942939758,
      "eval_runtime": 1269.2357,
      "eval_samples_per_second": 28.356,
      "eval_steps_per_second": 14.179,
      "step": 38500
    },
    {
      "acc": 0.76533337,
      "epoch": 0.8984674608845273,
      "grad_norm": 5.875,
      "learning_rate": 6.056859916034621e-06,
      "loss": 0.85079603,
      "memory(GiB)": 147.13,
      "step": 38510,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.80131989,
      "epoch": 0.8987007684568162,
      "grad_norm": 5.15625,
      "learning_rate": 6.055013412179732e-06,
      "loss": 0.73174257,
      "memory(GiB)": 147.13,
      "step": 38520,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.77752538,
      "epoch": 0.8989340760291051,
      "grad_norm": 5.59375,
      "learning_rate": 6.053166757721984e-06,
      "loss": 0.79698305,
      "memory(GiB)": 147.13,
      "step": 38530,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.78054886,
      "epoch": 0.899167383601394,
      "grad_norm": 5.03125,
      "learning_rate": 6.051319952924987e-06,
      "loss": 0.77370358,
      "memory(GiB)": 147.13,
      "step": 38540,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.79228392,
      "epoch": 0.8994006911736829,
      "grad_norm": 6.125,
      "learning_rate": 6.049472998052371e-06,
      "loss": 0.73808813,
      "memory(GiB)": 147.13,
      "step": 38550,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.78893967,
      "epoch": 0.8996339987459718,
      "grad_norm": 11.9375,
      "learning_rate": 6.047625893367791e-06,
      "loss": 0.74676943,
      "memory(GiB)": 147.13,
      "step": 38560,
      "train_speed(iter/s)": 0.200807
    },
    {
      "acc": 0.77075801,
      "epoch": 0.8998673063182607,
      "grad_norm": 6.25,
      "learning_rate": 6.0457786391349195e-06,
      "loss": 0.83967133,
      "memory(GiB)": 147.13,
      "step": 38570,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.77375088,
      "epoch": 0.9001006138905496,
      "grad_norm": 35.5,
      "learning_rate": 6.0439312356174495e-06,
      "loss": 0.83246117,
      "memory(GiB)": 147.13,
      "step": 38580,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.77322402,
      "epoch": 0.9003339214628385,
      "grad_norm": 4.71875,
      "learning_rate": 6.042083683079099e-06,
      "loss": 0.80312271,
      "memory(GiB)": 147.13,
      "step": 38590,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.78943686,
      "epoch": 0.9005672290351274,
      "grad_norm": 7.0,
      "learning_rate": 6.0402359817836065e-06,
      "loss": 0.76622558,
      "memory(GiB)": 147.13,
      "step": 38600,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.78096523,
      "epoch": 0.9008005366074162,
      "grad_norm": 6.9375,
      "learning_rate": 6.038388131994729e-06,
      "loss": 0.78436923,
      "memory(GiB)": 147.13,
      "step": 38610,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.7831727,
      "epoch": 0.9010338441797051,
      "grad_norm": 6.21875,
      "learning_rate": 6.036540133976247e-06,
      "loss": 0.77339869,
      "memory(GiB)": 147.13,
      "step": 38620,
      "train_speed(iter/s)": 0.200965
    },
    {
      "acc": 0.77260571,
      "epoch": 0.901267151751994,
      "grad_norm": 5.3125,
      "learning_rate": 6.034691987991963e-06,
      "loss": 0.81804924,
      "memory(GiB)": 147.13,
      "step": 38630,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.7917419,
      "epoch": 0.9015004593242829,
      "grad_norm": 3.796875,
      "learning_rate": 6.032843694305698e-06,
      "loss": 0.74714317,
      "memory(GiB)": 147.13,
      "step": 38640,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.78762455,
      "epoch": 0.9017337668965718,
      "grad_norm": 7.6875,
      "learning_rate": 6.0309952531812955e-06,
      "loss": 0.74814758,
      "memory(GiB)": 147.13,
      "step": 38650,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.78616996,
      "epoch": 0.9019670744688607,
      "grad_norm": 6.15625,
      "learning_rate": 6.029146664882619e-06,
      "loss": 0.77387581,
      "memory(GiB)": 147.13,
      "step": 38660,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.7877409,
      "epoch": 0.9022003820411496,
      "grad_norm": 4.65625,
      "learning_rate": 6.027297929673557e-06,
      "loss": 0.75647035,
      "memory(GiB)": 147.13,
      "step": 38670,
      "train_speed(iter/s)": 0.201104
    },
    {
      "acc": 0.77807121,
      "epoch": 0.9024336896134385,
      "grad_norm": 5.65625,
      "learning_rate": 6.025449047818012e-06,
      "loss": 0.80177183,
      "memory(GiB)": 147.13,
      "step": 38680,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.79884205,
      "epoch": 0.9026669971857274,
      "grad_norm": 3.53125,
      "learning_rate": 6.0236000195799164e-06,
      "loss": 0.73010259,
      "memory(GiB)": 147.13,
      "step": 38690,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.78179312,
      "epoch": 0.9029003047580163,
      "grad_norm": 6.6875,
      "learning_rate": 6.0217508452232135e-06,
      "loss": 0.76933064,
      "memory(GiB)": 147.13,
      "step": 38700,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.77845345,
      "epoch": 0.9031336123303052,
      "grad_norm": 8.9375,
      "learning_rate": 6.019901525011873e-06,
      "loss": 0.78353348,
      "memory(GiB)": 147.13,
      "step": 38710,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.78779402,
      "epoch": 0.9033669199025941,
      "grad_norm": 3.765625,
      "learning_rate": 6.018052059209887e-06,
      "loss": 0.76179886,
      "memory(GiB)": 147.13,
      "step": 38720,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.75972013,
      "epoch": 0.903600227474883,
      "grad_norm": 5.5625,
      "learning_rate": 6.016202448081266e-06,
      "loss": 0.8852212,
      "memory(GiB)": 147.13,
      "step": 38730,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.77685909,
      "epoch": 0.9038335350471719,
      "grad_norm": 7.03125,
      "learning_rate": 6.014352691890041e-06,
      "loss": 0.79779468,
      "memory(GiB)": 147.13,
      "step": 38740,
      "train_speed(iter/s)": 0.201293
    },
    {
      "acc": 0.79123516,
      "epoch": 0.9040668426194608,
      "grad_norm": 4.5625,
      "learning_rate": 6.012502790900263e-06,
      "loss": 0.76187844,
      "memory(GiB)": 147.13,
      "step": 38750,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.78490562,
      "epoch": 0.9043001501917497,
      "grad_norm": 5.65625,
      "learning_rate": 6.010652745376006e-06,
      "loss": 0.77625852,
      "memory(GiB)": 147.13,
      "step": 38760,
      "train_speed(iter/s)": 0.201347
    },
    {
      "acc": 0.77033277,
      "epoch": 0.9045334577640386,
      "grad_norm": 5.15625,
      "learning_rate": 6.008802555581364e-06,
      "loss": 0.81603584,
      "memory(GiB)": 147.13,
      "step": 38770,
      "train_speed(iter/s)": 0.201374
    },
    {
      "acc": 0.76615601,
      "epoch": 0.9047667653363275,
      "grad_norm": 6.5,
      "learning_rate": 6.006952221780447e-06,
      "loss": 0.86489754,
      "memory(GiB)": 147.13,
      "step": 38780,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.77902718,
      "epoch": 0.9050000729086164,
      "grad_norm": 6.21875,
      "learning_rate": 6.005101744237396e-06,
      "loss": 0.79199033,
      "memory(GiB)": 147.13,
      "step": 38790,
      "train_speed(iter/s)": 0.20143
    },
    {
      "acc": 0.78239946,
      "epoch": 0.9052333804809052,
      "grad_norm": 6.84375,
      "learning_rate": 6.003251123216362e-06,
      "loss": 0.76015949,
      "memory(GiB)": 147.13,
      "step": 38800,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.79029064,
      "epoch": 0.9054666880531941,
      "grad_norm": 5.625,
      "learning_rate": 6.001400358981522e-06,
      "loss": 0.73135719,
      "memory(GiB)": 147.13,
      "step": 38810,
      "train_speed(iter/s)": 0.201484
    },
    {
      "acc": 0.75774117,
      "epoch": 0.905699995625483,
      "grad_norm": 8.25,
      "learning_rate": 5.999549451797073e-06,
      "loss": 0.88968811,
      "memory(GiB)": 147.13,
      "step": 38820,
      "train_speed(iter/s)": 0.201513
    },
    {
      "acc": 0.78999863,
      "epoch": 0.9059333031977719,
      "grad_norm": 5.9375,
      "learning_rate": 5.997698401927228e-06,
      "loss": 0.77355022,
      "memory(GiB)": 147.13,
      "step": 38830,
      "train_speed(iter/s)": 0.201538
    },
    {
      "acc": 0.76618147,
      "epoch": 0.9061666107700608,
      "grad_norm": 5.78125,
      "learning_rate": 5.995847209636227e-06,
      "loss": 0.85498562,
      "memory(GiB)": 147.13,
      "step": 38840,
      "train_speed(iter/s)": 0.201565
    },
    {
      "acc": 0.7809248,
      "epoch": 0.9063999183423497,
      "grad_norm": 6.0625,
      "learning_rate": 5.993995875188324e-06,
      "loss": 0.76933775,
      "memory(GiB)": 147.13,
      "step": 38850,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.77357092,
      "epoch": 0.9066332259146386,
      "grad_norm": 4.6875,
      "learning_rate": 5.992144398847801e-06,
      "loss": 0.81348381,
      "memory(GiB)": 147.13,
      "step": 38860,
      "train_speed(iter/s)": 0.201618
    },
    {
      "acc": 0.77968731,
      "epoch": 0.9068665334869275,
      "grad_norm": 5.40625,
      "learning_rate": 5.990292780878952e-06,
      "loss": 0.79122777,
      "memory(GiB)": 147.13,
      "step": 38870,
      "train_speed(iter/s)": 0.201646
    },
    {
      "acc": 0.77165852,
      "epoch": 0.9070998410592164,
      "grad_norm": 7.46875,
      "learning_rate": 5.988441021546097e-06,
      "loss": 0.82253704,
      "memory(GiB)": 147.13,
      "step": 38880,
      "train_speed(iter/s)": 0.201674
    },
    {
      "acc": 0.798492,
      "epoch": 0.9073331486315053,
      "grad_norm": 5.15625,
      "learning_rate": 5.986589121113574e-06,
      "loss": 0.71198359,
      "memory(GiB)": 147.13,
      "step": 38890,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.78536911,
      "epoch": 0.9075664562037942,
      "grad_norm": 4.78125,
      "learning_rate": 5.9847370798457395e-06,
      "loss": 0.77795515,
      "memory(GiB)": 147.13,
      "step": 38900,
      "train_speed(iter/s)": 0.201727
    },
    {
      "acc": 0.80710964,
      "epoch": 0.9077997637760831,
      "grad_norm": 5.375,
      "learning_rate": 5.982884898006973e-06,
      "loss": 0.67784872,
      "memory(GiB)": 147.13,
      "step": 38910,
      "train_speed(iter/s)": 0.201755
    },
    {
      "acc": 0.76790614,
      "epoch": 0.908033071348372,
      "grad_norm": 4.5,
      "learning_rate": 5.981032575861674e-06,
      "loss": 0.83228683,
      "memory(GiB)": 147.13,
      "step": 38920,
      "train_speed(iter/s)": 0.201782
    },
    {
      "acc": 0.78043723,
      "epoch": 0.9082663789206609,
      "grad_norm": 8.8125,
      "learning_rate": 5.979180113674258e-06,
      "loss": 0.79833183,
      "memory(GiB)": 147.13,
      "step": 38930,
      "train_speed(iter/s)": 0.201808
    },
    {
      "acc": 0.78106966,
      "epoch": 0.9084996864929498,
      "grad_norm": 4.65625,
      "learning_rate": 5.9773275117091655e-06,
      "loss": 0.76774416,
      "memory(GiB)": 147.13,
      "step": 38940,
      "train_speed(iter/s)": 0.201834
    },
    {
      "acc": 0.77814817,
      "epoch": 0.9087329940652387,
      "grad_norm": 6.0,
      "learning_rate": 5.975474770230856e-06,
      "loss": 0.80169296,
      "memory(GiB)": 147.13,
      "step": 38950,
      "train_speed(iter/s)": 0.201858
    },
    {
      "acc": 0.79056687,
      "epoch": 0.9089663016375276,
      "grad_norm": 4.59375,
      "learning_rate": 5.973621889503804e-06,
      "loss": 0.74093547,
      "memory(GiB)": 147.13,
      "step": 38960,
      "train_speed(iter/s)": 0.201886
    },
    {
      "acc": 0.75867233,
      "epoch": 0.9091996092098165,
      "grad_norm": 10.8125,
      "learning_rate": 5.9717688697925134e-06,
      "loss": 0.87084265,
      "memory(GiB)": 147.13,
      "step": 38970,
      "train_speed(iter/s)": 0.201913
    },
    {
      "acc": 0.77255449,
      "epoch": 0.9094329167821054,
      "grad_norm": 4.78125,
      "learning_rate": 5.969915711361497e-06,
      "loss": 0.82182369,
      "memory(GiB)": 147.13,
      "step": 38980,
      "train_speed(iter/s)": 0.201939
    },
    {
      "acc": 0.77857056,
      "epoch": 0.9096662243543941,
      "grad_norm": 5.15625,
      "learning_rate": 5.968062414475294e-06,
      "loss": 0.80886345,
      "memory(GiB)": 147.13,
      "step": 38990,
      "train_speed(iter/s)": 0.201965
    },
    {
      "acc": 0.77122827,
      "epoch": 0.909899531926683,
      "grad_norm": 8.625,
      "learning_rate": 5.966208979398462e-06,
      "loss": 0.82270679,
      "memory(GiB)": 147.13,
      "step": 39000,
      "train_speed(iter/s)": 0.201991
    },
    {
      "epoch": 0.909899531926683,
      "eval_acc": 0.7434351674492115,
      "eval_loss": 0.8086429834365845,
      "eval_runtime": 1270.6235,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "step": 39000
    },
    {
      "acc": 0.78865824,
      "epoch": 0.910132839498972,
      "grad_norm": 5.09375,
      "learning_rate": 5.964355406395581e-06,
      "loss": 0.73996811,
      "memory(GiB)": 147.13,
      "step": 39010,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.77401848,
      "epoch": 0.9103661470712608,
      "grad_norm": 7.21875,
      "learning_rate": 5.962501695731245e-06,
      "loss": 0.81361609,
      "memory(GiB)": 147.13,
      "step": 39020,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.76471219,
      "epoch": 0.9105994546435497,
      "grad_norm": 6.40625,
      "learning_rate": 5.9606478476700714e-06,
      "loss": 0.85037766,
      "memory(GiB)": 147.13,
      "step": 39030,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.79004974,
      "epoch": 0.9108327622158386,
      "grad_norm": 7.375,
      "learning_rate": 5.958793862476699e-06,
      "loss": 0.75956354,
      "memory(GiB)": 147.13,
      "step": 39040,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.78604002,
      "epoch": 0.9110660697881275,
      "grad_norm": 11.625,
      "learning_rate": 5.956939740415778e-06,
      "loss": 0.77915764,
      "memory(GiB)": 147.13,
      "step": 39050,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.78488879,
      "epoch": 0.9112993773604164,
      "grad_norm": 4.65625,
      "learning_rate": 5.9550854817519875e-06,
      "loss": 0.75412149,
      "memory(GiB)": 147.13,
      "step": 39060,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.79100518,
      "epoch": 0.9115326849327053,
      "grad_norm": 3.953125,
      "learning_rate": 5.953231086750022e-06,
      "loss": 0.7392344,
      "memory(GiB)": 147.13,
      "step": 39070,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.76816645,
      "epoch": 0.9117659925049942,
      "grad_norm": 7.78125,
      "learning_rate": 5.951376555674596e-06,
      "loss": 0.84252634,
      "memory(GiB)": 147.13,
      "step": 39080,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.7637064,
      "epoch": 0.9119993000772831,
      "grad_norm": 6.3125,
      "learning_rate": 5.949521888790444e-06,
      "loss": 0.85107002,
      "memory(GiB)": 147.13,
      "step": 39090,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.77783241,
      "epoch": 0.912232607649572,
      "grad_norm": 5.8125,
      "learning_rate": 5.947667086362318e-06,
      "loss": 0.78836126,
      "memory(GiB)": 147.13,
      "step": 39100,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.79197435,
      "epoch": 0.9124659152218609,
      "grad_norm": 5.75,
      "learning_rate": 5.945812148654991e-06,
      "loss": 0.74622936,
      "memory(GiB)": 147.13,
      "step": 39110,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.78090453,
      "epoch": 0.9126992227941498,
      "grad_norm": 6.84375,
      "learning_rate": 5.943957075933253e-06,
      "loss": 0.77442908,
      "memory(GiB)": 147.13,
      "step": 39120,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.80023155,
      "epoch": 0.9129325303664387,
      "grad_norm": 4.4375,
      "learning_rate": 5.9421018684619165e-06,
      "loss": 0.71343641,
      "memory(GiB)": 147.13,
      "step": 39130,
      "train_speed(iter/s)": 0.200995
    },
    {
      "acc": 0.7824955,
      "epoch": 0.9131658379387276,
      "grad_norm": 4.6875,
      "learning_rate": 5.940246526505814e-06,
      "loss": 0.78219862,
      "memory(GiB)": 147.13,
      "step": 39140,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.78558354,
      "epoch": 0.9133991455110165,
      "grad_norm": 7.34375,
      "learning_rate": 5.9383910503297915e-06,
      "loss": 0.77994785,
      "memory(GiB)": 147.13,
      "step": 39150,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.77539968,
      "epoch": 0.9136324530833054,
      "grad_norm": 5.75,
      "learning_rate": 5.9365354401987195e-06,
      "loss": 0.80972853,
      "memory(GiB)": 147.13,
      "step": 39160,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.78717179,
      "epoch": 0.9138657606555943,
      "grad_norm": 5.78125,
      "learning_rate": 5.934679696377486e-06,
      "loss": 0.77642155,
      "memory(GiB)": 147.13,
      "step": 39170,
      "train_speed(iter/s)": 0.2011
    },
    {
      "acc": 0.76665878,
      "epoch": 0.9140990682278831,
      "grad_norm": 4.21875,
      "learning_rate": 5.932823819130997e-06,
      "loss": 0.84620333,
      "memory(GiB)": 147.13,
      "step": 39180,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.77948418,
      "epoch": 0.914332375800172,
      "grad_norm": 4.78125,
      "learning_rate": 5.930967808724178e-06,
      "loss": 0.79968572,
      "memory(GiB)": 147.13,
      "step": 39190,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.79536867,
      "epoch": 0.9145656833724609,
      "grad_norm": 4.875,
      "learning_rate": 5.929111665421976e-06,
      "loss": 0.7362793,
      "memory(GiB)": 147.13,
      "step": 39200,
      "train_speed(iter/s)": 0.201179
    },
    {
      "acc": 0.76817694,
      "epoch": 0.9147989909447498,
      "grad_norm": 5.4375,
      "learning_rate": 5.927255389489354e-06,
      "loss": 0.81463871,
      "memory(GiB)": 147.13,
      "step": 39210,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78327427,
      "epoch": 0.9150322985170387,
      "grad_norm": 7.375,
      "learning_rate": 5.925398981191293e-06,
      "loss": 0.75333605,
      "memory(GiB)": 147.13,
      "step": 39220,
      "train_speed(iter/s)": 0.201233
    },
    {
      "acc": 0.78863673,
      "epoch": 0.9152656060893276,
      "grad_norm": 4.53125,
      "learning_rate": 5.9235424407927965e-06,
      "loss": 0.749193,
      "memory(GiB)": 147.13,
      "step": 39230,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.76760859,
      "epoch": 0.9154989136616165,
      "grad_norm": 6.9375,
      "learning_rate": 5.9216857685588855e-06,
      "loss": 0.82770119,
      "memory(GiB)": 147.13,
      "step": 39240,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.78653831,
      "epoch": 0.9157322212339054,
      "grad_norm": 5.09375,
      "learning_rate": 5.919828964754599e-06,
      "loss": 0.76294737,
      "memory(GiB)": 147.13,
      "step": 39250,
      "train_speed(iter/s)": 0.201312
    },
    {
      "acc": 0.76094551,
      "epoch": 0.9159655288061943,
      "grad_norm": 6.21875,
      "learning_rate": 5.917972029644995e-06,
      "loss": 0.88056946,
      "memory(GiB)": 147.13,
      "step": 39260,
      "train_speed(iter/s)": 0.201338
    },
    {
      "acc": 0.7901597,
      "epoch": 0.9161988363784832,
      "grad_norm": 6.28125,
      "learning_rate": 5.91611496349515e-06,
      "loss": 0.76451311,
      "memory(GiB)": 147.13,
      "step": 39270,
      "train_speed(iter/s)": 0.201364
    },
    {
      "acc": 0.78220377,
      "epoch": 0.9164321439507721,
      "grad_norm": 4.46875,
      "learning_rate": 5.91425776657016e-06,
      "loss": 0.78790097,
      "memory(GiB)": 147.13,
      "step": 39280,
      "train_speed(iter/s)": 0.201389
    },
    {
      "acc": 0.76608167,
      "epoch": 0.916665451523061,
      "grad_norm": 5.03125,
      "learning_rate": 5.912400439135139e-06,
      "loss": 0.8713728,
      "memory(GiB)": 147.13,
      "step": 39290,
      "train_speed(iter/s)": 0.201416
    },
    {
      "acc": 0.77654085,
      "epoch": 0.9168987590953499,
      "grad_norm": 5.5625,
      "learning_rate": 5.9105429814552204e-06,
      "loss": 0.81738625,
      "memory(GiB)": 147.13,
      "step": 39300,
      "train_speed(iter/s)": 0.201442
    },
    {
      "acc": 0.79101648,
      "epoch": 0.9171320666676388,
      "grad_norm": 7.65625,
      "learning_rate": 5.908685393795557e-06,
      "loss": 0.74483733,
      "memory(GiB)": 147.13,
      "step": 39310,
      "train_speed(iter/s)": 0.201467
    },
    {
      "acc": 0.79285164,
      "epoch": 0.9173653742399277,
      "grad_norm": 11.4375,
      "learning_rate": 5.9068276764213175e-06,
      "loss": 0.73478308,
      "memory(GiB)": 147.13,
      "step": 39320,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.78043051,
      "epoch": 0.9175986818122166,
      "grad_norm": 6.78125,
      "learning_rate": 5.90496982959769e-06,
      "loss": 0.80263834,
      "memory(GiB)": 147.13,
      "step": 39330,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.77075157,
      "epoch": 0.9178319893845055,
      "grad_norm": 6.25,
      "learning_rate": 5.903111853589881e-06,
      "loss": 0.81982746,
      "memory(GiB)": 147.13,
      "step": 39340,
      "train_speed(iter/s)": 0.201548
    },
    {
      "acc": 0.78020802,
      "epoch": 0.9180652969567944,
      "grad_norm": 4.84375,
      "learning_rate": 5.9012537486631185e-06,
      "loss": 0.78267994,
      "memory(GiB)": 147.13,
      "step": 39350,
      "train_speed(iter/s)": 0.201574
    },
    {
      "acc": 0.77198472,
      "epoch": 0.9182986045290833,
      "grad_norm": 5.0625,
      "learning_rate": 5.899395515082644e-06,
      "loss": 0.82205286,
      "memory(GiB)": 147.13,
      "step": 39360,
      "train_speed(iter/s)": 0.2016
    },
    {
      "acc": 0.77900686,
      "epoch": 0.9185319121013722,
      "grad_norm": 4.40625,
      "learning_rate": 5.897537153113724e-06,
      "loss": 0.77138987,
      "memory(GiB)": 147.13,
      "step": 39370,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.7678503,
      "epoch": 0.918765219673661,
      "grad_norm": 5.96875,
      "learning_rate": 5.895678663021634e-06,
      "loss": 0.83069754,
      "memory(GiB)": 147.13,
      "step": 39380,
      "train_speed(iter/s)": 0.201654
    },
    {
      "acc": 0.775527,
      "epoch": 0.9189985272459499,
      "grad_norm": 5.5,
      "learning_rate": 5.893820045071675e-06,
      "loss": 0.80461073,
      "memory(GiB)": 147.13,
      "step": 39390,
      "train_speed(iter/s)": 0.201678
    },
    {
      "acc": 0.77663894,
      "epoch": 0.9192318348182388,
      "grad_norm": 6.15625,
      "learning_rate": 5.891961299529165e-06,
      "loss": 0.79502859,
      "memory(GiB)": 147.13,
      "step": 39400,
      "train_speed(iter/s)": 0.201704
    },
    {
      "acc": 0.76121426,
      "epoch": 0.9194651423905277,
      "grad_norm": 5.75,
      "learning_rate": 5.890102426659438e-06,
      "loss": 0.8578661,
      "memory(GiB)": 147.13,
      "step": 39410,
      "train_speed(iter/s)": 0.20173
    },
    {
      "acc": 0.78363619,
      "epoch": 0.9196984499628166,
      "grad_norm": 9.0625,
      "learning_rate": 5.888243426727847e-06,
      "loss": 0.78227863,
      "memory(GiB)": 147.13,
      "step": 39420,
      "train_speed(iter/s)": 0.201755
    },
    {
      "acc": 0.7768178,
      "epoch": 0.9199317575351055,
      "grad_norm": 6.5625,
      "learning_rate": 5.886384299999767e-06,
      "loss": 0.81008139,
      "memory(GiB)": 147.13,
      "step": 39430,
      "train_speed(iter/s)": 0.20178
    },
    {
      "acc": 0.77163954,
      "epoch": 0.9201650651073944,
      "grad_norm": 6.0,
      "learning_rate": 5.884525046740586e-06,
      "loss": 0.83757038,
      "memory(GiB)": 147.13,
      "step": 39440,
      "train_speed(iter/s)": 0.201805
    },
    {
      "acc": 0.76802855,
      "epoch": 0.9203983726796833,
      "grad_norm": 4.625,
      "learning_rate": 5.882665667215709e-06,
      "loss": 0.84138384,
      "memory(GiB)": 147.13,
      "step": 39450,
      "train_speed(iter/s)": 0.201832
    },
    {
      "acc": 0.77422495,
      "epoch": 0.9206316802519722,
      "grad_norm": 6.0,
      "learning_rate": 5.880806161690567e-06,
      "loss": 0.82321262,
      "memory(GiB)": 147.13,
      "step": 39460,
      "train_speed(iter/s)": 0.201859
    },
    {
      "acc": 0.7739378,
      "epoch": 0.9208649878242611,
      "grad_norm": 6.25,
      "learning_rate": 5.878946530430599e-06,
      "loss": 0.81013727,
      "memory(GiB)": 147.13,
      "step": 39470,
      "train_speed(iter/s)": 0.201885
    },
    {
      "acc": 0.77839308,
      "epoch": 0.92109829539655,
      "grad_norm": 10.25,
      "learning_rate": 5.877086773701271e-06,
      "loss": 0.80809059,
      "memory(GiB)": 147.13,
      "step": 39480,
      "train_speed(iter/s)": 0.201909
    },
    {
      "acc": 0.78637233,
      "epoch": 0.9213316029688389,
      "grad_norm": 8.4375,
      "learning_rate": 5.87522689176806e-06,
      "loss": 0.78325071,
      "memory(GiB)": 147.13,
      "step": 39490,
      "train_speed(iter/s)": 0.201935
    },
    {
      "acc": 0.78502064,
      "epoch": 0.9215649105411278,
      "grad_norm": 19.625,
      "learning_rate": 5.873366884896464e-06,
      "loss": 0.78469729,
      "memory(GiB)": 147.13,
      "step": 39500,
      "train_speed(iter/s)": 0.20196
    },
    {
      "epoch": 0.9215649105411278,
      "eval_acc": 0.7435581582196189,
      "eval_loss": 0.8083301782608032,
      "eval_runtime": 1269.8332,
      "eval_samples_per_second": 28.343,
      "eval_steps_per_second": 14.172,
      "step": 39500
    },
    {
      "acc": 0.758465,
      "epoch": 0.9217982181134167,
      "grad_norm": 4.4375,
      "learning_rate": 5.871506753352e-06,
      "loss": 0.88601971,
      "memory(GiB)": 147.13,
      "step": 39510,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.77248645,
      "epoch": 0.9220315256857056,
      "grad_norm": 5.03125,
      "learning_rate": 5.869646497400199e-06,
      "loss": 0.81327343,
      "memory(GiB)": 147.13,
      "step": 39520,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.78898239,
      "epoch": 0.9222648332579945,
      "grad_norm": 4.875,
      "learning_rate": 5.867786117306614e-06,
      "loss": 0.73193016,
      "memory(GiB)": 147.13,
      "step": 39530,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.77655249,
      "epoch": 0.9224981408302834,
      "grad_norm": 8.25,
      "learning_rate": 5.865925613336814e-06,
      "loss": 0.80876665,
      "memory(GiB)": 147.13,
      "step": 39540,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.79803247,
      "epoch": 0.9227314484025723,
      "grad_norm": 4.625,
      "learning_rate": 5.864064985756382e-06,
      "loss": 0.73142552,
      "memory(GiB)": 147.13,
      "step": 39550,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.78445044,
      "epoch": 0.9229647559748612,
      "grad_norm": 5.0,
      "learning_rate": 5.862204234830925e-06,
      "loss": 0.78575296,
      "memory(GiB)": 147.13,
      "step": 39560,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.77423639,
      "epoch": 0.92319806354715,
      "grad_norm": 7.25,
      "learning_rate": 5.860343360826063e-06,
      "loss": 0.79959049,
      "memory(GiB)": 147.13,
      "step": 39570,
      "train_speed(iter/s)": 0.200826
    },
    {
      "acc": 0.75270905,
      "epoch": 0.9234313711194388,
      "grad_norm": 5.40625,
      "learning_rate": 5.858482364007438e-06,
      "loss": 0.88813477,
      "memory(GiB)": 147.13,
      "step": 39580,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.77825441,
      "epoch": 0.9236646786917277,
      "grad_norm": 11.4375,
      "learning_rate": 5.856621244640704e-06,
      "loss": 0.79800529,
      "memory(GiB)": 147.13,
      "step": 39590,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.78405609,
      "epoch": 0.9238979862640166,
      "grad_norm": 5.34375,
      "learning_rate": 5.8547600029915366e-06,
      "loss": 0.77924662,
      "memory(GiB)": 147.13,
      "step": 39600,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.76475625,
      "epoch": 0.9241312938363055,
      "grad_norm": 5.25,
      "learning_rate": 5.852898639325627e-06,
      "loss": 0.84154186,
      "memory(GiB)": 147.13,
      "step": 39610,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.79988174,
      "epoch": 0.9243646014085944,
      "grad_norm": 5.5625,
      "learning_rate": 5.851037153908684e-06,
      "loss": 0.71289425,
      "memory(GiB)": 147.13,
      "step": 39620,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.77636728,
      "epoch": 0.9245979089808833,
      "grad_norm": 5.5625,
      "learning_rate": 5.849175547006433e-06,
      "loss": 0.8403513,
      "memory(GiB)": 147.13,
      "step": 39630,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.76683617,
      "epoch": 0.9248312165531722,
      "grad_norm": 5.75,
      "learning_rate": 5.8473138188846216e-06,
      "loss": 0.8414938,
      "memory(GiB)": 147.13,
      "step": 39640,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78084307,
      "epoch": 0.9250645241254611,
      "grad_norm": 4.65625,
      "learning_rate": 5.845451969809009e-06,
      "loss": 0.79839873,
      "memory(GiB)": 147.13,
      "step": 39650,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.79516888,
      "epoch": 0.92529783169775,
      "grad_norm": 3.96875,
      "learning_rate": 5.843590000045372e-06,
      "loss": 0.71090298,
      "memory(GiB)": 147.13,
      "step": 39660,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.7792943,
      "epoch": 0.9255311392700389,
      "grad_norm": 5.03125,
      "learning_rate": 5.841727909859508e-06,
      "loss": 0.78876328,
      "memory(GiB)": 147.13,
      "step": 39670,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.76696959,
      "epoch": 0.9257644468423278,
      "grad_norm": 8.625,
      "learning_rate": 5.83986569951723e-06,
      "loss": 0.85549431,
      "memory(GiB)": 147.13,
      "step": 39680,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.78547182,
      "epoch": 0.9259977544146167,
      "grad_norm": 4.59375,
      "learning_rate": 5.838003369284366e-06,
      "loss": 0.7880579,
      "memory(GiB)": 147.13,
      "step": 39690,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.7796061,
      "epoch": 0.9262310619869056,
      "grad_norm": 4.0625,
      "learning_rate": 5.836140919426765e-06,
      "loss": 0.78260422,
      "memory(GiB)": 147.13,
      "step": 39700,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.77718563,
      "epoch": 0.9264643695591945,
      "grad_norm": 4.15625,
      "learning_rate": 5.834278350210292e-06,
      "loss": 0.81985626,
      "memory(GiB)": 147.13,
      "step": 39710,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.78238964,
      "epoch": 0.9266976771314834,
      "grad_norm": 6.65625,
      "learning_rate": 5.832415661900826e-06,
      "loss": 0.7647913,
      "memory(GiB)": 147.13,
      "step": 39720,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.78833084,
      "epoch": 0.9269309847037723,
      "grad_norm": 4.90625,
      "learning_rate": 5.830552854764265e-06,
      "loss": 0.73814125,
      "memory(GiB)": 147.13,
      "step": 39730,
      "train_speed(iter/s)": 0.201244
    },
    {
      "acc": 0.77150917,
      "epoch": 0.9271642922760612,
      "grad_norm": 5.3125,
      "learning_rate": 5.828689929066526e-06,
      "loss": 0.81474161,
      "memory(GiB)": 147.13,
      "step": 39740,
      "train_speed(iter/s)": 0.20127
    },
    {
      "acc": 0.78371716,
      "epoch": 0.9273975998483501,
      "grad_norm": 5.34375,
      "learning_rate": 5.826826885073541e-06,
      "loss": 0.77574019,
      "memory(GiB)": 147.13,
      "step": 39750,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.79163008,
      "epoch": 0.9276309074206389,
      "grad_norm": 6.09375,
      "learning_rate": 5.824963723051258e-06,
      "loss": 0.75398407,
      "memory(GiB)": 147.13,
      "step": 39760,
      "train_speed(iter/s)": 0.201322
    },
    {
      "acc": 0.77016096,
      "epoch": 0.9278642149929278,
      "grad_norm": 5.71875,
      "learning_rate": 5.823100443265643e-06,
      "loss": 0.84416866,
      "memory(GiB)": 147.13,
      "step": 39770,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.7718441,
      "epoch": 0.9280975225652167,
      "grad_norm": 6.71875,
      "learning_rate": 5.821237045982679e-06,
      "loss": 0.82021675,
      "memory(GiB)": 147.13,
      "step": 39780,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.76817198,
      "epoch": 0.9283308301375056,
      "grad_norm": 7.84375,
      "learning_rate": 5.819373531468364e-06,
      "loss": 0.82225189,
      "memory(GiB)": 147.13,
      "step": 39790,
      "train_speed(iter/s)": 0.201401
    },
    {
      "acc": 0.77029276,
      "epoch": 0.9285641377097945,
      "grad_norm": 10.125,
      "learning_rate": 5.817509899988717e-06,
      "loss": 0.81980934,
      "memory(GiB)": 147.13,
      "step": 39800,
      "train_speed(iter/s)": 0.201428
    },
    {
      "acc": 0.77408419,
      "epoch": 0.9287974452820834,
      "grad_norm": 5.0625,
      "learning_rate": 5.8156461518097695e-06,
      "loss": 0.81796417,
      "memory(GiB)": 147.13,
      "step": 39810,
      "train_speed(iter/s)": 0.201454
    },
    {
      "acc": 0.78498955,
      "epoch": 0.9290307528543723,
      "grad_norm": 20.0,
      "learning_rate": 5.813782287197569e-06,
      "loss": 0.78012381,
      "memory(GiB)": 147.13,
      "step": 39820,
      "train_speed(iter/s)": 0.201481
    },
    {
      "acc": 0.76299672,
      "epoch": 0.9292640604266612,
      "grad_norm": 8.3125,
      "learning_rate": 5.8119183064181864e-06,
      "loss": 0.84655218,
      "memory(GiB)": 147.13,
      "step": 39830,
      "train_speed(iter/s)": 0.201508
    },
    {
      "acc": 0.77199259,
      "epoch": 0.9294973679989501,
      "grad_norm": 8.125,
      "learning_rate": 5.810054209737699e-06,
      "loss": 0.807763,
      "memory(GiB)": 147.13,
      "step": 39840,
      "train_speed(iter/s)": 0.201533
    },
    {
      "acc": 0.77625284,
      "epoch": 0.929730675571239,
      "grad_norm": 4.28125,
      "learning_rate": 5.8081899974222076e-06,
      "loss": 0.8012434,
      "memory(GiB)": 147.13,
      "step": 39850,
      "train_speed(iter/s)": 0.20156
    },
    {
      "acc": 0.79133472,
      "epoch": 0.9299639831435279,
      "grad_norm": 4.84375,
      "learning_rate": 5.80632566973783e-06,
      "loss": 0.75545387,
      "memory(GiB)": 147.13,
      "step": 39860,
      "train_speed(iter/s)": 0.201585
    },
    {
      "acc": 0.78476238,
      "epoch": 0.9301972907158168,
      "grad_norm": 6.0,
      "learning_rate": 5.804461226950697e-06,
      "loss": 0.75581503,
      "memory(GiB)": 147.13,
      "step": 39870,
      "train_speed(iter/s)": 0.201611
    },
    {
      "acc": 0.75795188,
      "epoch": 0.9304305982881057,
      "grad_norm": 6.6875,
      "learning_rate": 5.80259666932696e-06,
      "loss": 0.87952824,
      "memory(GiB)": 147.13,
      "step": 39880,
      "train_speed(iter/s)": 0.201638
    },
    {
      "acc": 0.75466013,
      "epoch": 0.9306639058603946,
      "grad_norm": 5.78125,
      "learning_rate": 5.800731997132779e-06,
      "loss": 0.8734972,
      "memory(GiB)": 147.13,
      "step": 39890,
      "train_speed(iter/s)": 0.201665
    },
    {
      "acc": 0.78315744,
      "epoch": 0.9308972134326835,
      "grad_norm": 6.375,
      "learning_rate": 5.7988672106343395e-06,
      "loss": 0.8057559,
      "memory(GiB)": 147.13,
      "step": 39900,
      "train_speed(iter/s)": 0.20169
    },
    {
      "acc": 0.76869955,
      "epoch": 0.9311305210049724,
      "grad_norm": 6.21875,
      "learning_rate": 5.797002310097836e-06,
      "loss": 0.82583561,
      "memory(GiB)": 147.13,
      "step": 39910,
      "train_speed(iter/s)": 0.201718
    },
    {
      "acc": 0.78581076,
      "epoch": 0.9313638285772613,
      "grad_norm": 5.0625,
      "learning_rate": 5.795137295789486e-06,
      "loss": 0.78243923,
      "memory(GiB)": 147.13,
      "step": 39920,
      "train_speed(iter/s)": 0.201744
    },
    {
      "acc": 0.77420025,
      "epoch": 0.9315971361495502,
      "grad_norm": 5.4375,
      "learning_rate": 5.7932721679755164e-06,
      "loss": 0.81065769,
      "memory(GiB)": 147.13,
      "step": 39930,
      "train_speed(iter/s)": 0.20177
    },
    {
      "acc": 0.76268339,
      "epoch": 0.9318304437218391,
      "grad_norm": 4.09375,
      "learning_rate": 5.791406926922176e-06,
      "loss": 0.88640518,
      "memory(GiB)": 147.13,
      "step": 39940,
      "train_speed(iter/s)": 0.201797
    },
    {
      "acc": 0.78374519,
      "epoch": 0.9320637512941279,
      "grad_norm": 5.59375,
      "learning_rate": 5.789541572895727e-06,
      "loss": 0.78053789,
      "memory(GiB)": 147.13,
      "step": 39950,
      "train_speed(iter/s)": 0.201824
    },
    {
      "acc": 0.79211001,
      "epoch": 0.9322970588664168,
      "grad_norm": 4.96875,
      "learning_rate": 5.787676106162449e-06,
      "loss": 0.74461236,
      "memory(GiB)": 147.13,
      "step": 39960,
      "train_speed(iter/s)": 0.20185
    },
    {
      "acc": 0.80399532,
      "epoch": 0.9325303664387057,
      "grad_norm": 5.0625,
      "learning_rate": 5.785810526988633e-06,
      "loss": 0.70544991,
      "memory(GiB)": 147.13,
      "step": 39970,
      "train_speed(iter/s)": 0.201875
    },
    {
      "acc": 0.76267509,
      "epoch": 0.9327636740109946,
      "grad_norm": 4.125,
      "learning_rate": 5.783944835640594e-06,
      "loss": 0.82315941,
      "memory(GiB)": 147.13,
      "step": 39980,
      "train_speed(iter/s)": 0.201902
    },
    {
      "acc": 0.76198874,
      "epoch": 0.9329969815832835,
      "grad_norm": 4.4375,
      "learning_rate": 5.7820790323846566e-06,
      "loss": 0.88518867,
      "memory(GiB)": 147.13,
      "step": 39990,
      "train_speed(iter/s)": 0.201927
    },
    {
      "acc": 0.76849947,
      "epoch": 0.9332302891555724,
      "grad_norm": 7.75,
      "learning_rate": 5.780213117487167e-06,
      "loss": 0.81474934,
      "memory(GiB)": 147.13,
      "step": 40000,
      "train_speed(iter/s)": 0.201955
    },
    {
      "epoch": 0.9332302891555724,
      "eval_acc": 0.7435186025942405,
      "eval_loss": 0.8081688284873962,
      "eval_runtime": 1269.5667,
      "eval_samples_per_second": 28.349,
      "eval_steps_per_second": 14.175,
      "step": 40000
    },
    {
      "acc": 0.77274613,
      "epoch": 0.9334635967278613,
      "grad_norm": 4.8125,
      "learning_rate": 5.778347091214479e-06,
      "loss": 0.81792908,
      "memory(GiB)": 147.13,
      "step": 40010,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.77955446,
      "epoch": 0.9336969043001502,
      "grad_norm": 4.9375,
      "learning_rate": 5.77648095383297e-06,
      "loss": 0.7724124,
      "memory(GiB)": 147.13,
      "step": 40020,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.76370978,
      "epoch": 0.9339302118724391,
      "grad_norm": 11.625,
      "learning_rate": 5.774614705609032e-06,
      "loss": 0.85831079,
      "memory(GiB)": 147.13,
      "step": 40030,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.77389021,
      "epoch": 0.934163519444728,
      "grad_norm": 4.5,
      "learning_rate": 5.7727483468090686e-06,
      "loss": 0.80834856,
      "memory(GiB)": 147.13,
      "step": 40040,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.77940512,
      "epoch": 0.9343968270170169,
      "grad_norm": 4.375,
      "learning_rate": 5.770881877699502e-06,
      "loss": 0.78713818,
      "memory(GiB)": 147.13,
      "step": 40050,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.77909374,
      "epoch": 0.9346301345893058,
      "grad_norm": 7.21875,
      "learning_rate": 5.769015298546774e-06,
      "loss": 0.79142208,
      "memory(GiB)": 147.13,
      "step": 40060,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.77855005,
      "epoch": 0.9348634421615947,
      "grad_norm": 4.84375,
      "learning_rate": 5.7671486096173336e-06,
      "loss": 0.78886166,
      "memory(GiB)": 147.13,
      "step": 40070,
      "train_speed(iter/s)": 0.200827
    },
    {
      "acc": 0.77905307,
      "epoch": 0.9350967497338836,
      "grad_norm": 5.90625,
      "learning_rate": 5.765281811177652e-06,
      "loss": 0.81017637,
      "memory(GiB)": 147.13,
      "step": 40080,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.79088497,
      "epoch": 0.9353300573061725,
      "grad_norm": 5.5625,
      "learning_rate": 5.763414903494216e-06,
      "loss": 0.75551167,
      "memory(GiB)": 147.13,
      "step": 40090,
      "train_speed(iter/s)": 0.200879
    },
    {
      "acc": 0.75535116,
      "epoch": 0.9355633648784614,
      "grad_norm": 5.75,
      "learning_rate": 5.761547886833523e-06,
      "loss": 0.89576149,
      "memory(GiB)": 147.13,
      "step": 40100,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.80530338,
      "epoch": 0.9357966724507503,
      "grad_norm": 6.34375,
      "learning_rate": 5.759680761462091e-06,
      "loss": 0.69478006,
      "memory(GiB)": 147.13,
      "step": 40110,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.77864857,
      "epoch": 0.9360299800230392,
      "grad_norm": 4.875,
      "learning_rate": 5.757813527646449e-06,
      "loss": 0.8089098,
      "memory(GiB)": 147.13,
      "step": 40120,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.79146218,
      "epoch": 0.9362632875953281,
      "grad_norm": 4.59375,
      "learning_rate": 5.755946185653148e-06,
      "loss": 0.74217372,
      "memory(GiB)": 147.13,
      "step": 40130,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.78336248,
      "epoch": 0.936496595167617,
      "grad_norm": 15.5625,
      "learning_rate": 5.7540787357487485e-06,
      "loss": 0.78665681,
      "memory(GiB)": 147.13,
      "step": 40140,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78126898,
      "epoch": 0.9367299027399058,
      "grad_norm": 5.59375,
      "learning_rate": 5.752211178199828e-06,
      "loss": 0.77836123,
      "memory(GiB)": 147.13,
      "step": 40150,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.78209891,
      "epoch": 0.9369632103121946,
      "grad_norm": 4.75,
      "learning_rate": 5.7503435132729805e-06,
      "loss": 0.76724977,
      "memory(GiB)": 147.13,
      "step": 40160,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.77962379,
      "epoch": 0.9371965178844835,
      "grad_norm": 6.0625,
      "learning_rate": 5.7484757412348146e-06,
      "loss": 0.78311558,
      "memory(GiB)": 147.13,
      "step": 40170,
      "train_speed(iter/s)": 0.201084
    },
    {
      "acc": 0.78048811,
      "epoch": 0.9374298254567724,
      "grad_norm": 7.125,
      "learning_rate": 5.746607862351955e-06,
      "loss": 0.78884706,
      "memory(GiB)": 147.13,
      "step": 40180,
      "train_speed(iter/s)": 0.201109
    },
    {
      "acc": 0.76181178,
      "epoch": 0.9376631330290613,
      "grad_norm": 5.09375,
      "learning_rate": 5.744739876891038e-06,
      "loss": 0.85814571,
      "memory(GiB)": 147.13,
      "step": 40190,
      "train_speed(iter/s)": 0.201136
    },
    {
      "acc": 0.77918482,
      "epoch": 0.9378964406013502,
      "grad_norm": 4.6875,
      "learning_rate": 5.742871785118721e-06,
      "loss": 0.76165657,
      "memory(GiB)": 147.13,
      "step": 40200,
      "train_speed(iter/s)": 0.201162
    },
    {
      "acc": 0.8000536,
      "epoch": 0.9381297481736391,
      "grad_norm": 13.25,
      "learning_rate": 5.741003587301673e-06,
      "loss": 0.71440544,
      "memory(GiB)": 147.13,
      "step": 40210,
      "train_speed(iter/s)": 0.201187
    },
    {
      "acc": 0.77322521,
      "epoch": 0.938363055745928,
      "grad_norm": 4.90625,
      "learning_rate": 5.739135283706576e-06,
      "loss": 0.82055111,
      "memory(GiB)": 147.13,
      "step": 40220,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.77559209,
      "epoch": 0.9385963633182169,
      "grad_norm": 5.875,
      "learning_rate": 5.737266874600134e-06,
      "loss": 0.79832792,
      "memory(GiB)": 147.13,
      "step": 40230,
      "train_speed(iter/s)": 0.20124
    },
    {
      "acc": 0.76249466,
      "epoch": 0.9388296708905058,
      "grad_norm": 4.59375,
      "learning_rate": 5.735398360249059e-06,
      "loss": 0.84215164,
      "memory(GiB)": 147.13,
      "step": 40240,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.79150157,
      "epoch": 0.9390629784627947,
      "grad_norm": 6.25,
      "learning_rate": 5.733529740920083e-06,
      "loss": 0.73628078,
      "memory(GiB)": 147.13,
      "step": 40250,
      "train_speed(iter/s)": 0.201291
    },
    {
      "acc": 0.77681289,
      "epoch": 0.9392962860350836,
      "grad_norm": 6.0625,
      "learning_rate": 5.731661016879948e-06,
      "loss": 0.79612532,
      "memory(GiB)": 147.13,
      "step": 40260,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.7825942,
      "epoch": 0.9395295936073725,
      "grad_norm": 4.9375,
      "learning_rate": 5.729792188395415e-06,
      "loss": 0.7699985,
      "memory(GiB)": 147.13,
      "step": 40270,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.76840849,
      "epoch": 0.9397629011796614,
      "grad_norm": 5.34375,
      "learning_rate": 5.7279232557332595e-06,
      "loss": 0.82999878,
      "memory(GiB)": 147.13,
      "step": 40280,
      "train_speed(iter/s)": 0.201367
    },
    {
      "acc": 0.77449684,
      "epoch": 0.9399962087519503,
      "grad_norm": 5.34375,
      "learning_rate": 5.726054219160273e-06,
      "loss": 0.81352596,
      "memory(GiB)": 147.13,
      "step": 40290,
      "train_speed(iter/s)": 0.201392
    },
    {
      "acc": 0.77998219,
      "epoch": 0.9402295163242392,
      "grad_norm": 4.65625,
      "learning_rate": 5.7241850789432555e-06,
      "loss": 0.79409661,
      "memory(GiB)": 147.13,
      "step": 40300,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.78303413,
      "epoch": 0.9404628238965281,
      "grad_norm": 5.8125,
      "learning_rate": 5.722315835349029e-06,
      "loss": 0.76099334,
      "memory(GiB)": 147.13,
      "step": 40310,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.78695984,
      "epoch": 0.940696131468817,
      "grad_norm": 5.59375,
      "learning_rate": 5.7204464886444265e-06,
      "loss": 0.76036892,
      "memory(GiB)": 147.13,
      "step": 40320,
      "train_speed(iter/s)": 0.201469
    },
    {
      "acc": 0.78353043,
      "epoch": 0.9409294390411059,
      "grad_norm": 5.25,
      "learning_rate": 5.718577039096297e-06,
      "loss": 0.78870249,
      "memory(GiB)": 147.13,
      "step": 40330,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.78347793,
      "epoch": 0.9411627466133947,
      "grad_norm": 5.25,
      "learning_rate": 5.7167074869715045e-06,
      "loss": 0.76787477,
      "memory(GiB)": 147.13,
      "step": 40340,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.79285727,
      "epoch": 0.9413960541856836,
      "grad_norm": 7.1875,
      "learning_rate": 5.714837832536926e-06,
      "loss": 0.73996181,
      "memory(GiB)": 147.13,
      "step": 40350,
      "train_speed(iter/s)": 0.201547
    },
    {
      "acc": 0.77512045,
      "epoch": 0.9416293617579725,
      "grad_norm": 5.78125,
      "learning_rate": 5.712968076059454e-06,
      "loss": 0.81619711,
      "memory(GiB)": 147.13,
      "step": 40360,
      "train_speed(iter/s)": 0.201572
    },
    {
      "acc": 0.78800678,
      "epoch": 0.9418626693302614,
      "grad_norm": 5.375,
      "learning_rate": 5.711098217805997e-06,
      "loss": 0.74943008,
      "memory(GiB)": 147.13,
      "step": 40370,
      "train_speed(iter/s)": 0.201597
    },
    {
      "acc": 0.77038002,
      "epoch": 0.9420959769025503,
      "grad_norm": 5.125,
      "learning_rate": 5.709228258043476e-06,
      "loss": 0.82832499,
      "memory(GiB)": 147.13,
      "step": 40380,
      "train_speed(iter/s)": 0.201622
    },
    {
      "acc": 0.7822566,
      "epoch": 0.9423292844748392,
      "grad_norm": 4.625,
      "learning_rate": 5.707358197038827e-06,
      "loss": 0.78541126,
      "memory(GiB)": 147.13,
      "step": 40390,
      "train_speed(iter/s)": 0.201646
    },
    {
      "acc": 0.77454891,
      "epoch": 0.9425625920471281,
      "grad_norm": 5.71875,
      "learning_rate": 5.7054880350590015e-06,
      "loss": 0.83365231,
      "memory(GiB)": 147.13,
      "step": 40400,
      "train_speed(iter/s)": 0.201673
    },
    {
      "acc": 0.77726865,
      "epoch": 0.942795899619417,
      "grad_norm": 4.96875,
      "learning_rate": 5.703617772370963e-06,
      "loss": 0.79805484,
      "memory(GiB)": 147.13,
      "step": 40410,
      "train_speed(iter/s)": 0.201698
    },
    {
      "acc": 0.7605464,
      "epoch": 0.9430292071917059,
      "grad_norm": 5.3125,
      "learning_rate": 5.701747409241691e-06,
      "loss": 0.85971928,
      "memory(GiB)": 147.13,
      "step": 40420,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.79386868,
      "epoch": 0.9432625147639948,
      "grad_norm": 6.03125,
      "learning_rate": 5.699876945938182e-06,
      "loss": 0.74754467,
      "memory(GiB)": 147.13,
      "step": 40430,
      "train_speed(iter/s)": 0.201748
    },
    {
      "acc": 0.7801425,
      "epoch": 0.9434958223362837,
      "grad_norm": 4.90625,
      "learning_rate": 5.698006382727441e-06,
      "loss": 0.7978281,
      "memory(GiB)": 147.13,
      "step": 40440,
      "train_speed(iter/s)": 0.201775
    },
    {
      "acc": 0.74835367,
      "epoch": 0.9437291299085726,
      "grad_norm": 8.5,
      "learning_rate": 5.696135719876492e-06,
      "loss": 0.92766171,
      "memory(GiB)": 147.13,
      "step": 40450,
      "train_speed(iter/s)": 0.201803
    },
    {
      "acc": 0.78347101,
      "epoch": 0.9439624374808615,
      "grad_norm": 5.0,
      "learning_rate": 5.694264957652373e-06,
      "loss": 0.75946207,
      "memory(GiB)": 147.13,
      "step": 40460,
      "train_speed(iter/s)": 0.201827
    },
    {
      "acc": 0.78046608,
      "epoch": 0.9441957450531504,
      "grad_norm": 5.59375,
      "learning_rate": 5.692394096322131e-06,
      "loss": 0.81162415,
      "memory(GiB)": 147.13,
      "step": 40470,
      "train_speed(iter/s)": 0.201854
    },
    {
      "acc": 0.77506027,
      "epoch": 0.9444290526254393,
      "grad_norm": 5.09375,
      "learning_rate": 5.690523136152834e-06,
      "loss": 0.83319483,
      "memory(GiB)": 147.13,
      "step": 40480,
      "train_speed(iter/s)": 0.201879
    },
    {
      "acc": 0.77044592,
      "epoch": 0.9446623601977282,
      "grad_norm": 5.75,
      "learning_rate": 5.688652077411558e-06,
      "loss": 0.82162943,
      "memory(GiB)": 147.13,
      "step": 40490,
      "train_speed(iter/s)": 0.201904
    },
    {
      "acc": 0.78238449,
      "epoch": 0.9448956677700171,
      "grad_norm": 6.28125,
      "learning_rate": 5.6867809203654004e-06,
      "loss": 0.78513522,
      "memory(GiB)": 147.13,
      "step": 40500,
      "train_speed(iter/s)": 0.201929
    },
    {
      "epoch": 0.9448956677700171,
      "eval_acc": 0.7436441556723647,
      "eval_loss": 0.8078497052192688,
      "eval_runtime": 1270.4605,
      "eval_samples_per_second": 28.329,
      "eval_steps_per_second": 14.165,
      "step": 40500
    },
    {
      "acc": 0.77993817,
      "epoch": 0.945128975342306,
      "grad_norm": 6.125,
      "learning_rate": 5.684909665281465e-06,
      "loss": 0.7786458,
      "memory(GiB)": 147.13,
      "step": 40510,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.7921401,
      "epoch": 0.9453622829145949,
      "grad_norm": 5.0,
      "learning_rate": 5.683038312426873e-06,
      "loss": 0.73767538,
      "memory(GiB)": 147.13,
      "step": 40520,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.77997236,
      "epoch": 0.9455955904868837,
      "grad_norm": 4.9375,
      "learning_rate": 5.681166862068761e-06,
      "loss": 0.80369387,
      "memory(GiB)": 147.13,
      "step": 40530,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.75641575,
      "epoch": 0.9458288980591726,
      "grad_norm": 4.34375,
      "learning_rate": 5.679295314474278e-06,
      "loss": 0.89748631,
      "memory(GiB)": 147.13,
      "step": 40540,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.77325706,
      "epoch": 0.9460622056314615,
      "grad_norm": 4.125,
      "learning_rate": 5.677423669910584e-06,
      "loss": 0.82863417,
      "memory(GiB)": 147.13,
      "step": 40550,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.7895503,
      "epoch": 0.9462955132037504,
      "grad_norm": 5.53125,
      "learning_rate": 5.67555192864486e-06,
      "loss": 0.74407339,
      "memory(GiB)": 147.13,
      "step": 40560,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.77496853,
      "epoch": 0.9465288207760393,
      "grad_norm": 11.0,
      "learning_rate": 5.673680090944294e-06,
      "loss": 0.81249332,
      "memory(GiB)": 147.13,
      "step": 40570,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.79530725,
      "epoch": 0.9467621283483282,
      "grad_norm": 9.1875,
      "learning_rate": 5.671808157076091e-06,
      "loss": 0.72718515,
      "memory(GiB)": 147.13,
      "step": 40580,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.76706448,
      "epoch": 0.9469954359206171,
      "grad_norm": 4.0,
      "learning_rate": 5.669936127307468e-06,
      "loss": 0.85577278,
      "memory(GiB)": 147.13,
      "step": 40590,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.77201781,
      "epoch": 0.947228743492906,
      "grad_norm": 6.84375,
      "learning_rate": 5.668064001905658e-06,
      "loss": 0.82298203,
      "memory(GiB)": 147.13,
      "step": 40600,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.76372414,
      "epoch": 0.9474620510651949,
      "grad_norm": 5.40625,
      "learning_rate": 5.666191781137905e-06,
      "loss": 0.85342846,
      "memory(GiB)": 147.13,
      "step": 40610,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77155447,
      "epoch": 0.9476953586374838,
      "grad_norm": 6.09375,
      "learning_rate": 5.66431946527147e-06,
      "loss": 0.80941982,
      "memory(GiB)": 147.13,
      "step": 40620,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.78885984,
      "epoch": 0.9479286662097727,
      "grad_norm": 4.8125,
      "learning_rate": 5.662447054573624e-06,
      "loss": 0.76518712,
      "memory(GiB)": 147.13,
      "step": 40630,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.7786191,
      "epoch": 0.9481619737820616,
      "grad_norm": 4.3125,
      "learning_rate": 5.660574549311653e-06,
      "loss": 0.80177717,
      "memory(GiB)": 147.13,
      "step": 40640,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.77824173,
      "epoch": 0.9483952813543505,
      "grad_norm": 5.71875,
      "learning_rate": 5.658701949752856e-06,
      "loss": 0.79966569,
      "memory(GiB)": 147.13,
      "step": 40650,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.77022877,
      "epoch": 0.9486285889266394,
      "grad_norm": 5.09375,
      "learning_rate": 5.656829256164549e-06,
      "loss": 0.85357208,
      "memory(GiB)": 147.13,
      "step": 40660,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.78125162,
      "epoch": 0.9488618964989283,
      "grad_norm": 5.28125,
      "learning_rate": 5.6549564688140555e-06,
      "loss": 0.76331964,
      "memory(GiB)": 147.13,
      "step": 40670,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.79808364,
      "epoch": 0.9490952040712172,
      "grad_norm": 4.8125,
      "learning_rate": 5.653083587968716e-06,
      "loss": 0.71971464,
      "memory(GiB)": 147.13,
      "step": 40680,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.77893338,
      "epoch": 0.9493285116435061,
      "grad_norm": 6.21875,
      "learning_rate": 5.651210613895885e-06,
      "loss": 0.78710918,
      "memory(GiB)": 147.13,
      "step": 40690,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.77154942,
      "epoch": 0.949561819215795,
      "grad_norm": 7.03125,
      "learning_rate": 5.649337546862927e-06,
      "loss": 0.81971178,
      "memory(GiB)": 147.13,
      "step": 40700,
      "train_speed(iter/s)": 0.201151
    },
    {
      "acc": 0.7884603,
      "epoch": 0.9497951267880839,
      "grad_norm": 4.5,
      "learning_rate": 5.647464387137224e-06,
      "loss": 0.75167475,
      "memory(GiB)": 147.13,
      "step": 40710,
      "train_speed(iter/s)": 0.201176
    },
    {
      "acc": 0.77500906,
      "epoch": 0.9500284343603727,
      "grad_norm": 6.5,
      "learning_rate": 5.645591134986166e-06,
      "loss": 0.81845722,
      "memory(GiB)": 147.13,
      "step": 40720,
      "train_speed(iter/s)": 0.201201
    },
    {
      "acc": 0.78605471,
      "epoch": 0.9502617419326616,
      "grad_norm": 5.09375,
      "learning_rate": 5.643717790677162e-06,
      "loss": 0.7811121,
      "memory(GiB)": 147.13,
      "step": 40730,
      "train_speed(iter/s)": 0.201226
    },
    {
      "acc": 0.77256351,
      "epoch": 0.9504950495049505,
      "grad_norm": 5.46875,
      "learning_rate": 5.641844354477631e-06,
      "loss": 0.84232235,
      "memory(GiB)": 147.13,
      "step": 40740,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.76221151,
      "epoch": 0.9507283570772393,
      "grad_norm": 6.34375,
      "learning_rate": 5.639970826655005e-06,
      "loss": 0.85408916,
      "memory(GiB)": 147.13,
      "step": 40750,
      "train_speed(iter/s)": 0.201277
    },
    {
      "acc": 0.78343468,
      "epoch": 0.9509616646495282,
      "grad_norm": 6.96875,
      "learning_rate": 5.63809720747673e-06,
      "loss": 0.76852407,
      "memory(GiB)": 147.13,
      "step": 40760,
      "train_speed(iter/s)": 0.201302
    },
    {
      "acc": 0.78977461,
      "epoch": 0.9511949722218171,
      "grad_norm": 7.5,
      "learning_rate": 5.636223497210261e-06,
      "loss": 0.74346333,
      "memory(GiB)": 147.13,
      "step": 40770,
      "train_speed(iter/s)": 0.201328
    },
    {
      "acc": 0.7746685,
      "epoch": 0.951428279794106,
      "grad_norm": 7.0625,
      "learning_rate": 5.634349696123075e-06,
      "loss": 0.81890717,
      "memory(GiB)": 147.13,
      "step": 40780,
      "train_speed(iter/s)": 0.201354
    },
    {
      "acc": 0.77995763,
      "epoch": 0.951661587366395,
      "grad_norm": 15.1875,
      "learning_rate": 5.6324758044826535e-06,
      "loss": 0.79575968,
      "memory(GiB)": 147.13,
      "step": 40790,
      "train_speed(iter/s)": 0.201378
    },
    {
      "acc": 0.78342981,
      "epoch": 0.9518948949386838,
      "grad_norm": 4.0,
      "learning_rate": 5.6306018225564955e-06,
      "loss": 0.79688606,
      "memory(GiB)": 147.13,
      "step": 40800,
      "train_speed(iter/s)": 0.201404
    },
    {
      "acc": 0.77644749,
      "epoch": 0.9521282025109727,
      "grad_norm": 7.59375,
      "learning_rate": 5.6287277506121084e-06,
      "loss": 0.80725603,
      "memory(GiB)": 147.13,
      "step": 40810,
      "train_speed(iter/s)": 0.20143
    },
    {
      "acc": 0.78804812,
      "epoch": 0.9523615100832616,
      "grad_norm": 5.0,
      "learning_rate": 5.626853588917021e-06,
      "loss": 0.7547946,
      "memory(GiB)": 147.13,
      "step": 40820,
      "train_speed(iter/s)": 0.201454
    },
    {
      "acc": 0.76683331,
      "epoch": 0.9525948176555505,
      "grad_norm": 4.78125,
      "learning_rate": 5.624979337738763e-06,
      "loss": 0.8316308,
      "memory(GiB)": 147.13,
      "step": 40830,
      "train_speed(iter/s)": 0.20148
    },
    {
      "acc": 0.77537379,
      "epoch": 0.9528281252278394,
      "grad_norm": 4.9375,
      "learning_rate": 5.623104997344886e-06,
      "loss": 0.80050964,
      "memory(GiB)": 147.13,
      "step": 40840,
      "train_speed(iter/s)": 0.201507
    },
    {
      "acc": 0.77048178,
      "epoch": 0.9530614328001283,
      "grad_norm": 7.84375,
      "learning_rate": 5.621230568002952e-06,
      "loss": 0.84840851,
      "memory(GiB)": 147.13,
      "step": 40850,
      "train_speed(iter/s)": 0.201531
    },
    {
      "acc": 0.77276115,
      "epoch": 0.9532947403724172,
      "grad_norm": 6.125,
      "learning_rate": 5.619356049980536e-06,
      "loss": 0.84306707,
      "memory(GiB)": 147.13,
      "step": 40860,
      "train_speed(iter/s)": 0.201556
    },
    {
      "acc": 0.75870538,
      "epoch": 0.9535280479447061,
      "grad_norm": 5.21875,
      "learning_rate": 5.617481443545223e-06,
      "loss": 0.88285389,
      "memory(GiB)": 147.13,
      "step": 40870,
      "train_speed(iter/s)": 0.201582
    },
    {
      "acc": 0.78115153,
      "epoch": 0.953761355516995,
      "grad_norm": 5.3125,
      "learning_rate": 5.615606748964613e-06,
      "loss": 0.77917652,
      "memory(GiB)": 147.13,
      "step": 40880,
      "train_speed(iter/s)": 0.201608
    },
    {
      "acc": 0.77769594,
      "epoch": 0.9539946630892839,
      "grad_norm": 7.40625,
      "learning_rate": 5.613731966506321e-06,
      "loss": 0.78834753,
      "memory(GiB)": 147.13,
      "step": 40890,
      "train_speed(iter/s)": 0.201632
    },
    {
      "acc": 0.78029628,
      "epoch": 0.9542279706615728,
      "grad_norm": 7.28125,
      "learning_rate": 5.611857096437966e-06,
      "loss": 0.80404196,
      "memory(GiB)": 147.13,
      "step": 40900,
      "train_speed(iter/s)": 0.201658
    },
    {
      "acc": 0.78504705,
      "epoch": 0.9544612782338617,
      "grad_norm": 5.34375,
      "learning_rate": 5.60998213902719e-06,
      "loss": 0.78944445,
      "memory(GiB)": 147.13,
      "step": 40910,
      "train_speed(iter/s)": 0.201684
    },
    {
      "acc": 0.78178701,
      "epoch": 0.9546945858061505,
      "grad_norm": 6.03125,
      "learning_rate": 5.60810709454164e-06,
      "loss": 0.791785,
      "memory(GiB)": 147.13,
      "step": 40920,
      "train_speed(iter/s)": 0.20171
    },
    {
      "acc": 0.77981234,
      "epoch": 0.9549278933784394,
      "grad_norm": 6.3125,
      "learning_rate": 5.606231963248978e-06,
      "loss": 0.80088806,
      "memory(GiB)": 147.13,
      "step": 40930,
      "train_speed(iter/s)": 0.201736
    },
    {
      "acc": 0.79295478,
      "epoch": 0.9551612009507283,
      "grad_norm": 3.71875,
      "learning_rate": 5.60435674541688e-06,
      "loss": 0.7574069,
      "memory(GiB)": 147.13,
      "step": 40940,
      "train_speed(iter/s)": 0.20176
    },
    {
      "acc": 0.77834711,
      "epoch": 0.9553945085230172,
      "grad_norm": 6.34375,
      "learning_rate": 5.602481441313032e-06,
      "loss": 0.78354673,
      "memory(GiB)": 147.13,
      "step": 40950,
      "train_speed(iter/s)": 0.201786
    },
    {
      "acc": 0.79002304,
      "epoch": 0.9556278160953061,
      "grad_norm": 4.09375,
      "learning_rate": 5.6006060512051355e-06,
      "loss": 0.75721602,
      "memory(GiB)": 147.13,
      "step": 40960,
      "train_speed(iter/s)": 0.20181
    },
    {
      "acc": 0.78958197,
      "epoch": 0.955861123667595,
      "grad_norm": 6.34375,
      "learning_rate": 5.598730575360898e-06,
      "loss": 0.77207351,
      "memory(GiB)": 147.13,
      "step": 40970,
      "train_speed(iter/s)": 0.201836
    },
    {
      "acc": 0.78223519,
      "epoch": 0.9560944312398839,
      "grad_norm": 8.1875,
      "learning_rate": 5.596855014048045e-06,
      "loss": 0.77432485,
      "memory(GiB)": 147.13,
      "step": 40980,
      "train_speed(iter/s)": 0.201861
    },
    {
      "acc": 0.76258755,
      "epoch": 0.9563277388121728,
      "grad_norm": 5.375,
      "learning_rate": 5.594979367534311e-06,
      "loss": 0.84754496,
      "memory(GiB)": 147.13,
      "step": 40990,
      "train_speed(iter/s)": 0.201888
    },
    {
      "acc": 0.79521761,
      "epoch": 0.9565610463844617,
      "grad_norm": 4.3125,
      "learning_rate": 5.593103636087446e-06,
      "loss": 0.74911346,
      "memory(GiB)": 147.13,
      "step": 41000,
      "train_speed(iter/s)": 0.201913
    },
    {
      "epoch": 0.9565610463844617,
      "eval_acc": 0.7435477488445194,
      "eval_loss": 0.8075899481773376,
      "eval_runtime": 1270.0956,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 41000
    },
    {
      "acc": 0.78175507,
      "epoch": 0.9567943539567506,
      "grad_norm": 5.21875,
      "learning_rate": 5.591227819975209e-06,
      "loss": 0.77779932,
      "memory(GiB)": 147.13,
      "step": 41010,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.77938576,
      "epoch": 0.9570276615290395,
      "grad_norm": 5.59375,
      "learning_rate": 5.589351919465373e-06,
      "loss": 0.80347643,
      "memory(GiB)": 147.13,
      "step": 41020,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.77089553,
      "epoch": 0.9572609691013284,
      "grad_norm": 5.03125,
      "learning_rate": 5.587475934825721e-06,
      "loss": 0.84662561,
      "memory(GiB)": 147.13,
      "step": 41030,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.77880993,
      "epoch": 0.9574942766736173,
      "grad_norm": 5.3125,
      "learning_rate": 5.585599866324052e-06,
      "loss": 0.77378798,
      "memory(GiB)": 147.13,
      "step": 41040,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.77963066,
      "epoch": 0.9577275842459062,
      "grad_norm": 5.46875,
      "learning_rate": 5.583723714228169e-06,
      "loss": 0.78790627,
      "memory(GiB)": 147.13,
      "step": 41050,
      "train_speed(iter/s)": 0.200765
    },
    {
      "acc": 0.78368587,
      "epoch": 0.9579608918181951,
      "grad_norm": 5.71875,
      "learning_rate": 5.581847478805898e-06,
      "loss": 0.78440642,
      "memory(GiB)": 147.13,
      "step": 41060,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.79512386,
      "epoch": 0.958194199390484,
      "grad_norm": 5.15625,
      "learning_rate": 5.579971160325066e-06,
      "loss": 0.73103809,
      "memory(GiB)": 147.13,
      "step": 41070,
      "train_speed(iter/s)": 0.200817
    },
    {
      "acc": 0.79529629,
      "epoch": 0.9584275069627729,
      "grad_norm": 5.90625,
      "learning_rate": 5.578094759053521e-06,
      "loss": 0.73167953,
      "memory(GiB)": 147.13,
      "step": 41080,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.77233238,
      "epoch": 0.9586608145350618,
      "grad_norm": 5.65625,
      "learning_rate": 5.576218275259116e-06,
      "loss": 0.83091927,
      "memory(GiB)": 147.13,
      "step": 41090,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.76945944,
      "epoch": 0.9588941221073507,
      "grad_norm": 7.71875,
      "learning_rate": 5.574341709209721e-06,
      "loss": 0.83968544,
      "memory(GiB)": 147.13,
      "step": 41100,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.78422494,
      "epoch": 0.9591274296796395,
      "grad_norm": 5.71875,
      "learning_rate": 5.572465061173215e-06,
      "loss": 0.76671977,
      "memory(GiB)": 147.13,
      "step": 41110,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.76734838,
      "epoch": 0.9593607372519284,
      "grad_norm": 4.75,
      "learning_rate": 5.5705883314174845e-06,
      "loss": 0.83322086,
      "memory(GiB)": 147.13,
      "step": 41120,
      "train_speed(iter/s)": 0.200946
    },
    {
      "acc": 0.77143869,
      "epoch": 0.9595940448242173,
      "grad_norm": 4.59375,
      "learning_rate": 5.568711520210437e-06,
      "loss": 0.83908682,
      "memory(GiB)": 147.13,
      "step": 41130,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.77202053,
      "epoch": 0.9598273523965062,
      "grad_norm": 5.34375,
      "learning_rate": 5.566834627819986e-06,
      "loss": 0.8184268,
      "memory(GiB)": 147.13,
      "step": 41140,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.77577133,
      "epoch": 0.9600606599687951,
      "grad_norm": 4.875,
      "learning_rate": 5.564957654514055e-06,
      "loss": 0.80979881,
      "memory(GiB)": 147.13,
      "step": 41150,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.77774215,
      "epoch": 0.960293967541084,
      "grad_norm": 3.46875,
      "learning_rate": 5.563080600560584e-06,
      "loss": 0.80543184,
      "memory(GiB)": 147.13,
      "step": 41160,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.77905116,
      "epoch": 0.9605272751133729,
      "grad_norm": 5.46875,
      "learning_rate": 5.5612034662275205e-06,
      "loss": 0.78646588,
      "memory(GiB)": 147.13,
      "step": 41170,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.77790079,
      "epoch": 0.9607605826856618,
      "grad_norm": 4.375,
      "learning_rate": 5.559326251782825e-06,
      "loss": 0.78109341,
      "memory(GiB)": 147.13,
      "step": 41180,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.7724268,
      "epoch": 0.9609938902579507,
      "grad_norm": 5.03125,
      "learning_rate": 5.55744895749447e-06,
      "loss": 0.82191772,
      "memory(GiB)": 147.13,
      "step": 41190,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.78639202,
      "epoch": 0.9612271978302396,
      "grad_norm": 4.53125,
      "learning_rate": 5.555571583630439e-06,
      "loss": 0.75079656,
      "memory(GiB)": 147.13,
      "step": 41200,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.78967714,
      "epoch": 0.9614605054025285,
      "grad_norm": 4.78125,
      "learning_rate": 5.553694130458725e-06,
      "loss": 0.77342958,
      "memory(GiB)": 147.13,
      "step": 41210,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.77171679,
      "epoch": 0.9616938129748174,
      "grad_norm": 6.59375,
      "learning_rate": 5.551816598247334e-06,
      "loss": 0.81496878,
      "memory(GiB)": 147.13,
      "step": 41220,
      "train_speed(iter/s)": 0.201198
    },
    {
      "acc": 0.79381685,
      "epoch": 0.9619271205471063,
      "grad_norm": 4.84375,
      "learning_rate": 5.549938987264284e-06,
      "loss": 0.73036442,
      "memory(GiB)": 147.13,
      "step": 41230,
      "train_speed(iter/s)": 0.201223
    },
    {
      "acc": 0.79161181,
      "epoch": 0.9621604281193952,
      "grad_norm": 4.71875,
      "learning_rate": 5.548061297777604e-06,
      "loss": 0.73119068,
      "memory(GiB)": 147.13,
      "step": 41240,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.8105217,
      "epoch": 0.9623937356916841,
      "grad_norm": 5.0,
      "learning_rate": 5.546183530055334e-06,
      "loss": 0.68218474,
      "memory(GiB)": 147.13,
      "step": 41250,
      "train_speed(iter/s)": 0.201274
    },
    {
      "acc": 0.75500202,
      "epoch": 0.962627043263973,
      "grad_norm": 7.15625,
      "learning_rate": 5.544305684365522e-06,
      "loss": 0.89543095,
      "memory(GiB)": 147.13,
      "step": 41260,
      "train_speed(iter/s)": 0.2013
    },
    {
      "acc": 0.77084942,
      "epoch": 0.9628603508362619,
      "grad_norm": 6.6875,
      "learning_rate": 5.542427760976232e-06,
      "loss": 0.82826271,
      "memory(GiB)": 147.13,
      "step": 41270,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.76677494,
      "epoch": 0.9630936584085508,
      "grad_norm": 8.4375,
      "learning_rate": 5.540549760155537e-06,
      "loss": 0.86246529,
      "memory(GiB)": 147.13,
      "step": 41280,
      "train_speed(iter/s)": 0.201348
    },
    {
      "acc": 0.79330645,
      "epoch": 0.9633269659808397,
      "grad_norm": 4.40625,
      "learning_rate": 5.53867168217152e-06,
      "loss": 0.74018192,
      "memory(GiB)": 147.13,
      "step": 41290,
      "train_speed(iter/s)": 0.201371
    },
    {
      "acc": 0.77103343,
      "epoch": 0.9635602735531285,
      "grad_norm": 5.84375,
      "learning_rate": 5.536793527292278e-06,
      "loss": 0.83305159,
      "memory(GiB)": 147.13,
      "step": 41300,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.76550531,
      "epoch": 0.9637935811254174,
      "grad_norm": 5.90625,
      "learning_rate": 5.5349152957859155e-06,
      "loss": 0.83599548,
      "memory(GiB)": 147.13,
      "step": 41310,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.77552824,
      "epoch": 0.9640268886977063,
      "grad_norm": 4.03125,
      "learning_rate": 5.53303698792055e-06,
      "loss": 0.82170572,
      "memory(GiB)": 147.13,
      "step": 41320,
      "train_speed(iter/s)": 0.201445
    },
    {
      "acc": 0.7870533,
      "epoch": 0.9642601962699952,
      "grad_norm": 8.375,
      "learning_rate": 5.531158603964309e-06,
      "loss": 0.75256863,
      "memory(GiB)": 147.13,
      "step": 41330,
      "train_speed(iter/s)": 0.20147
    },
    {
      "acc": 0.77868299,
      "epoch": 0.964493503842284,
      "grad_norm": 5.25,
      "learning_rate": 5.529280144185331e-06,
      "loss": 0.79275427,
      "memory(GiB)": 147.13,
      "step": 41340,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.77070007,
      "epoch": 0.964726811414573,
      "grad_norm": 4.53125,
      "learning_rate": 5.5274016088517676e-06,
      "loss": 0.82578239,
      "memory(GiB)": 147.13,
      "step": 41350,
      "train_speed(iter/s)": 0.201519
    },
    {
      "acc": 0.78871584,
      "epoch": 0.9649601189868618,
      "grad_norm": 5.125,
      "learning_rate": 5.525522998231777e-06,
      "loss": 0.76853447,
      "memory(GiB)": 147.13,
      "step": 41360,
      "train_speed(iter/s)": 0.201544
    },
    {
      "acc": 0.78402672,
      "epoch": 0.9651934265591507,
      "grad_norm": 4.65625,
      "learning_rate": 5.523644312593533e-06,
      "loss": 0.76367111,
      "memory(GiB)": 147.13,
      "step": 41370,
      "train_speed(iter/s)": 0.20157
    },
    {
      "acc": 0.79334326,
      "epoch": 0.9654267341314396,
      "grad_norm": 5.09375,
      "learning_rate": 5.521765552205213e-06,
      "loss": 0.734587,
      "memory(GiB)": 147.13,
      "step": 41380,
      "train_speed(iter/s)": 0.201596
    },
    {
      "acc": 0.7742136,
      "epoch": 0.9656600417037285,
      "grad_norm": 4.78125,
      "learning_rate": 5.519886717335012e-06,
      "loss": 0.82144842,
      "memory(GiB)": 147.13,
      "step": 41390,
      "train_speed(iter/s)": 0.201618
    },
    {
      "acc": 0.75570812,
      "epoch": 0.9658933492760174,
      "grad_norm": 5.75,
      "learning_rate": 5.518007808251135e-06,
      "loss": 0.89175406,
      "memory(GiB)": 147.13,
      "step": 41400,
      "train_speed(iter/s)": 0.201643
    },
    {
      "acc": 0.7830318,
      "epoch": 0.9661266568483063,
      "grad_norm": 5.34375,
      "learning_rate": 5.516128825221792e-06,
      "loss": 0.77546549,
      "memory(GiB)": 147.13,
      "step": 41410,
      "train_speed(iter/s)": 0.201668
    },
    {
      "acc": 0.77321138,
      "epoch": 0.9663599644205952,
      "grad_norm": 4.625,
      "learning_rate": 5.514249768515209e-06,
      "loss": 0.80514297,
      "memory(GiB)": 147.13,
      "step": 41420,
      "train_speed(iter/s)": 0.201691
    },
    {
      "acc": 0.78639116,
      "epoch": 0.9665932719928841,
      "grad_norm": 6.0625,
      "learning_rate": 5.512370638399622e-06,
      "loss": 0.77311621,
      "memory(GiB)": 147.13,
      "step": 41430,
      "train_speed(iter/s)": 0.201717
    },
    {
      "acc": 0.78260765,
      "epoch": 0.966826579565173,
      "grad_norm": 5.40625,
      "learning_rate": 5.510491435143275e-06,
      "loss": 0.76550779,
      "memory(GiB)": 147.13,
      "step": 41440,
      "train_speed(iter/s)": 0.201741
    },
    {
      "acc": 0.79104238,
      "epoch": 0.9670598871374619,
      "grad_norm": 7.21875,
      "learning_rate": 5.508612159014424e-06,
      "loss": 0.7574954,
      "memory(GiB)": 147.13,
      "step": 41450,
      "train_speed(iter/s)": 0.201766
    },
    {
      "acc": 0.77882171,
      "epoch": 0.9672931947097508,
      "grad_norm": 5.34375,
      "learning_rate": 5.506732810281335e-06,
      "loss": 0.80786209,
      "memory(GiB)": 147.13,
      "step": 41460,
      "train_speed(iter/s)": 0.201792
    },
    {
      "acc": 0.80258312,
      "epoch": 0.9675265022820397,
      "grad_norm": 3.875,
      "learning_rate": 5.504853389212285e-06,
      "loss": 0.72713561,
      "memory(GiB)": 147.13,
      "step": 41470,
      "train_speed(iter/s)": 0.201817
    },
    {
      "acc": 0.79807262,
      "epoch": 0.9677598098543286,
      "grad_norm": 5.125,
      "learning_rate": 5.502973896075559e-06,
      "loss": 0.72810946,
      "memory(GiB)": 147.13,
      "step": 41480,
      "train_speed(iter/s)": 0.201842
    },
    {
      "acc": 0.77658319,
      "epoch": 0.9679931174266175,
      "grad_norm": 6.90625,
      "learning_rate": 5.501094331139457e-06,
      "loss": 0.79982195,
      "memory(GiB)": 147.13,
      "step": 41490,
      "train_speed(iter/s)": 0.201866
    },
    {
      "acc": 0.75606503,
      "epoch": 0.9682264249989063,
      "grad_norm": 5.1875,
      "learning_rate": 5.499214694672283e-06,
      "loss": 0.88801994,
      "memory(GiB)": 147.13,
      "step": 41500,
      "train_speed(iter/s)": 0.201891
    },
    {
      "epoch": 0.9682264249989063,
      "eval_acc": 0.7436978039901857,
      "eval_loss": 0.8076462745666504,
      "eval_runtime": 1271.1744,
      "eval_samples_per_second": 28.313,
      "eval_steps_per_second": 14.157,
      "step": 41500
    },
    {
      "acc": 0.78003526,
      "epoch": 0.9684597325711952,
      "grad_norm": 5.21875,
      "learning_rate": 5.497334986942358e-06,
      "loss": 0.77953706,
      "memory(GiB)": 147.13,
      "step": 41510,
      "train_speed(iter/s)": 0.20066
    },
    {
      "acc": 0.78291988,
      "epoch": 0.9686930401434841,
      "grad_norm": 5.1875,
      "learning_rate": 5.495455208218008e-06,
      "loss": 0.7734293,
      "memory(GiB)": 147.13,
      "step": 41520,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.76620455,
      "epoch": 0.968926347715773,
      "grad_norm": 5.96875,
      "learning_rate": 5.493575358767571e-06,
      "loss": 0.82494164,
      "memory(GiB)": 147.13,
      "step": 41530,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.77570481,
      "epoch": 0.9691596552880619,
      "grad_norm": 6.125,
      "learning_rate": 5.491695438859394e-06,
      "loss": 0.79977193,
      "memory(GiB)": 147.13,
      "step": 41540,
      "train_speed(iter/s)": 0.200734
    },
    {
      "acc": 0.79774528,
      "epoch": 0.9693929628603508,
      "grad_norm": 4.875,
      "learning_rate": 5.489815448761837e-06,
      "loss": 0.72081814,
      "memory(GiB)": 147.13,
      "step": 41550,
      "train_speed(iter/s)": 0.200759
    },
    {
      "acc": 0.78871336,
      "epoch": 0.9696262704326397,
      "grad_norm": 6.8125,
      "learning_rate": 5.487935388743266e-06,
      "loss": 0.74919105,
      "memory(GiB)": 147.13,
      "step": 41560,
      "train_speed(iter/s)": 0.200783
    },
    {
      "acc": 0.76116972,
      "epoch": 0.9698595780049286,
      "grad_norm": 7.3125,
      "learning_rate": 5.486055259072059e-06,
      "loss": 0.86874237,
      "memory(GiB)": 147.13,
      "step": 41570,
      "train_speed(iter/s)": 0.200808
    },
    {
      "acc": 0.77257109,
      "epoch": 0.9700928855772175,
      "grad_norm": 5.21875,
      "learning_rate": 5.484175060016607e-06,
      "loss": 0.820648,
      "memory(GiB)": 147.13,
      "step": 41580,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.78412466,
      "epoch": 0.9703261931495064,
      "grad_norm": 5.78125,
      "learning_rate": 5.482294791845305e-06,
      "loss": 0.78075528,
      "memory(GiB)": 147.13,
      "step": 41590,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.79700346,
      "epoch": 0.9705595007217953,
      "grad_norm": 5.15625,
      "learning_rate": 5.480414454826563e-06,
      "loss": 0.7259798,
      "memory(GiB)": 147.13,
      "step": 41600,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.7838769,
      "epoch": 0.9707928082940842,
      "grad_norm": 4.65625,
      "learning_rate": 5.478534049228794e-06,
      "loss": 0.78670282,
      "memory(GiB)": 147.13,
      "step": 41610,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.77976408,
      "epoch": 0.9710261158663731,
      "grad_norm": 6.71875,
      "learning_rate": 5.476653575320432e-06,
      "loss": 0.79632206,
      "memory(GiB)": 147.13,
      "step": 41620,
      "train_speed(iter/s)": 0.200935
    },
    {
      "acc": 0.78164692,
      "epoch": 0.971259423438662,
      "grad_norm": 6.96875,
      "learning_rate": 5.474773033369908e-06,
      "loss": 0.77794366,
      "memory(GiB)": 147.13,
      "step": 41630,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.81677799,
      "epoch": 0.9714927310109509,
      "grad_norm": 5.84375,
      "learning_rate": 5.472892423645673e-06,
      "loss": 0.63997331,
      "memory(GiB)": 147.13,
      "step": 41640,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.77521987,
      "epoch": 0.9717260385832398,
      "grad_norm": 8.3125,
      "learning_rate": 5.47101174641618e-06,
      "loss": 0.81140451,
      "memory(GiB)": 147.13,
      "step": 41650,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.77916822,
      "epoch": 0.9719593461555287,
      "grad_norm": 7.21875,
      "learning_rate": 5.469131001949899e-06,
      "loss": 0.80082006,
      "memory(GiB)": 147.13,
      "step": 41660,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.76695161,
      "epoch": 0.9721926537278176,
      "grad_norm": 6.375,
      "learning_rate": 5.467250190515303e-06,
      "loss": 0.8524477,
      "memory(GiB)": 147.13,
      "step": 41670,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.79857035,
      "epoch": 0.9724259613001065,
      "grad_norm": 7.375,
      "learning_rate": 5.465369312380879e-06,
      "loss": 0.72482586,
      "memory(GiB)": 147.13,
      "step": 41680,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.77130175,
      "epoch": 0.9726592688723953,
      "grad_norm": 8.1875,
      "learning_rate": 5.463488367815119e-06,
      "loss": 0.8281208,
      "memory(GiB)": 147.13,
      "step": 41690,
      "train_speed(iter/s)": 0.201113
    },
    {
      "acc": 0.78064547,
      "epoch": 0.9728925764446842,
      "grad_norm": 5.65625,
      "learning_rate": 5.46160735708653e-06,
      "loss": 0.78882265,
      "memory(GiB)": 147.13,
      "step": 41700,
      "train_speed(iter/s)": 0.201139
    },
    {
      "acc": 0.79099216,
      "epoch": 0.9731258840169731,
      "grad_norm": 5.8125,
      "learning_rate": 5.459726280463625e-06,
      "loss": 0.75952692,
      "memory(GiB)": 147.13,
      "step": 41710,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.77816892,
      "epoch": 0.973359191589262,
      "grad_norm": 3.90625,
      "learning_rate": 5.4578451382149275e-06,
      "loss": 0.7955409,
      "memory(GiB)": 147.13,
      "step": 41720,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.79420767,
      "epoch": 0.9735924991615509,
      "grad_norm": 8.5625,
      "learning_rate": 5.455963930608969e-06,
      "loss": 0.75089536,
      "memory(GiB)": 147.13,
      "step": 41730,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.7890696,
      "epoch": 0.9738258067338398,
      "grad_norm": 7.65625,
      "learning_rate": 5.454082657914292e-06,
      "loss": 0.75170484,
      "memory(GiB)": 147.13,
      "step": 41740,
      "train_speed(iter/s)": 0.201236
    },
    {
      "acc": 0.79078021,
      "epoch": 0.9740591143061287,
      "grad_norm": 16.625,
      "learning_rate": 5.452201320399447e-06,
      "loss": 0.7524622,
      "memory(GiB)": 147.13,
      "step": 41750,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.80642824,
      "epoch": 0.9742924218784176,
      "grad_norm": 6.5,
      "learning_rate": 5.450319918332995e-06,
      "loss": 0.69903097,
      "memory(GiB)": 147.13,
      "step": 41760,
      "train_speed(iter/s)": 0.201285
    },
    {
      "acc": 0.79178162,
      "epoch": 0.9745257294507065,
      "grad_norm": 4.34375,
      "learning_rate": 5.448438451983507e-06,
      "loss": 0.74413743,
      "memory(GiB)": 147.13,
      "step": 41770,
      "train_speed(iter/s)": 0.201311
    },
    {
      "acc": 0.77486391,
      "epoch": 0.9747590370229954,
      "grad_norm": 4.0625,
      "learning_rate": 5.4465569216195576e-06,
      "loss": 0.82327709,
      "memory(GiB)": 147.13,
      "step": 41780,
      "train_speed(iter/s)": 0.201336
    },
    {
      "acc": 0.78499551,
      "epoch": 0.9749923445952843,
      "grad_norm": 6.21875,
      "learning_rate": 5.444675327509738e-06,
      "loss": 0.7698328,
      "memory(GiB)": 147.13,
      "step": 41790,
      "train_speed(iter/s)": 0.201362
    },
    {
      "acc": 0.78951769,
      "epoch": 0.9752256521675732,
      "grad_norm": 6.375,
      "learning_rate": 5.4427936699226455e-06,
      "loss": 0.75837417,
      "memory(GiB)": 147.13,
      "step": 41800,
      "train_speed(iter/s)": 0.201387
    },
    {
      "acc": 0.7888494,
      "epoch": 0.9754589597398621,
      "grad_norm": 9.125,
      "learning_rate": 5.440911949126885e-06,
      "loss": 0.75651331,
      "memory(GiB)": 147.13,
      "step": 41810,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.79928493,
      "epoch": 0.975692267312151,
      "grad_norm": 7.96875,
      "learning_rate": 5.4390301653910726e-06,
      "loss": 0.71884995,
      "memory(GiB)": 147.13,
      "step": 41820,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.76790247,
      "epoch": 0.9759255748844399,
      "grad_norm": 5.84375,
      "learning_rate": 5.4371483189838315e-06,
      "loss": 0.83689575,
      "memory(GiB)": 147.13,
      "step": 41830,
      "train_speed(iter/s)": 0.201459
    },
    {
      "acc": 0.78529429,
      "epoch": 0.9761588824567288,
      "grad_norm": 10.25,
      "learning_rate": 5.435266410173794e-06,
      "loss": 0.77676764,
      "memory(GiB)": 147.13,
      "step": 41840,
      "train_speed(iter/s)": 0.201483
    },
    {
      "acc": 0.75995874,
      "epoch": 0.9763921900290177,
      "grad_norm": 4.875,
      "learning_rate": 5.433384439229603e-06,
      "loss": 0.85082378,
      "memory(GiB)": 147.13,
      "step": 41850,
      "train_speed(iter/s)": 0.201509
    },
    {
      "acc": 0.7614418,
      "epoch": 0.9766254976013066,
      "grad_norm": 5.625,
      "learning_rate": 5.431502406419908e-06,
      "loss": 0.87694912,
      "memory(GiB)": 147.13,
      "step": 41860,
      "train_speed(iter/s)": 0.201535
    },
    {
      "acc": 0.80430679,
      "epoch": 0.9768588051735955,
      "grad_norm": 9.625,
      "learning_rate": 5.429620312013372e-06,
      "loss": 0.68969584,
      "memory(GiB)": 147.13,
      "step": 41870,
      "train_speed(iter/s)": 0.201561
    },
    {
      "acc": 0.76424179,
      "epoch": 0.9770921127458843,
      "grad_norm": 6.75,
      "learning_rate": 5.427738156278662e-06,
      "loss": 0.86263018,
      "memory(GiB)": 147.13,
      "step": 41880,
      "train_speed(iter/s)": 0.201587
    },
    {
      "acc": 0.77523551,
      "epoch": 0.9773254203181732,
      "grad_norm": 4.25,
      "learning_rate": 5.4258559394844515e-06,
      "loss": 0.80997353,
      "memory(GiB)": 147.13,
      "step": 41890,
      "train_speed(iter/s)": 0.201613
    },
    {
      "acc": 0.76072025,
      "epoch": 0.977558727890462,
      "grad_norm": 4.78125,
      "learning_rate": 5.423973661899431e-06,
      "loss": 0.87792006,
      "memory(GiB)": 147.13,
      "step": 41900,
      "train_speed(iter/s)": 0.201639
    },
    {
      "acc": 0.78689508,
      "epoch": 0.977792035462751,
      "grad_norm": 6.0,
      "learning_rate": 5.4220913237922936e-06,
      "loss": 0.76750278,
      "memory(GiB)": 147.13,
      "step": 41910,
      "train_speed(iter/s)": 0.201663
    },
    {
      "acc": 0.77254424,
      "epoch": 0.9780253430350399,
      "grad_norm": 11.1875,
      "learning_rate": 5.4202089254317415e-06,
      "loss": 0.83702393,
      "memory(GiB)": 147.13,
      "step": 41920,
      "train_speed(iter/s)": 0.20169
    },
    {
      "acc": 0.77660599,
      "epoch": 0.9782586506073288,
      "grad_norm": 5.21875,
      "learning_rate": 5.418326467086488e-06,
      "loss": 0.79623809,
      "memory(GiB)": 147.13,
      "step": 41930,
      "train_speed(iter/s)": 0.201717
    },
    {
      "acc": 0.76378241,
      "epoch": 0.9784919581796176,
      "grad_norm": 5.40625,
      "learning_rate": 5.416443949025253e-06,
      "loss": 0.84370394,
      "memory(GiB)": 147.13,
      "step": 41940,
      "train_speed(iter/s)": 0.201744
    },
    {
      "acc": 0.76264544,
      "epoch": 0.9787252657519065,
      "grad_norm": 4.8125,
      "learning_rate": 5.414561371516764e-06,
      "loss": 0.86721058,
      "memory(GiB)": 147.13,
      "step": 41950,
      "train_speed(iter/s)": 0.20177
    },
    {
      "acc": 0.81001492,
      "epoch": 0.9789585733241954,
      "grad_norm": 6.375,
      "learning_rate": 5.41267873482976e-06,
      "loss": 0.70287361,
      "memory(GiB)": 147.13,
      "step": 41960,
      "train_speed(iter/s)": 0.201794
    },
    {
      "acc": 0.78907146,
      "epoch": 0.9791918808964843,
      "grad_norm": 7.46875,
      "learning_rate": 5.410796039232989e-06,
      "loss": 0.73023872,
      "memory(GiB)": 147.13,
      "step": 41970,
      "train_speed(iter/s)": 0.201819
    },
    {
      "acc": 0.77236247,
      "epoch": 0.9794251884687732,
      "grad_norm": 6.0,
      "learning_rate": 5.4089132849952e-06,
      "loss": 0.83747387,
      "memory(GiB)": 147.13,
      "step": 41980,
      "train_speed(iter/s)": 0.201845
    },
    {
      "acc": 0.78121223,
      "epoch": 0.9796584960410621,
      "grad_norm": 4.875,
      "learning_rate": 5.407030472385158e-06,
      "loss": 0.78266506,
      "memory(GiB)": 147.13,
      "step": 41990,
      "train_speed(iter/s)": 0.201871
    },
    {
      "acc": 0.76154213,
      "epoch": 0.979891803613351,
      "grad_norm": 4.875,
      "learning_rate": 5.4051476016716365e-06,
      "loss": 0.85525732,
      "memory(GiB)": 147.13,
      "step": 42000,
      "train_speed(iter/s)": 0.201896
    },
    {
      "epoch": 0.979891803613351,
      "eval_acc": 0.743733035721292,
      "eval_loss": 0.8074895739555359,
      "eval_runtime": 1270.6849,
      "eval_samples_per_second": 28.324,
      "eval_steps_per_second": 14.162,
      "step": 42000
    },
    {
      "acc": 0.79927797,
      "epoch": 0.9801251111856399,
      "grad_norm": 9.125,
      "learning_rate": 5.4032646731234115e-06,
      "loss": 0.72784662,
      "memory(GiB)": 147.13,
      "step": 42010,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.78354425,
      "epoch": 0.9803584187579288,
      "grad_norm": 6.53125,
      "learning_rate": 5.401381687009271e-06,
      "loss": 0.78162694,
      "memory(GiB)": 147.13,
      "step": 42020,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.77189026,
      "epoch": 0.9805917263302177,
      "grad_norm": 5.625,
      "learning_rate": 5.399498643598011e-06,
      "loss": 0.81935215,
      "memory(GiB)": 147.13,
      "step": 42030,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.77528439,
      "epoch": 0.9808250339025066,
      "grad_norm": 6.21875,
      "learning_rate": 5.3976155431584375e-06,
      "loss": 0.82163353,
      "memory(GiB)": 147.13,
      "step": 42040,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.78729029,
      "epoch": 0.9810583414747955,
      "grad_norm": 4.9375,
      "learning_rate": 5.3957323859593604e-06,
      "loss": 0.77468286,
      "memory(GiB)": 147.13,
      "step": 42050,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.77076778,
      "epoch": 0.9812916490470844,
      "grad_norm": 4.4375,
      "learning_rate": 5.3938491722695996e-06,
      "loss": 0.82037773,
      "memory(GiB)": 147.13,
      "step": 42060,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.75968246,
      "epoch": 0.9815249566193732,
      "grad_norm": 6.15625,
      "learning_rate": 5.391965902357983e-06,
      "loss": 0.88906879,
      "memory(GiB)": 147.13,
      "step": 42070,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.78322248,
      "epoch": 0.9817582641916621,
      "grad_norm": 5.40625,
      "learning_rate": 5.390082576493348e-06,
      "loss": 0.76822419,
      "memory(GiB)": 147.13,
      "step": 42080,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.77662072,
      "epoch": 0.981991571763951,
      "grad_norm": 6.71875,
      "learning_rate": 5.388199194944539e-06,
      "loss": 0.80375099,
      "memory(GiB)": 147.13,
      "step": 42090,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.79028854,
      "epoch": 0.9822248793362399,
      "grad_norm": 4.5625,
      "learning_rate": 5.3863157579804075e-06,
      "loss": 0.72856355,
      "memory(GiB)": 147.13,
      "step": 42100,
      "train_speed(iter/s)": 0.200899
    },
    {
      "acc": 0.76783481,
      "epoch": 0.9824581869085288,
      "grad_norm": 5.09375,
      "learning_rate": 5.384432265869815e-06,
      "loss": 0.84242039,
      "memory(GiB)": 147.13,
      "step": 42110,
      "train_speed(iter/s)": 0.200924
    },
    {
      "acc": 0.7857872,
      "epoch": 0.9826914944808177,
      "grad_norm": 6.96875,
      "learning_rate": 5.382548718881627e-06,
      "loss": 0.77328167,
      "memory(GiB)": 147.13,
      "step": 42120,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.78016534,
      "epoch": 0.9829248020531066,
      "grad_norm": 4.59375,
      "learning_rate": 5.380665117284721e-06,
      "loss": 0.78974218,
      "memory(GiB)": 147.13,
      "step": 42130,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.78708715,
      "epoch": 0.9831581096253955,
      "grad_norm": 5.0625,
      "learning_rate": 5.378781461347979e-06,
      "loss": 0.76465178,
      "memory(GiB)": 147.13,
      "step": 42140,
      "train_speed(iter/s)": 0.200997
    },
    {
      "acc": 0.77445755,
      "epoch": 0.9833914171976844,
      "grad_norm": 4.6875,
      "learning_rate": 5.376897751340294e-06,
      "loss": 0.82658815,
      "memory(GiB)": 147.13,
      "step": 42150,
      "train_speed(iter/s)": 0.201023
    },
    {
      "acc": 0.78046808,
      "epoch": 0.9836247247699733,
      "grad_norm": 8.625,
      "learning_rate": 5.375013987530565e-06,
      "loss": 0.78468456,
      "memory(GiB)": 147.13,
      "step": 42160,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.77549314,
      "epoch": 0.9838580323422622,
      "grad_norm": 7.78125,
      "learning_rate": 5.3731301701876985e-06,
      "loss": 0.81372795,
      "memory(GiB)": 147.13,
      "step": 42170,
      "train_speed(iter/s)": 0.201073
    },
    {
      "acc": 0.77141991,
      "epoch": 0.9840913399145511,
      "grad_norm": 4.96875,
      "learning_rate": 5.371246299580608e-06,
      "loss": 0.82050037,
      "memory(GiB)": 147.13,
      "step": 42180,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.79150095,
      "epoch": 0.98432464748684,
      "grad_norm": 6.59375,
      "learning_rate": 5.3693623759782165e-06,
      "loss": 0.73675623,
      "memory(GiB)": 147.13,
      "step": 42190,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.75877752,
      "epoch": 0.9845579550591289,
      "grad_norm": 4.71875,
      "learning_rate": 5.367478399649453e-06,
      "loss": 0.85596247,
      "memory(GiB)": 147.13,
      "step": 42200,
      "train_speed(iter/s)": 0.201147
    },
    {
      "acc": 0.77118468,
      "epoch": 0.9847912626314178,
      "grad_norm": 6.4375,
      "learning_rate": 5.365594370863254e-06,
      "loss": 0.81483583,
      "memory(GiB)": 147.13,
      "step": 42210,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.77900705,
      "epoch": 0.9850245702037067,
      "grad_norm": 6.0,
      "learning_rate": 5.363710289888564e-06,
      "loss": 0.79868593,
      "memory(GiB)": 147.13,
      "step": 42220,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.77433815,
      "epoch": 0.9852578777759956,
      "grad_norm": 5.5625,
      "learning_rate": 5.361826156994338e-06,
      "loss": 0.81312532,
      "memory(GiB)": 147.13,
      "step": 42230,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.79729567,
      "epoch": 0.9854911853482845,
      "grad_norm": 5.875,
      "learning_rate": 5.359941972449532e-06,
      "loss": 0.73013325,
      "memory(GiB)": 147.13,
      "step": 42240,
      "train_speed(iter/s)": 0.201245
    },
    {
      "acc": 0.78885775,
      "epoch": 0.9857244929205734,
      "grad_norm": 15.5625,
      "learning_rate": 5.358057736523114e-06,
      "loss": 0.75900469,
      "memory(GiB)": 147.13,
      "step": 42250,
      "train_speed(iter/s)": 0.20127
    },
    {
      "acc": 0.76385336,
      "epoch": 0.9859578004928623,
      "grad_norm": 6.15625,
      "learning_rate": 5.356173449484059e-06,
      "loss": 0.87608242,
      "memory(GiB)": 147.13,
      "step": 42260,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.78156281,
      "epoch": 0.9861911080651511,
      "grad_norm": 6.4375,
      "learning_rate": 5.3542891116013465e-06,
      "loss": 0.75918655,
      "memory(GiB)": 147.13,
      "step": 42270,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.78791151,
      "epoch": 0.98642441563744,
      "grad_norm": 5.1875,
      "learning_rate": 5.352404723143968e-06,
      "loss": 0.74543934,
      "memory(GiB)": 147.13,
      "step": 42280,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.7900342,
      "epoch": 0.9866577232097289,
      "grad_norm": 4.71875,
      "learning_rate": 5.350520284380916e-06,
      "loss": 0.74275174,
      "memory(GiB)": 147.13,
      "step": 42290,
      "train_speed(iter/s)": 0.201366
    },
    {
      "acc": 0.78988333,
      "epoch": 0.9868910307820178,
      "grad_norm": 5.375,
      "learning_rate": 5.3486357955811945e-06,
      "loss": 0.75508475,
      "memory(GiB)": 147.13,
      "step": 42300,
      "train_speed(iter/s)": 0.20139
    },
    {
      "acc": 0.76127229,
      "epoch": 0.9871243383543067,
      "grad_norm": 10.625,
      "learning_rate": 5.346751257013815e-06,
      "loss": 0.85649261,
      "memory(GiB)": 147.13,
      "step": 42310,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.76956244,
      "epoch": 0.9873576459265956,
      "grad_norm": 7.53125,
      "learning_rate": 5.344866668947794e-06,
      "loss": 0.8190443,
      "memory(GiB)": 147.13,
      "step": 42320,
      "train_speed(iter/s)": 0.201439
    },
    {
      "acc": 0.76760712,
      "epoch": 0.9875909534988845,
      "grad_norm": 5.03125,
      "learning_rate": 5.342982031652159e-06,
      "loss": 0.84836092,
      "memory(GiB)": 147.13,
      "step": 42330,
      "train_speed(iter/s)": 0.201462
    },
    {
      "acc": 0.77389174,
      "epoch": 0.9878242610711734,
      "grad_norm": 8.375,
      "learning_rate": 5.341097345395937e-06,
      "loss": 0.7971365,
      "memory(GiB)": 147.13,
      "step": 42340,
      "train_speed(iter/s)": 0.201487
    },
    {
      "acc": 0.78406467,
      "epoch": 0.9880575686434623,
      "grad_norm": 7.625,
      "learning_rate": 5.339212610448167e-06,
      "loss": 0.76403122,
      "memory(GiB)": 147.13,
      "step": 42350,
      "train_speed(iter/s)": 0.201509
    },
    {
      "acc": 0.7485312,
      "epoch": 0.9882908762157512,
      "grad_norm": 4.40625,
      "learning_rate": 5.3373278270778965e-06,
      "loss": 0.91936789,
      "memory(GiB)": 147.13,
      "step": 42360,
      "train_speed(iter/s)": 0.201534
    },
    {
      "acc": 0.78323016,
      "epoch": 0.9885241837880401,
      "grad_norm": 4.46875,
      "learning_rate": 5.3354429955541755e-06,
      "loss": 0.76532645,
      "memory(GiB)": 147.13,
      "step": 42370,
      "train_speed(iter/s)": 0.201559
    },
    {
      "acc": 0.774506,
      "epoch": 0.988757491360329,
      "grad_norm": 6.3125,
      "learning_rate": 5.333558116146063e-06,
      "loss": 0.78331833,
      "memory(GiB)": 147.13,
      "step": 42380,
      "train_speed(iter/s)": 0.201585
    },
    {
      "acc": 0.77661877,
      "epoch": 0.9889907989326179,
      "grad_norm": 4.90625,
      "learning_rate": 5.33167318912263e-06,
      "loss": 0.79144692,
      "memory(GiB)": 147.13,
      "step": 42390,
      "train_speed(iter/s)": 0.201608
    },
    {
      "acc": 0.79478455,
      "epoch": 0.9892241065049068,
      "grad_norm": 5.21875,
      "learning_rate": 5.329788214752944e-06,
      "loss": 0.72164011,
      "memory(GiB)": 147.13,
      "step": 42400,
      "train_speed(iter/s)": 0.201632
    },
    {
      "acc": 0.77908754,
      "epoch": 0.9894574140771957,
      "grad_norm": 10.3125,
      "learning_rate": 5.327903193306087e-06,
      "loss": 0.77795801,
      "memory(GiB)": 147.13,
      "step": 42410,
      "train_speed(iter/s)": 0.201655
    },
    {
      "acc": 0.76483593,
      "epoch": 0.9896907216494846,
      "grad_norm": 8.0,
      "learning_rate": 5.326018125051142e-06,
      "loss": 0.84128876,
      "memory(GiB)": 147.13,
      "step": 42420,
      "train_speed(iter/s)": 0.20168
    },
    {
      "acc": 0.79741845,
      "epoch": 0.9899240292217735,
      "grad_norm": 4.0625,
      "learning_rate": 5.324133010257206e-06,
      "loss": 0.73309522,
      "memory(GiB)": 147.13,
      "step": 42430,
      "train_speed(iter/s)": 0.201704
    },
    {
      "acc": 0.78365011,
      "epoch": 0.9901573367940624,
      "grad_norm": 5.78125,
      "learning_rate": 5.3222478491933775e-06,
      "loss": 0.78707895,
      "memory(GiB)": 147.13,
      "step": 42440,
      "train_speed(iter/s)": 0.201729
    },
    {
      "acc": 0.77475691,
      "epoch": 0.9903906443663513,
      "grad_norm": 8.5,
      "learning_rate": 5.320362642128761e-06,
      "loss": 0.82604122,
      "memory(GiB)": 147.13,
      "step": 42450,
      "train_speed(iter/s)": 0.201754
    },
    {
      "acc": 0.80049133,
      "epoch": 0.9906239519386401,
      "grad_norm": 5.875,
      "learning_rate": 5.318477389332471e-06,
      "loss": 0.72088904,
      "memory(GiB)": 147.13,
      "step": 42460,
      "train_speed(iter/s)": 0.201781
    },
    {
      "acc": 0.75600042,
      "epoch": 0.990857259510929,
      "grad_norm": 5.875,
      "learning_rate": 5.316592091073626e-06,
      "loss": 0.88159542,
      "memory(GiB)": 147.13,
      "step": 42470,
      "train_speed(iter/s)": 0.201805
    },
    {
      "acc": 0.7931004,
      "epoch": 0.9910905670832179,
      "grad_norm": 4.0625,
      "learning_rate": 5.314706747621352e-06,
      "loss": 0.73830881,
      "memory(GiB)": 147.13,
      "step": 42480,
      "train_speed(iter/s)": 0.201829
    },
    {
      "acc": 0.77918282,
      "epoch": 0.9913238746555068,
      "grad_norm": 4.65625,
      "learning_rate": 5.312821359244781e-06,
      "loss": 0.81584482,
      "memory(GiB)": 147.13,
      "step": 42490,
      "train_speed(iter/s)": 0.201854
    },
    {
      "acc": 0.76876907,
      "epoch": 0.9915571822277957,
      "grad_norm": 16.625,
      "learning_rate": 5.310935926213052e-06,
      "loss": 0.82897224,
      "memory(GiB)": 147.13,
      "step": 42500,
      "train_speed(iter/s)": 0.201878
    },
    {
      "epoch": 0.9915571822277957,
      "eval_acc": 0.7438225563471484,
      "eval_loss": 0.8072462677955627,
      "eval_runtime": 1270.4976,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.165,
      "step": 42500
    },
    {
      "acc": 0.76333456,
      "epoch": 0.9917904898000846,
      "grad_norm": 4.8125,
      "learning_rate": 5.309050448795311e-06,
      "loss": 0.84974442,
      "memory(GiB)": 147.13,
      "step": 42510,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.77313266,
      "epoch": 0.9920237973723735,
      "grad_norm": 5.6875,
      "learning_rate": 5.307164927260706e-06,
      "loss": 0.82176523,
      "memory(GiB)": 147.13,
      "step": 42520,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.78784313,
      "epoch": 0.9922571049446623,
      "grad_norm": 4.4375,
      "learning_rate": 5.305279361878398e-06,
      "loss": 0.75282192,
      "memory(GiB)": 147.13,
      "step": 42530,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.79874549,
      "epoch": 0.9924904125169512,
      "grad_norm": 4.5,
      "learning_rate": 5.30339375291755e-06,
      "loss": 0.71279907,
      "memory(GiB)": 147.13,
      "step": 42540,
      "train_speed(iter/s)": 0.200753
    },
    {
      "acc": 0.77315574,
      "epoch": 0.9927237200892401,
      "grad_norm": 5.25,
      "learning_rate": 5.3015081006473315e-06,
      "loss": 0.81193829,
      "memory(GiB)": 147.13,
      "step": 42550,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.77819738,
      "epoch": 0.992957027661529,
      "grad_norm": 5.375,
      "learning_rate": 5.299622405336919e-06,
      "loss": 0.78049955,
      "memory(GiB)": 147.13,
      "step": 42560,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.78241773,
      "epoch": 0.993190335233818,
      "grad_norm": 6.71875,
      "learning_rate": 5.297736667255497e-06,
      "loss": 0.79108753,
      "memory(GiB)": 147.13,
      "step": 42570,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.78949213,
      "epoch": 0.9934236428061068,
      "grad_norm": 5.5625,
      "learning_rate": 5.2958508866722506e-06,
      "loss": 0.75022783,
      "memory(GiB)": 147.13,
      "step": 42580,
      "train_speed(iter/s)": 0.200849
    },
    {
      "acc": 0.75920568,
      "epoch": 0.9936569503783957,
      "grad_norm": 6.25,
      "learning_rate": 5.293965063856375e-06,
      "loss": 0.8708231,
      "memory(GiB)": 147.13,
      "step": 42590,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.76359487,
      "epoch": 0.9938902579506846,
      "grad_norm": 7.6875,
      "learning_rate": 5.292079199077073e-06,
      "loss": 0.85245781,
      "memory(GiB)": 147.13,
      "step": 42600,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.77966127,
      "epoch": 0.9941235655229735,
      "grad_norm": 5.9375,
      "learning_rate": 5.290193292603551e-06,
      "loss": 0.7903245,
      "memory(GiB)": 147.13,
      "step": 42610,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.78285179,
      "epoch": 0.9943568730952624,
      "grad_norm": 7.59375,
      "learning_rate": 5.2883073447050205e-06,
      "loss": 0.78474703,
      "memory(GiB)": 147.13,
      "step": 42620,
      "train_speed(iter/s)": 0.200946
    },
    {
      "acc": 0.78321028,
      "epoch": 0.9945901806675513,
      "grad_norm": 5.09375,
      "learning_rate": 5.2864213556507e-06,
      "loss": 0.79646807,
      "memory(GiB)": 147.13,
      "step": 42630,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.76653428,
      "epoch": 0.9948234882398402,
      "grad_norm": 7.8125,
      "learning_rate": 5.2845353257098146e-06,
      "loss": 0.84334564,
      "memory(GiB)": 147.13,
      "step": 42640,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.77440205,
      "epoch": 0.995056795812129,
      "grad_norm": 5.8125,
      "learning_rate": 5.282649255151593e-06,
      "loss": 0.80458755,
      "memory(GiB)": 147.13,
      "step": 42650,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.77449627,
      "epoch": 0.9952901033844179,
      "grad_norm": 4.875,
      "learning_rate": 5.280763144245272e-06,
      "loss": 0.81251926,
      "memory(GiB)": 147.13,
      "step": 42660,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.77925887,
      "epoch": 0.9955234109567068,
      "grad_norm": 6.125,
      "learning_rate": 5.2788769932600944e-06,
      "loss": 0.79605002,
      "memory(GiB)": 147.13,
      "step": 42670,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.77389841,
      "epoch": 0.9957567185289957,
      "grad_norm": 6.71875,
      "learning_rate": 5.276990802465309e-06,
      "loss": 0.80099325,
      "memory(GiB)": 147.13,
      "step": 42680,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.77797318,
      "epoch": 0.9959900261012846,
      "grad_norm": 7.3125,
      "learning_rate": 5.275104572130167e-06,
      "loss": 0.78075852,
      "memory(GiB)": 147.13,
      "step": 42690,
      "train_speed(iter/s)": 0.201116
    },
    {
      "acc": 0.78155584,
      "epoch": 0.9962233336735735,
      "grad_norm": 5.125,
      "learning_rate": 5.273218302523925e-06,
      "loss": 0.78047514,
      "memory(GiB)": 147.13,
      "step": 42700,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.78246479,
      "epoch": 0.9964566412458624,
      "grad_norm": 5.4375,
      "learning_rate": 5.2713319939158494e-06,
      "loss": 0.79133334,
      "memory(GiB)": 147.13,
      "step": 42710,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.78450847,
      "epoch": 0.9966899488181513,
      "grad_norm": 4.625,
      "learning_rate": 5.2694456465752104e-06,
      "loss": 0.76527557,
      "memory(GiB)": 147.13,
      "step": 42720,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.77672338,
      "epoch": 0.9969232563904402,
      "grad_norm": 5.40625,
      "learning_rate": 5.267559260771285e-06,
      "loss": 0.78869748,
      "memory(GiB)": 147.13,
      "step": 42730,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.79831858,
      "epoch": 0.9971565639627291,
      "grad_norm": 5.4375,
      "learning_rate": 5.265672836773353e-06,
      "loss": 0.73061762,
      "memory(GiB)": 147.13,
      "step": 42740,
      "train_speed(iter/s)": 0.201234
    },
    {
      "acc": 0.76231604,
      "epoch": 0.997389871535018,
      "grad_norm": 6.9375,
      "learning_rate": 5.2637863748507e-06,
      "loss": 0.87828217,
      "memory(GiB)": 147.13,
      "step": 42750,
      "train_speed(iter/s)": 0.201257
    },
    {
      "acc": 0.77685285,
      "epoch": 0.9976231791073069,
      "grad_norm": 4.75,
      "learning_rate": 5.261899875272619e-06,
      "loss": 0.79581809,
      "memory(GiB)": 147.13,
      "step": 42760,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.77012229,
      "epoch": 0.9978564866795958,
      "grad_norm": 5.59375,
      "learning_rate": 5.260013338308408e-06,
      "loss": 0.84313641,
      "memory(GiB)": 147.13,
      "step": 42770,
      "train_speed(iter/s)": 0.201305
    },
    {
      "acc": 0.7849412,
      "epoch": 0.9980897942518847,
      "grad_norm": 3.953125,
      "learning_rate": 5.258126764227366e-06,
      "loss": 0.76588078,
      "memory(GiB)": 147.13,
      "step": 42780,
      "train_speed(iter/s)": 0.201328
    },
    {
      "acc": 0.77974372,
      "epoch": 0.9983231018241736,
      "grad_norm": 6.03125,
      "learning_rate": 5.256240153298804e-06,
      "loss": 0.78483295,
      "memory(GiB)": 147.13,
      "step": 42790,
      "train_speed(iter/s)": 0.201355
    },
    {
      "acc": 0.78103266,
      "epoch": 0.9985564093964625,
      "grad_norm": 5.21875,
      "learning_rate": 5.254353505792036e-06,
      "loss": 0.78471365,
      "memory(GiB)": 147.13,
      "step": 42800,
      "train_speed(iter/s)": 0.201379
    },
    {
      "acc": 0.76989121,
      "epoch": 0.9987897169687514,
      "grad_norm": 6.0625,
      "learning_rate": 5.252466821976377e-06,
      "loss": 0.82569427,
      "memory(GiB)": 147.13,
      "step": 42810,
      "train_speed(iter/s)": 0.201404
    },
    {
      "acc": 0.77489061,
      "epoch": 0.9990230245410403,
      "grad_norm": 4.59375,
      "learning_rate": 5.250580102121153e-06,
      "loss": 0.81538868,
      "memory(GiB)": 147.13,
      "step": 42820,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.78204842,
      "epoch": 0.9992563321133292,
      "grad_norm": 4.625,
      "learning_rate": 5.248693346495694e-06,
      "loss": 0.78283563,
      "memory(GiB)": 147.13,
      "step": 42830,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.79220905,
      "epoch": 0.999489639685618,
      "grad_norm": 4.78125,
      "learning_rate": 5.2468065553693306e-06,
      "loss": 0.73160219,
      "memory(GiB)": 147.13,
      "step": 42840,
      "train_speed(iter/s)": 0.201476
    },
    {
      "acc": 0.77306366,
      "epoch": 0.9997229472579069,
      "grad_norm": 5.90625,
      "learning_rate": 5.244919729011403e-06,
      "loss": 0.8217906,
      "memory(GiB)": 147.13,
      "step": 42850,
      "train_speed(iter/s)": 0.201501
    },
    {
      "acc": 0.78262877,
      "epoch": 0.9999562548301958,
      "grad_norm": 6.875,
      "learning_rate": 5.243032867691257e-06,
      "loss": 0.78975267,
      "memory(GiB)": 147.13,
      "step": 42860,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.78576965,
      "epoch": 1.0001895624024848,
      "grad_norm": 3.8125,
      "learning_rate": 5.241145971678238e-06,
      "loss": 0.76854234,
      "memory(GiB)": 147.13,
      "step": 42870,
      "train_speed(iter/s)": 0.201543
    },
    {
      "acc": 0.7733151,
      "epoch": 1.0004228699747737,
      "grad_norm": 4.6875,
      "learning_rate": 5.239259041241701e-06,
      "loss": 0.82483797,
      "memory(GiB)": 147.13,
      "step": 42880,
      "train_speed(iter/s)": 0.201568
    },
    {
      "acc": 0.80650272,
      "epoch": 1.0006561775470626,
      "grad_norm": 4.28125,
      "learning_rate": 5.237372076651006e-06,
      "loss": 0.67205696,
      "memory(GiB)": 147.13,
      "step": 42890,
      "train_speed(iter/s)": 0.20159
    },
    {
      "acc": 0.78537331,
      "epoch": 1.0008894851193515,
      "grad_norm": 6.03125,
      "learning_rate": 5.2354850781755175e-06,
      "loss": 0.76955976,
      "memory(GiB)": 147.13,
      "step": 42900,
      "train_speed(iter/s)": 0.201612
    },
    {
      "acc": 0.76812172,
      "epoch": 1.0011227926916404,
      "grad_norm": 3.875,
      "learning_rate": 5.233598046084602e-06,
      "loss": 0.85648232,
      "memory(GiB)": 147.13,
      "step": 42910,
      "train_speed(iter/s)": 0.201637
    },
    {
      "acc": 0.77023239,
      "epoch": 1.0013561002639293,
      "grad_norm": 5.125,
      "learning_rate": 5.231710980647632e-06,
      "loss": 0.8251276,
      "memory(GiB)": 147.13,
      "step": 42920,
      "train_speed(iter/s)": 0.201662
    },
    {
      "acc": 0.78217449,
      "epoch": 1.0015894078362182,
      "grad_norm": 5.625,
      "learning_rate": 5.229823882133987e-06,
      "loss": 0.78995705,
      "memory(GiB)": 147.13,
      "step": 42930,
      "train_speed(iter/s)": 0.201687
    },
    {
      "acc": 0.77538042,
      "epoch": 1.0018227154085069,
      "grad_norm": 6.96875,
      "learning_rate": 5.22793675081305e-06,
      "loss": 0.78225121,
      "memory(GiB)": 147.13,
      "step": 42940,
      "train_speed(iter/s)": 0.201711
    },
    {
      "acc": 0.77648048,
      "epoch": 1.0020560229807958,
      "grad_norm": 6.875,
      "learning_rate": 5.226049586954207e-06,
      "loss": 0.81052818,
      "memory(GiB)": 147.13,
      "step": 42950,
      "train_speed(iter/s)": 0.201735
    },
    {
      "acc": 0.78938169,
      "epoch": 1.0022893305530847,
      "grad_norm": 4.78125,
      "learning_rate": 5.2241623908268524e-06,
      "loss": 0.77443514,
      "memory(GiB)": 147.13,
      "step": 42960,
      "train_speed(iter/s)": 0.201759
    },
    {
      "acc": 0.76893907,
      "epoch": 1.0025226381253736,
      "grad_norm": 4.75,
      "learning_rate": 5.222275162700382e-06,
      "loss": 0.84441872,
      "memory(GiB)": 147.13,
      "step": 42970,
      "train_speed(iter/s)": 0.201781
    },
    {
      "acc": 0.77256842,
      "epoch": 1.0027559456976625,
      "grad_norm": 8.75,
      "learning_rate": 5.2203879028441975e-06,
      "loss": 0.83643208,
      "memory(GiB)": 147.13,
      "step": 42980,
      "train_speed(iter/s)": 0.201807
    },
    {
      "acc": 0.76644096,
      "epoch": 1.0029892532699514,
      "grad_norm": 5.375,
      "learning_rate": 5.218500611527701e-06,
      "loss": 0.82999525,
      "memory(GiB)": 147.13,
      "step": 42990,
      "train_speed(iter/s)": 0.201831
    },
    {
      "acc": 0.77953863,
      "epoch": 1.0032225608422403,
      "grad_norm": 5.1875,
      "learning_rate": 5.216613289020307e-06,
      "loss": 0.77615538,
      "memory(GiB)": 147.13,
      "step": 43000,
      "train_speed(iter/s)": 0.201856
    },
    {
      "epoch": 1.0032225608422403,
      "eval_acc": 0.7438829307227262,
      "eval_loss": 0.8073763847351074,
      "eval_runtime": 1269.4805,
      "eval_samples_per_second": 28.351,
      "eval_steps_per_second": 14.176,
      "step": 43000
    },
    {
      "acc": 0.77727814,
      "epoch": 1.0034558684145292,
      "grad_norm": 4.84375,
      "learning_rate": 5.214725935591429e-06,
      "loss": 0.80646038,
      "memory(GiB)": 147.13,
      "step": 43010,
      "train_speed(iter/s)": 0.200669
    },
    {
      "acc": 0.7740561,
      "epoch": 1.003689175986818,
      "grad_norm": 6.75,
      "learning_rate": 5.2128385515104865e-06,
      "loss": 0.81611605,
      "memory(GiB)": 147.13,
      "step": 43020,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.77674236,
      "epoch": 1.003922483559107,
      "grad_norm": 6.1875,
      "learning_rate": 5.210951137046903e-06,
      "loss": 0.80263958,
      "memory(GiB)": 147.13,
      "step": 43030,
      "train_speed(iter/s)": 0.200718
    },
    {
      "acc": 0.77732944,
      "epoch": 1.0041557911313959,
      "grad_norm": 6.78125,
      "learning_rate": 5.209063692470104e-06,
      "loss": 0.80889053,
      "memory(GiB)": 147.13,
      "step": 43040,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.75513368,
      "epoch": 1.0043890987036848,
      "grad_norm": 5.40625,
      "learning_rate": 5.207176218049526e-06,
      "loss": 0.88391256,
      "memory(GiB)": 147.13,
      "step": 43050,
      "train_speed(iter/s)": 0.200767
    },
    {
      "acc": 0.77179451,
      "epoch": 1.0046224062759737,
      "grad_norm": 4.71875,
      "learning_rate": 5.205288714054602e-06,
      "loss": 0.80851946,
      "memory(GiB)": 147.13,
      "step": 43060,
      "train_speed(iter/s)": 0.200791
    },
    {
      "acc": 0.77560835,
      "epoch": 1.0048557138482626,
      "grad_norm": 5.1875,
      "learning_rate": 5.203401180754772e-06,
      "loss": 0.79950652,
      "memory(GiB)": 147.13,
      "step": 43070,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.76174564,
      "epoch": 1.0050890214205515,
      "grad_norm": 4.53125,
      "learning_rate": 5.201513618419486e-06,
      "loss": 0.84228649,
      "memory(GiB)": 147.13,
      "step": 43080,
      "train_speed(iter/s)": 0.200839
    },
    {
      "acc": 0.78287921,
      "epoch": 1.0053223289928404,
      "grad_norm": 5.9375,
      "learning_rate": 5.199626027318188e-06,
      "loss": 0.78160782,
      "memory(GiB)": 147.13,
      "step": 43090,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.78439932,
      "epoch": 1.0055556365651293,
      "grad_norm": 5.09375,
      "learning_rate": 5.197738407720331e-06,
      "loss": 0.76160207,
      "memory(GiB)": 147.13,
      "step": 43100,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.77529421,
      "epoch": 1.0057889441374182,
      "grad_norm": 5.96875,
      "learning_rate": 5.195850759895374e-06,
      "loss": 0.8274806,
      "memory(GiB)": 147.13,
      "step": 43110,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.79326754,
      "epoch": 1.006022251709707,
      "grad_norm": 5.96875,
      "learning_rate": 5.193963084112781e-06,
      "loss": 0.73483434,
      "memory(GiB)": 147.13,
      "step": 43120,
      "train_speed(iter/s)": 0.200935
    },
    {
      "acc": 0.78251967,
      "epoch": 1.006255559281996,
      "grad_norm": 4.96875,
      "learning_rate": 5.192075380642011e-06,
      "loss": 0.76476879,
      "memory(GiB)": 147.13,
      "step": 43130,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.78560572,
      "epoch": 1.0064888668542848,
      "grad_norm": 6.3125,
      "learning_rate": 5.190187649752538e-06,
      "loss": 0.76422467,
      "memory(GiB)": 147.13,
      "step": 43140,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.77720866,
      "epoch": 1.0067221744265737,
      "grad_norm": 6.53125,
      "learning_rate": 5.1882998917138324e-06,
      "loss": 0.79875774,
      "memory(GiB)": 147.13,
      "step": 43150,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.78121185,
      "epoch": 1.0069554819988626,
      "grad_norm": 5.28125,
      "learning_rate": 5.186412106795371e-06,
      "loss": 0.78524699,
      "memory(GiB)": 147.13,
      "step": 43160,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.7672533,
      "epoch": 1.0071887895711515,
      "grad_norm": 5.0,
      "learning_rate": 5.1845242952666365e-06,
      "loss": 0.84109287,
      "memory(GiB)": 147.13,
      "step": 43170,
      "train_speed(iter/s)": 0.201057
    },
    {
      "acc": 0.79040389,
      "epoch": 1.0074220971434404,
      "grad_norm": 4.96875,
      "learning_rate": 5.1826364573971125e-06,
      "loss": 0.75233536,
      "memory(GiB)": 147.13,
      "step": 43180,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.77585616,
      "epoch": 1.0076554047157293,
      "grad_norm": 4.53125,
      "learning_rate": 5.180748593456289e-06,
      "loss": 0.8176096,
      "memory(GiB)": 147.13,
      "step": 43190,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.76432118,
      "epoch": 1.0078887122880182,
      "grad_norm": 4.84375,
      "learning_rate": 5.178860703713654e-06,
      "loss": 0.84310722,
      "memory(GiB)": 147.13,
      "step": 43200,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.78325276,
      "epoch": 1.0081220198603071,
      "grad_norm": 5.34375,
      "learning_rate": 5.176972788438705e-06,
      "loss": 0.7665885,
      "memory(GiB)": 147.13,
      "step": 43210,
      "train_speed(iter/s)": 0.201154
    },
    {
      "acc": 0.77819371,
      "epoch": 1.008355327432596,
      "grad_norm": 6.0,
      "learning_rate": 5.175084847900943e-06,
      "loss": 0.79968734,
      "memory(GiB)": 147.13,
      "step": 43220,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.78666801,
      "epoch": 1.008588635004885,
      "grad_norm": 4.5625,
      "learning_rate": 5.17319688236987e-06,
      "loss": 0.77615905,
      "memory(GiB)": 147.13,
      "step": 43230,
      "train_speed(iter/s)": 0.201203
    },
    {
      "acc": 0.78133011,
      "epoch": 1.0088219425771738,
      "grad_norm": 5.28125,
      "learning_rate": 5.171308892114991e-06,
      "loss": 0.80085545,
      "memory(GiB)": 147.13,
      "step": 43240,
      "train_speed(iter/s)": 0.201226
    },
    {
      "acc": 0.79328427,
      "epoch": 1.0090552501494627,
      "grad_norm": 4.25,
      "learning_rate": 5.16942087740582e-06,
      "loss": 0.73711023,
      "memory(GiB)": 147.13,
      "step": 43250,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.78356128,
      "epoch": 1.0092885577217516,
      "grad_norm": 4.71875,
      "learning_rate": 5.167532838511866e-06,
      "loss": 0.76406059,
      "memory(GiB)": 147.13,
      "step": 43260,
      "train_speed(iter/s)": 0.201274
    },
    {
      "acc": 0.76661291,
      "epoch": 1.0095218652940405,
      "grad_norm": 4.9375,
      "learning_rate": 5.16564477570265e-06,
      "loss": 0.81206608,
      "memory(GiB)": 147.13,
      "step": 43270,
      "train_speed(iter/s)": 0.201297
    },
    {
      "acc": 0.78339691,
      "epoch": 1.0097551728663294,
      "grad_norm": 5.65625,
      "learning_rate": 5.163756689247687e-06,
      "loss": 0.76879001,
      "memory(GiB)": 147.13,
      "step": 43280,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.78117151,
      "epoch": 1.0099884804386183,
      "grad_norm": 4.625,
      "learning_rate": 5.1618685794165066e-06,
      "loss": 0.78589344,
      "memory(GiB)": 147.13,
      "step": 43290,
      "train_speed(iter/s)": 0.201347
    },
    {
      "acc": 0.75611968,
      "epoch": 1.0102217880109072,
      "grad_norm": 5.75,
      "learning_rate": 5.159980446478633e-06,
      "loss": 0.88358593,
      "memory(GiB)": 147.13,
      "step": 43300,
      "train_speed(iter/s)": 0.201372
    },
    {
      "acc": 0.77154584,
      "epoch": 1.0104550955831961,
      "grad_norm": 5.375,
      "learning_rate": 5.158092290703597e-06,
      "loss": 0.81572247,
      "memory(GiB)": 147.13,
      "step": 43310,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.76835303,
      "epoch": 1.010688403155485,
      "grad_norm": 5.9375,
      "learning_rate": 5.156204112360933e-06,
      "loss": 0.80885944,
      "memory(GiB)": 147.13,
      "step": 43320,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.77416339,
      "epoch": 1.0109217107277737,
      "grad_norm": 5.03125,
      "learning_rate": 5.154315911720176e-06,
      "loss": 0.81178932,
      "memory(GiB)": 147.13,
      "step": 43330,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.78630424,
      "epoch": 1.0111550183000626,
      "grad_norm": 6.625,
      "learning_rate": 5.152427689050869e-06,
      "loss": 0.77561946,
      "memory(GiB)": 147.13,
      "step": 43340,
      "train_speed(iter/s)": 0.201467
    },
    {
      "acc": 0.7755363,
      "epoch": 1.0113883258723515,
      "grad_norm": 3.859375,
      "learning_rate": 5.150539444622552e-06,
      "loss": 0.80117035,
      "memory(GiB)": 147.13,
      "step": 43350,
      "train_speed(iter/s)": 0.201491
    },
    {
      "acc": 0.78446379,
      "epoch": 1.0116216334446404,
      "grad_norm": 8.1875,
      "learning_rate": 5.148651178704775e-06,
      "loss": 0.82833748,
      "memory(GiB)": 147.13,
      "step": 43360,
      "train_speed(iter/s)": 0.201516
    },
    {
      "acc": 0.76374173,
      "epoch": 1.0118549410169293,
      "grad_norm": 7.96875,
      "learning_rate": 5.146762891567084e-06,
      "loss": 0.88345985,
      "memory(GiB)": 147.13,
      "step": 43370,
      "train_speed(iter/s)": 0.20154
    },
    {
      "acc": 0.79155569,
      "epoch": 1.0120882485892182,
      "grad_norm": 5.8125,
      "learning_rate": 5.144874583479034e-06,
      "loss": 0.74518204,
      "memory(GiB)": 147.13,
      "step": 43380,
      "train_speed(iter/s)": 0.201564
    },
    {
      "acc": 0.77843351,
      "epoch": 1.012321556161507,
      "grad_norm": 5.59375,
      "learning_rate": 5.142986254710177e-06,
      "loss": 0.78075495,
      "memory(GiB)": 147.13,
      "step": 43390,
      "train_speed(iter/s)": 0.201587
    },
    {
      "acc": 0.77403011,
      "epoch": 1.012554863733796,
      "grad_norm": 5.9375,
      "learning_rate": 5.141097905530077e-06,
      "loss": 0.80180321,
      "memory(GiB)": 147.13,
      "step": 43400,
      "train_speed(iter/s)": 0.201608
    },
    {
      "acc": 0.79151011,
      "epoch": 1.012788171306085,
      "grad_norm": 5.53125,
      "learning_rate": 5.139209536208289e-06,
      "loss": 0.74089231,
      "memory(GiB)": 147.13,
      "step": 43410,
      "train_speed(iter/s)": 0.20163
    },
    {
      "acc": 0.76592178,
      "epoch": 1.0130214788783738,
      "grad_norm": 8.125,
      "learning_rate": 5.1373211470143814e-06,
      "loss": 0.83816414,
      "memory(GiB)": 147.13,
      "step": 43420,
      "train_speed(iter/s)": 0.201653
    },
    {
      "acc": 0.76648312,
      "epoch": 1.0132547864506627,
      "grad_norm": 5.625,
      "learning_rate": 5.13543273821792e-06,
      "loss": 0.85734253,
      "memory(GiB)": 147.13,
      "step": 43430,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.76693745,
      "epoch": 1.0134880940229516,
      "grad_norm": 4.875,
      "learning_rate": 5.133544310088474e-06,
      "loss": 0.84319305,
      "memory(GiB)": 147.13,
      "step": 43440,
      "train_speed(iter/s)": 0.2017
    },
    {
      "acc": 0.78325624,
      "epoch": 1.0137214015952405,
      "grad_norm": 5.4375,
      "learning_rate": 5.131655862895617e-06,
      "loss": 0.77048082,
      "memory(GiB)": 147.13,
      "step": 43450,
      "train_speed(iter/s)": 0.201723
    },
    {
      "acc": 0.79426832,
      "epoch": 1.0139547091675294,
      "grad_norm": 4.09375,
      "learning_rate": 5.129767396908923e-06,
      "loss": 0.73330755,
      "memory(GiB)": 147.13,
      "step": 43460,
      "train_speed(iter/s)": 0.201747
    },
    {
      "acc": 0.78500352,
      "epoch": 1.0141880167398183,
      "grad_norm": 4.5625,
      "learning_rate": 5.1278789123979736e-06,
      "loss": 0.76361828,
      "memory(GiB)": 147.13,
      "step": 43470,
      "train_speed(iter/s)": 0.201772
    },
    {
      "acc": 0.77640676,
      "epoch": 1.0144213243121072,
      "grad_norm": 5.6875,
      "learning_rate": 5.125990409632344e-06,
      "loss": 0.82779522,
      "memory(GiB)": 147.13,
      "step": 43480,
      "train_speed(iter/s)": 0.201796
    },
    {
      "acc": 0.76425214,
      "epoch": 1.014654631884396,
      "grad_norm": 4.3125,
      "learning_rate": 5.1241018888816205e-06,
      "loss": 0.862677,
      "memory(GiB)": 147.13,
      "step": 43490,
      "train_speed(iter/s)": 0.20182
    },
    {
      "acc": 0.78395214,
      "epoch": 1.014887939456685,
      "grad_norm": 6.28125,
      "learning_rate": 5.122213350415389e-06,
      "loss": 0.76862893,
      "memory(GiB)": 147.13,
      "step": 43500,
      "train_speed(iter/s)": 0.201844
    },
    {
      "epoch": 1.014887939456685,
      "eval_acc": 0.743926970386609,
      "eval_loss": 0.8070799708366394,
      "eval_runtime": 1270.2738,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 43500
    },
    {
      "acc": 0.77859073,
      "epoch": 1.0151212470289739,
      "grad_norm": 5.625,
      "learning_rate": 5.1203247945032365e-06,
      "loss": 0.78678293,
      "memory(GiB)": 147.13,
      "step": 43510,
      "train_speed(iter/s)": 0.200669
    },
    {
      "acc": 0.77705793,
      "epoch": 1.0153545546012628,
      "grad_norm": 6.6875,
      "learning_rate": 5.118436221414754e-06,
      "loss": 0.80260468,
      "memory(GiB)": 147.13,
      "step": 43520,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.76431408,
      "epoch": 1.0155878621735517,
      "grad_norm": 7.75,
      "learning_rate": 5.116547631419536e-06,
      "loss": 0.85512829,
      "memory(GiB)": 147.13,
      "step": 43530,
      "train_speed(iter/s)": 0.200717
    },
    {
      "acc": 0.7825469,
      "epoch": 1.0158211697458406,
      "grad_norm": 5.0625,
      "learning_rate": 5.114659024787179e-06,
      "loss": 0.76410046,
      "memory(GiB)": 147.13,
      "step": 43540,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.76927857,
      "epoch": 1.0160544773181295,
      "grad_norm": 5.875,
      "learning_rate": 5.112770401787278e-06,
      "loss": 0.81999454,
      "memory(GiB)": 147.13,
      "step": 43550,
      "train_speed(iter/s)": 0.200765
    },
    {
      "acc": 0.76841311,
      "epoch": 1.0162877848904184,
      "grad_norm": 5.0625,
      "learning_rate": 5.110881762689435e-06,
      "loss": 0.84414501,
      "memory(GiB)": 147.13,
      "step": 43560,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.77150702,
      "epoch": 1.0165210924627073,
      "grad_norm": 4.28125,
      "learning_rate": 5.1089931077632514e-06,
      "loss": 0.81745987,
      "memory(GiB)": 147.13,
      "step": 43570,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.7604578,
      "epoch": 1.0167544000349962,
      "grad_norm": 7.5,
      "learning_rate": 5.1071044372783355e-06,
      "loss": 0.86848364,
      "memory(GiB)": 147.13,
      "step": 43580,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.77862034,
      "epoch": 1.016987707607285,
      "grad_norm": 6.6875,
      "learning_rate": 5.10521575150429e-06,
      "loss": 0.80064793,
      "memory(GiB)": 147.13,
      "step": 43590,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.77443581,
      "epoch": 1.017221015179574,
      "grad_norm": 5.75,
      "learning_rate": 5.103327050710729e-06,
      "loss": 0.83215542,
      "memory(GiB)": 147.13,
      "step": 43600,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.7929235,
      "epoch": 1.0174543227518629,
      "grad_norm": 6.40625,
      "learning_rate": 5.10143833516726e-06,
      "loss": 0.74127417,
      "memory(GiB)": 147.13,
      "step": 43610,
      "train_speed(iter/s)": 0.200907
    },
    {
      "acc": 0.79245968,
      "epoch": 1.0176876303241518,
      "grad_norm": 6.25,
      "learning_rate": 5.099549605143499e-06,
      "loss": 0.75981627,
      "memory(GiB)": 147.13,
      "step": 43620,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.79105101,
      "epoch": 1.0179209378964407,
      "grad_norm": 5.4375,
      "learning_rate": 5.0976608609090606e-06,
      "loss": 0.74999475,
      "memory(GiB)": 147.13,
      "step": 43630,
      "train_speed(iter/s)": 0.200955
    },
    {
      "acc": 0.77970495,
      "epoch": 1.0181542454687296,
      "grad_norm": 6.1875,
      "learning_rate": 5.095772102733561e-06,
      "loss": 0.77877584,
      "memory(GiB)": 147.13,
      "step": 43640,
      "train_speed(iter/s)": 0.200979
    },
    {
      "acc": 0.78451672,
      "epoch": 1.0183875530410185,
      "grad_norm": 5.4375,
      "learning_rate": 5.093883330886623e-06,
      "loss": 0.75853043,
      "memory(GiB)": 147.13,
      "step": 43650,
      "train_speed(iter/s)": 0.201002
    },
    {
      "acc": 0.79117913,
      "epoch": 1.0186208606133074,
      "grad_norm": 7.15625,
      "learning_rate": 5.091994545637867e-06,
      "loss": 0.74103169,
      "memory(GiB)": 147.13,
      "step": 43660,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.79023709,
      "epoch": 1.0188541681855963,
      "grad_norm": 5.75,
      "learning_rate": 5.090105747256916e-06,
      "loss": 0.7264926,
      "memory(GiB)": 147.13,
      "step": 43670,
      "train_speed(iter/s)": 0.201049
    },
    {
      "acc": 0.78450155,
      "epoch": 1.0190874757578852,
      "grad_norm": 4.6875,
      "learning_rate": 5.088216936013396e-06,
      "loss": 0.77232718,
      "memory(GiB)": 147.13,
      "step": 43680,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.77247133,
      "epoch": 1.019320783330174,
      "grad_norm": 5.84375,
      "learning_rate": 5.086328112176934e-06,
      "loss": 0.82542839,
      "memory(GiB)": 147.13,
      "step": 43690,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.77573242,
      "epoch": 1.019554090902463,
      "grad_norm": 5.3125,
      "learning_rate": 5.084439276017159e-06,
      "loss": 0.81169853,
      "memory(GiB)": 147.13,
      "step": 43700,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.79919705,
      "epoch": 1.0197873984747516,
      "grad_norm": 4.34375,
      "learning_rate": 5.082550427803702e-06,
      "loss": 0.72000685,
      "memory(GiB)": 147.13,
      "step": 43710,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.79065461,
      "epoch": 1.0200207060470405,
      "grad_norm": 10.6875,
      "learning_rate": 5.080661567806195e-06,
      "loss": 0.746523,
      "memory(GiB)": 147.13,
      "step": 43720,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.77126436,
      "epoch": 1.0202540136193294,
      "grad_norm": 5.0625,
      "learning_rate": 5.078772696294273e-06,
      "loss": 0.81962471,
      "memory(GiB)": 147.13,
      "step": 43730,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.78477058,
      "epoch": 1.0204873211916183,
      "grad_norm": 6.4375,
      "learning_rate": 5.076883813537571e-06,
      "loss": 0.77737856,
      "memory(GiB)": 147.13,
      "step": 43740,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.78322659,
      "epoch": 1.0207206287639072,
      "grad_norm": 6.875,
      "learning_rate": 5.074994919805727e-06,
      "loss": 0.75895319,
      "memory(GiB)": 147.13,
      "step": 43750,
      "train_speed(iter/s)": 0.201235
    },
    {
      "acc": 0.77741756,
      "epoch": 1.0209539363361961,
      "grad_norm": 5.28125,
      "learning_rate": 5.073106015368381e-06,
      "loss": 0.81184292,
      "memory(GiB)": 147.13,
      "step": 43760,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.78123074,
      "epoch": 1.021187243908485,
      "grad_norm": 5.65625,
      "learning_rate": 5.071217100495172e-06,
      "loss": 0.78832426,
      "memory(GiB)": 147.13,
      "step": 43770,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.76711731,
      "epoch": 1.021420551480774,
      "grad_norm": 7.5,
      "learning_rate": 5.069328175455742e-06,
      "loss": 0.82933445,
      "memory(GiB)": 147.13,
      "step": 43780,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.7838244,
      "epoch": 1.0216538590530628,
      "grad_norm": 5.3125,
      "learning_rate": 5.067439240519735e-06,
      "loss": 0.78824844,
      "memory(GiB)": 147.13,
      "step": 43790,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.77966681,
      "epoch": 1.0218871666253517,
      "grad_norm": 4.8125,
      "learning_rate": 5.065550295956796e-06,
      "loss": 0.78713975,
      "memory(GiB)": 147.13,
      "step": 43800,
      "train_speed(iter/s)": 0.201355
    },
    {
      "acc": 0.78061857,
      "epoch": 1.0221204741976406,
      "grad_norm": 5.25,
      "learning_rate": 5.063661342036571e-06,
      "loss": 0.76254168,
      "memory(GiB)": 147.13,
      "step": 43810,
      "train_speed(iter/s)": 0.201379
    },
    {
      "acc": 0.76380448,
      "epoch": 1.0223537817699295,
      "grad_norm": 5.84375,
      "learning_rate": 5.061772379028709e-06,
      "loss": 0.87199421,
      "memory(GiB)": 147.13,
      "step": 43820,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.79298944,
      "epoch": 1.0225870893422184,
      "grad_norm": 8.4375,
      "learning_rate": 5.059883407202858e-06,
      "loss": 0.7629261,
      "memory(GiB)": 147.13,
      "step": 43830,
      "train_speed(iter/s)": 0.201426
    },
    {
      "acc": 0.78520899,
      "epoch": 1.0228203969145073,
      "grad_norm": 7.0625,
      "learning_rate": 5.057994426828669e-06,
      "loss": 0.76913924,
      "memory(GiB)": 147.13,
      "step": 43840,
      "train_speed(iter/s)": 0.20145
    },
    {
      "acc": 0.76806917,
      "epoch": 1.0230537044867962,
      "grad_norm": 6.25,
      "learning_rate": 5.05610543817579e-06,
      "loss": 0.8473484,
      "memory(GiB)": 147.13,
      "step": 43850,
      "train_speed(iter/s)": 0.201474
    },
    {
      "acc": 0.79321308,
      "epoch": 1.0232870120590851,
      "grad_norm": 3.625,
      "learning_rate": 5.054216441513876e-06,
      "loss": 0.7350997,
      "memory(GiB)": 147.13,
      "step": 43860,
      "train_speed(iter/s)": 0.201496
    },
    {
      "acc": 0.77993116,
      "epoch": 1.023520319631374,
      "grad_norm": 4.59375,
      "learning_rate": 5.052327437112582e-06,
      "loss": 0.77887244,
      "memory(GiB)": 147.13,
      "step": 43870,
      "train_speed(iter/s)": 0.20152
    },
    {
      "acc": 0.7738163,
      "epoch": 1.023753627203663,
      "grad_norm": 6.71875,
      "learning_rate": 5.050438425241562e-06,
      "loss": 0.82020864,
      "memory(GiB)": 147.13,
      "step": 43880,
      "train_speed(iter/s)": 0.201543
    },
    {
      "acc": 0.78634443,
      "epoch": 1.0239869347759518,
      "grad_norm": 5.8125,
      "learning_rate": 5.0485494061704695e-06,
      "loss": 0.76208029,
      "memory(GiB)": 147.13,
      "step": 43890,
      "train_speed(iter/s)": 0.201566
    },
    {
      "acc": 0.76844702,
      "epoch": 1.0242202423482407,
      "grad_norm": 6.96875,
      "learning_rate": 5.0466603801689655e-06,
      "loss": 0.82169828,
      "memory(GiB)": 147.13,
      "step": 43900,
      "train_speed(iter/s)": 0.201589
    },
    {
      "acc": 0.78849974,
      "epoch": 1.0244535499205296,
      "grad_norm": 4.34375,
      "learning_rate": 5.044771347506705e-06,
      "loss": 0.78166656,
      "memory(GiB)": 147.13,
      "step": 43910,
      "train_speed(iter/s)": 0.201611
    },
    {
      "acc": 0.79453068,
      "epoch": 1.0246868574928185,
      "grad_norm": 4.71875,
      "learning_rate": 5.0428823084533475e-06,
      "loss": 0.72507105,
      "memory(GiB)": 147.13,
      "step": 43920,
      "train_speed(iter/s)": 0.201635
    },
    {
      "acc": 0.77169046,
      "epoch": 1.0249201650651074,
      "grad_norm": 5.9375,
      "learning_rate": 5.040993263278552e-06,
      "loss": 0.83655701,
      "memory(GiB)": 147.13,
      "step": 43930,
      "train_speed(iter/s)": 0.201657
    },
    {
      "acc": 0.77949696,
      "epoch": 1.0251534726373963,
      "grad_norm": 6.3125,
      "learning_rate": 5.0391042122519815e-06,
      "loss": 0.81380482,
      "memory(GiB)": 147.13,
      "step": 43940,
      "train_speed(iter/s)": 0.201681
    },
    {
      "acc": 0.79480991,
      "epoch": 1.0253867802096852,
      "grad_norm": 5.375,
      "learning_rate": 5.037215155643296e-06,
      "loss": 0.74450397,
      "memory(GiB)": 147.13,
      "step": 43950,
      "train_speed(iter/s)": 0.201704
    },
    {
      "acc": 0.79364281,
      "epoch": 1.025620087781974,
      "grad_norm": 5.71875,
      "learning_rate": 5.035326093722157e-06,
      "loss": 0.73111277,
      "memory(GiB)": 147.13,
      "step": 43960,
      "train_speed(iter/s)": 0.201728
    },
    {
      "acc": 0.77582326,
      "epoch": 1.025853395354263,
      "grad_norm": 6.65625,
      "learning_rate": 5.033437026758228e-06,
      "loss": 0.83863983,
      "memory(GiB)": 147.13,
      "step": 43970,
      "train_speed(iter/s)": 0.201752
    },
    {
      "acc": 0.78823829,
      "epoch": 1.026086702926552,
      "grad_norm": 6.75,
      "learning_rate": 5.0315479550211746e-06,
      "loss": 0.75079851,
      "memory(GiB)": 147.13,
      "step": 43980,
      "train_speed(iter/s)": 0.201776
    },
    {
      "acc": 0.7844943,
      "epoch": 1.0263200104988408,
      "grad_norm": 6.3125,
      "learning_rate": 5.029658878780659e-06,
      "loss": 0.77733545,
      "memory(GiB)": 147.13,
      "step": 43990,
      "train_speed(iter/s)": 0.201798
    },
    {
      "acc": 0.78340597,
      "epoch": 1.0265533180711297,
      "grad_norm": 4.6875,
      "learning_rate": 5.0277697983063476e-06,
      "loss": 0.76800084,
      "memory(GiB)": 147.13,
      "step": 44000,
      "train_speed(iter/s)": 0.201823
    },
    {
      "epoch": 1.0265533180711297,
      "eval_acc": 0.7438331258664804,
      "eval_loss": 0.8070977330207825,
      "eval_runtime": 1271.2669,
      "eval_samples_per_second": 28.311,
      "eval_steps_per_second": 14.156,
      "step": 44000
    },
    {
      "acc": 0.77107697,
      "epoch": 1.0267866256434186,
      "grad_norm": 7.34375,
      "learning_rate": 5.025880713867904e-06,
      "loss": 0.85529652,
      "memory(GiB)": 147.13,
      "step": 44010,
      "train_speed(iter/s)": 0.20066
    },
    {
      "acc": 0.78703861,
      "epoch": 1.0270199332157075,
      "grad_norm": 5.375,
      "learning_rate": 5.023991625734998e-06,
      "loss": 0.76970444,
      "memory(GiB)": 147.13,
      "step": 44020,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.76558523,
      "epoch": 1.0272532407879964,
      "grad_norm": 4.15625,
      "learning_rate": 5.022102534177293e-06,
      "loss": 0.83571815,
      "memory(GiB)": 147.13,
      "step": 44030,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.77779908,
      "epoch": 1.0274865483602853,
      "grad_norm": 5.03125,
      "learning_rate": 5.020213439464458e-06,
      "loss": 0.78558512,
      "memory(GiB)": 147.13,
      "step": 44040,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.79692812,
      "epoch": 1.0277198559325742,
      "grad_norm": 6.3125,
      "learning_rate": 5.018324341866161e-06,
      "loss": 0.71697149,
      "memory(GiB)": 147.13,
      "step": 44050,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.80139132,
      "epoch": 1.027953163504863,
      "grad_norm": 5.9375,
      "learning_rate": 5.01643524165207e-06,
      "loss": 0.68105035,
      "memory(GiB)": 147.13,
      "step": 44060,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.77024827,
      "epoch": 1.028186471077152,
      "grad_norm": 6.34375,
      "learning_rate": 5.014546139091851e-06,
      "loss": 0.83891077,
      "memory(GiB)": 147.13,
      "step": 44070,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.78477793,
      "epoch": 1.028419778649441,
      "grad_norm": 4.0,
      "learning_rate": 5.012657034455176e-06,
      "loss": 0.80206022,
      "memory(GiB)": 147.13,
      "step": 44080,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.79112525,
      "epoch": 1.0286530862217296,
      "grad_norm": 6.1875,
      "learning_rate": 5.010767928011713e-06,
      "loss": 0.7436903,
      "memory(GiB)": 147.13,
      "step": 44090,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.78853774,
      "epoch": 1.0288863937940185,
      "grad_norm": 5.28125,
      "learning_rate": 5.008878820031131e-06,
      "loss": 0.76734447,
      "memory(GiB)": 147.13,
      "step": 44100,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.78617477,
      "epoch": 1.0291197013663074,
      "grad_norm": 6.09375,
      "learning_rate": 5.006989710783101e-06,
      "loss": 0.75627079,
      "memory(GiB)": 147.13,
      "step": 44110,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.78956699,
      "epoch": 1.0293530089385963,
      "grad_norm": 6.1875,
      "learning_rate": 5.005100600537292e-06,
      "loss": 0.74087925,
      "memory(GiB)": 147.13,
      "step": 44120,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.77857628,
      "epoch": 1.0295863165108852,
      "grad_norm": 5.25,
      "learning_rate": 5.003211489563373e-06,
      "loss": 0.80192738,
      "memory(GiB)": 147.13,
      "step": 44130,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.79740629,
      "epoch": 1.029819624083174,
      "grad_norm": 7.25,
      "learning_rate": 5.001322378131015e-06,
      "loss": 0.72210054,
      "memory(GiB)": 147.13,
      "step": 44140,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.79078064,
      "epoch": 1.030052931655463,
      "grad_norm": 5.0625,
      "learning_rate": 4.9994332665098885e-06,
      "loss": 0.74621177,
      "memory(GiB)": 147.13,
      "step": 44150,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.7691617,
      "epoch": 1.0302862392277519,
      "grad_norm": 4.96875,
      "learning_rate": 4.997544154969661e-06,
      "loss": 0.84756422,
      "memory(GiB)": 147.13,
      "step": 44160,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.78140469,
      "epoch": 1.0305195468000408,
      "grad_norm": 4.9375,
      "learning_rate": 4.995655043780006e-06,
      "loss": 0.80409288,
      "memory(GiB)": 147.13,
      "step": 44170,
      "train_speed(iter/s)": 0.201029
    },
    {
      "acc": 0.78131447,
      "epoch": 1.0307528543723297,
      "grad_norm": 7.96875,
      "learning_rate": 4.993765933210592e-06,
      "loss": 0.79957089,
      "memory(GiB)": 147.13,
      "step": 44180,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.78027859,
      "epoch": 1.0309861619446186,
      "grad_norm": 11.5625,
      "learning_rate": 4.991876823531089e-06,
      "loss": 0.79265738,
      "memory(GiB)": 147.13,
      "step": 44190,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.77979045,
      "epoch": 1.0312194695169075,
      "grad_norm": 10.0625,
      "learning_rate": 4.989987715011168e-06,
      "loss": 0.80041943,
      "memory(GiB)": 147.13,
      "step": 44200,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.77964592,
      "epoch": 1.0314527770891964,
      "grad_norm": 4.09375,
      "learning_rate": 4.988098607920497e-06,
      "loss": 0.79669185,
      "memory(GiB)": 147.13,
      "step": 44210,
      "train_speed(iter/s)": 0.201122
    },
    {
      "acc": 0.76723943,
      "epoch": 1.0316860846614853,
      "grad_norm": 6.15625,
      "learning_rate": 4.986209502528746e-06,
      "loss": 0.83204918,
      "memory(GiB)": 147.13,
      "step": 44220,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.782441,
      "epoch": 1.0319193922337742,
      "grad_norm": 4.09375,
      "learning_rate": 4.984320399105585e-06,
      "loss": 0.77679338,
      "memory(GiB)": 147.13,
      "step": 44230,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.78049326,
      "epoch": 1.032152699806063,
      "grad_norm": 3.6875,
      "learning_rate": 4.982431297920682e-06,
      "loss": 0.77765503,
      "memory(GiB)": 147.13,
      "step": 44240,
      "train_speed(iter/s)": 0.201191
    },
    {
      "acc": 0.79742098,
      "epoch": 1.032386007378352,
      "grad_norm": 16.75,
      "learning_rate": 4.980542199243709e-06,
      "loss": 0.72890673,
      "memory(GiB)": 147.13,
      "step": 44250,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.7838129,
      "epoch": 1.0326193149506409,
      "grad_norm": 5.90625,
      "learning_rate": 4.978653103344328e-06,
      "loss": 0.79265904,
      "memory(GiB)": 147.13,
      "step": 44260,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.78945446,
      "epoch": 1.0328526225229298,
      "grad_norm": 6.09375,
      "learning_rate": 4.976764010492211e-06,
      "loss": 0.7581913,
      "memory(GiB)": 147.13,
      "step": 44270,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.77482014,
      "epoch": 1.0330859300952187,
      "grad_norm": 7.21875,
      "learning_rate": 4.974874920957025e-06,
      "loss": 0.81373167,
      "memory(GiB)": 147.13,
      "step": 44280,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.79341946,
      "epoch": 1.0333192376675076,
      "grad_norm": 7.125,
      "learning_rate": 4.972985835008437e-06,
      "loss": 0.75202971,
      "memory(GiB)": 147.13,
      "step": 44290,
      "train_speed(iter/s)": 0.201306
    },
    {
      "acc": 0.79722977,
      "epoch": 1.0335525452397964,
      "grad_norm": 5.0625,
      "learning_rate": 4.971096752916113e-06,
      "loss": 0.73516812,
      "memory(GiB)": 147.13,
      "step": 44300,
      "train_speed(iter/s)": 0.201329
    },
    {
      "acc": 0.79803047,
      "epoch": 1.0337858528120853,
      "grad_norm": 8.5,
      "learning_rate": 4.969207674949719e-06,
      "loss": 0.71913786,
      "memory(GiB)": 147.13,
      "step": 44310,
      "train_speed(iter/s)": 0.201351
    },
    {
      "acc": 0.77934179,
      "epoch": 1.0340191603843742,
      "grad_norm": 6.28125,
      "learning_rate": 4.96731860137892e-06,
      "loss": 0.8005085,
      "memory(GiB)": 147.13,
      "step": 44320,
      "train_speed(iter/s)": 0.201374
    },
    {
      "acc": 0.76904945,
      "epoch": 1.0342524679566631,
      "grad_norm": 5.71875,
      "learning_rate": 4.965429532473383e-06,
      "loss": 0.84740162,
      "memory(GiB)": 147.13,
      "step": 44330,
      "train_speed(iter/s)": 0.201398
    },
    {
      "acc": 0.78627586,
      "epoch": 1.034485775528952,
      "grad_norm": 7.21875,
      "learning_rate": 4.963540468502768e-06,
      "loss": 0.75449095,
      "memory(GiB)": 147.13,
      "step": 44340,
      "train_speed(iter/s)": 0.201421
    },
    {
      "acc": 0.7668519,
      "epoch": 1.034719083101241,
      "grad_norm": 6.375,
      "learning_rate": 4.961651409736741e-06,
      "loss": 0.82782583,
      "memory(GiB)": 147.13,
      "step": 44350,
      "train_speed(iter/s)": 0.201445
    },
    {
      "acc": 0.77557964,
      "epoch": 1.0349523906735298,
      "grad_norm": 6.5,
      "learning_rate": 4.959762356444964e-06,
      "loss": 0.81699305,
      "memory(GiB)": 147.13,
      "step": 44360,
      "train_speed(iter/s)": 0.201469
    },
    {
      "acc": 0.77146072,
      "epoch": 1.0351856982458187,
      "grad_norm": 6.84375,
      "learning_rate": 4.957873308897102e-06,
      "loss": 0.83753395,
      "memory(GiB)": 147.13,
      "step": 44370,
      "train_speed(iter/s)": 0.201493
    },
    {
      "acc": 0.77347927,
      "epoch": 1.0354190058181076,
      "grad_norm": 4.1875,
      "learning_rate": 4.95598426736281e-06,
      "loss": 0.82432508,
      "memory(GiB)": 147.13,
      "step": 44380,
      "train_speed(iter/s)": 0.201518
    },
    {
      "acc": 0.78271379,
      "epoch": 1.0356523133903965,
      "grad_norm": 4.5,
      "learning_rate": 4.954095232111751e-06,
      "loss": 0.80724306,
      "memory(GiB)": 147.13,
      "step": 44390,
      "train_speed(iter/s)": 0.201542
    },
    {
      "acc": 0.78393545,
      "epoch": 1.0358856209626854,
      "grad_norm": 7.78125,
      "learning_rate": 4.9522062034135845e-06,
      "loss": 0.76437645,
      "memory(GiB)": 147.13,
      "step": 44400,
      "train_speed(iter/s)": 0.201561
    },
    {
      "acc": 0.77311645,
      "epoch": 1.0361189285349743,
      "grad_norm": 5.6875,
      "learning_rate": 4.9503171815379695e-06,
      "loss": 0.82633915,
      "memory(GiB)": 147.13,
      "step": 44410,
      "train_speed(iter/s)": 0.201584
    },
    {
      "acc": 0.78500485,
      "epoch": 1.0363522361072632,
      "grad_norm": 5.65625,
      "learning_rate": 4.948428166754561e-06,
      "loss": 0.76627178,
      "memory(GiB)": 147.13,
      "step": 44420,
      "train_speed(iter/s)": 0.201605
    },
    {
      "acc": 0.77328243,
      "epoch": 1.0365855436795521,
      "grad_norm": 4.15625,
      "learning_rate": 4.946539159333017e-06,
      "loss": 0.81605453,
      "memory(GiB)": 147.13,
      "step": 44430,
      "train_speed(iter/s)": 0.201628
    },
    {
      "acc": 0.78427544,
      "epoch": 1.036818851251841,
      "grad_norm": 4.25,
      "learning_rate": 4.944650159542993e-06,
      "loss": 0.76791339,
      "memory(GiB)": 147.13,
      "step": 44440,
      "train_speed(iter/s)": 0.201651
    },
    {
      "acc": 0.77760768,
      "epoch": 1.03705215882413,
      "grad_norm": 5.4375,
      "learning_rate": 4.942761167654142e-06,
      "loss": 0.7886517,
      "memory(GiB)": 147.13,
      "step": 44450,
      "train_speed(iter/s)": 0.201675
    },
    {
      "acc": 0.7762989,
      "epoch": 1.0372854663964188,
      "grad_norm": 6.46875,
      "learning_rate": 4.940872183936118e-06,
      "loss": 0.82030087,
      "memory(GiB)": 147.13,
      "step": 44460,
      "train_speed(iter/s)": 0.201699
    },
    {
      "acc": 0.77942324,
      "epoch": 1.0375187739687077,
      "grad_norm": 5.625,
      "learning_rate": 4.938983208658574e-06,
      "loss": 0.7806747,
      "memory(GiB)": 147.13,
      "step": 44470,
      "train_speed(iter/s)": 0.201721
    },
    {
      "acc": 0.76620932,
      "epoch": 1.0377520815409964,
      "grad_norm": 4.5,
      "learning_rate": 4.937094242091158e-06,
      "loss": 0.83421717,
      "memory(GiB)": 147.13,
      "step": 44480,
      "train_speed(iter/s)": 0.201745
    },
    {
      "acc": 0.79280834,
      "epoch": 1.0379853891132853,
      "grad_norm": 5.53125,
      "learning_rate": 4.935205284503522e-06,
      "loss": 0.74198709,
      "memory(GiB)": 147.13,
      "step": 44490,
      "train_speed(iter/s)": 0.201768
    },
    {
      "acc": 0.79166679,
      "epoch": 1.0382186966855742,
      "grad_norm": 7.03125,
      "learning_rate": 4.933316336165311e-06,
      "loss": 0.7469903,
      "memory(GiB)": 147.13,
      "step": 44500,
      "train_speed(iter/s)": 0.201791
    },
    {
      "epoch": 1.0382186966855742,
      "eval_acc": 0.743963643415806,
      "eval_loss": 0.8068258166313171,
      "eval_runtime": 1270.5076,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.164,
      "step": 44500
    },
    {
      "acc": 0.78040266,
      "epoch": 1.038452004257863,
      "grad_norm": 4.4375,
      "learning_rate": 4.931427397346174e-06,
      "loss": 0.77377152,
      "memory(GiB)": 147.13,
      "step": 44510,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.76233931,
      "epoch": 1.038685311830152,
      "grad_norm": 5.5625,
      "learning_rate": 4.929538468315756e-06,
      "loss": 0.8748394,
      "memory(GiB)": 147.13,
      "step": 44520,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.77354889,
      "epoch": 1.038918619402441,
      "grad_norm": 5.25,
      "learning_rate": 4.927649549343701e-06,
      "loss": 0.81373053,
      "memory(GiB)": 147.13,
      "step": 44530,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.77879763,
      "epoch": 1.0391519269747298,
      "grad_norm": 4.9375,
      "learning_rate": 4.9257606406996525e-06,
      "loss": 0.78499241,
      "memory(GiB)": 147.13,
      "step": 44540,
      "train_speed(iter/s)": 0.200714
    },
    {
      "acc": 0.77786951,
      "epoch": 1.0393852345470187,
      "grad_norm": 4.625,
      "learning_rate": 4.923871742653251e-06,
      "loss": 0.80249662,
      "memory(GiB)": 147.13,
      "step": 44550,
      "train_speed(iter/s)": 0.200736
    },
    {
      "acc": 0.76491838,
      "epoch": 1.0396185421193076,
      "grad_norm": 5.5625,
      "learning_rate": 4.921982855474136e-06,
      "loss": 0.84366484,
      "memory(GiB)": 147.13,
      "step": 44560,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.80446472,
      "epoch": 1.0398518496915965,
      "grad_norm": 5.78125,
      "learning_rate": 4.9200939794319444e-06,
      "loss": 0.70642214,
      "memory(GiB)": 147.13,
      "step": 44570,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.78261833,
      "epoch": 1.0400851572638854,
      "grad_norm": 6.09375,
      "learning_rate": 4.918205114796315e-06,
      "loss": 0.7593545,
      "memory(GiB)": 147.13,
      "step": 44580,
      "train_speed(iter/s)": 0.200805
    },
    {
      "acc": 0.77353225,
      "epoch": 1.0403184648361743,
      "grad_norm": 6.625,
      "learning_rate": 4.916316261836882e-06,
      "loss": 0.84205828,
      "memory(GiB)": 147.13,
      "step": 44590,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.77843504,
      "epoch": 1.0405517724084632,
      "grad_norm": 8.75,
      "learning_rate": 4.91442742082328e-06,
      "loss": 0.79012375,
      "memory(GiB)": 147.13,
      "step": 44600,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.79202127,
      "epoch": 1.040785079980752,
      "grad_norm": 5.5,
      "learning_rate": 4.912538592025137e-06,
      "loss": 0.75029149,
      "memory(GiB)": 147.13,
      "step": 44610,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.79142704,
      "epoch": 1.041018387553041,
      "grad_norm": 4.90625,
      "learning_rate": 4.910649775712084e-06,
      "loss": 0.77472105,
      "memory(GiB)": 147.13,
      "step": 44620,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.79433427,
      "epoch": 1.04125169512533,
      "grad_norm": 5.4375,
      "learning_rate": 4.908760972153751e-06,
      "loss": 0.7298358,
      "memory(GiB)": 147.13,
      "step": 44630,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77725754,
      "epoch": 1.0414850026976188,
      "grad_norm": 6.46875,
      "learning_rate": 4.9068721816197615e-06,
      "loss": 0.83918152,
      "memory(GiB)": 147.13,
      "step": 44640,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.79220066,
      "epoch": 1.0417183102699077,
      "grad_norm": 5.34375,
      "learning_rate": 4.904983404379741e-06,
      "loss": 0.7297925,
      "memory(GiB)": 147.13,
      "step": 44650,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.77652454,
      "epoch": 1.0419516178421966,
      "grad_norm": 4.75,
      "learning_rate": 4.903094640703312e-06,
      "loss": 0.78495092,
      "memory(GiB)": 147.13,
      "step": 44660,
      "train_speed(iter/s)": 0.200988
    },
    {
      "acc": 0.79377813,
      "epoch": 1.0421849254144855,
      "grad_norm": 4.9375,
      "learning_rate": 4.901205890860095e-06,
      "loss": 0.73209939,
      "memory(GiB)": 147.13,
      "step": 44670,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.79054551,
      "epoch": 1.0424182329867744,
      "grad_norm": 5.34375,
      "learning_rate": 4.899317155119708e-06,
      "loss": 0.7544507,
      "memory(GiB)": 147.13,
      "step": 44680,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.79657516,
      "epoch": 1.0426515405590633,
      "grad_norm": 4.5625,
      "learning_rate": 4.89742843375177e-06,
      "loss": 0.75656972,
      "memory(GiB)": 147.13,
      "step": 44690,
      "train_speed(iter/s)": 0.201056
    },
    {
      "acc": 0.78189001,
      "epoch": 1.0428848481313522,
      "grad_norm": 6.46875,
      "learning_rate": 4.895539727025891e-06,
      "loss": 0.77084103,
      "memory(GiB)": 147.13,
      "step": 44700,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.78460836,
      "epoch": 1.043118155703641,
      "grad_norm": 4.6875,
      "learning_rate": 4.8936510352116886e-06,
      "loss": 0.77186456,
      "memory(GiB)": 147.13,
      "step": 44710,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.75842123,
      "epoch": 1.04335146327593,
      "grad_norm": 7.78125,
      "learning_rate": 4.891762358578767e-06,
      "loss": 0.86604462,
      "memory(GiB)": 147.13,
      "step": 44720,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.78901138,
      "epoch": 1.0435847708482189,
      "grad_norm": 4.5,
      "learning_rate": 4.889873697396738e-06,
      "loss": 0.76355696,
      "memory(GiB)": 147.13,
      "step": 44730,
      "train_speed(iter/s)": 0.201147
    },
    {
      "acc": 0.78258648,
      "epoch": 1.0438180784205078,
      "grad_norm": 5.46875,
      "learning_rate": 4.887985051935206e-06,
      "loss": 0.77364783,
      "memory(GiB)": 147.13,
      "step": 44740,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.7676549,
      "epoch": 1.0440513859927967,
      "grad_norm": 5.21875,
      "learning_rate": 4.8860964224637756e-06,
      "loss": 0.83399296,
      "memory(GiB)": 147.13,
      "step": 44750,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.76923347,
      "epoch": 1.0442846935650856,
      "grad_norm": 3.890625,
      "learning_rate": 4.884207809252049e-06,
      "loss": 0.81878109,
      "memory(GiB)": 147.13,
      "step": 44760,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.76931305,
      "epoch": 1.0445180011373745,
      "grad_norm": 3.890625,
      "learning_rate": 4.882319212569623e-06,
      "loss": 0.84117594,
      "memory(GiB)": 147.13,
      "step": 44770,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.78014016,
      "epoch": 1.0447513087096634,
      "grad_norm": 5.15625,
      "learning_rate": 4.880430632686096e-06,
      "loss": 0.78205185,
      "memory(GiB)": 147.13,
      "step": 44780,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.7773912,
      "epoch": 1.0449846162819523,
      "grad_norm": 4.84375,
      "learning_rate": 4.87854206987106e-06,
      "loss": 0.80879612,
      "memory(GiB)": 147.13,
      "step": 44790,
      "train_speed(iter/s)": 0.201287
    },
    {
      "acc": 0.78483715,
      "epoch": 1.0452179238542412,
      "grad_norm": 4.5625,
      "learning_rate": 4.876653524394109e-06,
      "loss": 0.77698545,
      "memory(GiB)": 147.13,
      "step": 44800,
      "train_speed(iter/s)": 0.20131
    },
    {
      "acc": 0.76030092,
      "epoch": 1.04545123142653,
      "grad_norm": 6.0625,
      "learning_rate": 4.874764996524831e-06,
      "loss": 0.86837893,
      "memory(GiB)": 147.13,
      "step": 44810,
      "train_speed(iter/s)": 0.201335
    },
    {
      "acc": 0.78192863,
      "epoch": 1.045684538998819,
      "grad_norm": 9.875,
      "learning_rate": 4.872876486532814e-06,
      "loss": 0.77877698,
      "memory(GiB)": 147.13,
      "step": 44820,
      "train_speed(iter/s)": 0.201359
    },
    {
      "acc": 0.77943573,
      "epoch": 1.0459178465711079,
      "grad_norm": 5.1875,
      "learning_rate": 4.870987994687644e-06,
      "loss": 0.78068628,
      "memory(GiB)": 147.13,
      "step": 44830,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.76940956,
      "epoch": 1.0461511541433968,
      "grad_norm": 4.8125,
      "learning_rate": 4.869099521258897e-06,
      "loss": 0.83160362,
      "memory(GiB)": 147.13,
      "step": 44840,
      "train_speed(iter/s)": 0.201403
    },
    {
      "acc": 0.77414241,
      "epoch": 1.0463844617156857,
      "grad_norm": 6.4375,
      "learning_rate": 4.867211066516157e-06,
      "loss": 0.82083912,
      "memory(GiB)": 147.13,
      "step": 44850,
      "train_speed(iter/s)": 0.201428
    },
    {
      "acc": 0.80341206,
      "epoch": 1.0466177692879746,
      "grad_norm": 5.40625,
      "learning_rate": 4.865322630728998e-06,
      "loss": 0.70446091,
      "memory(GiB)": 147.13,
      "step": 44860,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.79915457,
      "epoch": 1.0468510768602632,
      "grad_norm": 5.59375,
      "learning_rate": 4.863434214166994e-06,
      "loss": 0.69466219,
      "memory(GiB)": 147.13,
      "step": 44870,
      "train_speed(iter/s)": 0.201475
    },
    {
      "acc": 0.77168198,
      "epoch": 1.0470843844325521,
      "grad_norm": 5.75,
      "learning_rate": 4.8615458170997166e-06,
      "loss": 0.82629623,
      "memory(GiB)": 147.13,
      "step": 44880,
      "train_speed(iter/s)": 0.201499
    },
    {
      "acc": 0.78714318,
      "epoch": 1.047317692004841,
      "grad_norm": 6.3125,
      "learning_rate": 4.8596574397967324e-06,
      "loss": 0.74908113,
      "memory(GiB)": 147.13,
      "step": 44890,
      "train_speed(iter/s)": 0.201522
    },
    {
      "acc": 0.79111013,
      "epoch": 1.04755099957713,
      "grad_norm": 6.3125,
      "learning_rate": 4.857769082527609e-06,
      "loss": 0.77098122,
      "memory(GiB)": 147.13,
      "step": 44900,
      "train_speed(iter/s)": 0.201545
    },
    {
      "acc": 0.75890102,
      "epoch": 1.0477843071494188,
      "grad_norm": 5.875,
      "learning_rate": 4.855880745561909e-06,
      "loss": 0.86920433,
      "memory(GiB)": 147.13,
      "step": 44910,
      "train_speed(iter/s)": 0.201569
    },
    {
      "acc": 0.78598218,
      "epoch": 1.0480176147217077,
      "grad_norm": 6.53125,
      "learning_rate": 4.853992429169189e-06,
      "loss": 0.76193895,
      "memory(GiB)": 147.13,
      "step": 44920,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.75896392,
      "epoch": 1.0482509222939966,
      "grad_norm": 6.5,
      "learning_rate": 4.852104133619008e-06,
      "loss": 0.88328629,
      "memory(GiB)": 147.13,
      "step": 44930,
      "train_speed(iter/s)": 0.201613
    },
    {
      "acc": 0.78158102,
      "epoch": 1.0484842298662855,
      "grad_norm": 8.125,
      "learning_rate": 4.85021585918092e-06,
      "loss": 0.77389903,
      "memory(GiB)": 147.13,
      "step": 44940,
      "train_speed(iter/s)": 0.201638
    },
    {
      "acc": 0.78255506,
      "epoch": 1.0487175374385744,
      "grad_norm": 6.21875,
      "learning_rate": 4.848327606124473e-06,
      "loss": 0.75741982,
      "memory(GiB)": 147.13,
      "step": 44950,
      "train_speed(iter/s)": 0.20166
    },
    {
      "acc": 0.78304787,
      "epoch": 1.0489508450108633,
      "grad_norm": 5.03125,
      "learning_rate": 4.846439374719217e-06,
      "loss": 0.79422493,
      "memory(GiB)": 147.13,
      "step": 44960,
      "train_speed(iter/s)": 0.201682
    },
    {
      "acc": 0.77571564,
      "epoch": 1.0491841525831522,
      "grad_norm": 4.875,
      "learning_rate": 4.844551165234694e-06,
      "loss": 0.80409365,
      "memory(GiB)": 147.13,
      "step": 44970,
      "train_speed(iter/s)": 0.201706
    },
    {
      "acc": 0.78164635,
      "epoch": 1.0494174601554411,
      "grad_norm": 4.65625,
      "learning_rate": 4.842662977940448e-06,
      "loss": 0.80020761,
      "memory(GiB)": 147.13,
      "step": 44980,
      "train_speed(iter/s)": 0.201728
    },
    {
      "acc": 0.78135223,
      "epoch": 1.04965076772773,
      "grad_norm": 5.46875,
      "learning_rate": 4.8407748131060175e-06,
      "loss": 0.76936512,
      "memory(GiB)": 147.13,
      "step": 44990,
      "train_speed(iter/s)": 0.201751
    },
    {
      "acc": 0.7863555,
      "epoch": 1.049884075300019,
      "grad_norm": 6.15625,
      "learning_rate": 4.838886671000934e-06,
      "loss": 0.7614399,
      "memory(GiB)": 147.13,
      "step": 45000,
      "train_speed(iter/s)": 0.201774
    },
    {
      "epoch": 1.049884075300019,
      "eval_acc": 0.7440030388969522,
      "eval_loss": 0.8069285750389099,
      "eval_runtime": 1270.5322,
      "eval_samples_per_second": 28.327,
      "eval_steps_per_second": 14.164,
      "step": 45000
    },
    {
      "acc": 0.77712164,
      "epoch": 1.0501173828723078,
      "grad_norm": 6.6875,
      "learning_rate": 4.8369985518947336e-06,
      "loss": 0.79156408,
      "memory(GiB)": 147.13,
      "step": 45010,
      "train_speed(iter/s)": 0.200639
    },
    {
      "acc": 0.78002462,
      "epoch": 1.0503506904445967,
      "grad_norm": 6.0,
      "learning_rate": 4.83511045605694e-06,
      "loss": 0.79216776,
      "memory(GiB)": 147.13,
      "step": 45020,
      "train_speed(iter/s)": 0.200662
    },
    {
      "acc": 0.78218923,
      "epoch": 1.0505839980168856,
      "grad_norm": 6.40625,
      "learning_rate": 4.8332223837570824e-06,
      "loss": 0.7720716,
      "memory(GiB)": 147.13,
      "step": 45030,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.78232145,
      "epoch": 1.0508173055891745,
      "grad_norm": 6.15625,
      "learning_rate": 4.831334335264682e-06,
      "loss": 0.77959723,
      "memory(GiB)": 147.13,
      "step": 45040,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.78184223,
      "epoch": 1.0510506131614634,
      "grad_norm": 5.75,
      "learning_rate": 4.829446310849256e-06,
      "loss": 0.76550894,
      "memory(GiB)": 147.13,
      "step": 45050,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.78313303,
      "epoch": 1.0512839207337523,
      "grad_norm": 5.9375,
      "learning_rate": 4.827558310780319e-06,
      "loss": 0.77661057,
      "memory(GiB)": 147.13,
      "step": 45060,
      "train_speed(iter/s)": 0.20075
    },
    {
      "acc": 0.79087114,
      "epoch": 1.0515172283060412,
      "grad_norm": 4.40625,
      "learning_rate": 4.825670335327383e-06,
      "loss": 0.74372959,
      "memory(GiB)": 147.13,
      "step": 45070,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.80854607,
      "epoch": 1.0517505358783301,
      "grad_norm": 4.5625,
      "learning_rate": 4.823782384759955e-06,
      "loss": 0.68139353,
      "memory(GiB)": 147.13,
      "step": 45080,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.78581572,
      "epoch": 1.051983843450619,
      "grad_norm": 4.9375,
      "learning_rate": 4.821894459347542e-06,
      "loss": 0.7871974,
      "memory(GiB)": 147.13,
      "step": 45090,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.78187127,
      "epoch": 1.052217151022908,
      "grad_norm": 9.375,
      "learning_rate": 4.820006559359642e-06,
      "loss": 0.78979106,
      "memory(GiB)": 147.13,
      "step": 45100,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.76723819,
      "epoch": 1.0524504585951968,
      "grad_norm": 6.875,
      "learning_rate": 4.818118685065754e-06,
      "loss": 0.84743042,
      "memory(GiB)": 147.13,
      "step": 45110,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.77373695,
      "epoch": 1.0526837661674857,
      "grad_norm": 9.5,
      "learning_rate": 4.8162308367353705e-06,
      "loss": 0.81677189,
      "memory(GiB)": 147.13,
      "step": 45120,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78724155,
      "epoch": 1.0529170737397746,
      "grad_norm": 5.21875,
      "learning_rate": 4.814343014637982e-06,
      "loss": 0.75338268,
      "memory(GiB)": 147.13,
      "step": 45130,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.77668324,
      "epoch": 1.0531503813120635,
      "grad_norm": 5.625,
      "learning_rate": 4.812455219043074e-06,
      "loss": 0.80819511,
      "memory(GiB)": 147.13,
      "step": 45140,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.77539773,
      "epoch": 1.0533836888843524,
      "grad_norm": 4.875,
      "learning_rate": 4.810567450220128e-06,
      "loss": 0.78674135,
      "memory(GiB)": 147.13,
      "step": 45150,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.7868578,
      "epoch": 1.0536169964566413,
      "grad_norm": 6.34375,
      "learning_rate": 4.808679708438624e-06,
      "loss": 0.76853762,
      "memory(GiB)": 147.13,
      "step": 45160,
      "train_speed(iter/s)": 0.200979
    },
    {
      "acc": 0.76111031,
      "epoch": 1.0538503040289302,
      "grad_norm": 5.375,
      "learning_rate": 4.806791993968039e-06,
      "loss": 0.86612825,
      "memory(GiB)": 147.13,
      "step": 45170,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.78183184,
      "epoch": 1.054083611601219,
      "grad_norm": 6.9375,
      "learning_rate": 4.804904307077838e-06,
      "loss": 0.76009259,
      "memory(GiB)": 147.13,
      "step": 45180,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.78163671,
      "epoch": 1.054316919173508,
      "grad_norm": 7.125,
      "learning_rate": 4.80301664803749e-06,
      "loss": 0.78068633,
      "memory(GiB)": 147.13,
      "step": 45190,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.78807421,
      "epoch": 1.054550226745797,
      "grad_norm": 5.625,
      "learning_rate": 4.80112901711646e-06,
      "loss": 0.75582342,
      "memory(GiB)": 147.13,
      "step": 45200,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.77289944,
      "epoch": 1.0547835343180858,
      "grad_norm": 6.28125,
      "learning_rate": 4.799241414584204e-06,
      "loss": 0.82699385,
      "memory(GiB)": 147.13,
      "step": 45210,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.78645868,
      "epoch": 1.0550168418903747,
      "grad_norm": 3.78125,
      "learning_rate": 4.797353840710178e-06,
      "loss": 0.7695097,
      "memory(GiB)": 147.13,
      "step": 45220,
      "train_speed(iter/s)": 0.201114
    },
    {
      "acc": 0.77393785,
      "epoch": 1.0552501494626636,
      "grad_norm": 6.125,
      "learning_rate": 4.795466295763832e-06,
      "loss": 0.81475391,
      "memory(GiB)": 147.13,
      "step": 45230,
      "train_speed(iter/s)": 0.201135
    },
    {
      "acc": 0.78816233,
      "epoch": 1.0554834570349523,
      "grad_norm": 6.1875,
      "learning_rate": 4.793578780014612e-06,
      "loss": 0.76425867,
      "memory(GiB)": 147.13,
      "step": 45240,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.76952739,
      "epoch": 1.0557167646072414,
      "grad_norm": 7.21875,
      "learning_rate": 4.791691293731962e-06,
      "loss": 0.81708698,
      "memory(GiB)": 147.13,
      "step": 45250,
      "train_speed(iter/s)": 0.201181
    },
    {
      "acc": 0.80263548,
      "epoch": 1.05595007217953,
      "grad_norm": 4.1875,
      "learning_rate": 4.78980383718532e-06,
      "loss": 0.68140831,
      "memory(GiB)": 147.13,
      "step": 45260,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78784084,
      "epoch": 1.056183379751819,
      "grad_norm": 4.78125,
      "learning_rate": 4.787916410644119e-06,
      "loss": 0.75726004,
      "memory(GiB)": 147.13,
      "step": 45270,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.76332178,
      "epoch": 1.0564166873241079,
      "grad_norm": 5.5625,
      "learning_rate": 4.786029014377789e-06,
      "loss": 0.82791462,
      "memory(GiB)": 147.13,
      "step": 45280,
      "train_speed(iter/s)": 0.201253
    },
    {
      "acc": 0.77394991,
      "epoch": 1.0566499948963968,
      "grad_norm": 7.125,
      "learning_rate": 4.784141648655756e-06,
      "loss": 0.81197491,
      "memory(GiB)": 147.13,
      "step": 45290,
      "train_speed(iter/s)": 0.201276
    },
    {
      "acc": 0.76897669,
      "epoch": 1.0568833024686857,
      "grad_norm": 6.375,
      "learning_rate": 4.782254313747438e-06,
      "loss": 0.82489567,
      "memory(GiB)": 147.13,
      "step": 45300,
      "train_speed(iter/s)": 0.201299
    },
    {
      "acc": 0.78103123,
      "epoch": 1.0571166100409746,
      "grad_norm": 4.9375,
      "learning_rate": 4.780367009922253e-06,
      "loss": 0.785322,
      "memory(GiB)": 147.13,
      "step": 45310,
      "train_speed(iter/s)": 0.201322
    },
    {
      "acc": 0.76720481,
      "epoch": 1.0573499176132635,
      "grad_norm": 5.4375,
      "learning_rate": 4.778479737449614e-06,
      "loss": 0.85537357,
      "memory(GiB)": 147.13,
      "step": 45320,
      "train_speed(iter/s)": 0.201346
    },
    {
      "acc": 0.79495482,
      "epoch": 1.0575832251855524,
      "grad_norm": 6.28125,
      "learning_rate": 4.7765924965989286e-06,
      "loss": 0.72112303,
      "memory(GiB)": 147.13,
      "step": 45330,
      "train_speed(iter/s)": 0.201369
    },
    {
      "acc": 0.7869606,
      "epoch": 1.0578165327578413,
      "grad_norm": 6.15625,
      "learning_rate": 4.7747052876396e-06,
      "loss": 0.75025473,
      "memory(GiB)": 147.13,
      "step": 45340,
      "train_speed(iter/s)": 0.201393
    },
    {
      "acc": 0.77279506,
      "epoch": 1.0580498403301302,
      "grad_norm": 4.59375,
      "learning_rate": 4.772818110841025e-06,
      "loss": 0.82170506,
      "memory(GiB)": 147.13,
      "step": 45350,
      "train_speed(iter/s)": 0.201418
    },
    {
      "acc": 0.7949152,
      "epoch": 1.058283147902419,
      "grad_norm": 5.375,
      "learning_rate": 4.7709309664726e-06,
      "loss": 0.73024497,
      "memory(GiB)": 147.13,
      "step": 45360,
      "train_speed(iter/s)": 0.201443
    },
    {
      "acc": 0.79297123,
      "epoch": 1.058516455474708,
      "grad_norm": 6.84375,
      "learning_rate": 4.769043854803712e-06,
      "loss": 0.72323437,
      "memory(GiB)": 147.13,
      "step": 45370,
      "train_speed(iter/s)": 0.201465
    },
    {
      "acc": 0.77630758,
      "epoch": 1.0587497630469969,
      "grad_norm": 6.03125,
      "learning_rate": 4.767156776103746e-06,
      "loss": 0.80375509,
      "memory(GiB)": 147.13,
      "step": 45380,
      "train_speed(iter/s)": 0.201489
    },
    {
      "acc": 0.79003696,
      "epoch": 1.0589830706192858,
      "grad_norm": 5.09375,
      "learning_rate": 4.765269730642083e-06,
      "loss": 0.73524923,
      "memory(GiB)": 147.13,
      "step": 45390,
      "train_speed(iter/s)": 0.201511
    },
    {
      "acc": 0.79320269,
      "epoch": 1.0592163781915747,
      "grad_norm": 4.5625,
      "learning_rate": 4.7633827186881e-06,
      "loss": 0.7441308,
      "memory(GiB)": 147.13,
      "step": 45400,
      "train_speed(iter/s)": 0.201535
    },
    {
      "acc": 0.77364416,
      "epoch": 1.0594496857638636,
      "grad_norm": 5.75,
      "learning_rate": 4.7614957405111635e-06,
      "loss": 0.80192499,
      "memory(GiB)": 147.13,
      "step": 45410,
      "train_speed(iter/s)": 0.201559
    },
    {
      "acc": 0.78601189,
      "epoch": 1.0596829933361525,
      "grad_norm": 6.78125,
      "learning_rate": 4.759608796380642e-06,
      "loss": 0.77241449,
      "memory(GiB)": 147.13,
      "step": 45420,
      "train_speed(iter/s)": 0.20158
    },
    {
      "acc": 0.75279102,
      "epoch": 1.0599163009084414,
      "grad_norm": 5.5,
      "learning_rate": 4.757721886565893e-06,
      "loss": 0.89964409,
      "memory(GiB)": 147.13,
      "step": 45430,
      "train_speed(iter/s)": 0.201603
    },
    {
      "acc": 0.76471562,
      "epoch": 1.0601496084807303,
      "grad_norm": 9.5625,
      "learning_rate": 4.755835011336274e-06,
      "loss": 0.85386238,
      "memory(GiB)": 147.13,
      "step": 45440,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.78351641,
      "epoch": 1.0603829160530192,
      "grad_norm": 6.0625,
      "learning_rate": 4.753948170961137e-06,
      "loss": 0.76716695,
      "memory(GiB)": 147.13,
      "step": 45450,
      "train_speed(iter/s)": 0.201649
    },
    {
      "acc": 0.78080029,
      "epoch": 1.060616223625308,
      "grad_norm": 4.96875,
      "learning_rate": 4.752061365709827e-06,
      "loss": 0.76200881,
      "memory(GiB)": 147.13,
      "step": 45460,
      "train_speed(iter/s)": 0.201672
    },
    {
      "acc": 0.7751379,
      "epoch": 1.060849531197597,
      "grad_norm": 5.84375,
      "learning_rate": 4.750174595851685e-06,
      "loss": 0.79694057,
      "memory(GiB)": 147.13,
      "step": 45470,
      "train_speed(iter/s)": 0.201697
    },
    {
      "acc": 0.7780129,
      "epoch": 1.0610828387698858,
      "grad_norm": 4.71875,
      "learning_rate": 4.748287861656047e-06,
      "loss": 0.80889072,
      "memory(GiB)": 147.13,
      "step": 45480,
      "train_speed(iter/s)": 0.20172
    },
    {
      "acc": 0.78614645,
      "epoch": 1.0613161463421747,
      "grad_norm": 5.4375,
      "learning_rate": 4.746401163392244e-06,
      "loss": 0.79196577,
      "memory(GiB)": 147.13,
      "step": 45490,
      "train_speed(iter/s)": 0.201743
    },
    {
      "acc": 0.80344467,
      "epoch": 1.0615494539144636,
      "grad_norm": 4.78125,
      "learning_rate": 4.744514501329601e-06,
      "loss": 0.71334877,
      "memory(GiB)": 147.13,
      "step": 45500,
      "train_speed(iter/s)": 0.201765
    },
    {
      "epoch": 1.0615494539144636,
      "eval_acc": 0.7440579683686315,
      "eval_loss": 0.8067649006843567,
      "eval_runtime": 1270.3852,
      "eval_samples_per_second": 28.331,
      "eval_steps_per_second": 14.166,
      "step": 45500
    },
    {
      "acc": 0.78397212,
      "epoch": 1.0617827614867525,
      "grad_norm": 5.28125,
      "learning_rate": 4.74262787573744e-06,
      "loss": 0.78342257,
      "memory(GiB)": 147.13,
      "step": 45510,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.78444176,
      "epoch": 1.0620160690590414,
      "grad_norm": 11.125,
      "learning_rate": 4.7407412868850734e-06,
      "loss": 0.77105117,
      "memory(GiB)": 147.13,
      "step": 45520,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.79020176,
      "epoch": 1.0622493766313303,
      "grad_norm": 5.9375,
      "learning_rate": 4.738854735041813e-06,
      "loss": 0.75314264,
      "memory(GiB)": 147.13,
      "step": 45530,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.7691103,
      "epoch": 1.0624826842036192,
      "grad_norm": 6.375,
      "learning_rate": 4.736968220476963e-06,
      "loss": 0.8225666,
      "memory(GiB)": 147.13,
      "step": 45540,
      "train_speed(iter/s)": 0.200709
    },
    {
      "acc": 0.77286139,
      "epoch": 1.0627159917759081,
      "grad_norm": 5.78125,
      "learning_rate": 4.735081743459823e-06,
      "loss": 0.82172585,
      "memory(GiB)": 147.13,
      "step": 45550,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.78380318,
      "epoch": 1.062949299348197,
      "grad_norm": 5.03125,
      "learning_rate": 4.733195304259689e-06,
      "loss": 0.75959673,
      "memory(GiB)": 147.13,
      "step": 45560,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.78299665,
      "epoch": 1.063182606920486,
      "grad_norm": 6.28125,
      "learning_rate": 4.731308903145846e-06,
      "loss": 0.77152033,
      "memory(GiB)": 147.13,
      "step": 45570,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.78575029,
      "epoch": 1.0634159144927748,
      "grad_norm": 6.75,
      "learning_rate": 4.729422540387579e-06,
      "loss": 0.75135193,
      "memory(GiB)": 147.13,
      "step": 45580,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.77882252,
      "epoch": 1.0636492220650637,
      "grad_norm": 4.59375,
      "learning_rate": 4.727536216254166e-06,
      "loss": 0.77502632,
      "memory(GiB)": 147.13,
      "step": 45590,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.78998165,
      "epoch": 1.0638825296373526,
      "grad_norm": 5.15625,
      "learning_rate": 4.725649931014879e-06,
      "loss": 0.75522022,
      "memory(GiB)": 147.13,
      "step": 45600,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.78265238,
      "epoch": 1.0641158372096415,
      "grad_norm": 6.28125,
      "learning_rate": 4.723763684938985e-06,
      "loss": 0.78737378,
      "memory(GiB)": 147.13,
      "step": 45610,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.78826542,
      "epoch": 1.0643491447819304,
      "grad_norm": 4.53125,
      "learning_rate": 4.721877478295745e-06,
      "loss": 0.77142153,
      "memory(GiB)": 147.13,
      "step": 45620,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.76623106,
      "epoch": 1.064582452354219,
      "grad_norm": 7.78125,
      "learning_rate": 4.719991311354415e-06,
      "loss": 0.84487782,
      "memory(GiB)": 147.13,
      "step": 45630,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.78780022,
      "epoch": 1.0648157599265082,
      "grad_norm": 5.65625,
      "learning_rate": 4.718105184384243e-06,
      "loss": 0.75018334,
      "memory(GiB)": 147.13,
      "step": 45640,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.77883091,
      "epoch": 1.065049067498797,
      "grad_norm": 6.3125,
      "learning_rate": 4.7162190976544735e-06,
      "loss": 0.80597057,
      "memory(GiB)": 147.13,
      "step": 45650,
      "train_speed(iter/s)": 0.20096
    },
    {
      "acc": 0.78427258,
      "epoch": 1.0652823750710858,
      "grad_norm": 7.40625,
      "learning_rate": 4.7143330514343446e-06,
      "loss": 0.78061914,
      "memory(GiB)": 147.13,
      "step": 45660,
      "train_speed(iter/s)": 0.200983
    },
    {
      "acc": 0.78091021,
      "epoch": 1.0655156826433747,
      "grad_norm": 5.53125,
      "learning_rate": 4.712447045993091e-06,
      "loss": 0.78055716,
      "memory(GiB)": 147.13,
      "step": 45670,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.75933809,
      "epoch": 1.0657489902156636,
      "grad_norm": 7.15625,
      "learning_rate": 4.710561081599937e-06,
      "loss": 0.8501317,
      "memory(GiB)": 147.13,
      "step": 45680,
      "train_speed(iter/s)": 0.201028
    },
    {
      "acc": 0.77189841,
      "epoch": 1.0659822977879525,
      "grad_norm": 4.21875,
      "learning_rate": 4.708675158524105e-06,
      "loss": 0.81637125,
      "memory(GiB)": 147.13,
      "step": 45690,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.78322716,
      "epoch": 1.0662156053602414,
      "grad_norm": 4.46875,
      "learning_rate": 4.706789277034811e-06,
      "loss": 0.78145242,
      "memory(GiB)": 147.13,
      "step": 45700,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.7969058,
      "epoch": 1.0664489129325303,
      "grad_norm": 4.96875,
      "learning_rate": 4.704903437401261e-06,
      "loss": 0.70635157,
      "memory(GiB)": 147.13,
      "step": 45710,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.77726212,
      "epoch": 1.0666822205048192,
      "grad_norm": 6.0625,
      "learning_rate": 4.703017639892659e-06,
      "loss": 0.77836952,
      "memory(GiB)": 147.13,
      "step": 45720,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.77621179,
      "epoch": 1.066915528077108,
      "grad_norm": 6.3125,
      "learning_rate": 4.701131884778204e-06,
      "loss": 0.80525446,
      "memory(GiB)": 147.13,
      "step": 45730,
      "train_speed(iter/s)": 0.201144
    },
    {
      "acc": 0.77437353,
      "epoch": 1.067148835649397,
      "grad_norm": 9.125,
      "learning_rate": 4.699246172327087e-06,
      "loss": 0.79244003,
      "memory(GiB)": 147.13,
      "step": 45740,
      "train_speed(iter/s)": 0.201167
    },
    {
      "acc": 0.77285023,
      "epoch": 1.067382143221686,
      "grad_norm": 19.625,
      "learning_rate": 4.697360502808488e-06,
      "loss": 0.83040085,
      "memory(GiB)": 147.13,
      "step": 45750,
      "train_speed(iter/s)": 0.201191
    },
    {
      "acc": 0.7729352,
      "epoch": 1.0676154507939748,
      "grad_norm": 6.96875,
      "learning_rate": 4.695474876491592e-06,
      "loss": 0.84784021,
      "memory(GiB)": 147.13,
      "step": 45760,
      "train_speed(iter/s)": 0.201213
    },
    {
      "acc": 0.77671528,
      "epoch": 1.0678487583662637,
      "grad_norm": 6.15625,
      "learning_rate": 4.6935892936455664e-06,
      "loss": 0.78750763,
      "memory(GiB)": 147.13,
      "step": 45770,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.77845559,
      "epoch": 1.0680820659385526,
      "grad_norm": 5.90625,
      "learning_rate": 4.691703754539583e-06,
      "loss": 0.81025581,
      "memory(GiB)": 147.13,
      "step": 45780,
      "train_speed(iter/s)": 0.20126
    },
    {
      "acc": 0.79248986,
      "epoch": 1.0683153735108415,
      "grad_norm": 6.1875,
      "learning_rate": 4.689818259442797e-06,
      "loss": 0.75349188,
      "memory(GiB)": 147.13,
      "step": 45790,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.77852826,
      "epoch": 1.0685486810831304,
      "grad_norm": 9.5625,
      "learning_rate": 4.687932808624365e-06,
      "loss": 0.79906149,
      "memory(GiB)": 147.13,
      "step": 45800,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.78552589,
      "epoch": 1.0687819886554193,
      "grad_norm": 4.46875,
      "learning_rate": 4.686047402353433e-06,
      "loss": 0.74826708,
      "memory(GiB)": 147.13,
      "step": 45810,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.7994451,
      "epoch": 1.0690152962277082,
      "grad_norm": 3.953125,
      "learning_rate": 4.684162040899144e-06,
      "loss": 0.70688953,
      "memory(GiB)": 147.13,
      "step": 45820,
      "train_speed(iter/s)": 0.201354
    },
    {
      "acc": 0.78053541,
      "epoch": 1.069248603799997,
      "grad_norm": 5.09375,
      "learning_rate": 4.682276724530633e-06,
      "loss": 0.79235182,
      "memory(GiB)": 147.13,
      "step": 45830,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.77587891,
      "epoch": 1.069481911372286,
      "grad_norm": 4.6875,
      "learning_rate": 4.680391453517026e-06,
      "loss": 0.81508999,
      "memory(GiB)": 147.13,
      "step": 45840,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.79230328,
      "epoch": 1.0697152189445749,
      "grad_norm": 7.84375,
      "learning_rate": 4.678506228127447e-06,
      "loss": 0.7369689,
      "memory(GiB)": 147.13,
      "step": 45850,
      "train_speed(iter/s)": 0.201421
    },
    {
      "acc": 0.78080015,
      "epoch": 1.0699485265168638,
      "grad_norm": 5.6875,
      "learning_rate": 4.67662104863101e-06,
      "loss": 0.80230227,
      "memory(GiB)": 147.13,
      "step": 45860,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.79010229,
      "epoch": 1.0701818340891527,
      "grad_norm": 5.09375,
      "learning_rate": 4.674735915296824e-06,
      "loss": 0.75055971,
      "memory(GiB)": 147.13,
      "step": 45870,
      "train_speed(iter/s)": 0.201467
    },
    {
      "acc": 0.75482688,
      "epoch": 1.0704151416614416,
      "grad_norm": 6.71875,
      "learning_rate": 4.672850828393992e-06,
      "loss": 0.89350119,
      "memory(GiB)": 147.13,
      "step": 45880,
      "train_speed(iter/s)": 0.20149
    },
    {
      "acc": 0.76217985,
      "epoch": 1.0706484492337305,
      "grad_norm": 5.5625,
      "learning_rate": 4.670965788191609e-06,
      "loss": 0.86493149,
      "memory(GiB)": 147.13,
      "step": 45890,
      "train_speed(iter/s)": 0.201512
    },
    {
      "acc": 0.7592453,
      "epoch": 1.0708817568060194,
      "grad_norm": 5.46875,
      "learning_rate": 4.669080794958764e-06,
      "loss": 0.85552521,
      "memory(GiB)": 147.13,
      "step": 45900,
      "train_speed(iter/s)": 0.201532
    },
    {
      "acc": 0.76834116,
      "epoch": 1.0711150643783083,
      "grad_norm": 4.40625,
      "learning_rate": 4.6671958489645394e-06,
      "loss": 0.85116701,
      "memory(GiB)": 147.13,
      "step": 45910,
      "train_speed(iter/s)": 0.201556
    },
    {
      "acc": 0.79082565,
      "epoch": 1.0713483719505972,
      "grad_norm": 6.0625,
      "learning_rate": 4.665310950478011e-06,
      "loss": 0.74479675,
      "memory(GiB)": 147.13,
      "step": 45920,
      "train_speed(iter/s)": 0.201579
    },
    {
      "acc": 0.7921772,
      "epoch": 1.071581679522886,
      "grad_norm": 5.5625,
      "learning_rate": 4.663426099768247e-06,
      "loss": 0.73003244,
      "memory(GiB)": 147.13,
      "step": 45930,
      "train_speed(iter/s)": 0.201601
    },
    {
      "acc": 0.76949821,
      "epoch": 1.071814987095175,
      "grad_norm": 5.375,
      "learning_rate": 4.661541297104309e-06,
      "loss": 0.83000374,
      "memory(GiB)": 147.13,
      "step": 45940,
      "train_speed(iter/s)": 0.201623
    },
    {
      "acc": 0.77961545,
      "epoch": 1.0720482946674639,
      "grad_norm": 6.34375,
      "learning_rate": 4.659656542755253e-06,
      "loss": 0.76818419,
      "memory(GiB)": 147.13,
      "step": 45950,
      "train_speed(iter/s)": 0.201647
    },
    {
      "acc": 0.78886552,
      "epoch": 1.0722816022397528,
      "grad_norm": 7.03125,
      "learning_rate": 4.657771836990127e-06,
      "loss": 0.73285408,
      "memory(GiB)": 147.13,
      "step": 45960,
      "train_speed(iter/s)": 0.20167
    },
    {
      "acc": 0.77267704,
      "epoch": 1.0725149098120417,
      "grad_norm": 5.125,
      "learning_rate": 4.655887180077973e-06,
      "loss": 0.80965977,
      "memory(GiB)": 147.13,
      "step": 45970,
      "train_speed(iter/s)": 0.201693
    },
    {
      "acc": 0.7777483,
      "epoch": 1.0727482173843306,
      "grad_norm": 5.625,
      "learning_rate": 4.654002572287822e-06,
      "loss": 0.79634504,
      "memory(GiB)": 147.13,
      "step": 45980,
      "train_speed(iter/s)": 0.201714
    },
    {
      "acc": 0.7671217,
      "epoch": 1.0729815249566195,
      "grad_norm": 5.125,
      "learning_rate": 4.652118013888704e-06,
      "loss": 0.84396877,
      "memory(GiB)": 147.13,
      "step": 45990,
      "train_speed(iter/s)": 0.201736
    },
    {
      "acc": 0.7591877,
      "epoch": 1.0732148325289084,
      "grad_norm": 6.5625,
      "learning_rate": 4.650233505149639e-06,
      "loss": 0.86070147,
      "memory(GiB)": 147.13,
      "step": 46000,
      "train_speed(iter/s)": 0.201761
    },
    {
      "epoch": 1.0732148325289084,
      "eval_acc": 0.7440973638497776,
      "eval_loss": 0.8066761493682861,
      "eval_runtime": 1270.8927,
      "eval_samples_per_second": 28.319,
      "eval_steps_per_second": 14.16,
      "step": 46000
    },
    {
      "acc": 0.77020998,
      "epoch": 1.0734481401011973,
      "grad_norm": 5.5,
      "learning_rate": 4.648349046339639e-06,
      "loss": 0.83288393,
      "memory(GiB)": 147.13,
      "step": 46010,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78606977,
      "epoch": 1.073681447673486,
      "grad_norm": 5.6875,
      "learning_rate": 4.64646463772771e-06,
      "loss": 0.76476479,
      "memory(GiB)": 147.13,
      "step": 46020,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.79527607,
      "epoch": 1.0739147552457748,
      "grad_norm": 9.0625,
      "learning_rate": 4.6445802795828515e-06,
      "loss": 0.72505322,
      "memory(GiB)": 147.13,
      "step": 46030,
      "train_speed(iter/s)": 0.200694
    },
    {
      "acc": 0.77043562,
      "epoch": 1.0741480628180637,
      "grad_norm": 5.59375,
      "learning_rate": 4.642695972174055e-06,
      "loss": 0.81447153,
      "memory(GiB)": 147.13,
      "step": 46040,
      "train_speed(iter/s)": 0.200717
    },
    {
      "acc": 0.78194885,
      "epoch": 1.0743813703903526,
      "grad_norm": 4.84375,
      "learning_rate": 4.640811715770305e-06,
      "loss": 0.78343391,
      "memory(GiB)": 147.13,
      "step": 46050,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.78372726,
      "epoch": 1.0746146779626415,
      "grad_norm": 4.78125,
      "learning_rate": 4.638927510640578e-06,
      "loss": 0.76459522,
      "memory(GiB)": 147.13,
      "step": 46060,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.79502363,
      "epoch": 1.0748479855349304,
      "grad_norm": 5.28125,
      "learning_rate": 4.637043357053844e-06,
      "loss": 0.72123518,
      "memory(GiB)": 147.13,
      "step": 46070,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.78148775,
      "epoch": 1.0750812931072193,
      "grad_norm": 6.90625,
      "learning_rate": 4.635159255279066e-06,
      "loss": 0.76372013,
      "memory(GiB)": 147.13,
      "step": 46080,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.7984148,
      "epoch": 1.0753146006795082,
      "grad_norm": 4.1875,
      "learning_rate": 4.633275205585198e-06,
      "loss": 0.69609632,
      "memory(GiB)": 147.13,
      "step": 46090,
      "train_speed(iter/s)": 0.200823
    },
    {
      "acc": 0.77495604,
      "epoch": 1.0755479082517971,
      "grad_norm": 6.375,
      "learning_rate": 4.631391208241187e-06,
      "loss": 0.80969296,
      "memory(GiB)": 147.13,
      "step": 46100,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.7904788,
      "epoch": 1.075781215824086,
      "grad_norm": 4.96875,
      "learning_rate": 4.6295072635159744e-06,
      "loss": 0.74659705,
      "memory(GiB)": 147.13,
      "step": 46110,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.77688103,
      "epoch": 1.076014523396375,
      "grad_norm": 4.75,
      "learning_rate": 4.627623371678492e-06,
      "loss": 0.8099329,
      "memory(GiB)": 147.13,
      "step": 46120,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.78068047,
      "epoch": 1.0762478309686638,
      "grad_norm": 4.375,
      "learning_rate": 4.625739532997665e-06,
      "loss": 0.7743185,
      "memory(GiB)": 147.13,
      "step": 46130,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.79984784,
      "epoch": 1.0764811385409527,
      "grad_norm": 5.28125,
      "learning_rate": 4.623855747742412e-06,
      "loss": 0.69627056,
      "memory(GiB)": 147.13,
      "step": 46140,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.77657719,
      "epoch": 1.0767144461132416,
      "grad_norm": 5.84375,
      "learning_rate": 4.62197201618164e-06,
      "loss": 0.78484745,
      "memory(GiB)": 147.13,
      "step": 46150,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.78479581,
      "epoch": 1.0769477536855305,
      "grad_norm": 4.75,
      "learning_rate": 4.620088338584254e-06,
      "loss": 0.76777401,
      "memory(GiB)": 147.13,
      "step": 46160,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.75793247,
      "epoch": 1.0771810612578194,
      "grad_norm": 5.03125,
      "learning_rate": 4.618204715219147e-06,
      "loss": 0.87096195,
      "memory(GiB)": 147.13,
      "step": 46170,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.78456974,
      "epoch": 1.0774143688301083,
      "grad_norm": 5.5,
      "learning_rate": 4.616321146355206e-06,
      "loss": 0.77649322,
      "memory(GiB)": 147.13,
      "step": 46180,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.76846437,
      "epoch": 1.0776476764023972,
      "grad_norm": 5.78125,
      "learning_rate": 4.614437632261311e-06,
      "loss": 0.82811508,
      "memory(GiB)": 147.13,
      "step": 46190,
      "train_speed(iter/s)": 0.201049
    },
    {
      "acc": 0.77970672,
      "epoch": 1.0778809839746861,
      "grad_norm": 10.0625,
      "learning_rate": 4.6125541732063315e-06,
      "loss": 0.80484219,
      "memory(GiB)": 147.13,
      "step": 46200,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.78650169,
      "epoch": 1.078114291546975,
      "grad_norm": 4.6875,
      "learning_rate": 4.6106707694591324e-06,
      "loss": 0.78840094,
      "memory(GiB)": 147.13,
      "step": 46210,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.77818255,
      "epoch": 1.078347599119264,
      "grad_norm": 5.28125,
      "learning_rate": 4.608787421288566e-06,
      "loss": 0.79254904,
      "memory(GiB)": 147.13,
      "step": 46220,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.77555122,
      "epoch": 1.0785809066915528,
      "grad_norm": 8.5,
      "learning_rate": 4.606904128963482e-06,
      "loss": 0.80696802,
      "memory(GiB)": 147.13,
      "step": 46230,
      "train_speed(iter/s)": 0.201142
    },
    {
      "acc": 0.77328091,
      "epoch": 1.0788142142638417,
      "grad_norm": 5.75,
      "learning_rate": 4.605020892752718e-06,
      "loss": 0.81441078,
      "memory(GiB)": 147.13,
      "step": 46240,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.7695261,
      "epoch": 1.0790475218361306,
      "grad_norm": 4.625,
      "learning_rate": 4.603137712925108e-06,
      "loss": 0.82396679,
      "memory(GiB)": 147.13,
      "step": 46250,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.77598724,
      "epoch": 1.0792808294084195,
      "grad_norm": 5.59375,
      "learning_rate": 4.601254589749474e-06,
      "loss": 0.8179862,
      "memory(GiB)": 147.13,
      "step": 46260,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78526249,
      "epoch": 1.0795141369807084,
      "grad_norm": 5.4375,
      "learning_rate": 4.599371523494632e-06,
      "loss": 0.77546606,
      "memory(GiB)": 147.13,
      "step": 46270,
      "train_speed(iter/s)": 0.201228
    },
    {
      "acc": 0.78726249,
      "epoch": 1.0797474445529973,
      "grad_norm": 4.15625,
      "learning_rate": 4.597488514429388e-06,
      "loss": 0.75475173,
      "memory(GiB)": 147.13,
      "step": 46280,
      "train_speed(iter/s)": 0.20125
    },
    {
      "acc": 0.78735552,
      "epoch": 1.0799807521252862,
      "grad_norm": 5.21875,
      "learning_rate": 4.595605562822542e-06,
      "loss": 0.75245075,
      "memory(GiB)": 147.13,
      "step": 46290,
      "train_speed(iter/s)": 0.201272
    },
    {
      "acc": 0.7679944,
      "epoch": 1.0802140596975751,
      "grad_norm": 6.5,
      "learning_rate": 4.593722668942884e-06,
      "loss": 0.83475809,
      "memory(GiB)": 147.13,
      "step": 46300,
      "train_speed(iter/s)": 0.201295
    },
    {
      "acc": 0.78632035,
      "epoch": 1.080447367269864,
      "grad_norm": 5.8125,
      "learning_rate": 4.5918398330592e-06,
      "loss": 0.75767183,
      "memory(GiB)": 147.13,
      "step": 46310,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.7837204,
      "epoch": 1.080680674842153,
      "grad_norm": 8.5625,
      "learning_rate": 4.589957055440259e-06,
      "loss": 0.77630086,
      "memory(GiB)": 147.13,
      "step": 46320,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.77561617,
      "epoch": 1.0809139824144418,
      "grad_norm": 6.125,
      "learning_rate": 4.588074336354828e-06,
      "loss": 0.80536242,
      "memory(GiB)": 147.13,
      "step": 46330,
      "train_speed(iter/s)": 0.201363
    },
    {
      "acc": 0.77097578,
      "epoch": 1.0811472899867307,
      "grad_norm": 4.96875,
      "learning_rate": 4.586191676071666e-06,
      "loss": 0.83300571,
      "memory(GiB)": 147.13,
      "step": 46340,
      "train_speed(iter/s)": 0.201385
    },
    {
      "acc": 0.78117247,
      "epoch": 1.0813805975590196,
      "grad_norm": 5.21875,
      "learning_rate": 4.584309074859524e-06,
      "loss": 0.80158958,
      "memory(GiB)": 147.13,
      "step": 46350,
      "train_speed(iter/s)": 0.201407
    },
    {
      "acc": 0.78816423,
      "epoch": 1.0816139051313085,
      "grad_norm": 4.625,
      "learning_rate": 4.5824265329871395e-06,
      "loss": 0.7538969,
      "memory(GiB)": 147.13,
      "step": 46360,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.77398424,
      "epoch": 1.0818472127035974,
      "grad_norm": 5.90625,
      "learning_rate": 4.580544050723246e-06,
      "loss": 0.8361598,
      "memory(GiB)": 147.13,
      "step": 46370,
      "train_speed(iter/s)": 0.201451
    },
    {
      "acc": 0.79671211,
      "epoch": 1.0820805202758863,
      "grad_norm": 5.40625,
      "learning_rate": 4.578661628336567e-06,
      "loss": 0.73220673,
      "memory(GiB)": 147.13,
      "step": 46380,
      "train_speed(iter/s)": 0.201472
    },
    {
      "acc": 0.7743166,
      "epoch": 1.082313827848175,
      "grad_norm": 4.96875,
      "learning_rate": 4.576779266095818e-06,
      "loss": 0.8441576,
      "memory(GiB)": 147.13,
      "step": 46390,
      "train_speed(iter/s)": 0.201494
    },
    {
      "acc": 0.76517391,
      "epoch": 1.082547135420464,
      "grad_norm": 6.0,
      "learning_rate": 4.574896964269707e-06,
      "loss": 0.84825592,
      "memory(GiB)": 147.13,
      "step": 46400,
      "train_speed(iter/s)": 0.201518
    },
    {
      "acc": 0.80378437,
      "epoch": 1.0827804429927528,
      "grad_norm": 4.28125,
      "learning_rate": 4.573014723126931e-06,
      "loss": 0.70220609,
      "memory(GiB)": 147.13,
      "step": 46410,
      "train_speed(iter/s)": 0.201541
    },
    {
      "acc": 0.78200688,
      "epoch": 1.0830137505650417,
      "grad_norm": 3.625,
      "learning_rate": 4.571132542936179e-06,
      "loss": 0.79165678,
      "memory(GiB)": 147.13,
      "step": 46420,
      "train_speed(iter/s)": 0.201563
    },
    {
      "acc": 0.79855556,
      "epoch": 1.0832470581373306,
      "grad_norm": 4.78125,
      "learning_rate": 4.569250423966132e-06,
      "loss": 0.72647519,
      "memory(GiB)": 147.13,
      "step": 46430,
      "train_speed(iter/s)": 0.201586
    },
    {
      "acc": 0.77577281,
      "epoch": 1.0834803657096195,
      "grad_norm": 4.875,
      "learning_rate": 4.567368366485462e-06,
      "loss": 0.7856792,
      "memory(GiB)": 147.13,
      "step": 46440,
      "train_speed(iter/s)": 0.201609
    },
    {
      "acc": 0.77464085,
      "epoch": 1.0837136732819084,
      "grad_norm": 5.59375,
      "learning_rate": 4.56548637076283e-06,
      "loss": 0.80171261,
      "memory(GiB)": 147.13,
      "step": 46450,
      "train_speed(iter/s)": 0.201632
    },
    {
      "acc": 0.79902134,
      "epoch": 1.0839469808541973,
      "grad_norm": 5.875,
      "learning_rate": 4.563604437066894e-06,
      "loss": 0.714886,
      "memory(GiB)": 147.13,
      "step": 46460,
      "train_speed(iter/s)": 0.201654
    },
    {
      "acc": 0.78830729,
      "epoch": 1.0841802884264862,
      "grad_norm": 4.71875,
      "learning_rate": 4.561722565666298e-06,
      "loss": 0.74691525,
      "memory(GiB)": 147.13,
      "step": 46470,
      "train_speed(iter/s)": 0.201676
    },
    {
      "acc": 0.77470942,
      "epoch": 1.084413595998775,
      "grad_norm": 5.34375,
      "learning_rate": 4.559840756829677e-06,
      "loss": 0.80980682,
      "memory(GiB)": 147.13,
      "step": 46480,
      "train_speed(iter/s)": 0.201698
    },
    {
      "acc": 0.79907804,
      "epoch": 1.084646903571064,
      "grad_norm": 4.375,
      "learning_rate": 4.557959010825662e-06,
      "loss": 0.70692854,
      "memory(GiB)": 147.13,
      "step": 46490,
      "train_speed(iter/s)": 0.201721
    },
    {
      "acc": 0.76710625,
      "epoch": 1.0848802111433529,
      "grad_norm": 5.40625,
      "learning_rate": 4.5560773279228686e-06,
      "loss": 0.82628555,
      "memory(GiB)": 147.13,
      "step": 46500,
      "train_speed(iter/s)": 0.201745
    },
    {
      "epoch": 1.0848802111433529,
      "eval_acc": 0.7440590893782576,
      "eval_loss": 0.8066306710243225,
      "eval_runtime": 1271.0341,
      "eval_samples_per_second": 28.316,
      "eval_steps_per_second": 14.159,
      "step": 46500
    },
    {
      "acc": 0.76087437,
      "epoch": 1.0851135187156418,
      "grad_norm": 7.875,
      "learning_rate": 4.5541957083899075e-06,
      "loss": 0.90284786,
      "memory(GiB)": 147.13,
      "step": 46510,
      "train_speed(iter/s)": 0.200645
    },
    {
      "acc": 0.78702526,
      "epoch": 1.0853468262879307,
      "grad_norm": 7.34375,
      "learning_rate": 4.55231415249538e-06,
      "loss": 0.76634626,
      "memory(GiB)": 147.13,
      "step": 46520,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.75945778,
      "epoch": 1.0855801338602196,
      "grad_norm": 5.5,
      "learning_rate": 4.550432660507877e-06,
      "loss": 0.8995718,
      "memory(GiB)": 147.13,
      "step": 46530,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.78146367,
      "epoch": 1.0858134414325085,
      "grad_norm": 5.5625,
      "learning_rate": 4.548551232695983e-06,
      "loss": 0.78113079,
      "memory(GiB)": 147.13,
      "step": 46540,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.77994041,
      "epoch": 1.0860467490047974,
      "grad_norm": 5.125,
      "learning_rate": 4.5466698693282675e-06,
      "loss": 0.79309092,
      "memory(GiB)": 147.13,
      "step": 46550,
      "train_speed(iter/s)": 0.200731
    },
    {
      "acc": 0.79029708,
      "epoch": 1.0862800565770863,
      "grad_norm": 6.3125,
      "learning_rate": 4.544788570673296e-06,
      "loss": 0.76324611,
      "memory(GiB)": 147.13,
      "step": 46560,
      "train_speed(iter/s)": 0.200754
    },
    {
      "acc": 0.77777467,
      "epoch": 1.0865133641493752,
      "grad_norm": 14.5625,
      "learning_rate": 4.542907336999625e-06,
      "loss": 0.81118917,
      "memory(GiB)": 147.13,
      "step": 46570,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.78099852,
      "epoch": 1.086746671721664,
      "grad_norm": 5.34375,
      "learning_rate": 4.541026168575798e-06,
      "loss": 0.76369138,
      "memory(GiB)": 147.13,
      "step": 46580,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.79265385,
      "epoch": 1.086979979293953,
      "grad_norm": 4.53125,
      "learning_rate": 4.539145065670353e-06,
      "loss": 0.74946361,
      "memory(GiB)": 147.13,
      "step": 46590,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.7763402,
      "epoch": 1.0872132868662419,
      "grad_norm": 7.15625,
      "learning_rate": 4.537264028551814e-06,
      "loss": 0.81608562,
      "memory(GiB)": 147.13,
      "step": 46600,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.79085789,
      "epoch": 1.0874465944385308,
      "grad_norm": 6.6875,
      "learning_rate": 4.535383057488702e-06,
      "loss": 0.75591068,
      "memory(GiB)": 147.13,
      "step": 46610,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.78365879,
      "epoch": 1.0876799020108197,
      "grad_norm": 5.75,
      "learning_rate": 4.533502152749523e-06,
      "loss": 0.78133993,
      "memory(GiB)": 147.13,
      "step": 46620,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.81641464,
      "epoch": 1.0879132095831086,
      "grad_norm": 4.9375,
      "learning_rate": 4.531621314602777e-06,
      "loss": 0.6454319,
      "memory(GiB)": 147.13,
      "step": 46630,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.7887289,
      "epoch": 1.0881465171553975,
      "grad_norm": 6.1875,
      "learning_rate": 4.529740543316952e-06,
      "loss": 0.75883055,
      "memory(GiB)": 147.13,
      "step": 46640,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.76232872,
      "epoch": 1.0883798247276864,
      "grad_norm": 4.0,
      "learning_rate": 4.52785983916053e-06,
      "loss": 0.8628355,
      "memory(GiB)": 147.13,
      "step": 46650,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.76253319,
      "epoch": 1.0886131322999753,
      "grad_norm": 6.25,
      "learning_rate": 4.525979202401976e-06,
      "loss": 0.85835104,
      "memory(GiB)": 147.13,
      "step": 46660,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.78178864,
      "epoch": 1.0888464398722641,
      "grad_norm": 5.8125,
      "learning_rate": 4.524098633309753e-06,
      "loss": 0.79391246,
      "memory(GiB)": 147.13,
      "step": 46670,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.7927947,
      "epoch": 1.089079747444553,
      "grad_norm": 8.4375,
      "learning_rate": 4.522218132152313e-06,
      "loss": 0.73775425,
      "memory(GiB)": 147.13,
      "step": 46680,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.77629023,
      "epoch": 1.089313055016842,
      "grad_norm": 5.25,
      "learning_rate": 4.520337699198095e-06,
      "loss": 0.80242271,
      "memory(GiB)": 147.13,
      "step": 46690,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.79058509,
      "epoch": 1.0895463625891308,
      "grad_norm": 4.4375,
      "learning_rate": 4.5184573347155316e-06,
      "loss": 0.73654985,
      "memory(GiB)": 147.13,
      "step": 46700,
      "train_speed(iter/s)": 0.201065
    },
    {
      "acc": 0.77961817,
      "epoch": 1.0897796701614197,
      "grad_norm": 6.5,
      "learning_rate": 4.516577038973044e-06,
      "loss": 0.7927855,
      "memory(GiB)": 147.13,
      "step": 46710,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.78442645,
      "epoch": 1.0900129777337086,
      "grad_norm": 5.34375,
      "learning_rate": 4.514696812239043e-06,
      "loss": 0.78609529,
      "memory(GiB)": 147.13,
      "step": 46720,
      "train_speed(iter/s)": 0.20111
    },
    {
      "acc": 0.79100533,
      "epoch": 1.0902462853059975,
      "grad_norm": 8.9375,
      "learning_rate": 4.512816654781931e-06,
      "loss": 0.74049482,
      "memory(GiB)": 147.13,
      "step": 46730,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.78928852,
      "epoch": 1.0904795928782864,
      "grad_norm": 8.875,
      "learning_rate": 4.5109365668701e-06,
      "loss": 0.73953748,
      "memory(GiB)": 147.13,
      "step": 46740,
      "train_speed(iter/s)": 0.201155
    },
    {
      "acc": 0.78580799,
      "epoch": 1.0907129004505753,
      "grad_norm": 6.53125,
      "learning_rate": 4.5090565487719326e-06,
      "loss": 0.75008278,
      "memory(GiB)": 147.13,
      "step": 46750,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.77508364,
      "epoch": 1.0909462080228642,
      "grad_norm": 5.125,
      "learning_rate": 4.5071766007558e-06,
      "loss": 0.81171427,
      "memory(GiB)": 147.13,
      "step": 46760,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.78406925,
      "epoch": 1.0911795155951531,
      "grad_norm": 5.96875,
      "learning_rate": 4.505296723090066e-06,
      "loss": 0.7646771,
      "memory(GiB)": 147.13,
      "step": 46770,
      "train_speed(iter/s)": 0.201222
    },
    {
      "acc": 0.78110132,
      "epoch": 1.0914128231674418,
      "grad_norm": 5.6875,
      "learning_rate": 4.503416916043079e-06,
      "loss": 0.76935401,
      "memory(GiB)": 147.13,
      "step": 46780,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.79187145,
      "epoch": 1.091646130739731,
      "grad_norm": 4.8125,
      "learning_rate": 4.501537179883184e-06,
      "loss": 0.74730501,
      "memory(GiB)": 147.13,
      "step": 46790,
      "train_speed(iter/s)": 0.201264
    },
    {
      "acc": 0.78256168,
      "epoch": 1.0918794383120196,
      "grad_norm": 4.96875,
      "learning_rate": 4.499657514878711e-06,
      "loss": 0.77862916,
      "memory(GiB)": 147.13,
      "step": 46800,
      "train_speed(iter/s)": 0.201287
    },
    {
      "acc": 0.77598977,
      "epoch": 1.0921127458843085,
      "grad_norm": 6.75,
      "learning_rate": 4.497777921297983e-06,
      "loss": 0.8359479,
      "memory(GiB)": 147.13,
      "step": 46810,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.77790146,
      "epoch": 1.0923460534565974,
      "grad_norm": 5.96875,
      "learning_rate": 4.49589839940931e-06,
      "loss": 0.81699162,
      "memory(GiB)": 147.13,
      "step": 46820,
      "train_speed(iter/s)": 0.201332
    },
    {
      "acc": 0.78884344,
      "epoch": 1.0925793610288863,
      "grad_norm": 4.9375,
      "learning_rate": 4.494018949480994e-06,
      "loss": 0.74211559,
      "memory(GiB)": 147.13,
      "step": 46830,
      "train_speed(iter/s)": 0.201354
    },
    {
      "acc": 0.77702522,
      "epoch": 1.0928126686011752,
      "grad_norm": 6.0625,
      "learning_rate": 4.492139571781328e-06,
      "loss": 0.79819546,
      "memory(GiB)": 147.13,
      "step": 46840,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.77434607,
      "epoch": 1.093045976173464,
      "grad_norm": 5.3125,
      "learning_rate": 4.490260266578589e-06,
      "loss": 0.81396513,
      "memory(GiB)": 147.13,
      "step": 46850,
      "train_speed(iter/s)": 0.201398
    },
    {
      "acc": 0.79454117,
      "epoch": 1.093279283745753,
      "grad_norm": 6.28125,
      "learning_rate": 4.4883810341410485e-06,
      "loss": 0.74300327,
      "memory(GiB)": 147.13,
      "step": 46860,
      "train_speed(iter/s)": 0.20142
    },
    {
      "acc": 0.77728949,
      "epoch": 1.093512591318042,
      "grad_norm": 4.96875,
      "learning_rate": 4.486501874736967e-06,
      "loss": 0.80336361,
      "memory(GiB)": 147.13,
      "step": 46870,
      "train_speed(iter/s)": 0.201442
    },
    {
      "acc": 0.79518509,
      "epoch": 1.0937458988903308,
      "grad_norm": 5.3125,
      "learning_rate": 4.484622788634596e-06,
      "loss": 0.7204237,
      "memory(GiB)": 147.13,
      "step": 46880,
      "train_speed(iter/s)": 0.201465
    },
    {
      "acc": 0.78115616,
      "epoch": 1.0939792064626197,
      "grad_norm": 5.8125,
      "learning_rate": 4.48274377610217e-06,
      "loss": 0.78942556,
      "memory(GiB)": 147.13,
      "step": 46890,
      "train_speed(iter/s)": 0.201487
    },
    {
      "acc": 0.78453555,
      "epoch": 1.0942125140349086,
      "grad_norm": 4.6875,
      "learning_rate": 4.480864837407919e-06,
      "loss": 0.76847844,
      "memory(GiB)": 147.13,
      "step": 46900,
      "train_speed(iter/s)": 0.201509
    },
    {
      "acc": 0.80208855,
      "epoch": 1.0944458216071975,
      "grad_norm": 5.125,
      "learning_rate": 4.478985972820063e-06,
      "loss": 0.69647503,
      "memory(GiB)": 147.13,
      "step": 46910,
      "train_speed(iter/s)": 0.20153
    },
    {
      "acc": 0.77438059,
      "epoch": 1.0946791291794864,
      "grad_norm": 5.25,
      "learning_rate": 4.477107182606807e-06,
      "loss": 0.82461748,
      "memory(GiB)": 147.13,
      "step": 46920,
      "train_speed(iter/s)": 0.201553
    },
    {
      "acc": 0.7742578,
      "epoch": 1.0949124367517753,
      "grad_norm": 6.0625,
      "learning_rate": 4.4752284670363495e-06,
      "loss": 0.80661144,
      "memory(GiB)": 147.13,
      "step": 46930,
      "train_speed(iter/s)": 0.201575
    },
    {
      "acc": 0.78175735,
      "epoch": 1.0951457443240642,
      "grad_norm": 5.75,
      "learning_rate": 4.473349826376876e-06,
      "loss": 0.80005064,
      "memory(GiB)": 147.13,
      "step": 46940,
      "train_speed(iter/s)": 0.201597
    },
    {
      "acc": 0.78072362,
      "epoch": 1.095379051896353,
      "grad_norm": 6.78125,
      "learning_rate": 4.471471260896561e-06,
      "loss": 0.80625315,
      "memory(GiB)": 147.13,
      "step": 46950,
      "train_speed(iter/s)": 0.20162
    },
    {
      "acc": 0.78413506,
      "epoch": 1.095612359468642,
      "grad_norm": 4.4375,
      "learning_rate": 4.46959277086357e-06,
      "loss": 0.76228724,
      "memory(GiB)": 147.13,
      "step": 46960,
      "train_speed(iter/s)": 0.201642
    },
    {
      "acc": 0.7963975,
      "epoch": 1.095845667040931,
      "grad_norm": 5.03125,
      "learning_rate": 4.467714356546057e-06,
      "loss": 0.71428413,
      "memory(GiB)": 147.13,
      "step": 46970,
      "train_speed(iter/s)": 0.201664
    },
    {
      "acc": 0.79295254,
      "epoch": 1.0960789746132198,
      "grad_norm": 5.09375,
      "learning_rate": 4.465836018212166e-06,
      "loss": 0.72961626,
      "memory(GiB)": 147.13,
      "step": 46980,
      "train_speed(iter/s)": 0.201686
    },
    {
      "acc": 0.76461864,
      "epoch": 1.0963122821855087,
      "grad_norm": 5.84375,
      "learning_rate": 4.463957756130028e-06,
      "loss": 0.82763643,
      "memory(GiB)": 147.13,
      "step": 46990,
      "train_speed(iter/s)": 0.201709
    },
    {
      "acc": 0.77132835,
      "epoch": 1.0965455897577976,
      "grad_norm": 6.9375,
      "learning_rate": 4.462079570567765e-06,
      "loss": 0.81077747,
      "memory(GiB)": 147.13,
      "step": 47000,
      "train_speed(iter/s)": 0.20173
    },
    {
      "epoch": 1.0965455897577976,
      "eval_acc": 0.7441737526485854,
      "eval_loss": 0.8062352538108826,
      "eval_runtime": 1270.2715,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 47000
    },
    {
      "acc": 0.77072382,
      "epoch": 1.0967788973300865,
      "grad_norm": 4.21875,
      "learning_rate": 4.460201461793486e-06,
      "loss": 0.83631325,
      "memory(GiB)": 147.13,
      "step": 47010,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.79384909,
      "epoch": 1.0970122049023754,
      "grad_norm": 6.21875,
      "learning_rate": 4.458323430075292e-06,
      "loss": 0.74116993,
      "memory(GiB)": 147.13,
      "step": 47020,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.7897253,
      "epoch": 1.0972455124746643,
      "grad_norm": 4.375,
      "learning_rate": 4.45644547568127e-06,
      "loss": 0.75881343,
      "memory(GiB)": 147.13,
      "step": 47030,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.78165913,
      "epoch": 1.0974788200469532,
      "grad_norm": 4.9375,
      "learning_rate": 4.4545675988795e-06,
      "loss": 0.79373956,
      "memory(GiB)": 147.13,
      "step": 47040,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.78019323,
      "epoch": 1.097712127619242,
      "grad_norm": 5.5625,
      "learning_rate": 4.452689799938045e-06,
      "loss": 0.79432139,
      "memory(GiB)": 147.13,
      "step": 47050,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.76499996,
      "epoch": 1.097945435191531,
      "grad_norm": 6.0625,
      "learning_rate": 4.450812079124964e-06,
      "loss": 0.85962629,
      "memory(GiB)": 147.13,
      "step": 47060,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.78296137,
      "epoch": 1.0981787427638199,
      "grad_norm": 4.84375,
      "learning_rate": 4.448934436708297e-06,
      "loss": 0.78475742,
      "memory(GiB)": 147.13,
      "step": 47070,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.78817444,
      "epoch": 1.0984120503361088,
      "grad_norm": 6.28125,
      "learning_rate": 4.44705687295608e-06,
      "loss": 0.77784424,
      "memory(GiB)": 147.13,
      "step": 47080,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.78499746,
      "epoch": 1.0986453579083977,
      "grad_norm": 7.09375,
      "learning_rate": 4.445179388136335e-06,
      "loss": 0.760812,
      "memory(GiB)": 147.13,
      "step": 47090,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.78877263,
      "epoch": 1.0988786654806866,
      "grad_norm": 6.59375,
      "learning_rate": 4.44330198251707e-06,
      "loss": 0.73901072,
      "memory(GiB)": 147.13,
      "step": 47100,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.75595312,
      "epoch": 1.0991119730529755,
      "grad_norm": 6.3125,
      "learning_rate": 4.441424656366287e-06,
      "loss": 0.85828323,
      "memory(GiB)": 147.13,
      "step": 47110,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.78338614,
      "epoch": 1.0993452806252644,
      "grad_norm": 6.90625,
      "learning_rate": 4.43954740995197e-06,
      "loss": 0.77686892,
      "memory(GiB)": 147.13,
      "step": 47120,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.79634991,
      "epoch": 1.0995785881975533,
      "grad_norm": 7.125,
      "learning_rate": 4.437670243542097e-06,
      "loss": 0.72251873,
      "memory(GiB)": 147.13,
      "step": 47130,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.7735898,
      "epoch": 1.0998118957698422,
      "grad_norm": 5.3125,
      "learning_rate": 4.435793157404636e-06,
      "loss": 0.81141148,
      "memory(GiB)": 147.13,
      "step": 47140,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.76752191,
      "epoch": 1.100045203342131,
      "grad_norm": 5.78125,
      "learning_rate": 4.433916151807535e-06,
      "loss": 0.83682642,
      "memory(GiB)": 147.13,
      "step": 47150,
      "train_speed(iter/s)": 0.200948
    },
    {
      "acc": 0.78204641,
      "epoch": 1.10027851091442,
      "grad_norm": 5.5625,
      "learning_rate": 4.43203922701874e-06,
      "loss": 0.77274141,
      "memory(GiB)": 147.13,
      "step": 47160,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.7881916,
      "epoch": 1.1005118184867086,
      "grad_norm": 4.9375,
      "learning_rate": 4.43016238330618e-06,
      "loss": 0.74384875,
      "memory(GiB)": 147.13,
      "step": 47170,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.78752022,
      "epoch": 1.1007451260589978,
      "grad_norm": 3.984375,
      "learning_rate": 4.428285620937774e-06,
      "loss": 0.76339617,
      "memory(GiB)": 147.13,
      "step": 47180,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.76733403,
      "epoch": 1.1009784336312864,
      "grad_norm": 5.78125,
      "learning_rate": 4.4264089401814306e-06,
      "loss": 0.84128456,
      "memory(GiB)": 147.13,
      "step": 47190,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.78480959,
      "epoch": 1.1012117412035753,
      "grad_norm": 4.625,
      "learning_rate": 4.4245323413050446e-06,
      "loss": 0.76962199,
      "memory(GiB)": 147.13,
      "step": 47200,
      "train_speed(iter/s)": 0.201057
    },
    {
      "acc": 0.7589119,
      "epoch": 1.1014450487758642,
      "grad_norm": 4.6875,
      "learning_rate": 4.422655824576499e-06,
      "loss": 0.88679562,
      "memory(GiB)": 147.13,
      "step": 47210,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.80034866,
      "epoch": 1.1016783563481531,
      "grad_norm": 4.90625,
      "learning_rate": 4.420779390263669e-06,
      "loss": 0.70025153,
      "memory(GiB)": 147.13,
      "step": 47220,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.78663492,
      "epoch": 1.101911663920442,
      "grad_norm": 4.90625,
      "learning_rate": 4.4189030386344094e-06,
      "loss": 0.76303225,
      "memory(GiB)": 147.13,
      "step": 47230,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.77694702,
      "epoch": 1.102144971492731,
      "grad_norm": 6.90625,
      "learning_rate": 4.417026769956573e-06,
      "loss": 0.79356718,
      "memory(GiB)": 147.13,
      "step": 47240,
      "train_speed(iter/s)": 0.201143
    },
    {
      "acc": 0.77413359,
      "epoch": 1.1023782790650198,
      "grad_norm": 4.9375,
      "learning_rate": 4.415150584497996e-06,
      "loss": 0.81121407,
      "memory(GiB)": 147.13,
      "step": 47250,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.74965243,
      "epoch": 1.1026115866373087,
      "grad_norm": 6.5,
      "learning_rate": 4.413274482526503e-06,
      "loss": 0.91974735,
      "memory(GiB)": 147.13,
      "step": 47260,
      "train_speed(iter/s)": 0.201187
    },
    {
      "acc": 0.78864617,
      "epoch": 1.1028448942095976,
      "grad_norm": 5.28125,
      "learning_rate": 4.4113984643099075e-06,
      "loss": 0.75029535,
      "memory(GiB)": 147.13,
      "step": 47270,
      "train_speed(iter/s)": 0.201207
    },
    {
      "acc": 0.79483843,
      "epoch": 1.1030782017818865,
      "grad_norm": 5.21875,
      "learning_rate": 4.409522530116011e-06,
      "loss": 0.73387432,
      "memory(GiB)": 147.13,
      "step": 47280,
      "train_speed(iter/s)": 0.20123
    },
    {
      "acc": 0.7674356,
      "epoch": 1.1033115093541754,
      "grad_norm": 4.6875,
      "learning_rate": 4.407646680212601e-06,
      "loss": 0.85647984,
      "memory(GiB)": 147.13,
      "step": 47290,
      "train_speed(iter/s)": 0.201252
    },
    {
      "acc": 0.78836451,
      "epoch": 1.1035448169264643,
      "grad_norm": 5.90625,
      "learning_rate": 4.405770914867455e-06,
      "loss": 0.74849033,
      "memory(GiB)": 147.13,
      "step": 47300,
      "train_speed(iter/s)": 0.201274
    },
    {
      "acc": 0.78828807,
      "epoch": 1.1037781244987532,
      "grad_norm": 7.03125,
      "learning_rate": 4.403895234348338e-06,
      "loss": 0.74482512,
      "memory(GiB)": 147.13,
      "step": 47310,
      "train_speed(iter/s)": 0.201297
    },
    {
      "acc": 0.77510147,
      "epoch": 1.1040114320710421,
      "grad_norm": 5.4375,
      "learning_rate": 4.402019638923003e-06,
      "loss": 0.82446518,
      "memory(GiB)": 147.13,
      "step": 47320,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.78557339,
      "epoch": 1.104244739643331,
      "grad_norm": 5.3125,
      "learning_rate": 4.400144128859192e-06,
      "loss": 0.76738482,
      "memory(GiB)": 147.13,
      "step": 47330,
      "train_speed(iter/s)": 0.201342
    },
    {
      "acc": 0.79229355,
      "epoch": 1.10447804721562,
      "grad_norm": 4.28125,
      "learning_rate": 4.3982687044246336e-06,
      "loss": 0.76299677,
      "memory(GiB)": 147.13,
      "step": 47340,
      "train_speed(iter/s)": 0.201363
    },
    {
      "acc": 0.79036603,
      "epoch": 1.1047113547879088,
      "grad_norm": 6.1875,
      "learning_rate": 4.396393365887041e-06,
      "loss": 0.74940066,
      "memory(GiB)": 147.13,
      "step": 47350,
      "train_speed(iter/s)": 0.201384
    },
    {
      "acc": 0.78618107,
      "epoch": 1.1049446623601977,
      "grad_norm": 6.125,
      "learning_rate": 4.394518113514121e-06,
      "loss": 0.74857254,
      "memory(GiB)": 147.13,
      "step": 47360,
      "train_speed(iter/s)": 0.201407
    },
    {
      "acc": 0.78618221,
      "epoch": 1.1051779699324866,
      "grad_norm": 5.65625,
      "learning_rate": 4.392642947573563e-06,
      "loss": 0.75713739,
      "memory(GiB)": 147.13,
      "step": 47370,
      "train_speed(iter/s)": 0.20143
    },
    {
      "acc": 0.77709045,
      "epoch": 1.1054112775047755,
      "grad_norm": 4.9375,
      "learning_rate": 4.3907678683330486e-06,
      "loss": 0.7963932,
      "memory(GiB)": 147.13,
      "step": 47380,
      "train_speed(iter/s)": 0.201452
    },
    {
      "acc": 0.78100262,
      "epoch": 1.1056445850770644,
      "grad_norm": 5.375,
      "learning_rate": 4.388892876060243e-06,
      "loss": 0.79459877,
      "memory(GiB)": 147.13,
      "step": 47390,
      "train_speed(iter/s)": 0.201473
    },
    {
      "acc": 0.78773174,
      "epoch": 1.1058778926493533,
      "grad_norm": 8.875,
      "learning_rate": 4.387017971022803e-06,
      "loss": 0.74201722,
      "memory(GiB)": 147.13,
      "step": 47400,
      "train_speed(iter/s)": 0.201496
    },
    {
      "acc": 0.7911149,
      "epoch": 1.1061112002216422,
      "grad_norm": 47.25,
      "learning_rate": 4.385143153488369e-06,
      "loss": 0.74900484,
      "memory(GiB)": 147.13,
      "step": 47410,
      "train_speed(iter/s)": 0.201518
    },
    {
      "acc": 0.77765427,
      "epoch": 1.1063445077939311,
      "grad_norm": 12.8125,
      "learning_rate": 4.383268423724572e-06,
      "loss": 0.78825626,
      "memory(GiB)": 147.13,
      "step": 47420,
      "train_speed(iter/s)": 0.201539
    },
    {
      "acc": 0.76809454,
      "epoch": 1.10657781536622,
      "grad_norm": 5.375,
      "learning_rate": 4.381393781999027e-06,
      "loss": 0.84127693,
      "memory(GiB)": 147.13,
      "step": 47430,
      "train_speed(iter/s)": 0.201561
    },
    {
      "acc": 0.80026913,
      "epoch": 1.106811122938509,
      "grad_norm": 4.75,
      "learning_rate": 4.379519228579342e-06,
      "loss": 0.69325418,
      "memory(GiB)": 147.13,
      "step": 47440,
      "train_speed(iter/s)": 0.201583
    },
    {
      "acc": 0.78321247,
      "epoch": 1.1070444305107978,
      "grad_norm": 7.125,
      "learning_rate": 4.377644763733106e-06,
      "loss": 0.79312048,
      "memory(GiB)": 147.13,
      "step": 47450,
      "train_speed(iter/s)": 0.201604
    },
    {
      "acc": 0.7822238,
      "epoch": 1.1072777380830867,
      "grad_norm": 4.625,
      "learning_rate": 4.375770387727899e-06,
      "loss": 0.79141903,
      "memory(GiB)": 147.13,
      "step": 47460,
      "train_speed(iter/s)": 0.201627
    },
    {
      "acc": 0.7735548,
      "epoch": 1.1075110456553756,
      "grad_norm": 5.5625,
      "learning_rate": 4.373896100831288e-06,
      "loss": 0.81294842,
      "memory(GiB)": 147.13,
      "step": 47470,
      "train_speed(iter/s)": 0.20165
    },
    {
      "acc": 0.77273912,
      "epoch": 1.1077443532276645,
      "grad_norm": 5.96875,
      "learning_rate": 4.372021903310826e-06,
      "loss": 0.81545506,
      "memory(GiB)": 147.13,
      "step": 47480,
      "train_speed(iter/s)": 0.201672
    },
    {
      "acc": 0.77078466,
      "epoch": 1.1079776607999534,
      "grad_norm": 6.375,
      "learning_rate": 4.370147795434054e-06,
      "loss": 0.84389324,
      "memory(GiB)": 147.13,
      "step": 47490,
      "train_speed(iter/s)": 0.201694
    },
    {
      "acc": 0.81073818,
      "epoch": 1.1082109683722423,
      "grad_norm": 6.0625,
      "learning_rate": 4.3682737774685035e-06,
      "loss": 0.68347592,
      "memory(GiB)": 147.13,
      "step": 47500,
      "train_speed(iter/s)": 0.201717
    },
    {
      "epoch": 1.1082109683722423,
      "eval_acc": 0.7442459776973533,
      "eval_loss": 0.8062799572944641,
      "eval_runtime": 1271.1815,
      "eval_samples_per_second": 28.313,
      "eval_steps_per_second": 14.157,
      "step": 47500
    },
    {
      "acc": 0.77623062,
      "epoch": 1.1084442759445312,
      "grad_norm": 5.125,
      "learning_rate": 4.366399849681686e-06,
      "loss": 0.78513613,
      "memory(GiB)": 147.13,
      "step": 47510,
      "train_speed(iter/s)": 0.200641
    },
    {
      "acc": 0.78764448,
      "epoch": 1.10867758351682,
      "grad_norm": 7.3125,
      "learning_rate": 4.364526012341107e-06,
      "loss": 0.74058805,
      "memory(GiB)": 147.13,
      "step": 47520,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.77476292,
      "epoch": 1.108910891089109,
      "grad_norm": 4.53125,
      "learning_rate": 4.362652265714254e-06,
      "loss": 0.79294233,
      "memory(GiB)": 147.13,
      "step": 47530,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.75863619,
      "epoch": 1.109144198661398,
      "grad_norm": 6.375,
      "learning_rate": 4.360778610068605e-06,
      "loss": 0.88454666,
      "memory(GiB)": 147.13,
      "step": 47540,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.78705368,
      "epoch": 1.1093775062336868,
      "grad_norm": 6.1875,
      "learning_rate": 4.3589050456716254e-06,
      "loss": 0.74554396,
      "memory(GiB)": 147.13,
      "step": 47550,
      "train_speed(iter/s)": 0.200727
    },
    {
      "acc": 0.79435959,
      "epoch": 1.1096108138059755,
      "grad_norm": 4.65625,
      "learning_rate": 4.357031572790763e-06,
      "loss": 0.74123554,
      "memory(GiB)": 147.13,
      "step": 47560,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76363263,
      "epoch": 1.1098441213782644,
      "grad_norm": 5.9375,
      "learning_rate": 4.355158191693458e-06,
      "loss": 0.87095814,
      "memory(GiB)": 147.13,
      "step": 47570,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.77314081,
      "epoch": 1.1100774289505533,
      "grad_norm": 6.1875,
      "learning_rate": 4.353284902647133e-06,
      "loss": 0.82467804,
      "memory(GiB)": 147.13,
      "step": 47580,
      "train_speed(iter/s)": 0.200788
    },
    {
      "acc": 0.7841754,
      "epoch": 1.1103107365228422,
      "grad_norm": 4.6875,
      "learning_rate": 4.351411705919201e-06,
      "loss": 0.78773432,
      "memory(GiB)": 147.13,
      "step": 47590,
      "train_speed(iter/s)": 0.200809
    },
    {
      "acc": 0.80135212,
      "epoch": 1.110544044095131,
      "grad_norm": 5.75,
      "learning_rate": 4.349538601777058e-06,
      "loss": 0.74643879,
      "memory(GiB)": 147.13,
      "step": 47600,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.77409286,
      "epoch": 1.11077735166742,
      "grad_norm": 4.28125,
      "learning_rate": 4.347665590488091e-06,
      "loss": 0.81729527,
      "memory(GiB)": 147.13,
      "step": 47610,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.80320225,
      "epoch": 1.1110106592397089,
      "grad_norm": 4.34375,
      "learning_rate": 4.3457926723196716e-06,
      "loss": 0.69598637,
      "memory(GiB)": 147.13,
      "step": 47620,
      "train_speed(iter/s)": 0.200873
    },
    {
      "acc": 0.79413824,
      "epoch": 1.1112439668119978,
      "grad_norm": 5.5,
      "learning_rate": 4.343919847539157e-06,
      "loss": 0.7392138,
      "memory(GiB)": 147.13,
      "step": 47630,
      "train_speed(iter/s)": 0.200895
    },
    {
      "acc": 0.76051364,
      "epoch": 1.1114772743842867,
      "grad_norm": 6.71875,
      "learning_rate": 4.342047116413897e-06,
      "loss": 0.88393078,
      "memory(GiB)": 147.13,
      "step": 47640,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.77696719,
      "epoch": 1.1117105819565756,
      "grad_norm": 4.84375,
      "learning_rate": 4.340174479211217e-06,
      "loss": 0.80170889,
      "memory(GiB)": 147.13,
      "step": 47650,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.77932134,
      "epoch": 1.1119438895288645,
      "grad_norm": 5.65625,
      "learning_rate": 4.338301936198439e-06,
      "loss": 0.79549494,
      "memory(GiB)": 147.13,
      "step": 47660,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.77386909,
      "epoch": 1.1121771971011534,
      "grad_norm": 5.4375,
      "learning_rate": 4.336429487642867e-06,
      "loss": 0.82580423,
      "memory(GiB)": 147.13,
      "step": 47670,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.79550128,
      "epoch": 1.1124105046734423,
      "grad_norm": 5.28125,
      "learning_rate": 4.334557133811796e-06,
      "loss": 0.71878395,
      "memory(GiB)": 147.13,
      "step": 47680,
      "train_speed(iter/s)": 0.201003
    },
    {
      "acc": 0.77494755,
      "epoch": 1.1126438122457312,
      "grad_norm": 5.9375,
      "learning_rate": 4.332684874972498e-06,
      "loss": 0.80830698,
      "memory(GiB)": 147.13,
      "step": 47690,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.80317068,
      "epoch": 1.11287711981802,
      "grad_norm": 5.78125,
      "learning_rate": 4.330812711392241e-06,
      "loss": 0.7105298,
      "memory(GiB)": 147.13,
      "step": 47700,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.78965735,
      "epoch": 1.113110427390309,
      "grad_norm": 3.765625,
      "learning_rate": 4.328940643338274e-06,
      "loss": 0.742593,
      "memory(GiB)": 147.13,
      "step": 47710,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.78511152,
      "epoch": 1.1133437349625979,
      "grad_norm": 5.40625,
      "learning_rate": 4.327068671077836e-06,
      "loss": 0.78913937,
      "memory(GiB)": 147.13,
      "step": 47720,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.78472223,
      "epoch": 1.1135770425348868,
      "grad_norm": 5.71875,
      "learning_rate": 4.32519679487815e-06,
      "loss": 0.77086167,
      "memory(GiB)": 147.13,
      "step": 47730,
      "train_speed(iter/s)": 0.20111
    },
    {
      "acc": 0.77630825,
      "epoch": 1.1138103501071757,
      "grad_norm": 6.6875,
      "learning_rate": 4.323325015006425e-06,
      "loss": 0.82653818,
      "memory(GiB)": 147.13,
      "step": 47740,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.76846933,
      "epoch": 1.1140436576794646,
      "grad_norm": 5.65625,
      "learning_rate": 4.321453331729857e-06,
      "loss": 0.82080832,
      "memory(GiB)": 147.13,
      "step": 47750,
      "train_speed(iter/s)": 0.201153
    },
    {
      "acc": 0.78095446,
      "epoch": 1.1142769652517535,
      "grad_norm": 3.890625,
      "learning_rate": 4.319581745315629e-06,
      "loss": 0.78204536,
      "memory(GiB)": 147.13,
      "step": 47760,
      "train_speed(iter/s)": 0.201175
    },
    {
      "acc": 0.77750607,
      "epoch": 1.1145102728240424,
      "grad_norm": 5.4375,
      "learning_rate": 4.317710256030911e-06,
      "loss": 0.82766047,
      "memory(GiB)": 147.13,
      "step": 47770,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.77918196,
      "epoch": 1.1147435803963313,
      "grad_norm": 7.03125,
      "learning_rate": 4.3158388641428536e-06,
      "loss": 0.7857625,
      "memory(GiB)": 147.13,
      "step": 47780,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.77556639,
      "epoch": 1.1149768879686202,
      "grad_norm": 5.21875,
      "learning_rate": 4.3139675699186e-06,
      "loss": 0.81662083,
      "memory(GiB)": 147.13,
      "step": 47790,
      "train_speed(iter/s)": 0.20124
    },
    {
      "acc": 0.79196119,
      "epoch": 1.115210195540909,
      "grad_norm": 6.40625,
      "learning_rate": 4.312096373625279e-06,
      "loss": 0.75588217,
      "memory(GiB)": 147.13,
      "step": 47800,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.77584,
      "epoch": 1.115443503113198,
      "grad_norm": 6.6875,
      "learning_rate": 4.310225275529998e-06,
      "loss": 0.79082928,
      "memory(GiB)": 147.13,
      "step": 47810,
      "train_speed(iter/s)": 0.201282
    },
    {
      "acc": 0.76462359,
      "epoch": 1.1156768106854869,
      "grad_norm": 4.78125,
      "learning_rate": 4.308354275899859e-06,
      "loss": 0.8640892,
      "memory(GiB)": 147.13,
      "step": 47820,
      "train_speed(iter/s)": 0.201304
    },
    {
      "acc": 0.8109581,
      "epoch": 1.1159101182577758,
      "grad_norm": 7.6875,
      "learning_rate": 4.306483375001946e-06,
      "loss": 0.65745254,
      "memory(GiB)": 147.13,
      "step": 47830,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.770755,
      "epoch": 1.1161434258300647,
      "grad_norm": 4.6875,
      "learning_rate": 4.30461257310333e-06,
      "loss": 0.84625854,
      "memory(GiB)": 147.13,
      "step": 47840,
      "train_speed(iter/s)": 0.201348
    },
    {
      "acc": 0.77912922,
      "epoch": 1.1163767334023535,
      "grad_norm": 4.78125,
      "learning_rate": 4.302741870471069e-06,
      "loss": 0.76887741,
      "memory(GiB)": 147.13,
      "step": 47850,
      "train_speed(iter/s)": 0.201369
    },
    {
      "acc": 0.78028278,
      "epoch": 1.1166100409746424,
      "grad_norm": 8.1875,
      "learning_rate": 4.3008712673722005e-06,
      "loss": 0.80194454,
      "memory(GiB)": 147.13,
      "step": 47860,
      "train_speed(iter/s)": 0.201392
    },
    {
      "acc": 0.79243369,
      "epoch": 1.1168433485469313,
      "grad_norm": 5.6875,
      "learning_rate": 4.299000764073757e-06,
      "loss": 0.733636,
      "memory(GiB)": 147.13,
      "step": 47870,
      "train_speed(iter/s)": 0.201412
    },
    {
      "acc": 0.78628654,
      "epoch": 1.1170766561192202,
      "grad_norm": 5.1875,
      "learning_rate": 4.29713036084275e-06,
      "loss": 0.76849709,
      "memory(GiB)": 147.13,
      "step": 47880,
      "train_speed(iter/s)": 0.201434
    },
    {
      "acc": 0.78145351,
      "epoch": 1.1173099636915091,
      "grad_norm": 5.53125,
      "learning_rate": 4.29526005794618e-06,
      "loss": 0.79676495,
      "memory(GiB)": 147.13,
      "step": 47890,
      "train_speed(iter/s)": 0.201454
    },
    {
      "acc": 0.7888092,
      "epoch": 1.117543271263798,
      "grad_norm": 6.4375,
      "learning_rate": 4.2933898556510325e-06,
      "loss": 0.7636147,
      "memory(GiB)": 147.13,
      "step": 47900,
      "train_speed(iter/s)": 0.201475
    },
    {
      "acc": 0.77102165,
      "epoch": 1.117776578836087,
      "grad_norm": 10.4375,
      "learning_rate": 4.29151975422428e-06,
      "loss": 0.8149704,
      "memory(GiB)": 147.13,
      "step": 47910,
      "train_speed(iter/s)": 0.201496
    },
    {
      "acc": 0.7934557,
      "epoch": 1.1180098864083758,
      "grad_norm": 5.0,
      "learning_rate": 4.289649753932874e-06,
      "loss": 0.72969637,
      "memory(GiB)": 147.13,
      "step": 47920,
      "train_speed(iter/s)": 0.201517
    },
    {
      "acc": 0.79851007,
      "epoch": 1.1182431939806645,
      "grad_norm": 5.03125,
      "learning_rate": 4.28777985504376e-06,
      "loss": 0.72625418,
      "memory(GiB)": 147.13,
      "step": 47930,
      "train_speed(iter/s)": 0.201538
    },
    {
      "acc": 0.77621775,
      "epoch": 1.1184765015529536,
      "grad_norm": 6.03125,
      "learning_rate": 4.285910057823864e-06,
      "loss": 0.83182402,
      "memory(GiB)": 147.13,
      "step": 47940,
      "train_speed(iter/s)": 0.201559
    },
    {
      "acc": 0.79367595,
      "epoch": 1.1187098091252423,
      "grad_norm": 5.25,
      "learning_rate": 4.284040362540101e-06,
      "loss": 0.73981581,
      "memory(GiB)": 147.13,
      "step": 47950,
      "train_speed(iter/s)": 0.201581
    },
    {
      "acc": 0.78099489,
      "epoch": 1.1189431166975312,
      "grad_norm": 4.21875,
      "learning_rate": 4.282170769459367e-06,
      "loss": 0.79089422,
      "memory(GiB)": 147.13,
      "step": 47960,
      "train_speed(iter/s)": 0.201603
    },
    {
      "acc": 0.79144993,
      "epoch": 1.1191764242698201,
      "grad_norm": 4.5625,
      "learning_rate": 4.2803012788485475e-06,
      "loss": 0.74069662,
      "memory(GiB)": 147.13,
      "step": 47970,
      "train_speed(iter/s)": 0.201622
    },
    {
      "acc": 0.77184448,
      "epoch": 1.119409731842109,
      "grad_norm": 5.5625,
      "learning_rate": 4.278431890974511e-06,
      "loss": 0.82890167,
      "memory(GiB)": 147.13,
      "step": 47980,
      "train_speed(iter/s)": 0.201646
    },
    {
      "acc": 0.77109141,
      "epoch": 1.119643039414398,
      "grad_norm": 8.125,
      "learning_rate": 4.276562606104114e-06,
      "loss": 0.80681925,
      "memory(GiB)": 147.13,
      "step": 47990,
      "train_speed(iter/s)": 0.201667
    },
    {
      "acc": 0.78441682,
      "epoch": 1.1198763469866868,
      "grad_norm": 4.875,
      "learning_rate": 4.274693424504194e-06,
      "loss": 0.78442831,
      "memory(GiB)": 147.13,
      "step": 48000,
      "train_speed(iter/s)": 0.201688
    },
    {
      "epoch": 1.1198763469866868,
      "eval_acc": 0.744195852552643,
      "eval_loss": 0.806307315826416,
      "eval_runtime": 1268.7398,
      "eval_samples_per_second": 28.368,
      "eval_steps_per_second": 14.184,
      "step": 48000
    },
    {
      "acc": 0.78372445,
      "epoch": 1.1201096545589757,
      "grad_norm": 11.0,
      "learning_rate": 4.272824346441576e-06,
      "loss": 0.77319117,
      "memory(GiB)": 147.13,
      "step": 48010,
      "train_speed(iter/s)": 0.200624
    },
    {
      "acc": 0.77591352,
      "epoch": 1.1203429621312646,
      "grad_norm": 7.625,
      "learning_rate": 4.270955372183074e-06,
      "loss": 0.80051012,
      "memory(GiB)": 147.13,
      "step": 48020,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.77649298,
      "epoch": 1.1205762697035535,
      "grad_norm": 4.59375,
      "learning_rate": 4.269086501995478e-06,
      "loss": 0.81262932,
      "memory(GiB)": 147.13,
      "step": 48030,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.79521112,
      "epoch": 1.1208095772758424,
      "grad_norm": 5.34375,
      "learning_rate": 4.267217736145573e-06,
      "loss": 0.7146986,
      "memory(GiB)": 147.13,
      "step": 48040,
      "train_speed(iter/s)": 0.200684
    },
    {
      "acc": 0.76602602,
      "epoch": 1.1210428848481313,
      "grad_norm": 5.75,
      "learning_rate": 4.265349074900123e-06,
      "loss": 0.83298635,
      "memory(GiB)": 147.13,
      "step": 48050,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.78246584,
      "epoch": 1.1212761924204202,
      "grad_norm": 5.21875,
      "learning_rate": 4.263480518525878e-06,
      "loss": 0.78094912,
      "memory(GiB)": 147.13,
      "step": 48060,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.77939157,
      "epoch": 1.121509499992709,
      "grad_norm": 4.78125,
      "learning_rate": 4.261612067289577e-06,
      "loss": 0.78741126,
      "memory(GiB)": 147.13,
      "step": 48070,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.78157773,
      "epoch": 1.121742807564998,
      "grad_norm": 5.25,
      "learning_rate": 4.259743721457937e-06,
      "loss": 0.76782036,
      "memory(GiB)": 147.13,
      "step": 48080,
      "train_speed(iter/s)": 0.200772
    },
    {
      "acc": 0.76486025,
      "epoch": 1.121976115137287,
      "grad_norm": 6.53125,
      "learning_rate": 4.257875481297667e-06,
      "loss": 0.86555462,
      "memory(GiB)": 147.13,
      "step": 48090,
      "train_speed(iter/s)": 0.200794
    },
    {
      "acc": 0.79323206,
      "epoch": 1.1222094227095758,
      "grad_norm": 5.40625,
      "learning_rate": 4.256007347075455e-06,
      "loss": 0.73063769,
      "memory(GiB)": 147.13,
      "step": 48100,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.76760402,
      "epoch": 1.1224427302818647,
      "grad_norm": 5.46875,
      "learning_rate": 4.254139319057979e-06,
      "loss": 0.84659023,
      "memory(GiB)": 147.13,
      "step": 48110,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.78787589,
      "epoch": 1.1226760378541536,
      "grad_norm": 8.625,
      "learning_rate": 4.252271397511898e-06,
      "loss": 0.76824799,
      "memory(GiB)": 147.13,
      "step": 48120,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.78889484,
      "epoch": 1.1229093454264425,
      "grad_norm": 5.4375,
      "learning_rate": 4.2504035827038595e-06,
      "loss": 0.75601206,
      "memory(GiB)": 147.13,
      "step": 48130,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.78154402,
      "epoch": 1.1231426529987314,
      "grad_norm": 6.90625,
      "learning_rate": 4.248535874900491e-06,
      "loss": 0.79978552,
      "memory(GiB)": 147.13,
      "step": 48140,
      "train_speed(iter/s)": 0.200899
    },
    {
      "acc": 0.79487672,
      "epoch": 1.1233759605710203,
      "grad_norm": 4.25,
      "learning_rate": 4.246668274368409e-06,
      "loss": 0.71404648,
      "memory(GiB)": 147.13,
      "step": 48150,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.7581543,
      "epoch": 1.1236092681433092,
      "grad_norm": 4.8125,
      "learning_rate": 4.24480078137421e-06,
      "loss": 0.88291979,
      "memory(GiB)": 147.13,
      "step": 48160,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.75657516,
      "epoch": 1.123842575715598,
      "grad_norm": 4.46875,
      "learning_rate": 4.2429333961844805e-06,
      "loss": 0.87030544,
      "memory(GiB)": 147.13,
      "step": 48170,
      "train_speed(iter/s)": 0.200963
    },
    {
      "acc": 0.79136295,
      "epoch": 1.124075883287887,
      "grad_norm": 3.890625,
      "learning_rate": 4.241066119065789e-06,
      "loss": 0.74944844,
      "memory(GiB)": 147.13,
      "step": 48180,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.77575951,
      "epoch": 1.1243091908601759,
      "grad_norm": 4.90625,
      "learning_rate": 4.239198950284688e-06,
      "loss": 0.80282288,
      "memory(GiB)": 147.13,
      "step": 48190,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.78090439,
      "epoch": 1.1245424984324648,
      "grad_norm": 5.28125,
      "learning_rate": 4.237331890107717e-06,
      "loss": 0.79569597,
      "memory(GiB)": 147.13,
      "step": 48200,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.78601589,
      "epoch": 1.1247758060047537,
      "grad_norm": 4.09375,
      "learning_rate": 4.2354649388013965e-06,
      "loss": 0.77665925,
      "memory(GiB)": 147.13,
      "step": 48210,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.77956362,
      "epoch": 1.1250091135770426,
      "grad_norm": 4.9375,
      "learning_rate": 4.233598096632234e-06,
      "loss": 0.77131734,
      "memory(GiB)": 147.13,
      "step": 48220,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.79384608,
      "epoch": 1.1252424211493315,
      "grad_norm": 6.09375,
      "learning_rate": 4.23173136386672e-06,
      "loss": 0.73389316,
      "memory(GiB)": 147.13,
      "step": 48230,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.78729687,
      "epoch": 1.1254757287216204,
      "grad_norm": 4.53125,
      "learning_rate": 4.2298647407713314e-06,
      "loss": 0.7819458,
      "memory(GiB)": 147.13,
      "step": 48240,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.78039856,
      "epoch": 1.1257090362939093,
      "grad_norm": 5.1875,
      "learning_rate": 4.227998227612529e-06,
      "loss": 0.78439426,
      "memory(GiB)": 147.13,
      "step": 48250,
      "train_speed(iter/s)": 0.201133
    },
    {
      "acc": 0.78524809,
      "epoch": 1.1259423438661982,
      "grad_norm": 4.25,
      "learning_rate": 4.226131824656752e-06,
      "loss": 0.7713275,
      "memory(GiB)": 147.13,
      "step": 48260,
      "train_speed(iter/s)": 0.201155
    },
    {
      "acc": 0.77486734,
      "epoch": 1.126175651438487,
      "grad_norm": 4.125,
      "learning_rate": 4.224265532170434e-06,
      "loss": 0.79942617,
      "memory(GiB)": 147.13,
      "step": 48270,
      "train_speed(iter/s)": 0.201175
    },
    {
      "acc": 0.78316307,
      "epoch": 1.126408959010776,
      "grad_norm": 6.03125,
      "learning_rate": 4.222399350419985e-06,
      "loss": 0.76972885,
      "memory(GiB)": 147.13,
      "step": 48280,
      "train_speed(iter/s)": 0.201196
    },
    {
      "acc": 0.7986887,
      "epoch": 1.1266422665830649,
      "grad_norm": 6.625,
      "learning_rate": 4.220533279671804e-06,
      "loss": 0.71824002,
      "memory(GiB)": 147.13,
      "step": 48290,
      "train_speed(iter/s)": 0.201218
    },
    {
      "acc": 0.79125404,
      "epoch": 1.1268755741553538,
      "grad_norm": 4.4375,
      "learning_rate": 4.21866732019227e-06,
      "loss": 0.74609642,
      "memory(GiB)": 147.13,
      "step": 48300,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.78295269,
      "epoch": 1.1271088817276427,
      "grad_norm": 6.53125,
      "learning_rate": 4.216801472247749e-06,
      "loss": 0.77721262,
      "memory(GiB)": 147.13,
      "step": 48310,
      "train_speed(iter/s)": 0.20126
    },
    {
      "acc": 0.76647134,
      "epoch": 1.1273421892999314,
      "grad_norm": 5.59375,
      "learning_rate": 4.214935736104591e-06,
      "loss": 0.85473747,
      "memory(GiB)": 147.13,
      "step": 48320,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.7766263,
      "epoch": 1.1275754968722205,
      "grad_norm": 6.625,
      "learning_rate": 4.213070112029127e-06,
      "loss": 0.80300884,
      "memory(GiB)": 147.13,
      "step": 48330,
      "train_speed(iter/s)": 0.201304
    },
    {
      "acc": 0.78820081,
      "epoch": 1.1278088044445091,
      "grad_norm": 5.8125,
      "learning_rate": 4.211204600287677e-06,
      "loss": 0.7805532,
      "memory(GiB)": 147.13,
      "step": 48340,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.78450546,
      "epoch": 1.128042112016798,
      "grad_norm": 7.84375,
      "learning_rate": 4.2093392011465425e-06,
      "loss": 0.77261581,
      "memory(GiB)": 147.13,
      "step": 48350,
      "train_speed(iter/s)": 0.201347
    },
    {
      "acc": 0.75979958,
      "epoch": 1.128275419589087,
      "grad_norm": 5.46875,
      "learning_rate": 4.207473914872006e-06,
      "loss": 0.86760702,
      "memory(GiB)": 147.13,
      "step": 48360,
      "train_speed(iter/s)": 0.201368
    },
    {
      "acc": 0.78485813,
      "epoch": 1.1285087271613758,
      "grad_norm": 6.1875,
      "learning_rate": 4.20560874173034e-06,
      "loss": 0.7671032,
      "memory(GiB)": 147.13,
      "step": 48370,
      "train_speed(iter/s)": 0.201391
    },
    {
      "acc": 0.77036834,
      "epoch": 1.1287420347336647,
      "grad_norm": 5.6875,
      "learning_rate": 4.203743681987793e-06,
      "loss": 0.8098217,
      "memory(GiB)": 147.13,
      "step": 48380,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.77560768,
      "epoch": 1.1289753423059536,
      "grad_norm": 4.3125,
      "learning_rate": 4.2018787359106045e-06,
      "loss": 0.81092434,
      "memory(GiB)": 147.13,
      "step": 48390,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.78146276,
      "epoch": 1.1292086498782425,
      "grad_norm": 5.15625,
      "learning_rate": 4.200013903764994e-06,
      "loss": 0.77416515,
      "memory(GiB)": 147.13,
      "step": 48400,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.77433596,
      "epoch": 1.1294419574505314,
      "grad_norm": 6.34375,
      "learning_rate": 4.198149185817167e-06,
      "loss": 0.80879793,
      "memory(GiB)": 147.13,
      "step": 48410,
      "train_speed(iter/s)": 0.201478
    },
    {
      "acc": 0.77343392,
      "epoch": 1.1296752650228203,
      "grad_norm": 6.15625,
      "learning_rate": 4.19628458233331e-06,
      "loss": 0.81420822,
      "memory(GiB)": 147.13,
      "step": 48420,
      "train_speed(iter/s)": 0.201501
    },
    {
      "acc": 0.78997188,
      "epoch": 1.1299085725951092,
      "grad_norm": 5.625,
      "learning_rate": 4.194420093579597e-06,
      "loss": 0.74577589,
      "memory(GiB)": 147.13,
      "step": 48430,
      "train_speed(iter/s)": 0.201523
    },
    {
      "acc": 0.79533091,
      "epoch": 1.1301418801673981,
      "grad_norm": 6.25,
      "learning_rate": 4.1925557198221805e-06,
      "loss": 0.72079945,
      "memory(GiB)": 147.13,
      "step": 48440,
      "train_speed(iter/s)": 0.201542
    },
    {
      "acc": 0.76862497,
      "epoch": 1.130375187739687,
      "grad_norm": 4.5625,
      "learning_rate": 4.1906914613272e-06,
      "loss": 0.86250811,
      "memory(GiB)": 147.13,
      "step": 48450,
      "train_speed(iter/s)": 0.201564
    },
    {
      "acc": 0.78376331,
      "epoch": 1.130608495311976,
      "grad_norm": 4.53125,
      "learning_rate": 4.188827318360779e-06,
      "loss": 0.76748829,
      "memory(GiB)": 147.13,
      "step": 48460,
      "train_speed(iter/s)": 0.201585
    },
    {
      "acc": 0.77294278,
      "epoch": 1.1308418028842648,
      "grad_norm": 6.09375,
      "learning_rate": 4.186963291189022e-06,
      "loss": 0.83114796,
      "memory(GiB)": 147.13,
      "step": 48470,
      "train_speed(iter/s)": 0.201606
    },
    {
      "acc": 0.78433104,
      "epoch": 1.1310751104565537,
      "grad_norm": 4.625,
      "learning_rate": 4.185099380078022e-06,
      "loss": 0.76472044,
      "memory(GiB)": 147.13,
      "step": 48480,
      "train_speed(iter/s)": 0.201626
    },
    {
      "acc": 0.78431997,
      "epoch": 1.1313084180288426,
      "grad_norm": 5.84375,
      "learning_rate": 4.183235585293846e-06,
      "loss": 0.79146395,
      "memory(GiB)": 147.13,
      "step": 48490,
      "train_speed(iter/s)": 0.201648
    },
    {
      "acc": 0.78068762,
      "epoch": 1.1315417256011315,
      "grad_norm": 5.96875,
      "learning_rate": 4.181371907102553e-06,
      "loss": 0.79927993,
      "memory(GiB)": 147.13,
      "step": 48500,
      "train_speed(iter/s)": 0.201669
    },
    {
      "epoch": 1.1315417256011315,
      "eval_acc": 0.7441143391384016,
      "eval_loss": 0.8060808777809143,
      "eval_runtime": 1270.8615,
      "eval_samples_per_second": 28.32,
      "eval_steps_per_second": 14.16,
      "step": 48500
    },
    {
      "acc": 0.7775506,
      "epoch": 1.1317750331734204,
      "grad_norm": 5.4375,
      "learning_rate": 4.179508345770184e-06,
      "loss": 0.77313948,
      "memory(GiB)": 147.13,
      "step": 48510,
      "train_speed(iter/s)": 0.200616
    },
    {
      "acc": 0.78219233,
      "epoch": 1.1320083407457093,
      "grad_norm": 5.0625,
      "learning_rate": 4.177644901562758e-06,
      "loss": 0.79809017,
      "memory(GiB)": 147.13,
      "step": 48520,
      "train_speed(iter/s)": 0.200637
    },
    {
      "acc": 0.7918745,
      "epoch": 1.1322416483179982,
      "grad_norm": 4.59375,
      "learning_rate": 4.1757815747462845e-06,
      "loss": 0.740306,
      "memory(GiB)": 147.13,
      "step": 48530,
      "train_speed(iter/s)": 0.200658
    },
    {
      "acc": 0.7779398,
      "epoch": 1.1324749558902871,
      "grad_norm": 7.3125,
      "learning_rate": 4.173918365586751e-06,
      "loss": 0.81543484,
      "memory(GiB)": 147.13,
      "step": 48540,
      "train_speed(iter/s)": 0.20068
    },
    {
      "acc": 0.7860157,
      "epoch": 1.132708263462576,
      "grad_norm": 4.875,
      "learning_rate": 4.172055274350132e-06,
      "loss": 0.75977716,
      "memory(GiB)": 147.13,
      "step": 48550,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.775599,
      "epoch": 1.132941571034865,
      "grad_norm": 5.4375,
      "learning_rate": 4.170192301302382e-06,
      "loss": 0.80197821,
      "memory(GiB)": 147.13,
      "step": 48560,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.79071817,
      "epoch": 1.1331748786071538,
      "grad_norm": 5.21875,
      "learning_rate": 4.168329446709439e-06,
      "loss": 0.73534966,
      "memory(GiB)": 147.13,
      "step": 48570,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.79963331,
      "epoch": 1.1334081861794427,
      "grad_norm": 4.5,
      "learning_rate": 4.166466710837226e-06,
      "loss": 0.71867814,
      "memory(GiB)": 147.13,
      "step": 48580,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.76818967,
      "epoch": 1.1336414937517316,
      "grad_norm": 4.40625,
      "learning_rate": 4.1646040939516485e-06,
      "loss": 0.83885851,
      "memory(GiB)": 147.13,
      "step": 48590,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.8015955,
      "epoch": 1.1338748013240205,
      "grad_norm": 4.53125,
      "learning_rate": 4.162741596318596e-06,
      "loss": 0.70797453,
      "memory(GiB)": 147.13,
      "step": 48600,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.78102293,
      "epoch": 1.1341081088963094,
      "grad_norm": 5.59375,
      "learning_rate": 4.160879218203935e-06,
      "loss": 0.79669533,
      "memory(GiB)": 147.13,
      "step": 48610,
      "train_speed(iter/s)": 0.200826
    },
    {
      "acc": 0.77543592,
      "epoch": 1.1343414164685983,
      "grad_norm": 3.890625,
      "learning_rate": 4.159016959873521e-06,
      "loss": 0.79444084,
      "memory(GiB)": 147.13,
      "step": 48620,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.76163797,
      "epoch": 1.1345747240408872,
      "grad_norm": 5.65625,
      "learning_rate": 4.1571548215931925e-06,
      "loss": 0.86727257,
      "memory(GiB)": 147.13,
      "step": 48630,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.78029289,
      "epoch": 1.1348080316131761,
      "grad_norm": 4.46875,
      "learning_rate": 4.155292803628768e-06,
      "loss": 0.79705906,
      "memory(GiB)": 147.13,
      "step": 48640,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.77660732,
      "epoch": 1.135041339185465,
      "grad_norm": 5.21875,
      "learning_rate": 4.153430906246052e-06,
      "loss": 0.80238552,
      "memory(GiB)": 147.13,
      "step": 48650,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.78615427,
      "epoch": 1.135274646757754,
      "grad_norm": 4.875,
      "learning_rate": 4.151569129710827e-06,
      "loss": 0.78471584,
      "memory(GiB)": 147.13,
      "step": 48660,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.77250824,
      "epoch": 1.1355079543300428,
      "grad_norm": 4.8125,
      "learning_rate": 4.149707474288862e-06,
      "loss": 0.81316957,
      "memory(GiB)": 147.13,
      "step": 48670,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.79985304,
      "epoch": 1.1357412619023317,
      "grad_norm": 4.53125,
      "learning_rate": 4.147845940245908e-06,
      "loss": 0.70010648,
      "memory(GiB)": 147.13,
      "step": 48680,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.77105551,
      "epoch": 1.1359745694746204,
      "grad_norm": 5.28125,
      "learning_rate": 4.145984527847699e-06,
      "loss": 0.81544399,
      "memory(GiB)": 147.13,
      "step": 48690,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.77478409,
      "epoch": 1.1362078770469095,
      "grad_norm": 6.03125,
      "learning_rate": 4.14412323735995e-06,
      "loss": 0.81853638,
      "memory(GiB)": 147.13,
      "step": 48700,
      "train_speed(iter/s)": 0.20102
    },
    {
      "acc": 0.78074069,
      "epoch": 1.1364411846191982,
      "grad_norm": 6.78125,
      "learning_rate": 4.142262069048362e-06,
      "loss": 0.78377013,
      "memory(GiB)": 147.13,
      "step": 48710,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.78800902,
      "epoch": 1.1366744921914873,
      "grad_norm": 5.03125,
      "learning_rate": 4.140401023178613e-06,
      "loss": 0.76176023,
      "memory(GiB)": 147.13,
      "step": 48720,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.78853445,
      "epoch": 1.136907799763776,
      "grad_norm": 4.4375,
      "learning_rate": 4.138540100016369e-06,
      "loss": 0.7591939,
      "memory(GiB)": 147.13,
      "step": 48730,
      "train_speed(iter/s)": 0.201084
    },
    {
      "acc": 0.78152742,
      "epoch": 1.1371411073360649,
      "grad_norm": 7.03125,
      "learning_rate": 4.136679299827275e-06,
      "loss": 0.77542849,
      "memory(GiB)": 147.13,
      "step": 48740,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.80343904,
      "epoch": 1.1373744149083538,
      "grad_norm": 5.21875,
      "learning_rate": 4.134818622876959e-06,
      "loss": 0.68836613,
      "memory(GiB)": 147.13,
      "step": 48750,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.77372913,
      "epoch": 1.1376077224806427,
      "grad_norm": 6.5625,
      "learning_rate": 4.132958069431034e-06,
      "loss": 0.78857126,
      "memory(GiB)": 147.13,
      "step": 48760,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.78540783,
      "epoch": 1.1378410300529316,
      "grad_norm": 4.6875,
      "learning_rate": 4.131097639755093e-06,
      "loss": 0.76112299,
      "memory(GiB)": 147.13,
      "step": 48770,
      "train_speed(iter/s)": 0.201169
    },
    {
      "acc": 0.77772455,
      "epoch": 1.1380743376252205,
      "grad_norm": 6.21875,
      "learning_rate": 4.129237334114712e-06,
      "loss": 0.81547289,
      "memory(GiB)": 147.13,
      "step": 48780,
      "train_speed(iter/s)": 0.20119
    },
    {
      "acc": 0.78677707,
      "epoch": 1.1383076451975094,
      "grad_norm": 4.40625,
      "learning_rate": 4.127377152775448e-06,
      "loss": 0.77611208,
      "memory(GiB)": 147.13,
      "step": 48790,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.79749675,
      "epoch": 1.1385409527697983,
      "grad_norm": 4.125,
      "learning_rate": 4.125517096002842e-06,
      "loss": 0.71030278,
      "memory(GiB)": 147.13,
      "step": 48800,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.79024119,
      "epoch": 1.1387742603420872,
      "grad_norm": 4.46875,
      "learning_rate": 4.123657164062415e-06,
      "loss": 0.74269834,
      "memory(GiB)": 147.13,
      "step": 48810,
      "train_speed(iter/s)": 0.201252
    },
    {
      "acc": 0.76520209,
      "epoch": 1.139007567914376,
      "grad_norm": 4.84375,
      "learning_rate": 4.121797357219678e-06,
      "loss": 0.84504023,
      "memory(GiB)": 147.13,
      "step": 48820,
      "train_speed(iter/s)": 0.201274
    },
    {
      "acc": 0.7725544,
      "epoch": 1.139240875486665,
      "grad_norm": 5.09375,
      "learning_rate": 4.119937675740109e-06,
      "loss": 0.81992121,
      "memory(GiB)": 147.13,
      "step": 48830,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.78341656,
      "epoch": 1.1394741830589539,
      "grad_norm": 5.65625,
      "learning_rate": 4.118078119889182e-06,
      "loss": 0.76839838,
      "memory(GiB)": 147.13,
      "step": 48840,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.80117064,
      "epoch": 1.1397074906312428,
      "grad_norm": 5.21875,
      "learning_rate": 4.116218689932346e-06,
      "loss": 0.71321936,
      "memory(GiB)": 147.13,
      "step": 48850,
      "train_speed(iter/s)": 0.201338
    },
    {
      "acc": 0.79148402,
      "epoch": 1.1399407982035317,
      "grad_norm": 5.53125,
      "learning_rate": 4.114359386135038e-06,
      "loss": 0.74038553,
      "memory(GiB)": 147.13,
      "step": 48860,
      "train_speed(iter/s)": 0.20136
    },
    {
      "acc": 0.77297692,
      "epoch": 1.1401741057758206,
      "grad_norm": 5.4375,
      "learning_rate": 4.112500208762668e-06,
      "loss": 0.80347652,
      "memory(GiB)": 147.13,
      "step": 48870,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.78603468,
      "epoch": 1.1404074133481095,
      "grad_norm": 4.40625,
      "learning_rate": 4.110641158080636e-06,
      "loss": 0.78780212,
      "memory(GiB)": 147.13,
      "step": 48880,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.80655193,
      "epoch": 1.1406407209203984,
      "grad_norm": 4.78125,
      "learning_rate": 4.108782234354321e-06,
      "loss": 0.7006731,
      "memory(GiB)": 147.13,
      "step": 48890,
      "train_speed(iter/s)": 0.201423
    },
    {
      "acc": 0.79013271,
      "epoch": 1.1408740284926873,
      "grad_norm": 5.5,
      "learning_rate": 4.106923437849082e-06,
      "loss": 0.75852995,
      "memory(GiB)": 147.13,
      "step": 48900,
      "train_speed(iter/s)": 0.201443
    },
    {
      "acc": 0.7818615,
      "epoch": 1.1411073360649762,
      "grad_norm": 4.40625,
      "learning_rate": 4.105064768830263e-06,
      "loss": 0.77103171,
      "memory(GiB)": 147.13,
      "step": 48910,
      "train_speed(iter/s)": 0.201463
    },
    {
      "acc": 0.76176796,
      "epoch": 1.141340643637265,
      "grad_norm": 4.90625,
      "learning_rate": 4.1032062275631894e-06,
      "loss": 0.87524958,
      "memory(GiB)": 147.13,
      "step": 48920,
      "train_speed(iter/s)": 0.201485
    },
    {
      "acc": 0.76909437,
      "epoch": 1.141573951209554,
      "grad_norm": 4.625,
      "learning_rate": 4.101347814313166e-06,
      "loss": 0.81877556,
      "memory(GiB)": 147.13,
      "step": 48930,
      "train_speed(iter/s)": 0.201506
    },
    {
      "acc": 0.80096893,
      "epoch": 1.1418072587818429,
      "grad_norm": 4.25,
      "learning_rate": 4.099489529345483e-06,
      "loss": 0.71625419,
      "memory(GiB)": 147.13,
      "step": 48940,
      "train_speed(iter/s)": 0.201527
    },
    {
      "acc": 0.77617383,
      "epoch": 1.1420405663541318,
      "grad_norm": 5.75,
      "learning_rate": 4.097631372925405e-06,
      "loss": 0.79977846,
      "memory(GiB)": 147.13,
      "step": 48950,
      "train_speed(iter/s)": 0.201549
    },
    {
      "acc": 0.78573585,
      "epoch": 1.1422738739264207,
      "grad_norm": 6.28125,
      "learning_rate": 4.095773345318186e-06,
      "loss": 0.77477055,
      "memory(GiB)": 147.13,
      "step": 48960,
      "train_speed(iter/s)": 0.201571
    },
    {
      "acc": 0.78423834,
      "epoch": 1.1425071814987096,
      "grad_norm": 15.9375,
      "learning_rate": 4.0939154467890605e-06,
      "loss": 0.76231022,
      "memory(GiB)": 147.13,
      "step": 48970,
      "train_speed(iter/s)": 0.201591
    },
    {
      "acc": 0.79738851,
      "epoch": 1.1427404890709985,
      "grad_norm": 8.875,
      "learning_rate": 4.0920576776032415e-06,
      "loss": 0.71518207,
      "memory(GiB)": 147.13,
      "step": 48980,
      "train_speed(iter/s)": 0.201613
    },
    {
      "acc": 0.77292356,
      "epoch": 1.1429737966432874,
      "grad_norm": 4.375,
      "learning_rate": 4.090200038025926e-06,
      "loss": 0.79759941,
      "memory(GiB)": 147.13,
      "step": 48990,
      "train_speed(iter/s)": 0.201634
    },
    {
      "acc": 0.8075819,
      "epoch": 1.1432071042155763,
      "grad_norm": 7.0625,
      "learning_rate": 4.08834252832229e-06,
      "loss": 0.67757139,
      "memory(GiB)": 147.13,
      "step": 49000,
      "train_speed(iter/s)": 0.201656
    },
    {
      "epoch": 1.1432071042155763,
      "eval_acc": 0.744222436495205,
      "eval_loss": 0.8061447739601135,
      "eval_runtime": 1270.2536,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "step": 49000
    },
    {
      "acc": 0.78951435,
      "epoch": 1.1434404117878652,
      "grad_norm": 6.3125,
      "learning_rate": 4.086485148757493e-06,
      "loss": 0.7389822,
      "memory(GiB)": 147.13,
      "step": 49010,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.78884678,
      "epoch": 1.143673719360154,
      "grad_norm": 5.875,
      "learning_rate": 4.084627899596676e-06,
      "loss": 0.75943174,
      "memory(GiB)": 147.13,
      "step": 49020,
      "train_speed(iter/s)": 0.200633
    },
    {
      "acc": 0.76504393,
      "epoch": 1.143907026932443,
      "grad_norm": 4.6875,
      "learning_rate": 4.082770781104961e-06,
      "loss": 0.84475698,
      "memory(GiB)": 147.13,
      "step": 49030,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.79473829,
      "epoch": 1.1441403345047318,
      "grad_norm": 6.03125,
      "learning_rate": 4.080913793547449e-06,
      "loss": 0.74036322,
      "memory(GiB)": 147.13,
      "step": 49040,
      "train_speed(iter/s)": 0.200673
    },
    {
      "acc": 0.77508821,
      "epoch": 1.1443736420770207,
      "grad_norm": 5.4375,
      "learning_rate": 4.079056937189229e-06,
      "loss": 0.80481701,
      "memory(GiB)": 147.13,
      "step": 49050,
      "train_speed(iter/s)": 0.200695
    },
    {
      "acc": 0.7855969,
      "epoch": 1.1446069496493096,
      "grad_norm": 5.5625,
      "learning_rate": 4.077200212295361e-06,
      "loss": 0.76578369,
      "memory(GiB)": 147.13,
      "step": 49060,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.7720355,
      "epoch": 1.1448402572215985,
      "grad_norm": 4.875,
      "learning_rate": 4.075343619130895e-06,
      "loss": 0.8170085,
      "memory(GiB)": 147.13,
      "step": 49070,
      "train_speed(iter/s)": 0.200737
    },
    {
      "acc": 0.7803793,
      "epoch": 1.1450735647938872,
      "grad_norm": 4.71875,
      "learning_rate": 4.0734871579608606e-06,
      "loss": 0.79507275,
      "memory(GiB)": 147.13,
      "step": 49080,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.78040457,
      "epoch": 1.1453068723661763,
      "grad_norm": 5.28125,
      "learning_rate": 4.071630829050263e-06,
      "loss": 0.79167814,
      "memory(GiB)": 147.13,
      "step": 49090,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.80139008,
      "epoch": 1.145540179938465,
      "grad_norm": 4.28125,
      "learning_rate": 4.069774632664095e-06,
      "loss": 0.69903708,
      "memory(GiB)": 147.13,
      "step": 49100,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.7990375,
      "epoch": 1.1457734875107541,
      "grad_norm": 6.46875,
      "learning_rate": 4.0679185690673285e-06,
      "loss": 0.707726,
      "memory(GiB)": 147.13,
      "step": 49110,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.78790112,
      "epoch": 1.1460067950830428,
      "grad_norm": 5.625,
      "learning_rate": 4.066062638524915e-06,
      "loss": 0.76991944,
      "memory(GiB)": 147.13,
      "step": 49120,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.77768164,
      "epoch": 1.1462401026553317,
      "grad_norm": 7.125,
      "learning_rate": 4.064206841301789e-06,
      "loss": 0.78967957,
      "memory(GiB)": 147.13,
      "step": 49130,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.78068905,
      "epoch": 1.1464734102276206,
      "grad_norm": 5.53125,
      "learning_rate": 4.062351177662866e-06,
      "loss": 0.77087574,
      "memory(GiB)": 147.13,
      "step": 49140,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.78477345,
      "epoch": 1.1467067177999095,
      "grad_norm": 4.6875,
      "learning_rate": 4.060495647873038e-06,
      "loss": 0.77597842,
      "memory(GiB)": 147.13,
      "step": 49150,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.77061286,
      "epoch": 1.1469400253721984,
      "grad_norm": 6.0625,
      "learning_rate": 4.058640252197184e-06,
      "loss": 0.81207161,
      "memory(GiB)": 147.13,
      "step": 49160,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.79196901,
      "epoch": 1.1471733329444873,
      "grad_norm": 6.90625,
      "learning_rate": 4.056784990900162e-06,
      "loss": 0.73144436,
      "memory(GiB)": 147.13,
      "step": 49170,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.76946406,
      "epoch": 1.1474066405167762,
      "grad_norm": 4.75,
      "learning_rate": 4.054929864246807e-06,
      "loss": 0.8269702,
      "memory(GiB)": 147.13,
      "step": 49180,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.79781704,
      "epoch": 1.147639948089065,
      "grad_norm": 5.03125,
      "learning_rate": 4.053074872501939e-06,
      "loss": 0.72426014,
      "memory(GiB)": 147.13,
      "step": 49190,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.79303007,
      "epoch": 1.147873255661354,
      "grad_norm": 3.4375,
      "learning_rate": 4.051220015930358e-06,
      "loss": 0.74754534,
      "memory(GiB)": 147.13,
      "step": 49200,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.78933134,
      "epoch": 1.148106563233643,
      "grad_norm": 7.0,
      "learning_rate": 4.049365294796844e-06,
      "loss": 0.74844933,
      "memory(GiB)": 147.13,
      "step": 49210,
      "train_speed(iter/s)": 0.20103
    },
    {
      "acc": 0.76711445,
      "epoch": 1.1483398708059318,
      "grad_norm": 3.390625,
      "learning_rate": 4.047510709366159e-06,
      "loss": 0.85391102,
      "memory(GiB)": 147.13,
      "step": 49220,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.77335253,
      "epoch": 1.1485731783782207,
      "grad_norm": 6.59375,
      "learning_rate": 4.045656259903042e-06,
      "loss": 0.81023026,
      "memory(GiB)": 147.13,
      "step": 49230,
      "train_speed(iter/s)": 0.201073
    },
    {
      "acc": 0.79616175,
      "epoch": 1.1488064859505096,
      "grad_norm": 5.15625,
      "learning_rate": 4.043801946672217e-06,
      "loss": 0.72714806,
      "memory(GiB)": 147.13,
      "step": 49240,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.77853937,
      "epoch": 1.1490397935227985,
      "grad_norm": 6.4375,
      "learning_rate": 4.041947769938387e-06,
      "loss": 0.81164007,
      "memory(GiB)": 147.13,
      "step": 49250,
      "train_speed(iter/s)": 0.201117
    },
    {
      "acc": 0.78679638,
      "epoch": 1.1492731010950874,
      "grad_norm": 4.6875,
      "learning_rate": 4.040093729966234e-06,
      "loss": 0.77023234,
      "memory(GiB)": 147.13,
      "step": 49260,
      "train_speed(iter/s)": 0.201139
    },
    {
      "acc": 0.79337749,
      "epoch": 1.1495064086673763,
      "grad_norm": 5.0625,
      "learning_rate": 4.038239827020424e-06,
      "loss": 0.72135153,
      "memory(GiB)": 147.13,
      "step": 49270,
      "train_speed(iter/s)": 0.20116
    },
    {
      "acc": 0.79230156,
      "epoch": 1.1497397162396652,
      "grad_norm": 6.0,
      "learning_rate": 4.036386061365598e-06,
      "loss": 0.74501095,
      "memory(GiB)": 147.13,
      "step": 49280,
      "train_speed(iter/s)": 0.201181
    },
    {
      "acc": 0.78310337,
      "epoch": 1.149973023811954,
      "grad_norm": 5.21875,
      "learning_rate": 4.034532433266382e-06,
      "loss": 0.78334503,
      "memory(GiB)": 147.13,
      "step": 49290,
      "train_speed(iter/s)": 0.201201
    },
    {
      "acc": 0.78788719,
      "epoch": 1.150206331384243,
      "grad_norm": 12.9375,
      "learning_rate": 4.032678942987382e-06,
      "loss": 0.74433508,
      "memory(GiB)": 147.13,
      "step": 49300,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.77241621,
      "epoch": 1.150439638956532,
      "grad_norm": 5.0625,
      "learning_rate": 4.030825590793179e-06,
      "loss": 0.81397333,
      "memory(GiB)": 147.13,
      "step": 49310,
      "train_speed(iter/s)": 0.201242
    },
    {
      "acc": 0.78942137,
      "epoch": 1.1506729465288208,
      "grad_norm": 4.75,
      "learning_rate": 4.028972376948343e-06,
      "loss": 0.75723009,
      "memory(GiB)": 147.13,
      "step": 49320,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.78589869,
      "epoch": 1.1509062541011097,
      "grad_norm": 6.0625,
      "learning_rate": 4.027119301717417e-06,
      "loss": 0.74469566,
      "memory(GiB)": 147.13,
      "step": 49330,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.79240808,
      "epoch": 1.1511395616733986,
      "grad_norm": 4.71875,
      "learning_rate": 4.025266365364928e-06,
      "loss": 0.72842922,
      "memory(GiB)": 147.13,
      "step": 49340,
      "train_speed(iter/s)": 0.201302
    },
    {
      "acc": 0.79350748,
      "epoch": 1.1513728692456875,
      "grad_norm": 8.3125,
      "learning_rate": 4.0234135681553835e-06,
      "loss": 0.74578514,
      "memory(GiB)": 147.13,
      "step": 49350,
      "train_speed(iter/s)": 0.201324
    },
    {
      "acc": 0.76579685,
      "epoch": 1.1516061768179764,
      "grad_norm": 5.5625,
      "learning_rate": 4.021560910353268e-06,
      "loss": 0.84622517,
      "memory(GiB)": 147.13,
      "step": 49360,
      "train_speed(iter/s)": 0.201344
    },
    {
      "acc": 0.77959232,
      "epoch": 1.1518394843902653,
      "grad_norm": 5.03125,
      "learning_rate": 4.019708392223048e-06,
      "loss": 0.79495625,
      "memory(GiB)": 147.13,
      "step": 49370,
      "train_speed(iter/s)": 0.201366
    },
    {
      "acc": 0.79989409,
      "epoch": 1.1520727919625542,
      "grad_norm": 4.875,
      "learning_rate": 4.017856014029171e-06,
      "loss": 0.71503725,
      "memory(GiB)": 147.13,
      "step": 49380,
      "train_speed(iter/s)": 0.201386
    },
    {
      "acc": 0.77668247,
      "epoch": 1.152306099534843,
      "grad_norm": 5.0,
      "learning_rate": 4.016003776036064e-06,
      "loss": 0.78333073,
      "memory(GiB)": 147.13,
      "step": 49390,
      "train_speed(iter/s)": 0.201405
    },
    {
      "acc": 0.77576346,
      "epoch": 1.152539407107132,
      "grad_norm": 6.34375,
      "learning_rate": 4.01415167850813e-06,
      "loss": 0.80841331,
      "memory(GiB)": 147.13,
      "step": 49400,
      "train_speed(iter/s)": 0.201426
    },
    {
      "acc": 0.77319818,
      "epoch": 1.1527727146794209,
      "grad_norm": 5.6875,
      "learning_rate": 4.012299721709757e-06,
      "loss": 0.81761799,
      "memory(GiB)": 147.13,
      "step": 49410,
      "train_speed(iter/s)": 0.201447
    },
    {
      "acc": 0.76485691,
      "epoch": 1.1530060222517098,
      "grad_norm": 8.375,
      "learning_rate": 4.010447905905312e-06,
      "loss": 0.85202141,
      "memory(GiB)": 147.13,
      "step": 49420,
      "train_speed(iter/s)": 0.201468
    },
    {
      "acc": 0.77609119,
      "epoch": 1.1532393298239987,
      "grad_norm": 6.0,
      "learning_rate": 4.0085962313591416e-06,
      "loss": 0.82331104,
      "memory(GiB)": 147.13,
      "step": 49430,
      "train_speed(iter/s)": 0.201488
    },
    {
      "acc": 0.76981869,
      "epoch": 1.1534726373962876,
      "grad_norm": 4.125,
      "learning_rate": 4.006744698335572e-06,
      "loss": 0.83070965,
      "memory(GiB)": 147.13,
      "step": 49440,
      "train_speed(iter/s)": 0.201508
    },
    {
      "acc": 0.77602968,
      "epoch": 1.1537059449685765,
      "grad_norm": 5.5,
      "learning_rate": 4.004893307098907e-06,
      "loss": 0.80735464,
      "memory(GiB)": 147.13,
      "step": 49450,
      "train_speed(iter/s)": 0.20153
    },
    {
      "acc": 0.79087119,
      "epoch": 1.1539392525408654,
      "grad_norm": 4.46875,
      "learning_rate": 4.003042057913434e-06,
      "loss": 0.74953089,
      "memory(GiB)": 147.13,
      "step": 49460,
      "train_speed(iter/s)": 0.201551
    },
    {
      "acc": 0.79258838,
      "epoch": 1.154172560113154,
      "grad_norm": 4.65625,
      "learning_rate": 4.001190951043416e-06,
      "loss": 0.73958902,
      "memory(GiB)": 147.13,
      "step": 49470,
      "train_speed(iter/s)": 0.201572
    },
    {
      "acc": 0.78888607,
      "epoch": 1.1544058676854432,
      "grad_norm": 5.125,
      "learning_rate": 3.9993399867531e-06,
      "loss": 0.74525504,
      "memory(GiB)": 147.13,
      "step": 49480,
      "train_speed(iter/s)": 0.201592
    },
    {
      "acc": 0.78004522,
      "epoch": 1.1546391752577319,
      "grad_norm": 7.15625,
      "learning_rate": 3.997489165306713e-06,
      "loss": 0.77018299,
      "memory(GiB)": 147.13,
      "step": 49490,
      "train_speed(iter/s)": 0.201612
    },
    {
      "acc": 0.78893394,
      "epoch": 1.154872482830021,
      "grad_norm": 7.375,
      "learning_rate": 3.995638486968453e-06,
      "loss": 0.74949064,
      "memory(GiB)": 147.13,
      "step": 49500,
      "train_speed(iter/s)": 0.201633
    },
    {
      "epoch": 1.154872482830021,
      "eval_acc": 0.7442927398131853,
      "eval_loss": 0.8059793710708618,
      "eval_runtime": 1270.1643,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 49500
    },
    {
      "acc": 0.77485609,
      "epoch": 1.1551057904023097,
      "grad_norm": 4.84375,
      "learning_rate": 3.99378795200251e-06,
      "loss": 0.8245182,
      "memory(GiB)": 147.13,
      "step": 49510,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.77423291,
      "epoch": 1.1553390979745986,
      "grad_norm": 6.34375,
      "learning_rate": 3.991937560673044e-06,
      "loss": 0.79736261,
      "memory(GiB)": 147.13,
      "step": 49520,
      "train_speed(iter/s)": 0.20062
    },
    {
      "acc": 0.77984371,
      "epoch": 1.1555724055468874,
      "grad_norm": 5.15625,
      "learning_rate": 3.990087313244197e-06,
      "loss": 0.77974262,
      "memory(GiB)": 147.13,
      "step": 49530,
      "train_speed(iter/s)": 0.200641
    },
    {
      "acc": 0.77480555,
      "epoch": 1.1558057131191763,
      "grad_norm": 6.34375,
      "learning_rate": 3.988237209980093e-06,
      "loss": 0.79339681,
      "memory(GiB)": 147.13,
      "step": 49540,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.79203959,
      "epoch": 1.1560390206914652,
      "grad_norm": 5.0,
      "learning_rate": 3.986387251144833e-06,
      "loss": 0.74149151,
      "memory(GiB)": 147.13,
      "step": 49550,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.78118038,
      "epoch": 1.1562723282637541,
      "grad_norm": 6.0,
      "learning_rate": 3.9845374370024995e-06,
      "loss": 0.78897324,
      "memory(GiB)": 147.13,
      "step": 49560,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.78257017,
      "epoch": 1.156505635836043,
      "grad_norm": 5.03125,
      "learning_rate": 3.9826877678171515e-06,
      "loss": 0.78667088,
      "memory(GiB)": 147.13,
      "step": 49570,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.77982354,
      "epoch": 1.156738943408332,
      "grad_norm": 5.1875,
      "learning_rate": 3.980838243852829e-06,
      "loss": 0.77447562,
      "memory(GiB)": 147.13,
      "step": 49580,
      "train_speed(iter/s)": 0.200745
    },
    {
      "acc": 0.7748292,
      "epoch": 1.1569722509806208,
      "grad_norm": 4.5,
      "learning_rate": 3.978988865373551e-06,
      "loss": 0.80467196,
      "memory(GiB)": 147.13,
      "step": 49590,
      "train_speed(iter/s)": 0.200766
    },
    {
      "acc": 0.78131981,
      "epoch": 1.1572055585529097,
      "grad_norm": 5.0625,
      "learning_rate": 3.977139632643316e-06,
      "loss": 0.7902432,
      "memory(GiB)": 147.13,
      "step": 49600,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.79091144,
      "epoch": 1.1574388661251986,
      "grad_norm": 5.3125,
      "learning_rate": 3.975290545926101e-06,
      "loss": 0.77418213,
      "memory(GiB)": 147.13,
      "step": 49610,
      "train_speed(iter/s)": 0.200807
    },
    {
      "acc": 0.77418609,
      "epoch": 1.1576721736974875,
      "grad_norm": 6.0625,
      "learning_rate": 3.973441605485864e-06,
      "loss": 0.81481934,
      "memory(GiB)": 147.13,
      "step": 49620,
      "train_speed(iter/s)": 0.200826
    },
    {
      "acc": 0.78058105,
      "epoch": 1.1579054812697764,
      "grad_norm": 8.375,
      "learning_rate": 3.971592811586539e-06,
      "loss": 0.78378577,
      "memory(GiB)": 147.13,
      "step": 49630,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.76909003,
      "epoch": 1.1581387888420653,
      "grad_norm": 6.09375,
      "learning_rate": 3.969744164492041e-06,
      "loss": 0.83589916,
      "memory(GiB)": 147.13,
      "step": 49640,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.77044625,
      "epoch": 1.1583720964143542,
      "grad_norm": 5.15625,
      "learning_rate": 3.967895664466265e-06,
      "loss": 0.81802549,
      "memory(GiB)": 147.13,
      "step": 49650,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.77337999,
      "epoch": 1.1586054039866431,
      "grad_norm": 5.375,
      "learning_rate": 3.966047311773083e-06,
      "loss": 0.82878065,
      "memory(GiB)": 147.13,
      "step": 49660,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.76443853,
      "epoch": 1.158838711558932,
      "grad_norm": 5.96875,
      "learning_rate": 3.964199106676345e-06,
      "loss": 0.85824375,
      "memory(GiB)": 147.13,
      "step": 49670,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.7868217,
      "epoch": 1.159072019131221,
      "grad_norm": 13.0625,
      "learning_rate": 3.962351049439885e-06,
      "loss": 0.77052197,
      "memory(GiB)": 147.13,
      "step": 49680,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.78962936,
      "epoch": 1.1593053267035098,
      "grad_norm": 5.96875,
      "learning_rate": 3.960503140327511e-06,
      "loss": 0.75969067,
      "memory(GiB)": 147.13,
      "step": 49690,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.76684074,
      "epoch": 1.1595386342757987,
      "grad_norm": 7.0625,
      "learning_rate": 3.958655379603011e-06,
      "loss": 0.85526161,
      "memory(GiB)": 147.13,
      "step": 49700,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.78498111,
      "epoch": 1.1597719418480876,
      "grad_norm": 5.0625,
      "learning_rate": 3.956807767530155e-06,
      "loss": 0.78701396,
      "memory(GiB)": 147.13,
      "step": 49710,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.7782752,
      "epoch": 1.1600052494203765,
      "grad_norm": 4.8125,
      "learning_rate": 3.954960304372686e-06,
      "loss": 0.82441578,
      "memory(GiB)": 147.13,
      "step": 49720,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.76695485,
      "epoch": 1.1602385569926654,
      "grad_norm": 4.90625,
      "learning_rate": 3.95311299039433e-06,
      "loss": 0.86624813,
      "memory(GiB)": 147.13,
      "step": 49730,
      "train_speed(iter/s)": 0.20106
    },
    {
      "acc": 0.77228241,
      "epoch": 1.1604718645649543,
      "grad_norm": 3.875,
      "learning_rate": 3.951265825858792e-06,
      "loss": 0.8170001,
      "memory(GiB)": 147.13,
      "step": 49740,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.77791176,
      "epoch": 1.1607051721372432,
      "grad_norm": 5.75,
      "learning_rate": 3.949418811029752e-06,
      "loss": 0.8135541,
      "memory(GiB)": 147.13,
      "step": 49750,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.7993576,
      "epoch": 1.1609384797095321,
      "grad_norm": 6.09375,
      "learning_rate": 3.94757194617087e-06,
      "loss": 0.70589614,
      "memory(GiB)": 147.13,
      "step": 49760,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.7687006,
      "epoch": 1.161171787281821,
      "grad_norm": 6.90625,
      "learning_rate": 3.945725231545787e-06,
      "loss": 0.8219142,
      "memory(GiB)": 147.13,
      "step": 49770,
      "train_speed(iter/s)": 0.201145
    },
    {
      "acc": 0.78447962,
      "epoch": 1.16140509485411,
      "grad_norm": 5.375,
      "learning_rate": 3.943878667418122e-06,
      "loss": 0.76722612,
      "memory(GiB)": 147.13,
      "step": 49780,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.78034124,
      "epoch": 1.1616384024263988,
      "grad_norm": 4.46875,
      "learning_rate": 3.942032254051471e-06,
      "loss": 0.78942509,
      "memory(GiB)": 147.13,
      "step": 49790,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.79355407,
      "epoch": 1.1618717099986877,
      "grad_norm": 5.78125,
      "learning_rate": 3.940185991709407e-06,
      "loss": 0.72849441,
      "memory(GiB)": 147.13,
      "step": 49800,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.7907186,
      "epoch": 1.1621050175709766,
      "grad_norm": 4.125,
      "learning_rate": 3.938339880655485e-06,
      "loss": 0.76254416,
      "memory(GiB)": 147.13,
      "step": 49810,
      "train_speed(iter/s)": 0.201226
    },
    {
      "acc": 0.78690691,
      "epoch": 1.1623383251432655,
      "grad_norm": 5.03125,
      "learning_rate": 3.9364939211532365e-06,
      "loss": 0.76221375,
      "memory(GiB)": 147.13,
      "step": 49820,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.77374029,
      "epoch": 1.1625716327155544,
      "grad_norm": 4.96875,
      "learning_rate": 3.934648113466172e-06,
      "loss": 0.80819073,
      "memory(GiB)": 147.13,
      "step": 49830,
      "train_speed(iter/s)": 0.201268
    },
    {
      "acc": 0.77530112,
      "epoch": 1.1628049402878433,
      "grad_norm": 5.0,
      "learning_rate": 3.93280245785778e-06,
      "loss": 0.78218956,
      "memory(GiB)": 147.13,
      "step": 49840,
      "train_speed(iter/s)": 0.201288
    },
    {
      "acc": 0.78257365,
      "epoch": 1.1630382478601322,
      "grad_norm": 5.3125,
      "learning_rate": 3.9309569545915285e-06,
      "loss": 0.77504187,
      "memory(GiB)": 147.13,
      "step": 49850,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.77935219,
      "epoch": 1.163271555432421,
      "grad_norm": 7.0625,
      "learning_rate": 3.9291116039308605e-06,
      "loss": 0.79420118,
      "memory(GiB)": 147.13,
      "step": 49860,
      "train_speed(iter/s)": 0.20133
    },
    {
      "acc": 0.78821473,
      "epoch": 1.16350486300471,
      "grad_norm": 5.625,
      "learning_rate": 3.9272664061392e-06,
      "loss": 0.74523487,
      "memory(GiB)": 147.13,
      "step": 49870,
      "train_speed(iter/s)": 0.201351
    },
    {
      "acc": 0.75341196,
      "epoch": 1.1637381705769987,
      "grad_norm": 6.125,
      "learning_rate": 3.925421361479947e-06,
      "loss": 0.88330593,
      "memory(GiB)": 147.13,
      "step": 49880,
      "train_speed(iter/s)": 0.20137
    },
    {
      "acc": 0.77374578,
      "epoch": 1.1639714781492876,
      "grad_norm": 5.25,
      "learning_rate": 3.923576470216483e-06,
      "loss": 0.83307543,
      "memory(GiB)": 147.13,
      "step": 49890,
      "train_speed(iter/s)": 0.201391
    },
    {
      "acc": 0.7787755,
      "epoch": 1.1642047857215765,
      "grad_norm": 8.25,
      "learning_rate": 3.9217317326121655e-06,
      "loss": 0.78595629,
      "memory(GiB)": 147.13,
      "step": 49900,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.79464808,
      "epoch": 1.1644380932938654,
      "grad_norm": 5.5625,
      "learning_rate": 3.919887148930329e-06,
      "loss": 0.72763758,
      "memory(GiB)": 147.13,
      "step": 49910,
      "train_speed(iter/s)": 0.201431
    },
    {
      "acc": 0.78488331,
      "epoch": 1.1646714008661543,
      "grad_norm": 5.21875,
      "learning_rate": 3.918042719434288e-06,
      "loss": 0.77773986,
      "memory(GiB)": 147.13,
      "step": 49920,
      "train_speed(iter/s)": 0.201453
    },
    {
      "acc": 0.79530802,
      "epoch": 1.1649047084384432,
      "grad_norm": 4.46875,
      "learning_rate": 3.916198444387337e-06,
      "loss": 0.71487732,
      "memory(GiB)": 147.13,
      "step": 49930,
      "train_speed(iter/s)": 0.201473
    },
    {
      "acc": 0.78052902,
      "epoch": 1.165138016010732,
      "grad_norm": 6.03125,
      "learning_rate": 3.914354324052741e-06,
      "loss": 0.77790031,
      "memory(GiB)": 147.13,
      "step": 49940,
      "train_speed(iter/s)": 0.201494
    },
    {
      "acc": 0.7807313,
      "epoch": 1.165371323583021,
      "grad_norm": 4.46875,
      "learning_rate": 3.91251035869375e-06,
      "loss": 0.77833533,
      "memory(GiB)": 147.13,
      "step": 49950,
      "train_speed(iter/s)": 0.201514
    },
    {
      "acc": 0.78631954,
      "epoch": 1.1656046311553099,
      "grad_norm": 5.5625,
      "learning_rate": 3.91066654857359e-06,
      "loss": 0.7649682,
      "memory(GiB)": 147.13,
      "step": 49960,
      "train_speed(iter/s)": 0.201533
    },
    {
      "acc": 0.78310833,
      "epoch": 1.1658379387275988,
      "grad_norm": 5.28125,
      "learning_rate": 3.908822893955466e-06,
      "loss": 0.78763723,
      "memory(GiB)": 147.13,
      "step": 49970,
      "train_speed(iter/s)": 0.201554
    },
    {
      "acc": 0.7878459,
      "epoch": 1.1660712462998877,
      "grad_norm": 7.6875,
      "learning_rate": 3.9069793951025544e-06,
      "loss": 0.74921327,
      "memory(GiB)": 147.13,
      "step": 49980,
      "train_speed(iter/s)": 0.201575
    },
    {
      "acc": 0.76945477,
      "epoch": 1.1663045538721766,
      "grad_norm": 4.125,
      "learning_rate": 3.9051360522780166e-06,
      "loss": 0.81270313,
      "memory(GiB)": 147.13,
      "step": 49990,
      "train_speed(iter/s)": 0.201596
    },
    {
      "acc": 0.76828485,
      "epoch": 1.1665378614444655,
      "grad_norm": 6.5625,
      "learning_rate": 3.903292865744989e-06,
      "loss": 0.81358232,
      "memory(GiB)": 147.13,
      "step": 50000,
      "train_speed(iter/s)": 0.201615
    },
    {
      "epoch": 1.1665378614444655,
      "eval_acc": 0.7442993057267097,
      "eval_loss": 0.8060175180435181,
      "eval_runtime": 1270.1101,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 50000
    },
    {
      "acc": 0.78241692,
      "epoch": 1.1667711690167544,
      "grad_norm": 5.375,
      "learning_rate": 3.901449835766588e-06,
      "loss": 0.79187217,
      "memory(GiB)": 147.13,
      "step": 50010,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.7830759,
      "epoch": 1.1670044765890433,
      "grad_norm": 6.4375,
      "learning_rate": 3.899606962605902e-06,
      "loss": 0.76939902,
      "memory(GiB)": 147.13,
      "step": 50020,
      "train_speed(iter/s)": 0.200614
    },
    {
      "acc": 0.77431717,
      "epoch": 1.1672377841613322,
      "grad_norm": 5.84375,
      "learning_rate": 3.897764246526003e-06,
      "loss": 0.80634575,
      "memory(GiB)": 147.13,
      "step": 50030,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.77010593,
      "epoch": 1.167471091733621,
      "grad_norm": 6.03125,
      "learning_rate": 3.895921687789936e-06,
      "loss": 0.83955479,
      "memory(GiB)": 147.13,
      "step": 50040,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.77356577,
      "epoch": 1.16770439930591,
      "grad_norm": 5.0625,
      "learning_rate": 3.894079286660729e-06,
      "loss": 0.80566463,
      "memory(GiB)": 147.13,
      "step": 50050,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.75591946,
      "epoch": 1.1679377068781989,
      "grad_norm": 25.0,
      "learning_rate": 3.892237043401382e-06,
      "loss": 0.87540159,
      "memory(GiB)": 147.13,
      "step": 50060,
      "train_speed(iter/s)": 0.200697
    },
    {
      "acc": 0.79346662,
      "epoch": 1.1681710144504878,
      "grad_norm": 9.4375,
      "learning_rate": 3.890394958274877e-06,
      "loss": 0.74789476,
      "memory(GiB)": 147.13,
      "step": 50070,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.78836994,
      "epoch": 1.1684043220227767,
      "grad_norm": 5.625,
      "learning_rate": 3.888553031544169e-06,
      "loss": 0.75674076,
      "memory(GiB)": 147.13,
      "step": 50080,
      "train_speed(iter/s)": 0.20074
    },
    {
      "acc": 0.78470092,
      "epoch": 1.1686376295950656,
      "grad_norm": 4.6875,
      "learning_rate": 3.886711263472192e-06,
      "loss": 0.78518977,
      "memory(GiB)": 147.13,
      "step": 50090,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.79191232,
      "epoch": 1.1688709371673545,
      "grad_norm": 3.90625,
      "learning_rate": 3.884869654321859e-06,
      "loss": 0.73618574,
      "memory(GiB)": 147.13,
      "step": 50100,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.78092413,
      "epoch": 1.1691042447396434,
      "grad_norm": 10.1875,
      "learning_rate": 3.883028204356058e-06,
      "loss": 0.77842164,
      "memory(GiB)": 147.13,
      "step": 50110,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.7757453,
      "epoch": 1.1693375523119323,
      "grad_norm": 26.125,
      "learning_rate": 3.881186913837657e-06,
      "loss": 0.79169908,
      "memory(GiB)": 147.13,
      "step": 50120,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.80329933,
      "epoch": 1.1695708598842212,
      "grad_norm": 6.96875,
      "learning_rate": 3.879345783029498e-06,
      "loss": 0.70160761,
      "memory(GiB)": 147.13,
      "step": 50130,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.77083302,
      "epoch": 1.16980416745651,
      "grad_norm": 6.34375,
      "learning_rate": 3.877504812194404e-06,
      "loss": 0.82752018,
      "memory(GiB)": 147.13,
      "step": 50140,
      "train_speed(iter/s)": 0.200867
    },
    {
      "acc": 0.79996233,
      "epoch": 1.170037475028799,
      "grad_norm": 4.78125,
      "learning_rate": 3.875664001595172e-06,
      "loss": 0.68956242,
      "memory(GiB)": 147.13,
      "step": 50150,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78814058,
      "epoch": 1.1702707826010879,
      "grad_norm": 5.25,
      "learning_rate": 3.873823351494576e-06,
      "loss": 0.74643149,
      "memory(GiB)": 147.13,
      "step": 50160,
      "train_speed(iter/s)": 0.200908
    },
    {
      "acc": 0.7647471,
      "epoch": 1.1705040901733768,
      "grad_norm": 5.71875,
      "learning_rate": 3.8719828621553715e-06,
      "loss": 0.85008469,
      "memory(GiB)": 147.13,
      "step": 50170,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.75985403,
      "epoch": 1.1707373977456657,
      "grad_norm": 4.875,
      "learning_rate": 3.870142533840283e-06,
      "loss": 0.88021393,
      "memory(GiB)": 147.13,
      "step": 50180,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.77487907,
      "epoch": 1.1709707053179546,
      "grad_norm": 5.34375,
      "learning_rate": 3.868302366812024e-06,
      "loss": 0.81592302,
      "memory(GiB)": 147.13,
      "step": 50190,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.77659655,
      "epoch": 1.1712040128902435,
      "grad_norm": 4.375,
      "learning_rate": 3.8664623613332705e-06,
      "loss": 0.81098127,
      "memory(GiB)": 147.13,
      "step": 50200,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.79070673,
      "epoch": 1.1714373204625323,
      "grad_norm": 4.65625,
      "learning_rate": 3.864622517666685e-06,
      "loss": 0.75920906,
      "memory(GiB)": 147.13,
      "step": 50210,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.79322906,
      "epoch": 1.1716706280348212,
      "grad_norm": 5.15625,
      "learning_rate": 3.862782836074906e-06,
      "loss": 0.72068224,
      "memory(GiB)": 147.13,
      "step": 50220,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.76726875,
      "epoch": 1.17190393560711,
      "grad_norm": 6.5,
      "learning_rate": 3.860943316820548e-06,
      "loss": 0.84628353,
      "memory(GiB)": 147.13,
      "step": 50230,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.79586458,
      "epoch": 1.172137243179399,
      "grad_norm": 7.03125,
      "learning_rate": 3.859103960166198e-06,
      "loss": 0.72394934,
      "memory(GiB)": 147.13,
      "step": 50240,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.77846246,
      "epoch": 1.1723705507516877,
      "grad_norm": 4.65625,
      "learning_rate": 3.857264766374428e-06,
      "loss": 0.79643097,
      "memory(GiB)": 147.13,
      "step": 50250,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.77642593,
      "epoch": 1.1726038583239768,
      "grad_norm": 4.5625,
      "learning_rate": 3.855425735707779e-06,
      "loss": 0.82001638,
      "memory(GiB)": 147.13,
      "step": 50260,
      "train_speed(iter/s)": 0.201116
    },
    {
      "acc": 0.76520214,
      "epoch": 1.1728371658962655,
      "grad_norm": 5.28125,
      "learning_rate": 3.853586868428775e-06,
      "loss": 0.84052525,
      "memory(GiB)": 147.13,
      "step": 50270,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.7804378,
      "epoch": 1.1730704734685544,
      "grad_norm": 5.375,
      "learning_rate": 3.851748164799914e-06,
      "loss": 0.77831354,
      "memory(GiB)": 147.13,
      "step": 50280,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.79932928,
      "epoch": 1.1733037810408433,
      "grad_norm": 13.5,
      "learning_rate": 3.849909625083666e-06,
      "loss": 0.70987091,
      "memory(GiB)": 147.13,
      "step": 50290,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.7968874,
      "epoch": 1.1735370886131322,
      "grad_norm": 4.96875,
      "learning_rate": 3.848071249542486e-06,
      "loss": 0.72468081,
      "memory(GiB)": 147.13,
      "step": 50300,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.76522064,
      "epoch": 1.1737703961854211,
      "grad_norm": 4.5,
      "learning_rate": 3.846233038438803e-06,
      "loss": 0.83323078,
      "memory(GiB)": 147.13,
      "step": 50310,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.76262903,
      "epoch": 1.17400370375771,
      "grad_norm": 5.59375,
      "learning_rate": 3.844394992035017e-06,
      "loss": 0.83975763,
      "memory(GiB)": 147.13,
      "step": 50320,
      "train_speed(iter/s)": 0.201241
    },
    {
      "acc": 0.79148641,
      "epoch": 1.174237011329999,
      "grad_norm": 4.34375,
      "learning_rate": 3.842557110593509e-06,
      "loss": 0.74630098,
      "memory(GiB)": 147.13,
      "step": 50330,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.78939476,
      "epoch": 1.1744703189022878,
      "grad_norm": 4.90625,
      "learning_rate": 3.840719394376638e-06,
      "loss": 0.74165897,
      "memory(GiB)": 147.13,
      "step": 50340,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.78680506,
      "epoch": 1.1747036264745767,
      "grad_norm": 5.5,
      "learning_rate": 3.838881843646736e-06,
      "loss": 0.77483697,
      "memory(GiB)": 147.13,
      "step": 50350,
      "train_speed(iter/s)": 0.2013
    },
    {
      "acc": 0.77357845,
      "epoch": 1.1749369340468656,
      "grad_norm": 5.125,
      "learning_rate": 3.8370444586661135e-06,
      "loss": 0.82649841,
      "memory(GiB)": 147.13,
      "step": 50360,
      "train_speed(iter/s)": 0.20132
    },
    {
      "acc": 0.77592096,
      "epoch": 1.1751702416191545,
      "grad_norm": 37.0,
      "learning_rate": 3.835207239697057e-06,
      "loss": 0.78373413,
      "memory(GiB)": 147.13,
      "step": 50370,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.80010624,
      "epoch": 1.1754035491914434,
      "grad_norm": 7.3125,
      "learning_rate": 3.8333701870018296e-06,
      "loss": 0.71716213,
      "memory(GiB)": 147.13,
      "step": 50380,
      "train_speed(iter/s)": 0.20136
    },
    {
      "acc": 0.78422251,
      "epoch": 1.1756368567637323,
      "grad_norm": 5.25,
      "learning_rate": 3.831533300842667e-06,
      "loss": 0.77883387,
      "memory(GiB)": 147.13,
      "step": 50390,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.77899208,
      "epoch": 1.1758701643360212,
      "grad_norm": 5.5,
      "learning_rate": 3.829696581481787e-06,
      "loss": 0.78697596,
      "memory(GiB)": 147.13,
      "step": 50400,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.77367115,
      "epoch": 1.17610347190831,
      "grad_norm": 4.1875,
      "learning_rate": 3.827860029181382e-06,
      "loss": 0.82581224,
      "memory(GiB)": 147.13,
      "step": 50410,
      "train_speed(iter/s)": 0.20142
    },
    {
      "acc": 0.76857705,
      "epoch": 1.176336779480599,
      "grad_norm": 6.0625,
      "learning_rate": 3.826023644203617e-06,
      "loss": 0.85204086,
      "memory(GiB)": 147.13,
      "step": 50420,
      "train_speed(iter/s)": 0.20144
    },
    {
      "acc": 0.79034348,
      "epoch": 1.176570087052888,
      "grad_norm": 6.96875,
      "learning_rate": 3.824187426810635e-06,
      "loss": 0.73557873,
      "memory(GiB)": 147.13,
      "step": 50430,
      "train_speed(iter/s)": 0.201461
    },
    {
      "acc": 0.7905654,
      "epoch": 1.1768033946251768,
      "grad_norm": 4.96875,
      "learning_rate": 3.822351377264555e-06,
      "loss": 0.75825205,
      "memory(GiB)": 147.13,
      "step": 50440,
      "train_speed(iter/s)": 0.201482
    },
    {
      "acc": 0.78993406,
      "epoch": 1.1770367021974657,
      "grad_norm": 4.5,
      "learning_rate": 3.820515495827476e-06,
      "loss": 0.77078609,
      "memory(GiB)": 147.13,
      "step": 50450,
      "train_speed(iter/s)": 0.201502
    },
    {
      "acc": 0.76711464,
      "epoch": 1.1772700097697546,
      "grad_norm": 5.4375,
      "learning_rate": 3.818679782761465e-06,
      "loss": 0.83489895,
      "memory(GiB)": 147.13,
      "step": 50460,
      "train_speed(iter/s)": 0.201523
    },
    {
      "acc": 0.76933699,
      "epoch": 1.1775033173420435,
      "grad_norm": 8.25,
      "learning_rate": 3.816844238328573e-06,
      "loss": 0.83145552,
      "memory(GiB)": 147.13,
      "step": 50470,
      "train_speed(iter/s)": 0.201543
    },
    {
      "acc": 0.77108583,
      "epoch": 1.1777366249143324,
      "grad_norm": 6.75,
      "learning_rate": 3.815008862790822e-06,
      "loss": 0.82043381,
      "memory(GiB)": 147.13,
      "step": 50480,
      "train_speed(iter/s)": 0.201563
    },
    {
      "acc": 0.7991888,
      "epoch": 1.1779699324866213,
      "grad_norm": 5.3125,
      "learning_rate": 3.813173656410211e-06,
      "loss": 0.74574351,
      "memory(GiB)": 147.13,
      "step": 50490,
      "train_speed(iter/s)": 0.201583
    },
    {
      "acc": 0.7735095,
      "epoch": 1.1782032400589102,
      "grad_norm": 6.46875,
      "learning_rate": 3.8113386194487177e-06,
      "loss": 0.8170289,
      "memory(GiB)": 147.13,
      "step": 50500,
      "train_speed(iter/s)": 0.201604
    },
    {
      "epoch": 1.1782032400589102,
      "eval_acc": 0.7443518330349046,
      "eval_loss": 0.8057728409767151,
      "eval_runtime": 1269.6788,
      "eval_samples_per_second": 28.347,
      "eval_steps_per_second": 14.174,
      "step": 50500
    },
    {
      "acc": 0.77446365,
      "epoch": 1.178436547631199,
      "grad_norm": 6.3125,
      "learning_rate": 3.80950375216829e-06,
      "loss": 0.80997276,
      "memory(GiB)": 147.13,
      "step": 50510,
      "train_speed(iter/s)": 0.200592
    },
    {
      "acc": 0.78115602,
      "epoch": 1.178669855203488,
      "grad_norm": 5.5625,
      "learning_rate": 3.807669054830855e-06,
      "loss": 0.77538776,
      "memory(GiB)": 147.13,
      "step": 50520,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.7817029,
      "epoch": 1.178903162775777,
      "grad_norm": 4.875,
      "learning_rate": 3.8058345276983165e-06,
      "loss": 0.78529196,
      "memory(GiB)": 147.13,
      "step": 50530,
      "train_speed(iter/s)": 0.200633
    },
    {
      "acc": 0.76126156,
      "epoch": 1.1791364703480658,
      "grad_norm": 5.46875,
      "learning_rate": 3.8040001710325547e-06,
      "loss": 0.87188816,
      "memory(GiB)": 147.13,
      "step": 50540,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.77631016,
      "epoch": 1.1793697779203547,
      "grad_norm": 6.09375,
      "learning_rate": 3.8021659850954186e-06,
      "loss": 0.79485397,
      "memory(GiB)": 147.13,
      "step": 50550,
      "train_speed(iter/s)": 0.200673
    },
    {
      "acc": 0.78766565,
      "epoch": 1.1796030854926436,
      "grad_norm": 4.78125,
      "learning_rate": 3.8003319701487407e-06,
      "loss": 0.77381859,
      "memory(GiB)": 147.13,
      "step": 50560,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.76473374,
      "epoch": 1.1798363930649325,
      "grad_norm": 4.59375,
      "learning_rate": 3.7984981264543247e-06,
      "loss": 0.8554389,
      "memory(GiB)": 147.13,
      "step": 50570,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.78639488,
      "epoch": 1.1800697006372214,
      "grad_norm": 6.28125,
      "learning_rate": 3.7966644542739538e-06,
      "loss": 0.76430273,
      "memory(GiB)": 147.13,
      "step": 50580,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.77539215,
      "epoch": 1.1803030082095103,
      "grad_norm": 4.5,
      "learning_rate": 3.794830953869381e-06,
      "loss": 0.81135311,
      "memory(GiB)": 147.13,
      "step": 50590,
      "train_speed(iter/s)": 0.200759
    },
    {
      "acc": 0.77136354,
      "epoch": 1.1805363157817992,
      "grad_norm": 4.875,
      "learning_rate": 3.7929976255023398e-06,
      "loss": 0.81377373,
      "memory(GiB)": 147.13,
      "step": 50600,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.77100811,
      "epoch": 1.180769623354088,
      "grad_norm": 7.5625,
      "learning_rate": 3.7911644694345368e-06,
      "loss": 0.80235748,
      "memory(GiB)": 147.13,
      "step": 50610,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.77771163,
      "epoch": 1.1810029309263768,
      "grad_norm": 4.5,
      "learning_rate": 3.789331485927654e-06,
      "loss": 0.78241501,
      "memory(GiB)": 147.13,
      "step": 50620,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.76570158,
      "epoch": 1.1812362384986659,
      "grad_norm": 4.71875,
      "learning_rate": 3.7874986752433506e-06,
      "loss": 0.8620245,
      "memory(GiB)": 147.13,
      "step": 50630,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.79454651,
      "epoch": 1.1814695460709546,
      "grad_norm": 6.4375,
      "learning_rate": 3.78566603764326e-06,
      "loss": 0.74558945,
      "memory(GiB)": 147.13,
      "step": 50640,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.78433299,
      "epoch": 1.1817028536432437,
      "grad_norm": 5.46875,
      "learning_rate": 3.7838335733889895e-06,
      "loss": 0.78314257,
      "memory(GiB)": 147.13,
      "step": 50650,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.77772627,
      "epoch": 1.1819361612155324,
      "grad_norm": 5.59375,
      "learning_rate": 3.782001282742124e-06,
      "loss": 0.80842457,
      "memory(GiB)": 147.13,
      "step": 50660,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.77893171,
      "epoch": 1.1821694687878213,
      "grad_norm": 6.9375,
      "learning_rate": 3.7801691659642196e-06,
      "loss": 0.79274435,
      "memory(GiB)": 147.13,
      "step": 50670,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.77441139,
      "epoch": 1.1824027763601102,
      "grad_norm": 7.625,
      "learning_rate": 3.7783372233168127e-06,
      "loss": 0.81310921,
      "memory(GiB)": 147.13,
      "step": 50680,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.78166981,
      "epoch": 1.182636083932399,
      "grad_norm": 5.875,
      "learning_rate": 3.776505455061412e-06,
      "loss": 0.78240037,
      "memory(GiB)": 147.13,
      "step": 50690,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.78959174,
      "epoch": 1.182869391504688,
      "grad_norm": 4.6875,
      "learning_rate": 3.7746738614595022e-06,
      "loss": 0.75032959,
      "memory(GiB)": 147.13,
      "step": 50700,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.78085089,
      "epoch": 1.1831026990769768,
      "grad_norm": 5.21875,
      "learning_rate": 3.772842442772543e-06,
      "loss": 0.78894525,
      "memory(GiB)": 147.13,
      "step": 50710,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.79085646,
      "epoch": 1.1833360066492657,
      "grad_norm": 6.75,
      "learning_rate": 3.7710111992619696e-06,
      "loss": 0.7664422,
      "memory(GiB)": 147.13,
      "step": 50720,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.77849669,
      "epoch": 1.1835693142215546,
      "grad_norm": 5.21875,
      "learning_rate": 3.7691801311891898e-06,
      "loss": 0.79077878,
      "memory(GiB)": 147.13,
      "step": 50730,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.77560053,
      "epoch": 1.1838026217938435,
      "grad_norm": 6.28125,
      "learning_rate": 3.767349238815588e-06,
      "loss": 0.8155261,
      "memory(GiB)": 147.13,
      "step": 50740,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.76759486,
      "epoch": 1.1840359293661324,
      "grad_norm": 5.09375,
      "learning_rate": 3.7655185224025247e-06,
      "loss": 0.84323349,
      "memory(GiB)": 147.13,
      "step": 50750,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.77860336,
      "epoch": 1.1842692369384213,
      "grad_norm": 5.9375,
      "learning_rate": 3.7636879822113338e-06,
      "loss": 0.78937483,
      "memory(GiB)": 147.13,
      "step": 50760,
      "train_speed(iter/s)": 0.201104
    },
    {
      "acc": 0.77916279,
      "epoch": 1.1845025445107102,
      "grad_norm": 8.8125,
      "learning_rate": 3.761857618503326e-06,
      "loss": 0.80653362,
      "memory(GiB)": 147.13,
      "step": 50770,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.75581121,
      "epoch": 1.1847358520829991,
      "grad_norm": 6.75,
      "learning_rate": 3.7600274315397816e-06,
      "loss": 0.90901031,
      "memory(GiB)": 147.13,
      "step": 50780,
      "train_speed(iter/s)": 0.201144
    },
    {
      "acc": 0.75556707,
      "epoch": 1.184969159655288,
      "grad_norm": 4.53125,
      "learning_rate": 3.758197421581961e-06,
      "loss": 0.90556011,
      "memory(GiB)": 147.13,
      "step": 50790,
      "train_speed(iter/s)": 0.201164
    },
    {
      "acc": 0.77783809,
      "epoch": 1.185202467227577,
      "grad_norm": 4.46875,
      "learning_rate": 3.756367588891099e-06,
      "loss": 0.79138689,
      "memory(GiB)": 147.13,
      "step": 50800,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.77093124,
      "epoch": 1.1854357747998658,
      "grad_norm": 6.28125,
      "learning_rate": 3.754537933728401e-06,
      "loss": 0.82287769,
      "memory(GiB)": 147.13,
      "step": 50810,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.78769693,
      "epoch": 1.1856690823721547,
      "grad_norm": 15.625,
      "learning_rate": 3.7527084563550515e-06,
      "loss": 0.75005512,
      "memory(GiB)": 147.13,
      "step": 50820,
      "train_speed(iter/s)": 0.201226
    },
    {
      "acc": 0.78668466,
      "epoch": 1.1859023899444436,
      "grad_norm": 5.9375,
      "learning_rate": 3.750879157032207e-06,
      "loss": 0.75284238,
      "memory(GiB)": 147.13,
      "step": 50830,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.78141642,
      "epoch": 1.1861356975167325,
      "grad_norm": 7.15625,
      "learning_rate": 3.7490500360210003e-06,
      "loss": 0.79019589,
      "memory(GiB)": 147.13,
      "step": 50840,
      "train_speed(iter/s)": 0.201266
    },
    {
      "acc": 0.76448064,
      "epoch": 1.1863690050890214,
      "grad_norm": 6.15625,
      "learning_rate": 3.747221093582538e-06,
      "loss": 0.86280327,
      "memory(GiB)": 147.13,
      "step": 50850,
      "train_speed(iter/s)": 0.201286
    },
    {
      "acc": 0.76448793,
      "epoch": 1.1866023126613103,
      "grad_norm": 5.28125,
      "learning_rate": 3.7453923299779014e-06,
      "loss": 0.86757164,
      "memory(GiB)": 147.13,
      "step": 50860,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.78806419,
      "epoch": 1.1868356202335992,
      "grad_norm": 4.3125,
      "learning_rate": 3.743563745468144e-06,
      "loss": 0.75735083,
      "memory(GiB)": 147.13,
      "step": 50870,
      "train_speed(iter/s)": 0.201327
    },
    {
      "acc": 0.78509789,
      "epoch": 1.1870689278058881,
      "grad_norm": 6.125,
      "learning_rate": 3.7417353403142988e-06,
      "loss": 0.77933402,
      "memory(GiB)": 147.13,
      "step": 50880,
      "train_speed(iter/s)": 0.201348
    },
    {
      "acc": 0.76924162,
      "epoch": 1.187302235378177,
      "grad_norm": 4.84375,
      "learning_rate": 3.7399071147773668e-06,
      "loss": 0.83347654,
      "memory(GiB)": 147.13,
      "step": 50890,
      "train_speed(iter/s)": 0.201369
    },
    {
      "acc": 0.76855621,
      "epoch": 1.187535542950466,
      "grad_norm": 4.1875,
      "learning_rate": 3.7380790691183276e-06,
      "loss": 0.82479677,
      "memory(GiB)": 147.13,
      "step": 50900,
      "train_speed(iter/s)": 0.201388
    },
    {
      "acc": 0.77282152,
      "epoch": 1.1877688505227548,
      "grad_norm": 5.0625,
      "learning_rate": 3.7362512035981347e-06,
      "loss": 0.81683245,
      "memory(GiB)": 147.13,
      "step": 50910,
      "train_speed(iter/s)": 0.201409
    },
    {
      "acc": 0.77848644,
      "epoch": 1.1880021580950437,
      "grad_norm": 5.59375,
      "learning_rate": 3.7344235184777157e-06,
      "loss": 0.78320317,
      "memory(GiB)": 147.13,
      "step": 50920,
      "train_speed(iter/s)": 0.20143
    },
    {
      "acc": 0.77635636,
      "epoch": 1.1882354656673326,
      "grad_norm": 4.5625,
      "learning_rate": 3.7325960140179717e-06,
      "loss": 0.82402401,
      "memory(GiB)": 147.13,
      "step": 50930,
      "train_speed(iter/s)": 0.201451
    },
    {
      "acc": 0.80113115,
      "epoch": 1.1884687732396215,
      "grad_norm": 3.828125,
      "learning_rate": 3.730768690479779e-06,
      "loss": 0.70224762,
      "memory(GiB)": 147.13,
      "step": 50940,
      "train_speed(iter/s)": 0.20147
    },
    {
      "acc": 0.78452883,
      "epoch": 1.1887020808119104,
      "grad_norm": 6.09375,
      "learning_rate": 3.7289415481239865e-06,
      "loss": 0.78695316,
      "memory(GiB)": 147.13,
      "step": 50950,
      "train_speed(iter/s)": 0.20149
    },
    {
      "acc": 0.79433985,
      "epoch": 1.1889353883841993,
      "grad_norm": 5.53125,
      "learning_rate": 3.727114587211419e-06,
      "loss": 0.7226675,
      "memory(GiB)": 147.13,
      "step": 50960,
      "train_speed(iter/s)": 0.201511
    },
    {
      "acc": 0.78255625,
      "epoch": 1.1891686959564882,
      "grad_norm": 4.46875,
      "learning_rate": 3.7252878080028744e-06,
      "loss": 0.77606573,
      "memory(GiB)": 147.13,
      "step": 50970,
      "train_speed(iter/s)": 0.201531
    },
    {
      "acc": 0.77274323,
      "epoch": 1.1894020035287771,
      "grad_norm": 5.75,
      "learning_rate": 3.7234612107591246e-06,
      "loss": 0.83783932,
      "memory(GiB)": 147.13,
      "step": 50980,
      "train_speed(iter/s)": 0.201551
    },
    {
      "acc": 0.79396834,
      "epoch": 1.189635311101066,
      "grad_norm": 3.640625,
      "learning_rate": 3.721634795740918e-06,
      "loss": 0.73608699,
      "memory(GiB)": 147.13,
      "step": 50990,
      "train_speed(iter/s)": 0.201572
    },
    {
      "acc": 0.78861499,
      "epoch": 1.189868618673355,
      "grad_norm": 4.125,
      "learning_rate": 3.719808563208971e-06,
      "loss": 0.75779295,
      "memory(GiB)": 147.13,
      "step": 51000,
      "train_speed(iter/s)": 0.201593
    },
    {
      "epoch": 1.189868618673355,
      "eval_acc": 0.7443236476500195,
      "eval_loss": 0.8055968880653381,
      "eval_runtime": 1269.0015,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "step": 51000
    },
    {
      "acc": 0.79447026,
      "epoch": 1.1901019262456436,
      "grad_norm": 5.28125,
      "learning_rate": 3.71798251342398e-06,
      "loss": 0.73775778,
      "memory(GiB)": 147.13,
      "step": 51010,
      "train_speed(iter/s)": 0.20059
    },
    {
      "acc": 0.78940754,
      "epoch": 1.1903352338179327,
      "grad_norm": 6.1875,
      "learning_rate": 3.7161566466466137e-06,
      "loss": 0.74937639,
      "memory(GiB)": 147.13,
      "step": 51020,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.78550367,
      "epoch": 1.1905685413902214,
      "grad_norm": 5.09375,
      "learning_rate": 3.714330963137512e-06,
      "loss": 0.78713713,
      "memory(GiB)": 147.13,
      "step": 51030,
      "train_speed(iter/s)": 0.200632
    },
    {
      "acc": 0.77606764,
      "epoch": 1.1908018489625105,
      "grad_norm": 7.4375,
      "learning_rate": 3.7125054631572915e-06,
      "loss": 0.8188735,
      "memory(GiB)": 147.13,
      "step": 51040,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.79022417,
      "epoch": 1.1910351565347992,
      "grad_norm": 5.5,
      "learning_rate": 3.710680146966542e-06,
      "loss": 0.75739288,
      "memory(GiB)": 147.13,
      "step": 51050,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.75823822,
      "epoch": 1.191268464107088,
      "grad_norm": 9.25,
      "learning_rate": 3.7088550148258277e-06,
      "loss": 0.87464418,
      "memory(GiB)": 147.13,
      "step": 51060,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.7662158,
      "epoch": 1.191501771679377,
      "grad_norm": 6.46875,
      "learning_rate": 3.707030066995685e-06,
      "loss": 0.84084873,
      "memory(GiB)": 147.13,
      "step": 51070,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.77479043,
      "epoch": 1.1917350792516659,
      "grad_norm": 5.4375,
      "learning_rate": 3.705205303736625e-06,
      "loss": 0.8108655,
      "memory(GiB)": 147.13,
      "step": 51080,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78514309,
      "epoch": 1.1919683868239548,
      "grad_norm": 5.40625,
      "learning_rate": 3.7033807253091313e-06,
      "loss": 0.77751369,
      "memory(GiB)": 147.13,
      "step": 51090,
      "train_speed(iter/s)": 0.200749
    },
    {
      "acc": 0.76177063,
      "epoch": 1.1922016943962437,
      "grad_norm": 5.46875,
      "learning_rate": 3.7015563319736618e-06,
      "loss": 0.8812314,
      "memory(GiB)": 147.13,
      "step": 51100,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.78570685,
      "epoch": 1.1924350019685326,
      "grad_norm": 4.375,
      "learning_rate": 3.6997321239906513e-06,
      "loss": 0.76931667,
      "memory(GiB)": 147.13,
      "step": 51110,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.78043647,
      "epoch": 1.1926683095408215,
      "grad_norm": 5.15625,
      "learning_rate": 3.6979081016204998e-06,
      "loss": 0.79293995,
      "memory(GiB)": 147.13,
      "step": 51120,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.7917635,
      "epoch": 1.1929016171131104,
      "grad_norm": 5.59375,
      "learning_rate": 3.6960842651235894e-06,
      "loss": 0.73350315,
      "memory(GiB)": 147.13,
      "step": 51130,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.78342714,
      "epoch": 1.1931349246853993,
      "grad_norm": 4.6875,
      "learning_rate": 3.6942606147602705e-06,
      "loss": 0.76641498,
      "memory(GiB)": 147.13,
      "step": 51140,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.77356782,
      "epoch": 1.1933682322576882,
      "grad_norm": 7.90625,
      "learning_rate": 3.6924371507908695e-06,
      "loss": 0.81099625,
      "memory(GiB)": 147.13,
      "step": 51150,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.8011385,
      "epoch": 1.193601539829977,
      "grad_norm": 5.09375,
      "learning_rate": 3.690613873475687e-06,
      "loss": 0.71837916,
      "memory(GiB)": 147.13,
      "step": 51160,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.78795547,
      "epoch": 1.193834847402266,
      "grad_norm": 5.6875,
      "learning_rate": 3.6887907830749923e-06,
      "loss": 0.75448265,
      "memory(GiB)": 147.13,
      "step": 51170,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.78816214,
      "epoch": 1.1940681549745549,
      "grad_norm": 4.46875,
      "learning_rate": 3.686967879849033e-06,
      "loss": 0.7446311,
      "memory(GiB)": 147.13,
      "step": 51180,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.78221531,
      "epoch": 1.1943014625468438,
      "grad_norm": 5.71875,
      "learning_rate": 3.6851451640580264e-06,
      "loss": 0.7772069,
      "memory(GiB)": 147.13,
      "step": 51190,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.76984034,
      "epoch": 1.1945347701191327,
      "grad_norm": 6.5625,
      "learning_rate": 3.6833226359621668e-06,
      "loss": 0.83698673,
      "memory(GiB)": 147.13,
      "step": 51200,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.77070379,
      "epoch": 1.1947680776914216,
      "grad_norm": 5.59375,
      "learning_rate": 3.6815002958216183e-06,
      "loss": 0.81270151,
      "memory(GiB)": 147.13,
      "step": 51210,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.79680748,
      "epoch": 1.1950013852637105,
      "grad_norm": 5.78125,
      "learning_rate": 3.67967814389652e-06,
      "loss": 0.71274958,
      "memory(GiB)": 147.13,
      "step": 51220,
      "train_speed(iter/s)": 0.201011
    },
    {
      "acc": 0.78539243,
      "epoch": 1.1952346928359994,
      "grad_norm": 4.84375,
      "learning_rate": 3.6778561804469825e-06,
      "loss": 0.78309517,
      "memory(GiB)": 147.13,
      "step": 51230,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.78247108,
      "epoch": 1.1954680004082883,
      "grad_norm": 5.625,
      "learning_rate": 3.676034405733092e-06,
      "loss": 0.7762372,
      "memory(GiB)": 147.13,
      "step": 51240,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.77460847,
      "epoch": 1.1957013079805772,
      "grad_norm": 5.46875,
      "learning_rate": 3.6742128200149042e-06,
      "loss": 0.82830582,
      "memory(GiB)": 147.13,
      "step": 51250,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.77715383,
      "epoch": 1.195934615552866,
      "grad_norm": 5.6875,
      "learning_rate": 3.672391423552451e-06,
      "loss": 0.81013908,
      "memory(GiB)": 147.13,
      "step": 51260,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.76946192,
      "epoch": 1.196167923125155,
      "grad_norm": 5.75,
      "learning_rate": 3.6705702166057366e-06,
      "loss": 0.83462391,
      "memory(GiB)": 147.13,
      "step": 51270,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.77984385,
      "epoch": 1.1964012306974439,
      "grad_norm": 5.75,
      "learning_rate": 3.668749199434738e-06,
      "loss": 0.80135126,
      "memory(GiB)": 147.13,
      "step": 51280,
      "train_speed(iter/s)": 0.20113
    },
    {
      "acc": 0.77329564,
      "epoch": 1.1966345382697328,
      "grad_norm": 5.625,
      "learning_rate": 3.6669283722994054e-06,
      "loss": 0.81273098,
      "memory(GiB)": 147.13,
      "step": 51290,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.78445587,
      "epoch": 1.1968678458420217,
      "grad_norm": 5.78125,
      "learning_rate": 3.6651077354596586e-06,
      "loss": 0.76524854,
      "memory(GiB)": 147.13,
      "step": 51300,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.79493475,
      "epoch": 1.1971011534143106,
      "grad_norm": 7.125,
      "learning_rate": 3.6632872891753956e-06,
      "loss": 0.73145099,
      "memory(GiB)": 147.13,
      "step": 51310,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.77402053,
      "epoch": 1.1973344609865995,
      "grad_norm": 3.96875,
      "learning_rate": 3.661467033706483e-06,
      "loss": 0.79744701,
      "memory(GiB)": 147.13,
      "step": 51320,
      "train_speed(iter/s)": 0.201213
    },
    {
      "acc": 0.78484211,
      "epoch": 1.1975677685588884,
      "grad_norm": 6.1875,
      "learning_rate": 3.6596469693127636e-06,
      "loss": 0.77361097,
      "memory(GiB)": 147.13,
      "step": 51330,
      "train_speed(iter/s)": 0.201234
    },
    {
      "acc": 0.80426512,
      "epoch": 1.1978010761311773,
      "grad_norm": 7.53125,
      "learning_rate": 3.6578270962540506e-06,
      "loss": 0.69536757,
      "memory(GiB)": 147.13,
      "step": 51340,
      "train_speed(iter/s)": 0.201254
    },
    {
      "acc": 0.75910811,
      "epoch": 1.1980343837034662,
      "grad_norm": 13.5,
      "learning_rate": 3.6560074147901287e-06,
      "loss": 0.85062847,
      "memory(GiB)": 147.13,
      "step": 51350,
      "train_speed(iter/s)": 0.201275
    },
    {
      "acc": 0.78548346,
      "epoch": 1.198267691275755,
      "grad_norm": 8.625,
      "learning_rate": 3.654187925180758e-06,
      "loss": 0.76579847,
      "memory(GiB)": 147.13,
      "step": 51360,
      "train_speed(iter/s)": 0.201297
    },
    {
      "acc": 0.78351984,
      "epoch": 1.198500998848044,
      "grad_norm": 5.25,
      "learning_rate": 3.65236862768567e-06,
      "loss": 0.77279978,
      "memory(GiB)": 147.13,
      "step": 51370,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.79646559,
      "epoch": 1.1987343064203329,
      "grad_norm": 5.90625,
      "learning_rate": 3.650549522564569e-06,
      "loss": 0.73050995,
      "memory(GiB)": 147.13,
      "step": 51380,
      "train_speed(iter/s)": 0.201338
    },
    {
      "acc": 0.77936029,
      "epoch": 1.1989676139926218,
      "grad_norm": 8.0625,
      "learning_rate": 3.648730610077131e-06,
      "loss": 0.79830408,
      "memory(GiB)": 147.13,
      "step": 51390,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.78159542,
      "epoch": 1.1992009215649104,
      "grad_norm": 4.71875,
      "learning_rate": 3.646911890483006e-06,
      "loss": 0.80945339,
      "memory(GiB)": 147.13,
      "step": 51400,
      "train_speed(iter/s)": 0.201379
    },
    {
      "acc": 0.77410226,
      "epoch": 1.1994342291371995,
      "grad_norm": 5.34375,
      "learning_rate": 3.645093364041815e-06,
      "loss": 0.81879921,
      "memory(GiB)": 147.13,
      "step": 51410,
      "train_speed(iter/s)": 0.201398
    },
    {
      "acc": 0.78253832,
      "epoch": 1.1996675367094882,
      "grad_norm": 4.40625,
      "learning_rate": 3.6432750310131537e-06,
      "loss": 0.78207197,
      "memory(GiB)": 147.13,
      "step": 51420,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.77488294,
      "epoch": 1.1999008442817771,
      "grad_norm": 6.0625,
      "learning_rate": 3.6414568916565884e-06,
      "loss": 0.791008,
      "memory(GiB)": 147.13,
      "step": 51430,
      "train_speed(iter/s)": 0.201438
    },
    {
      "acc": 0.78896904,
      "epoch": 1.200134151854066,
      "grad_norm": 6.5,
      "learning_rate": 3.6396389462316558e-06,
      "loss": 0.76538782,
      "memory(GiB)": 147.13,
      "step": 51440,
      "train_speed(iter/s)": 0.201457
    },
    {
      "acc": 0.77477431,
      "epoch": 1.200367459426355,
      "grad_norm": 5.03125,
      "learning_rate": 3.6378211949978693e-06,
      "loss": 0.8104063,
      "memory(GiB)": 147.13,
      "step": 51450,
      "train_speed(iter/s)": 0.201477
    },
    {
      "acc": 0.77967916,
      "epoch": 1.2006007669986438,
      "grad_norm": 5.40625,
      "learning_rate": 3.6360036382147117e-06,
      "loss": 0.7820168,
      "memory(GiB)": 147.13,
      "step": 51460,
      "train_speed(iter/s)": 0.201498
    },
    {
      "acc": 0.79494371,
      "epoch": 1.2008340745709327,
      "grad_norm": 6.40625,
      "learning_rate": 3.634186276141638e-06,
      "loss": 0.73448949,
      "memory(GiB)": 147.13,
      "step": 51470,
      "train_speed(iter/s)": 0.201519
    },
    {
      "acc": 0.77666898,
      "epoch": 1.2010673821432216,
      "grad_norm": 5.84375,
      "learning_rate": 3.6323691090380756e-06,
      "loss": 0.79062681,
      "memory(GiB)": 147.13,
      "step": 51480,
      "train_speed(iter/s)": 0.20154
    },
    {
      "acc": 0.78186355,
      "epoch": 1.2013006897155105,
      "grad_norm": 4.625,
      "learning_rate": 3.630552137163427e-06,
      "loss": 0.7758111,
      "memory(GiB)": 147.13,
      "step": 51490,
      "train_speed(iter/s)": 0.20156
    },
    {
      "acc": 0.78854303,
      "epoch": 1.2015339972877994,
      "grad_norm": 4.375,
      "learning_rate": 3.6287353607770613e-06,
      "loss": 0.75728836,
      "memory(GiB)": 147.13,
      "step": 51500,
      "train_speed(iter/s)": 0.20158
    },
    {
      "epoch": 1.2015339972877994,
      "eval_acc": 0.7444104458239269,
      "eval_loss": 0.8055254817008972,
      "eval_runtime": 1269.0052,
      "eval_samples_per_second": 28.362,
      "eval_steps_per_second": 14.181,
      "step": 51500
    },
    {
      "acc": 0.79272242,
      "epoch": 1.2017673048600883,
      "grad_norm": 4.34375,
      "learning_rate": 3.6269187801383267e-06,
      "loss": 0.74171906,
      "memory(GiB)": 147.13,
      "step": 51510,
      "train_speed(iter/s)": 0.20059
    },
    {
      "acc": 0.77594776,
      "epoch": 1.2020006124323772,
      "grad_norm": 6.90625,
      "learning_rate": 3.6251023955065356e-06,
      "loss": 0.79424891,
      "memory(GiB)": 147.13,
      "step": 51520,
      "train_speed(iter/s)": 0.200612
    },
    {
      "acc": 0.78154349,
      "epoch": 1.2022339200046661,
      "grad_norm": 4.5,
      "learning_rate": 3.623286207140979e-06,
      "loss": 0.7680542,
      "memory(GiB)": 147.13,
      "step": 51530,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.7791707,
      "epoch": 1.202467227576955,
      "grad_norm": 5.25,
      "learning_rate": 3.6214702153009157e-06,
      "loss": 0.79780264,
      "memory(GiB)": 147.13,
      "step": 51540,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.7682992,
      "epoch": 1.202700535149244,
      "grad_norm": 8.75,
      "learning_rate": 3.6196544202455787e-06,
      "loss": 0.84575796,
      "memory(GiB)": 147.13,
      "step": 51550,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.77210584,
      "epoch": 1.2029338427215328,
      "grad_norm": 4.5,
      "learning_rate": 3.617838822234175e-06,
      "loss": 0.82075291,
      "memory(GiB)": 147.13,
      "step": 51560,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.77745943,
      "epoch": 1.2031671502938217,
      "grad_norm": 6.71875,
      "learning_rate": 3.616023421525875e-06,
      "loss": 0.81119947,
      "memory(GiB)": 147.13,
      "step": 51570,
      "train_speed(iter/s)": 0.200713
    },
    {
      "acc": 0.78472643,
      "epoch": 1.2034004578661106,
      "grad_norm": 5.53125,
      "learning_rate": 3.61420821837983e-06,
      "loss": 0.76066618,
      "memory(GiB)": 147.13,
      "step": 51580,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.79031987,
      "epoch": 1.2036337654383995,
      "grad_norm": 4.8125,
      "learning_rate": 3.61239321305516e-06,
      "loss": 0.73573508,
      "memory(GiB)": 147.13,
      "step": 51590,
      "train_speed(iter/s)": 0.200754
    },
    {
      "acc": 0.78004036,
      "epoch": 1.2038670730106884,
      "grad_norm": 6.9375,
      "learning_rate": 3.610578405810955e-06,
      "loss": 0.80121174,
      "memory(GiB)": 147.13,
      "step": 51600,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.78116407,
      "epoch": 1.2041003805829773,
      "grad_norm": 5.15625,
      "learning_rate": 3.6087637969062783e-06,
      "loss": 0.77282748,
      "memory(GiB)": 147.13,
      "step": 51610,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.79813347,
      "epoch": 1.2043336881552662,
      "grad_norm": 7.09375,
      "learning_rate": 3.606949386600166e-06,
      "loss": 0.71265526,
      "memory(GiB)": 147.13,
      "step": 51620,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.76442337,
      "epoch": 1.204566995727555,
      "grad_norm": 7.375,
      "learning_rate": 3.605135175151624e-06,
      "loss": 0.8571002,
      "memory(GiB)": 147.13,
      "step": 51630,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.78852344,
      "epoch": 1.204800303299844,
      "grad_norm": 4.78125,
      "learning_rate": 3.6033211628196308e-06,
      "loss": 0.74523458,
      "memory(GiB)": 147.13,
      "step": 51640,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.77451267,
      "epoch": 1.205033610872133,
      "grad_norm": 6.15625,
      "learning_rate": 3.601507349863137e-06,
      "loss": 0.79993911,
      "memory(GiB)": 147.13,
      "step": 51650,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.77563667,
      "epoch": 1.2052669184444218,
      "grad_norm": 6.34375,
      "learning_rate": 3.599693736541061e-06,
      "loss": 0.81237659,
      "memory(GiB)": 147.13,
      "step": 51660,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.76454887,
      "epoch": 1.2055002260167107,
      "grad_norm": 5.71875,
      "learning_rate": 3.5978803231122977e-06,
      "loss": 0.86580725,
      "memory(GiB)": 147.13,
      "step": 51670,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.77970967,
      "epoch": 1.2057335335889996,
      "grad_norm": 5.125,
      "learning_rate": 3.596067109835713e-06,
      "loss": 0.78640342,
      "memory(GiB)": 147.13,
      "step": 51680,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.77567797,
      "epoch": 1.2059668411612885,
      "grad_norm": 3.65625,
      "learning_rate": 3.5942540969701386e-06,
      "loss": 0.79667077,
      "memory(GiB)": 147.13,
      "step": 51690,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.77737932,
      "epoch": 1.2062001487335774,
      "grad_norm": 4.84375,
      "learning_rate": 3.592441284774383e-06,
      "loss": 0.79587197,
      "memory(GiB)": 147.13,
      "step": 51700,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.770786,
      "epoch": 1.2064334563058663,
      "grad_norm": 6.46875,
      "learning_rate": 3.5906286735072255e-06,
      "loss": 0.82086658,
      "memory(GiB)": 147.13,
      "step": 51710,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.78262281,
      "epoch": 1.2066667638781552,
      "grad_norm": 4.4375,
      "learning_rate": 3.5888162634274154e-06,
      "loss": 0.77802362,
      "memory(GiB)": 147.13,
      "step": 51720,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.7677485,
      "epoch": 1.206900071450444,
      "grad_norm": 4.5625,
      "learning_rate": 3.5870040547936748e-06,
      "loss": 0.83916025,
      "memory(GiB)": 147.13,
      "step": 51730,
      "train_speed(iter/s)": 0.201034
    },
    {
      "acc": 0.77800913,
      "epoch": 1.207133379022733,
      "grad_norm": 6.09375,
      "learning_rate": 3.585192047864694e-06,
      "loss": 0.80633202,
      "memory(GiB)": 147.13,
      "step": 51740,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.75977621,
      "epoch": 1.2073666865950219,
      "grad_norm": 6.5,
      "learning_rate": 3.5833802428991373e-06,
      "loss": 0.86888771,
      "memory(GiB)": 147.13,
      "step": 51750,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.76785235,
      "epoch": 1.2075999941673108,
      "grad_norm": 5.3125,
      "learning_rate": 3.581568640155639e-06,
      "loss": 0.82541256,
      "memory(GiB)": 147.13,
      "step": 51760,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.78889942,
      "epoch": 1.2078333017395995,
      "grad_norm": 5.21875,
      "learning_rate": 3.5797572398928053e-06,
      "loss": 0.73905168,
      "memory(GiB)": 147.13,
      "step": 51770,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.78549714,
      "epoch": 1.2080666093118886,
      "grad_norm": 5.1875,
      "learning_rate": 3.5779460423692136e-06,
      "loss": 0.77860193,
      "memory(GiB)": 147.13,
      "step": 51780,
      "train_speed(iter/s)": 0.201131
    },
    {
      "acc": 0.77393236,
      "epoch": 1.2082999168841773,
      "grad_norm": 4.46875,
      "learning_rate": 3.5761350478434133e-06,
      "loss": 0.82066584,
      "memory(GiB)": 147.13,
      "step": 51790,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.76424036,
      "epoch": 1.2085332244564664,
      "grad_norm": 5.03125,
      "learning_rate": 3.5743242565739183e-06,
      "loss": 0.86626759,
      "memory(GiB)": 147.13,
      "step": 51800,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.77947626,
      "epoch": 1.208766532028755,
      "grad_norm": 5.71875,
      "learning_rate": 3.572513668819223e-06,
      "loss": 0.79311905,
      "memory(GiB)": 147.13,
      "step": 51810,
      "train_speed(iter/s)": 0.20119
    },
    {
      "acc": 0.77670922,
      "epoch": 1.208999839601044,
      "grad_norm": 6.03125,
      "learning_rate": 3.570703284837786e-06,
      "loss": 0.81235285,
      "memory(GiB)": 147.13,
      "step": 51820,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.79082127,
      "epoch": 1.2092331471733329,
      "grad_norm": 5.375,
      "learning_rate": 3.5688931048880397e-06,
      "loss": 0.75774207,
      "memory(GiB)": 147.13,
      "step": 51830,
      "train_speed(iter/s)": 0.20123
    },
    {
      "acc": 0.78081455,
      "epoch": 1.2094664547456218,
      "grad_norm": 5.375,
      "learning_rate": 3.567083129228387e-06,
      "loss": 0.76074839,
      "memory(GiB)": 147.13,
      "step": 51840,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.77981596,
      "epoch": 1.2096997623179107,
      "grad_norm": 6.0,
      "learning_rate": 3.5652733581172015e-06,
      "loss": 0.7848402,
      "memory(GiB)": 147.13,
      "step": 51850,
      "train_speed(iter/s)": 0.201272
    },
    {
      "acc": 0.78686342,
      "epoch": 1.2099330698901996,
      "grad_norm": 6.28125,
      "learning_rate": 3.5634637918128267e-06,
      "loss": 0.77134657,
      "memory(GiB)": 147.13,
      "step": 51860,
      "train_speed(iter/s)": 0.201291
    },
    {
      "acc": 0.7602006,
      "epoch": 1.2101663774624885,
      "grad_norm": 6.46875,
      "learning_rate": 3.56165443057358e-06,
      "loss": 0.87410669,
      "memory(GiB)": 147.13,
      "step": 51870,
      "train_speed(iter/s)": 0.20131
    },
    {
      "acc": 0.78266001,
      "epoch": 1.2103996850347774,
      "grad_norm": 3.96875,
      "learning_rate": 3.5598452746577443e-06,
      "loss": 0.76107483,
      "memory(GiB)": 147.13,
      "step": 51880,
      "train_speed(iter/s)": 0.20133
    },
    {
      "acc": 0.7808578,
      "epoch": 1.2106329926070662,
      "grad_norm": 5.40625,
      "learning_rate": 3.5580363243235773e-06,
      "loss": 0.76968455,
      "memory(GiB)": 147.13,
      "step": 51890,
      "train_speed(iter/s)": 0.20135
    },
    {
      "acc": 0.78987422,
      "epoch": 1.2108663001793551,
      "grad_norm": 5.3125,
      "learning_rate": 3.556227579829306e-06,
      "loss": 0.75738258,
      "memory(GiB)": 147.13,
      "step": 51900,
      "train_speed(iter/s)": 0.201369
    },
    {
      "acc": 0.78409944,
      "epoch": 1.211099607751644,
      "grad_norm": 7.03125,
      "learning_rate": 3.5544190414331305e-06,
      "loss": 0.77000666,
      "memory(GiB)": 147.13,
      "step": 51910,
      "train_speed(iter/s)": 0.201389
    },
    {
      "acc": 0.76755562,
      "epoch": 1.211332915323933,
      "grad_norm": 6.5625,
      "learning_rate": 3.552610709393215e-06,
      "loss": 0.83009071,
      "memory(GiB)": 147.13,
      "step": 51920,
      "train_speed(iter/s)": 0.201409
    },
    {
      "acc": 0.77854753,
      "epoch": 1.2115662228962218,
      "grad_norm": 4.71875,
      "learning_rate": 3.5508025839676997e-06,
      "loss": 0.79306674,
      "memory(GiB)": 147.13,
      "step": 51930,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.78213997,
      "epoch": 1.2117995304685107,
      "grad_norm": 7.90625,
      "learning_rate": 3.5489946654146945e-06,
      "loss": 0.77478514,
      "memory(GiB)": 147.13,
      "step": 51940,
      "train_speed(iter/s)": 0.20145
    },
    {
      "acc": 0.79072771,
      "epoch": 1.2120328380407996,
      "grad_norm": 6.75,
      "learning_rate": 3.547186953992281e-06,
      "loss": 0.74444256,
      "memory(GiB)": 147.13,
      "step": 51950,
      "train_speed(iter/s)": 0.20147
    },
    {
      "acc": 0.78262076,
      "epoch": 1.2122661456130885,
      "grad_norm": 9.25,
      "learning_rate": 3.5453794499585057e-06,
      "loss": 0.77354512,
      "memory(GiB)": 147.13,
      "step": 51960,
      "train_speed(iter/s)": 0.20149
    },
    {
      "acc": 0.79070244,
      "epoch": 1.2124994531853774,
      "grad_norm": 4.3125,
      "learning_rate": 3.543572153571393e-06,
      "loss": 0.73883228,
      "memory(GiB)": 147.13,
      "step": 51970,
      "train_speed(iter/s)": 0.20151
    },
    {
      "acc": 0.78741131,
      "epoch": 1.2127327607576663,
      "grad_norm": 5.40625,
      "learning_rate": 3.541765065088931e-06,
      "loss": 0.76005692,
      "memory(GiB)": 147.13,
      "step": 51980,
      "train_speed(iter/s)": 0.20153
    },
    {
      "acc": 0.77303057,
      "epoch": 1.2129660683299552,
      "grad_norm": 6.53125,
      "learning_rate": 3.539958184769082e-06,
      "loss": 0.80659695,
      "memory(GiB)": 147.13,
      "step": 51990,
      "train_speed(iter/s)": 0.20155
    },
    {
      "acc": 0.78834496,
      "epoch": 1.2131993759022441,
      "grad_norm": 5.28125,
      "learning_rate": 3.53815151286978e-06,
      "loss": 0.74369602,
      "memory(GiB)": 147.13,
      "step": 52000,
      "train_speed(iter/s)": 0.20157
    },
    {
      "epoch": 1.2131993759022441,
      "eval_acc": 0.7444085240931393,
      "eval_loss": 0.8053334355354309,
      "eval_runtime": 1270.0,
      "eval_samples_per_second": 28.339,
      "eval_steps_per_second": 14.17,
      "step": 52000
    },
    {
      "acc": 0.77881718,
      "epoch": 1.213432683474533,
      "grad_norm": 5.96875,
      "learning_rate": 3.536345049648924e-06,
      "loss": 0.78626618,
      "memory(GiB)": 147.13,
      "step": 52010,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.77282524,
      "epoch": 1.213665991046822,
      "grad_norm": 6.09375,
      "learning_rate": 3.5345387953643872e-06,
      "loss": 0.82502003,
      "memory(GiB)": 147.13,
      "step": 52020,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.79441557,
      "epoch": 1.2138992986191108,
      "grad_norm": 4.40625,
      "learning_rate": 3.5327327502740114e-06,
      "loss": 0.73947439,
      "memory(GiB)": 147.13,
      "step": 52030,
      "train_speed(iter/s)": 0.200626
    },
    {
      "acc": 0.7964479,
      "epoch": 1.2141326061913997,
      "grad_norm": 5.375,
      "learning_rate": 3.5309269146356097e-06,
      "loss": 0.72861004,
      "memory(GiB)": 147.13,
      "step": 52040,
      "train_speed(iter/s)": 0.200645
    },
    {
      "acc": 0.78275375,
      "epoch": 1.2143659137636886,
      "grad_norm": 4.9375,
      "learning_rate": 3.5291212887069624e-06,
      "loss": 0.77264972,
      "memory(GiB)": 147.13,
      "step": 52050,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.77427526,
      "epoch": 1.2145992213359775,
      "grad_norm": 8.25,
      "learning_rate": 3.5273158727458253e-06,
      "loss": 0.82076292,
      "memory(GiB)": 147.13,
      "step": 52060,
      "train_speed(iter/s)": 0.200684
    },
    {
      "acc": 0.78050809,
      "epoch": 1.2148325289082664,
      "grad_norm": 5.0625,
      "learning_rate": 3.5255106670099186e-06,
      "loss": 0.79774323,
      "memory(GiB)": 147.13,
      "step": 52070,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.77670012,
      "epoch": 1.2150658364805553,
      "grad_norm": 5.375,
      "learning_rate": 3.5237056717569363e-06,
      "loss": 0.80653028,
      "memory(GiB)": 147.13,
      "step": 52080,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.77818155,
      "epoch": 1.2152991440528442,
      "grad_norm": 5.09375,
      "learning_rate": 3.5219008872445414e-06,
      "loss": 0.80688896,
      "memory(GiB)": 147.13,
      "step": 52090,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.79195747,
      "epoch": 1.2155324516251331,
      "grad_norm": 4.40625,
      "learning_rate": 3.5200963137303644e-06,
      "loss": 0.73866539,
      "memory(GiB)": 147.13,
      "step": 52100,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.77116313,
      "epoch": 1.215765759197422,
      "grad_norm": 4.375,
      "learning_rate": 3.5182919514720087e-06,
      "loss": 0.81336851,
      "memory(GiB)": 147.13,
      "step": 52110,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.79589829,
      "epoch": 1.215999066769711,
      "grad_norm": 4.9375,
      "learning_rate": 3.5164878007270464e-06,
      "loss": 0.72459707,
      "memory(GiB)": 147.13,
      "step": 52120,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.7925993,
      "epoch": 1.2162323743419998,
      "grad_norm": 8.5625,
      "learning_rate": 3.5146838617530197e-06,
      "loss": 0.7503953,
      "memory(GiB)": 147.13,
      "step": 52130,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.77429123,
      "epoch": 1.2164656819142887,
      "grad_norm": 5.34375,
      "learning_rate": 3.5128801348074426e-06,
      "loss": 0.80790281,
      "memory(GiB)": 147.13,
      "step": 52140,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.79118776,
      "epoch": 1.2166989894865776,
      "grad_norm": 7.59375,
      "learning_rate": 3.511076620147792e-06,
      "loss": 0.74122844,
      "memory(GiB)": 147.13,
      "step": 52150,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.79350772,
      "epoch": 1.2169322970588663,
      "grad_norm": 4.96875,
      "learning_rate": 3.5092733180315206e-06,
      "loss": 0.75740881,
      "memory(GiB)": 147.13,
      "step": 52160,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.80063057,
      "epoch": 1.2171656046311554,
      "grad_norm": 6.40625,
      "learning_rate": 3.5074702287160523e-06,
      "loss": 0.70421405,
      "memory(GiB)": 147.13,
      "step": 52170,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.78242369,
      "epoch": 1.217398912203444,
      "grad_norm": 5.9375,
      "learning_rate": 3.5056673524587733e-06,
      "loss": 0.79225969,
      "memory(GiB)": 147.13,
      "step": 52180,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.77852802,
      "epoch": 1.2176322197757332,
      "grad_norm": 4.40625,
      "learning_rate": 3.503864689517046e-06,
      "loss": 0.80126581,
      "memory(GiB)": 147.13,
      "step": 52190,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.78113604,
      "epoch": 1.217865527348022,
      "grad_norm": 4.71875,
      "learning_rate": 3.5020622401481996e-06,
      "loss": 0.8105751,
      "memory(GiB)": 147.13,
      "step": 52200,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.76554623,
      "epoch": 1.2180988349203108,
      "grad_norm": 5.71875,
      "learning_rate": 3.500260004609533e-06,
      "loss": 0.83915272,
      "memory(GiB)": 147.13,
      "step": 52210,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.76399746,
      "epoch": 1.2183321424925997,
      "grad_norm": 5.4375,
      "learning_rate": 3.4984579831583166e-06,
      "loss": 0.82813606,
      "memory(GiB)": 147.13,
      "step": 52220,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.79804144,
      "epoch": 1.2185654500648886,
      "grad_norm": 14.8125,
      "learning_rate": 3.4966561760517852e-06,
      "loss": 0.71784048,
      "memory(GiB)": 147.13,
      "step": 52230,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.77001858,
      "epoch": 1.2187987576371775,
      "grad_norm": 4.96875,
      "learning_rate": 3.494854583547148e-06,
      "loss": 0.8265419,
      "memory(GiB)": 147.13,
      "step": 52240,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.80135403,
      "epoch": 1.2190320652094664,
      "grad_norm": 5.8125,
      "learning_rate": 3.4930532059015845e-06,
      "loss": 0.70065279,
      "memory(GiB)": 147.13,
      "step": 52250,
      "train_speed(iter/s)": 0.201066
    },
    {
      "acc": 0.77459936,
      "epoch": 1.2192653727817553,
      "grad_norm": 7.46875,
      "learning_rate": 3.491252043372236e-06,
      "loss": 0.81747303,
      "memory(GiB)": 147.13,
      "step": 52260,
      "train_speed(iter/s)": 0.201087
    },
    {
      "acc": 0.79471178,
      "epoch": 1.2194986803540442,
      "grad_norm": 7.625,
      "learning_rate": 3.4894510962162194e-06,
      "loss": 0.72956462,
      "memory(GiB)": 147.13,
      "step": 52270,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.76919899,
      "epoch": 1.219731987926333,
      "grad_norm": 6.25,
      "learning_rate": 3.4876503646906203e-06,
      "loss": 0.82097034,
      "memory(GiB)": 147.13,
      "step": 52280,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.78185291,
      "epoch": 1.219965295498622,
      "grad_norm": 4.71875,
      "learning_rate": 3.4858498490524924e-06,
      "loss": 0.80037918,
      "memory(GiB)": 147.13,
      "step": 52290,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.7733079,
      "epoch": 1.2201986030709109,
      "grad_norm": 4.65625,
      "learning_rate": 3.4840495495588593e-06,
      "loss": 0.78827391,
      "memory(GiB)": 147.13,
      "step": 52300,
      "train_speed(iter/s)": 0.201166
    },
    {
      "acc": 0.78070698,
      "epoch": 1.2204319106431998,
      "grad_norm": 4.59375,
      "learning_rate": 3.4822494664667117e-06,
      "loss": 0.80994854,
      "memory(GiB)": 147.13,
      "step": 52310,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.78385916,
      "epoch": 1.2206652182154887,
      "grad_norm": 5.4375,
      "learning_rate": 3.4804496000330124e-06,
      "loss": 0.75973425,
      "memory(GiB)": 147.13,
      "step": 52320,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78199615,
      "epoch": 1.2208985257877776,
      "grad_norm": 6.09375,
      "learning_rate": 3.478649950514691e-06,
      "loss": 0.76241202,
      "memory(GiB)": 147.13,
      "step": 52330,
      "train_speed(iter/s)": 0.201227
    },
    {
      "acc": 0.79231329,
      "epoch": 1.2211318333600665,
      "grad_norm": 8.8125,
      "learning_rate": 3.4768505181686468e-06,
      "loss": 0.71977391,
      "memory(GiB)": 147.13,
      "step": 52340,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.77791686,
      "epoch": 1.2213651409323554,
      "grad_norm": 6.875,
      "learning_rate": 3.4750513032517493e-06,
      "loss": 0.79545641,
      "memory(GiB)": 147.13,
      "step": 52350,
      "train_speed(iter/s)": 0.201267
    },
    {
      "acc": 0.78212147,
      "epoch": 1.2215984485046443,
      "grad_norm": 4.8125,
      "learning_rate": 3.473252306020837e-06,
      "loss": 0.7947474,
      "memory(GiB)": 147.13,
      "step": 52360,
      "train_speed(iter/s)": 0.201287
    },
    {
      "acc": 0.77962971,
      "epoch": 1.2218317560769332,
      "grad_norm": 6.71875,
      "learning_rate": 3.471453526732712e-06,
      "loss": 0.79630547,
      "memory(GiB)": 147.13,
      "step": 52370,
      "train_speed(iter/s)": 0.201308
    },
    {
      "acc": 0.78212519,
      "epoch": 1.222065063649222,
      "grad_norm": 5.90625,
      "learning_rate": 3.4696549656441537e-06,
      "loss": 0.77139959,
      "memory(GiB)": 147.13,
      "step": 52380,
      "train_speed(iter/s)": 0.201328
    },
    {
      "acc": 0.76968145,
      "epoch": 1.222298371221511,
      "grad_norm": 6.375,
      "learning_rate": 3.467856623011903e-06,
      "loss": 0.83397465,
      "memory(GiB)": 147.13,
      "step": 52390,
      "train_speed(iter/s)": 0.201348
    },
    {
      "acc": 0.79080944,
      "epoch": 1.2225316787937999,
      "grad_norm": 4.09375,
      "learning_rate": 3.4660584990926748e-06,
      "loss": 0.73990412,
      "memory(GiB)": 147.13,
      "step": 52400,
      "train_speed(iter/s)": 0.201367
    },
    {
      "acc": 0.77898731,
      "epoch": 1.2227649863660888,
      "grad_norm": 6.5,
      "learning_rate": 3.4642605941431494e-06,
      "loss": 0.78911119,
      "memory(GiB)": 147.13,
      "step": 52410,
      "train_speed(iter/s)": 0.201387
    },
    {
      "acc": 0.76685104,
      "epoch": 1.2229982939383777,
      "grad_norm": 4.1875,
      "learning_rate": 3.462462908419979e-06,
      "loss": 0.82818241,
      "memory(GiB)": 147.13,
      "step": 52420,
      "train_speed(iter/s)": 0.201405
    },
    {
      "acc": 0.78912749,
      "epoch": 1.2232316015106666,
      "grad_norm": 4.4375,
      "learning_rate": 3.4606654421797814e-06,
      "loss": 0.75287247,
      "memory(GiB)": 147.13,
      "step": 52430,
      "train_speed(iter/s)": 0.201424
    },
    {
      "acc": 0.77771249,
      "epoch": 1.2234649090829555,
      "grad_norm": 5.40625,
      "learning_rate": 3.458868195679146e-06,
      "loss": 0.81555805,
      "memory(GiB)": 147.13,
      "step": 52440,
      "train_speed(iter/s)": 0.201444
    },
    {
      "acc": 0.78952436,
      "epoch": 1.2236982166552444,
      "grad_norm": 4.28125,
      "learning_rate": 3.4570711691746262e-06,
      "loss": 0.772684,
      "memory(GiB)": 147.13,
      "step": 52450,
      "train_speed(iter/s)": 0.201465
    },
    {
      "acc": 0.77552791,
      "epoch": 1.2239315242275333,
      "grad_norm": 4.65625,
      "learning_rate": 3.4552743629227494e-06,
      "loss": 0.80711937,
      "memory(GiB)": 147.13,
      "step": 52460,
      "train_speed(iter/s)": 0.201485
    },
    {
      "acc": 0.7582983,
      "epoch": 1.2241648317998222,
      "grad_norm": 7.4375,
      "learning_rate": 3.4534777771800083e-06,
      "loss": 0.85153837,
      "memory(GiB)": 147.13,
      "step": 52470,
      "train_speed(iter/s)": 0.201505
    },
    {
      "acc": 0.77164292,
      "epoch": 1.224398139372111,
      "grad_norm": 5.0625,
      "learning_rate": 3.4516814122028676e-06,
      "loss": 0.84448309,
      "memory(GiB)": 147.13,
      "step": 52480,
      "train_speed(iter/s)": 0.201523
    },
    {
      "acc": 0.76962109,
      "epoch": 1.2246314469444,
      "grad_norm": 6.625,
      "learning_rate": 3.449885268247753e-06,
      "loss": 0.81848526,
      "memory(GiB)": 147.13,
      "step": 52490,
      "train_speed(iter/s)": 0.201542
    },
    {
      "acc": 0.78164902,
      "epoch": 1.2248647545166889,
      "grad_norm": 5.28125,
      "learning_rate": 3.448089345571066e-06,
      "loss": 0.78736157,
      "memory(GiB)": 147.13,
      "step": 52500,
      "train_speed(iter/s)": 0.201562
    },
    {
      "epoch": 1.2248647545166889,
      "eval_acc": 0.7444283819779446,
      "eval_loss": 0.8053562045097351,
      "eval_runtime": 1270.1507,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 52500
    },
    {
      "acc": 0.77160788,
      "epoch": 1.2250980620889778,
      "grad_norm": 4.78125,
      "learning_rate": 3.4462936444291744e-06,
      "loss": 0.80924244,
      "memory(GiB)": 147.13,
      "step": 52510,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.78200493,
      "epoch": 1.2253313696612667,
      "grad_norm": 4.40625,
      "learning_rate": 3.4444981650784147e-06,
      "loss": 0.78650246,
      "memory(GiB)": 147.13,
      "step": 52520,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.77312975,
      "epoch": 1.2255646772335556,
      "grad_norm": 5.5,
      "learning_rate": 3.4427029077750895e-06,
      "loss": 0.81262064,
      "memory(GiB)": 147.13,
      "step": 52530,
      "train_speed(iter/s)": 0.200627
    },
    {
      "acc": 0.78849087,
      "epoch": 1.2257979848058445,
      "grad_norm": 9.75,
      "learning_rate": 3.4409078727754707e-06,
      "loss": 0.75306044,
      "memory(GiB)": 147.13,
      "step": 52540,
      "train_speed(iter/s)": 0.200647
    },
    {
      "acc": 0.78517804,
      "epoch": 1.2260312923781331,
      "grad_norm": 4.9375,
      "learning_rate": 3.4391130603358013e-06,
      "loss": 0.7650878,
      "memory(GiB)": 147.13,
      "step": 52550,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.81452112,
      "epoch": 1.2262645999504223,
      "grad_norm": 4.4375,
      "learning_rate": 3.4373184707122886e-06,
      "loss": 0.67917662,
      "memory(GiB)": 147.13,
      "step": 52560,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.80121727,
      "epoch": 1.226497907522711,
      "grad_norm": 4.375,
      "learning_rate": 3.4355241041611096e-06,
      "loss": 0.69461622,
      "memory(GiB)": 147.13,
      "step": 52570,
      "train_speed(iter/s)": 0.200706
    },
    {
      "acc": 0.78457727,
      "epoch": 1.226731215095,
      "grad_norm": 5.125,
      "learning_rate": 3.4337299609384122e-06,
      "loss": 0.76386065,
      "memory(GiB)": 147.13,
      "step": 52580,
      "train_speed(iter/s)": 0.200725
    },
    {
      "acc": 0.7923645,
      "epoch": 1.2269645226672887,
      "grad_norm": 6.375,
      "learning_rate": 3.431936041300308e-06,
      "loss": 0.75402551,
      "memory(GiB)": 147.13,
      "step": 52590,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.78272791,
      "epoch": 1.2271978302395776,
      "grad_norm": 4.96875,
      "learning_rate": 3.4301423455028777e-06,
      "loss": 0.77303209,
      "memory(GiB)": 147.13,
      "step": 52600,
      "train_speed(iter/s)": 0.200763
    },
    {
      "acc": 0.77151518,
      "epoch": 1.2274311378118665,
      "grad_norm": 4.4375,
      "learning_rate": 3.4283488738021707e-06,
      "loss": 0.83244991,
      "memory(GiB)": 147.13,
      "step": 52610,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.79330654,
      "epoch": 1.2276644453841554,
      "grad_norm": 5.3125,
      "learning_rate": 3.4265556264542054e-06,
      "loss": 0.72327337,
      "memory(GiB)": 147.13,
      "step": 52620,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.7816957,
      "epoch": 1.2278977529564443,
      "grad_norm": 6.5,
      "learning_rate": 3.424762603714967e-06,
      "loss": 0.794205,
      "memory(GiB)": 147.13,
      "step": 52630,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.7910058,
      "epoch": 1.2281310605287332,
      "grad_norm": 4.5625,
      "learning_rate": 3.4229698058404106e-06,
      "loss": 0.7762475,
      "memory(GiB)": 147.13,
      "step": 52640,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.7753264,
      "epoch": 1.2283643681010221,
      "grad_norm": 8.4375,
      "learning_rate": 3.4211772330864552e-06,
      "loss": 0.81560392,
      "memory(GiB)": 147.13,
      "step": 52650,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.76706119,
      "epoch": 1.228597675673311,
      "grad_norm": 7.78125,
      "learning_rate": 3.4193848857089924e-06,
      "loss": 0.85505123,
      "memory(GiB)": 147.13,
      "step": 52660,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.78135409,
      "epoch": 1.2288309832456,
      "grad_norm": 5.03125,
      "learning_rate": 3.4175927639638767e-06,
      "loss": 0.76477842,
      "memory(GiB)": 147.13,
      "step": 52670,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.79110146,
      "epoch": 1.2290642908178888,
      "grad_norm": 5.3125,
      "learning_rate": 3.4158008681069343e-06,
      "loss": 0.76938314,
      "memory(GiB)": 147.13,
      "step": 52680,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.7797955,
      "epoch": 1.2292975983901777,
      "grad_norm": 6.0,
      "learning_rate": 3.4140091983939584e-06,
      "loss": 0.78467655,
      "memory(GiB)": 147.13,
      "step": 52690,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.78217092,
      "epoch": 1.2295309059624666,
      "grad_norm": 5.25,
      "learning_rate": 3.4122177550807077e-06,
      "loss": 0.79160671,
      "memory(GiB)": 147.13,
      "step": 52700,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.79885192,
      "epoch": 1.2297642135347555,
      "grad_norm": 5.25,
      "learning_rate": 3.410426538422914e-06,
      "loss": 0.71652822,
      "memory(GiB)": 147.13,
      "step": 52710,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.78403473,
      "epoch": 1.2299975211070444,
      "grad_norm": 6.59375,
      "learning_rate": 3.4086355486762678e-06,
      "loss": 0.78251448,
      "memory(GiB)": 147.13,
      "step": 52720,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.76983075,
      "epoch": 1.2302308286793333,
      "grad_norm": 7.3125,
      "learning_rate": 3.406844786096435e-06,
      "loss": 0.8267252,
      "memory(GiB)": 147.13,
      "step": 52730,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.77623463,
      "epoch": 1.2304641362516222,
      "grad_norm": 5.375,
      "learning_rate": 3.405054250939047e-06,
      "loss": 0.81395969,
      "memory(GiB)": 147.13,
      "step": 52740,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.76780348,
      "epoch": 1.230697443823911,
      "grad_norm": 7.53125,
      "learning_rate": 3.4032639434597003e-06,
      "loss": 0.83625088,
      "memory(GiB)": 147.13,
      "step": 52750,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.79240489,
      "epoch": 1.2309307513962,
      "grad_norm": 5.5,
      "learning_rate": 3.4014738639139622e-06,
      "loss": 0.75073404,
      "memory(GiB)": 147.13,
      "step": 52760,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.79068785,
      "epoch": 1.231164058968489,
      "grad_norm": 5.4375,
      "learning_rate": 3.399684012557365e-06,
      "loss": 0.75344276,
      "memory(GiB)": 147.13,
      "step": 52770,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.81060543,
      "epoch": 1.2313973665407778,
      "grad_norm": 4.625,
      "learning_rate": 3.3978943896454107e-06,
      "loss": 0.66089754,
      "memory(GiB)": 147.13,
      "step": 52780,
      "train_speed(iter/s)": 0.201117
    },
    {
      "acc": 0.79075351,
      "epoch": 1.2316306741130667,
      "grad_norm": 6.84375,
      "learning_rate": 3.396104995433567e-06,
      "loss": 0.73719893,
      "memory(GiB)": 147.13,
      "step": 52790,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.78999639,
      "epoch": 1.2318639816853556,
      "grad_norm": 6.25,
      "learning_rate": 3.3943158301772695e-06,
      "loss": 0.74546919,
      "memory(GiB)": 147.13,
      "step": 52800,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.79035387,
      "epoch": 1.2320972892576445,
      "grad_norm": 3.6875,
      "learning_rate": 3.39252689413192e-06,
      "loss": 0.75522537,
      "memory(GiB)": 147.13,
      "step": 52810,
      "train_speed(iter/s)": 0.201176
    },
    {
      "acc": 0.77618947,
      "epoch": 1.2323305968299334,
      "grad_norm": 4.9375,
      "learning_rate": 3.3907381875528916e-06,
      "loss": 0.81478825,
      "memory(GiB)": 147.13,
      "step": 52820,
      "train_speed(iter/s)": 0.201194
    },
    {
      "acc": 0.75931473,
      "epoch": 1.2325639044022223,
      "grad_norm": 5.125,
      "learning_rate": 3.388949710695517e-06,
      "loss": 0.86494942,
      "memory(GiB)": 147.13,
      "step": 52830,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.7774549,
      "epoch": 1.2327972119745112,
      "grad_norm": 8.5625,
      "learning_rate": 3.387161463815104e-06,
      "loss": 0.80290031,
      "memory(GiB)": 147.13,
      "step": 52840,
      "train_speed(iter/s)": 0.201234
    },
    {
      "acc": 0.7622488,
      "epoch": 1.2330305195468,
      "grad_norm": 5.0,
      "learning_rate": 3.3853734471669232e-06,
      "loss": 0.86460876,
      "memory(GiB)": 147.13,
      "step": 52850,
      "train_speed(iter/s)": 0.201254
    },
    {
      "acc": 0.78175044,
      "epoch": 1.233263827119089,
      "grad_norm": 7.09375,
      "learning_rate": 3.3835856610062135e-06,
      "loss": 0.76835246,
      "memory(GiB)": 147.13,
      "step": 52860,
      "train_speed(iter/s)": 0.201275
    },
    {
      "acc": 0.77455969,
      "epoch": 1.233497134691378,
      "grad_norm": 6.3125,
      "learning_rate": 3.381798105588181e-06,
      "loss": 0.79920435,
      "memory(GiB)": 147.13,
      "step": 52870,
      "train_speed(iter/s)": 0.201295
    },
    {
      "acc": 0.7796454,
      "epoch": 1.2337304422636668,
      "grad_norm": 6.21875,
      "learning_rate": 3.3800107811680004e-06,
      "loss": 0.78245277,
      "memory(GiB)": 147.13,
      "step": 52880,
      "train_speed(iter/s)": 0.201315
    },
    {
      "acc": 0.77907124,
      "epoch": 1.2339637498359557,
      "grad_norm": 5.46875,
      "learning_rate": 3.378223688000809e-06,
      "loss": 0.78705425,
      "memory(GiB)": 147.13,
      "step": 52890,
      "train_speed(iter/s)": 0.201334
    },
    {
      "acc": 0.7818182,
      "epoch": 1.2341970574082446,
      "grad_norm": 4.96875,
      "learning_rate": 3.3764368263417146e-06,
      "loss": 0.76016235,
      "memory(GiB)": 147.13,
      "step": 52900,
      "train_speed(iter/s)": 0.201355
    },
    {
      "acc": 0.78806572,
      "epoch": 1.2344303649805335,
      "grad_norm": 8.0,
      "learning_rate": 3.3746501964457916e-06,
      "loss": 0.75710969,
      "memory(GiB)": 147.13,
      "step": 52910,
      "train_speed(iter/s)": 0.201375
    },
    {
      "acc": 0.78649473,
      "epoch": 1.2346636725528224,
      "grad_norm": 5.28125,
      "learning_rate": 3.3728637985680814e-06,
      "loss": 0.7589705,
      "memory(GiB)": 147.13,
      "step": 52920,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.79017825,
      "epoch": 1.2348969801251113,
      "grad_norm": 6.28125,
      "learning_rate": 3.371077632963592e-06,
      "loss": 0.74173431,
      "memory(GiB)": 147.13,
      "step": 52930,
      "train_speed(iter/s)": 0.201413
    },
    {
      "acc": 0.77001181,
      "epoch": 1.2351302876974,
      "grad_norm": 9.25,
      "learning_rate": 3.3692916998872972e-06,
      "loss": 0.82475138,
      "memory(GiB)": 147.13,
      "step": 52940,
      "train_speed(iter/s)": 0.201433
    },
    {
      "acc": 0.78582697,
      "epoch": 1.235363595269689,
      "grad_norm": 5.40625,
      "learning_rate": 3.367505999594138e-06,
      "loss": 0.76669807,
      "memory(GiB)": 147.13,
      "step": 52950,
      "train_speed(iter/s)": 0.201453
    },
    {
      "acc": 0.7884789,
      "epoch": 1.2355969028419778,
      "grad_norm": 7.0,
      "learning_rate": 3.3657205323390234e-06,
      "loss": 0.76776428,
      "memory(GiB)": 147.13,
      "step": 52960,
      "train_speed(iter/s)": 0.201474
    },
    {
      "acc": 0.77210593,
      "epoch": 1.2358302104142667,
      "grad_norm": 4.9375,
      "learning_rate": 3.3639352983768276e-06,
      "loss": 0.8105154,
      "memory(GiB)": 147.13,
      "step": 52970,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.80149565,
      "epoch": 1.2360635179865556,
      "grad_norm": 9.625,
      "learning_rate": 3.3621502979623923e-06,
      "loss": 0.69630108,
      "memory(GiB)": 147.13,
      "step": 52980,
      "train_speed(iter/s)": 0.201515
    },
    {
      "acc": 0.77142386,
      "epoch": 1.2362968255588445,
      "grad_norm": 10.125,
      "learning_rate": 3.360365531350527e-06,
      "loss": 0.830303,
      "memory(GiB)": 147.13,
      "step": 52990,
      "train_speed(iter/s)": 0.201536
    },
    {
      "acc": 0.78944068,
      "epoch": 1.2365301331311334,
      "grad_norm": 5.59375,
      "learning_rate": 3.358580998796005e-06,
      "loss": 0.75569763,
      "memory(GiB)": 147.13,
      "step": 53000,
      "train_speed(iter/s)": 0.201555
    },
    {
      "epoch": 1.2365301331311334,
      "eval_acc": 0.7444263001029247,
      "eval_loss": 0.8053351044654846,
      "eval_runtime": 1270.5258,
      "eval_samples_per_second": 28.328,
      "eval_steps_per_second": 14.164,
      "step": 53000
    },
    {
      "acc": 0.7961484,
      "epoch": 1.2367634407034223,
      "grad_norm": 5.375,
      "learning_rate": 3.3567967005535696e-06,
      "loss": 0.73044991,
      "memory(GiB)": 147.13,
      "step": 53010,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.77993622,
      "epoch": 1.2369967482757112,
      "grad_norm": 4.75,
      "learning_rate": 3.355012636877927e-06,
      "loss": 0.77450337,
      "memory(GiB)": 147.13,
      "step": 53020,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.78480597,
      "epoch": 1.237230055848,
      "grad_norm": 5.0,
      "learning_rate": 3.353228808023752e-06,
      "loss": 0.77071924,
      "memory(GiB)": 147.13,
      "step": 53030,
      "train_speed(iter/s)": 0.200626
    },
    {
      "acc": 0.8039917,
      "epoch": 1.237463363420289,
      "grad_norm": 4.46875,
      "learning_rate": 3.351445214245687e-06,
      "loss": 0.68680391,
      "memory(GiB)": 147.13,
      "step": 53040,
      "train_speed(iter/s)": 0.200645
    },
    {
      "acc": 0.78955479,
      "epoch": 1.2376966709925779,
      "grad_norm": 5.40625,
      "learning_rate": 3.3496618557983405e-06,
      "loss": 0.73595805,
      "memory(GiB)": 147.13,
      "step": 53050,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.79907274,
      "epoch": 1.2379299785648668,
      "grad_norm": 4.46875,
      "learning_rate": 3.347878732936283e-06,
      "loss": 0.73365273,
      "memory(GiB)": 147.13,
      "step": 53060,
      "train_speed(iter/s)": 0.200684
    },
    {
      "acc": 0.77731886,
      "epoch": 1.2381632861371556,
      "grad_norm": 4.65625,
      "learning_rate": 3.346095845914056e-06,
      "loss": 0.81012163,
      "memory(GiB)": 147.13,
      "step": 53070,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.76856656,
      "epoch": 1.2383965937094445,
      "grad_norm": 5.40625,
      "learning_rate": 3.3443131949861667e-06,
      "loss": 0.84078465,
      "memory(GiB)": 147.13,
      "step": 53080,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.78171597,
      "epoch": 1.2386299012817334,
      "grad_norm": 12.3125,
      "learning_rate": 3.3425307804070896e-06,
      "loss": 0.79482059,
      "memory(GiB)": 147.13,
      "step": 53090,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.78333645,
      "epoch": 1.2388632088540223,
      "grad_norm": 6.78125,
      "learning_rate": 3.3407486024312596e-06,
      "loss": 0.78734274,
      "memory(GiB)": 147.13,
      "step": 53100,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.7784162,
      "epoch": 1.2390965164263112,
      "grad_norm": 6.375,
      "learning_rate": 3.3389666613130856e-06,
      "loss": 0.79651308,
      "memory(GiB)": 147.13,
      "step": 53110,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.81201239,
      "epoch": 1.2393298239986001,
      "grad_norm": 3.671875,
      "learning_rate": 3.337184957306938e-06,
      "loss": 0.6528636,
      "memory(GiB)": 147.13,
      "step": 53120,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.79003553,
      "epoch": 1.239563131570889,
      "grad_norm": 7.8125,
      "learning_rate": 3.3354034906671545e-06,
      "loss": 0.77086411,
      "memory(GiB)": 147.13,
      "step": 53130,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.78376465,
      "epoch": 1.239796439143178,
      "grad_norm": 7.25,
      "learning_rate": 3.333622261648039e-06,
      "loss": 0.76185646,
      "memory(GiB)": 147.13,
      "step": 53140,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.7993577,
      "epoch": 1.2400297467154668,
      "grad_norm": 5.9375,
      "learning_rate": 3.3318412705038626e-06,
      "loss": 0.70314708,
      "memory(GiB)": 147.13,
      "step": 53150,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.77541742,
      "epoch": 1.2402630542877557,
      "grad_norm": 6.5625,
      "learning_rate": 3.330060517488861e-06,
      "loss": 0.81968384,
      "memory(GiB)": 147.13,
      "step": 53160,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.78192739,
      "epoch": 1.2404963618600446,
      "grad_norm": 6.625,
      "learning_rate": 3.328280002857234e-06,
      "loss": 0.77632504,
      "memory(GiB)": 147.13,
      "step": 53170,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.79778252,
      "epoch": 1.2407296694323335,
      "grad_norm": 6.5625,
      "learning_rate": 3.3264997268631515e-06,
      "loss": 0.75389705,
      "memory(GiB)": 147.13,
      "step": 53180,
      "train_speed(iter/s)": 0.200921
    },
    {
      "acc": 0.77507801,
      "epoch": 1.2409629770046224,
      "grad_norm": 5.0,
      "learning_rate": 3.324719689760746e-06,
      "loss": 0.80741787,
      "memory(GiB)": 147.13,
      "step": 53190,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.78459139,
      "epoch": 1.2411962845769113,
      "grad_norm": 5.78125,
      "learning_rate": 3.3229398918041184e-06,
      "loss": 0.76652284,
      "memory(GiB)": 147.13,
      "step": 53200,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.7865819,
      "epoch": 1.2414295921492002,
      "grad_norm": 4.125,
      "learning_rate": 3.321160333247334e-06,
      "loss": 0.76798935,
      "memory(GiB)": 147.13,
      "step": 53210,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.77011223,
      "epoch": 1.2416628997214891,
      "grad_norm": 8.125,
      "learning_rate": 3.319381014344424e-06,
      "loss": 0.81456013,
      "memory(GiB)": 147.13,
      "step": 53220,
      "train_speed(iter/s)": 0.200995
    },
    {
      "acc": 0.78842793,
      "epoch": 1.241896207293778,
      "grad_norm": 5.65625,
      "learning_rate": 3.3176019353493873e-06,
      "loss": 0.74759521,
      "memory(GiB)": 147.13,
      "step": 53230,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.75929575,
      "epoch": 1.242129514866067,
      "grad_norm": 5.09375,
      "learning_rate": 3.315823096516184e-06,
      "loss": 0.8730217,
      "memory(GiB)": 147.13,
      "step": 53240,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.78181305,
      "epoch": 1.2423628224383558,
      "grad_norm": 7.375,
      "learning_rate": 3.314044498098745e-06,
      "loss": 0.77940979,
      "memory(GiB)": 147.13,
      "step": 53250,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.79055223,
      "epoch": 1.2425961300106447,
      "grad_norm": 5.09375,
      "learning_rate": 3.3122661403509643e-06,
      "loss": 0.73277559,
      "memory(GiB)": 147.13,
      "step": 53260,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.78648615,
      "epoch": 1.2428294375829336,
      "grad_norm": 7.40625,
      "learning_rate": 3.3104880235267014e-06,
      "loss": 0.7687428,
      "memory(GiB)": 147.13,
      "step": 53270,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.77741222,
      "epoch": 1.2430627451552225,
      "grad_norm": 7.5625,
      "learning_rate": 3.3087101478797846e-06,
      "loss": 0.80210009,
      "memory(GiB)": 147.13,
      "step": 53280,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.77117157,
      "epoch": 1.2432960527275114,
      "grad_norm": 4.46875,
      "learning_rate": 3.3069325136640007e-06,
      "loss": 0.81199608,
      "memory(GiB)": 147.13,
      "step": 53290,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.77286162,
      "epoch": 1.2435293602998003,
      "grad_norm": 6.21875,
      "learning_rate": 3.305155121133109e-06,
      "loss": 0.80843086,
      "memory(GiB)": 147.13,
      "step": 53300,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.78854752,
      "epoch": 1.243762667872089,
      "grad_norm": 5.5625,
      "learning_rate": 3.303377970540832e-06,
      "loss": 0.76747618,
      "memory(GiB)": 147.13,
      "step": 53310,
      "train_speed(iter/s)": 0.201164
    },
    {
      "acc": 0.77600183,
      "epoch": 1.2439959754443781,
      "grad_norm": 5.75,
      "learning_rate": 3.3016010621408558e-06,
      "loss": 0.7977457,
      "memory(GiB)": 147.13,
      "step": 53320,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.79537611,
      "epoch": 1.2442292830166668,
      "grad_norm": 9.9375,
      "learning_rate": 3.299824396186835e-06,
      "loss": 0.73316031,
      "memory(GiB)": 147.13,
      "step": 53330,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.78379049,
      "epoch": 1.244462590588956,
      "grad_norm": 4.90625,
      "learning_rate": 3.2980479729323867e-06,
      "loss": 0.77379413,
      "memory(GiB)": 147.13,
      "step": 53340,
      "train_speed(iter/s)": 0.201222
    },
    {
      "acc": 0.75966711,
      "epoch": 1.2446958981612446,
      "grad_norm": 4.625,
      "learning_rate": 3.2962717926310966e-06,
      "loss": 0.87534847,
      "memory(GiB)": 147.13,
      "step": 53350,
      "train_speed(iter/s)": 0.20124
    },
    {
      "acc": 0.79759836,
      "epoch": 1.2449292057335335,
      "grad_norm": 3.828125,
      "learning_rate": 3.2944958555365135e-06,
      "loss": 0.71614137,
      "memory(GiB)": 147.13,
      "step": 53360,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.80230551,
      "epoch": 1.2451625133058224,
      "grad_norm": 5.625,
      "learning_rate": 3.292720161902152e-06,
      "loss": 0.69462061,
      "memory(GiB)": 147.13,
      "step": 53370,
      "train_speed(iter/s)": 0.201278
    },
    {
      "acc": 0.79200816,
      "epoch": 1.2453958208781113,
      "grad_norm": 6.5625,
      "learning_rate": 3.2909447119814907e-06,
      "loss": 0.74009314,
      "memory(GiB)": 147.13,
      "step": 53380,
      "train_speed(iter/s)": 0.201299
    },
    {
      "acc": 0.78173537,
      "epoch": 1.2456291284504002,
      "grad_norm": 4.8125,
      "learning_rate": 3.289169506027977e-06,
      "loss": 0.769735,
      "memory(GiB)": 147.13,
      "step": 53390,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.78138733,
      "epoch": 1.245862436022689,
      "grad_norm": 4.84375,
      "learning_rate": 3.287394544295018e-06,
      "loss": 0.79002943,
      "memory(GiB)": 147.13,
      "step": 53400,
      "train_speed(iter/s)": 0.201338
    },
    {
      "acc": 0.77280626,
      "epoch": 1.246095743594978,
      "grad_norm": 6.21875,
      "learning_rate": 3.2856198270359895e-06,
      "loss": 0.82154875,
      "memory(GiB)": 147.13,
      "step": 53410,
      "train_speed(iter/s)": 0.201357
    },
    {
      "acc": 0.79443498,
      "epoch": 1.2463290511672669,
      "grad_norm": 4.90625,
      "learning_rate": 3.2838453545042326e-06,
      "loss": 0.72624025,
      "memory(GiB)": 147.13,
      "step": 53420,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.78037777,
      "epoch": 1.2465623587395558,
      "grad_norm": 4.71875,
      "learning_rate": 3.2820711269530535e-06,
      "loss": 0.7707921,
      "memory(GiB)": 147.13,
      "step": 53430,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.78268867,
      "epoch": 1.2467956663118447,
      "grad_norm": 5.90625,
      "learning_rate": 3.280297144635721e-06,
      "loss": 0.76652794,
      "memory(GiB)": 147.13,
      "step": 53440,
      "train_speed(iter/s)": 0.201415
    },
    {
      "acc": 0.76735125,
      "epoch": 1.2470289738841336,
      "grad_norm": 4.125,
      "learning_rate": 3.278523407805474e-06,
      "loss": 0.83181791,
      "memory(GiB)": 147.13,
      "step": 53450,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.77343864,
      "epoch": 1.2472622814564225,
      "grad_norm": 6.125,
      "learning_rate": 3.276749916715508e-06,
      "loss": 0.81363611,
      "memory(GiB)": 147.13,
      "step": 53460,
      "train_speed(iter/s)": 0.201455
    },
    {
      "acc": 0.7868227,
      "epoch": 1.2474955890287114,
      "grad_norm": 5.28125,
      "learning_rate": 3.274976671618992e-06,
      "loss": 0.75913687,
      "memory(GiB)": 147.13,
      "step": 53470,
      "train_speed(iter/s)": 0.201475
    },
    {
      "acc": 0.7895021,
      "epoch": 1.2477288966010003,
      "grad_norm": 4.40625,
      "learning_rate": 3.2732036727690543e-06,
      "loss": 0.75617399,
      "memory(GiB)": 147.13,
      "step": 53480,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.78445072,
      "epoch": 1.2479622041732892,
      "grad_norm": 5.25,
      "learning_rate": 3.2714309204187905e-06,
      "loss": 0.77543073,
      "memory(GiB)": 147.13,
      "step": 53490,
      "train_speed(iter/s)": 0.201513
    },
    {
      "acc": 0.76975474,
      "epoch": 1.248195511745578,
      "grad_norm": 5.25,
      "learning_rate": 3.2696584148212606e-06,
      "loss": 0.8373764,
      "memory(GiB)": 147.13,
      "step": 53500,
      "train_speed(iter/s)": 0.201533
    },
    {
      "epoch": 1.248195511745578,
      "eval_acc": 0.744429663131803,
      "eval_loss": 0.8053871393203735,
      "eval_runtime": 1269.4157,
      "eval_samples_per_second": 28.352,
      "eval_steps_per_second": 14.177,
      "step": 53500
    },
    {
      "acc": 0.7716815,
      "epoch": 1.248428819317867,
      "grad_norm": 4.34375,
      "learning_rate": 3.2678861562294916e-06,
      "loss": 0.81136675,
      "memory(GiB)": 147.13,
      "step": 53510,
      "train_speed(iter/s)": 0.20058
    },
    {
      "acc": 0.77584715,
      "epoch": 1.2486621268901559,
      "grad_norm": 5.5,
      "learning_rate": 3.2661141448964688e-06,
      "loss": 0.78764925,
      "memory(GiB)": 147.13,
      "step": 53520,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.78310165,
      "epoch": 1.2488954344624448,
      "grad_norm": 4.25,
      "learning_rate": 3.2643423810751497e-06,
      "loss": 0.76774969,
      "memory(GiB)": 147.13,
      "step": 53530,
      "train_speed(iter/s)": 0.200618
    },
    {
      "acc": 0.76813574,
      "epoch": 1.2491287420347337,
      "grad_norm": 5.78125,
      "learning_rate": 3.2625708650184496e-06,
      "loss": 0.83579845,
      "memory(GiB)": 147.13,
      "step": 53540,
      "train_speed(iter/s)": 0.200637
    },
    {
      "acc": 0.77997684,
      "epoch": 1.2493620496070226,
      "grad_norm": 6.34375,
      "learning_rate": 3.260799596979254e-06,
      "loss": 0.79803414,
      "memory(GiB)": 147.13,
      "step": 53550,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.78893514,
      "epoch": 1.2495953571793115,
      "grad_norm": 4.4375,
      "learning_rate": 3.25902857721041e-06,
      "loss": 0.76923342,
      "memory(GiB)": 147.13,
      "step": 53560,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.76379528,
      "epoch": 1.2498286647516004,
      "grad_norm": 4.34375,
      "learning_rate": 3.257257805964732e-06,
      "loss": 0.85577602,
      "memory(GiB)": 147.13,
      "step": 53570,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.7713376,
      "epoch": 1.2500619723238893,
      "grad_norm": 5.34375,
      "learning_rate": 3.255487283494995e-06,
      "loss": 0.83352032,
      "memory(GiB)": 147.13,
      "step": 53580,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.78212161,
      "epoch": 1.2502952798961782,
      "grad_norm": 6.6875,
      "learning_rate": 3.253717010053943e-06,
      "loss": 0.7555759,
      "memory(GiB)": 147.13,
      "step": 53590,
      "train_speed(iter/s)": 0.200734
    },
    {
      "acc": 0.7807415,
      "epoch": 1.250528587468467,
      "grad_norm": 7.53125,
      "learning_rate": 3.25194698589428e-06,
      "loss": 0.7950223,
      "memory(GiB)": 147.13,
      "step": 53600,
      "train_speed(iter/s)": 0.200754
    },
    {
      "acc": 0.76310434,
      "epoch": 1.250761895040756,
      "grad_norm": 4.8125,
      "learning_rate": 3.2501772112686757e-06,
      "loss": 0.87415972,
      "memory(GiB)": 147.13,
      "step": 53610,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.7808527,
      "epoch": 1.2509952026130449,
      "grad_norm": 5.59375,
      "learning_rate": 3.2484076864297687e-06,
      "loss": 0.78480973,
      "memory(GiB)": 147.13,
      "step": 53620,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.77754536,
      "epoch": 1.2512285101853338,
      "grad_norm": 5.0,
      "learning_rate": 3.246638411630154e-06,
      "loss": 0.80107412,
      "memory(GiB)": 147.13,
      "step": 53630,
      "train_speed(iter/s)": 0.200812
    },
    {
      "acc": 0.76862297,
      "epoch": 1.2514618177576227,
      "grad_norm": 4.3125,
      "learning_rate": 3.2448693871223968e-06,
      "loss": 0.85014858,
      "memory(GiB)": 147.13,
      "step": 53640,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.78168688,
      "epoch": 1.2516951253299116,
      "grad_norm": 7.71875,
      "learning_rate": 3.2431006131590244e-06,
      "loss": 0.77133665,
      "memory(GiB)": 147.13,
      "step": 53650,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.77619853,
      "epoch": 1.2519284329022005,
      "grad_norm": 4.78125,
      "learning_rate": 3.2413320899925287e-06,
      "loss": 0.782967,
      "memory(GiB)": 147.13,
      "step": 53660,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.78052235,
      "epoch": 1.2521617404744894,
      "grad_norm": 6.5,
      "learning_rate": 3.2395638178753673e-06,
      "loss": 0.77000246,
      "memory(GiB)": 147.13,
      "step": 53670,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.76536398,
      "epoch": 1.252395048046778,
      "grad_norm": 5.96875,
      "learning_rate": 3.2377957970599594e-06,
      "loss": 0.83446522,
      "memory(GiB)": 147.13,
      "step": 53680,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.79172969,
      "epoch": 1.2526283556190672,
      "grad_norm": 6.09375,
      "learning_rate": 3.2360280277986887e-06,
      "loss": 0.73861761,
      "memory(GiB)": 147.13,
      "step": 53690,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.76554937,
      "epoch": 1.2528616631913558,
      "grad_norm": 6.1875,
      "learning_rate": 3.234260510343905e-06,
      "loss": 0.84061546,
      "memory(GiB)": 147.13,
      "step": 53700,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.77454815,
      "epoch": 1.253094970763645,
      "grad_norm": 4.78125,
      "learning_rate": 3.23249324494792e-06,
      "loss": 0.82709484,
      "memory(GiB)": 147.13,
      "step": 53710,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.7816751,
      "epoch": 1.2533282783359336,
      "grad_norm": 5.3125,
      "learning_rate": 3.230726231863013e-06,
      "loss": 0.78138809,
      "memory(GiB)": 147.13,
      "step": 53720,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.78555498,
      "epoch": 1.2535615859082228,
      "grad_norm": 6.78125,
      "learning_rate": 3.2289594713414207e-06,
      "loss": 0.78290424,
      "memory(GiB)": 147.13,
      "step": 53730,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.79429498,
      "epoch": 1.2537948934805114,
      "grad_norm": 5.59375,
      "learning_rate": 3.2271929636353494e-06,
      "loss": 0.72328415,
      "memory(GiB)": 147.13,
      "step": 53740,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.78629532,
      "epoch": 1.2540282010528006,
      "grad_norm": 4.15625,
      "learning_rate": 3.2254267089969688e-06,
      "loss": 0.76473923,
      "memory(GiB)": 147.13,
      "step": 53750,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.78774004,
      "epoch": 1.2542615086250892,
      "grad_norm": 7.0,
      "learning_rate": 3.2236607076784086e-06,
      "loss": 0.76563559,
      "memory(GiB)": 147.13,
      "step": 53760,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.78109026,
      "epoch": 1.2544948161973781,
      "grad_norm": 6.8125,
      "learning_rate": 3.2218949599317664e-06,
      "loss": 0.79619112,
      "memory(GiB)": 147.13,
      "step": 53770,
      "train_speed(iter/s)": 0.201082
    },
    {
      "acc": 0.77077236,
      "epoch": 1.254728123769667,
      "grad_norm": 5.03125,
      "learning_rate": 3.220129466009102e-06,
      "loss": 0.83952332,
      "memory(GiB)": 147.13,
      "step": 53780,
      "train_speed(iter/s)": 0.201101
    },
    {
      "acc": 0.7884572,
      "epoch": 1.254961431341956,
      "grad_norm": 5.625,
      "learning_rate": 3.2183642261624393e-06,
      "loss": 0.75626945,
      "memory(GiB)": 147.13,
      "step": 53790,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.75669336,
      "epoch": 1.2551947389142448,
      "grad_norm": 5.0,
      "learning_rate": 3.216599240643765e-06,
      "loss": 0.88252993,
      "memory(GiB)": 147.13,
      "step": 53800,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.77265863,
      "epoch": 1.2554280464865337,
      "grad_norm": 9.5,
      "learning_rate": 3.2148345097050332e-06,
      "loss": 0.83709316,
      "memory(GiB)": 147.13,
      "step": 53810,
      "train_speed(iter/s)": 0.201159
    },
    {
      "acc": 0.76388264,
      "epoch": 1.2556613540588226,
      "grad_norm": 15.5625,
      "learning_rate": 3.213070033598155e-06,
      "loss": 0.85438976,
      "memory(GiB)": 147.13,
      "step": 53820,
      "train_speed(iter/s)": 0.20118
    },
    {
      "acc": 0.80244446,
      "epoch": 1.2558946616311115,
      "grad_norm": 5.0625,
      "learning_rate": 3.211305812575011e-06,
      "loss": 0.70825286,
      "memory(GiB)": 147.13,
      "step": 53830,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.79701781,
      "epoch": 1.2561279692034004,
      "grad_norm": 5.4375,
      "learning_rate": 3.209541846887442e-06,
      "loss": 0.73176861,
      "memory(GiB)": 147.13,
      "step": 53840,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.79489713,
      "epoch": 1.2563612767756893,
      "grad_norm": 4.0,
      "learning_rate": 3.207778136787256e-06,
      "loss": 0.70806804,
      "memory(GiB)": 147.13,
      "step": 53850,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.78195076,
      "epoch": 1.2565945843479782,
      "grad_norm": 4.5625,
      "learning_rate": 3.2060146825262196e-06,
      "loss": 0.79273672,
      "memory(GiB)": 147.13,
      "step": 53860,
      "train_speed(iter/s)": 0.201258
    },
    {
      "acc": 0.79850488,
      "epoch": 1.2568278919202671,
      "grad_norm": 4.46875,
      "learning_rate": 3.2042514843560644e-06,
      "loss": 0.71020679,
      "memory(GiB)": 147.13,
      "step": 53870,
      "train_speed(iter/s)": 0.201279
    },
    {
      "acc": 0.77073278,
      "epoch": 1.257061199492556,
      "grad_norm": 5.84375,
      "learning_rate": 3.2024885425284893e-06,
      "loss": 0.81626215,
      "memory(GiB)": 147.13,
      "step": 53880,
      "train_speed(iter/s)": 0.201299
    },
    {
      "acc": 0.7916851,
      "epoch": 1.257294507064845,
      "grad_norm": 4.28125,
      "learning_rate": 3.200725857295153e-06,
      "loss": 0.73321352,
      "memory(GiB)": 147.13,
      "step": 53890,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.77081394,
      "epoch": 1.2575278146371338,
      "grad_norm": 4.78125,
      "learning_rate": 3.1989634289076776e-06,
      "loss": 0.81450729,
      "memory(GiB)": 147.13,
      "step": 53900,
      "train_speed(iter/s)": 0.201336
    },
    {
      "acc": 0.77993765,
      "epoch": 1.2577611222094227,
      "grad_norm": 4.625,
      "learning_rate": 3.197201257617649e-06,
      "loss": 0.78307796,
      "memory(GiB)": 147.13,
      "step": 53910,
      "train_speed(iter/s)": 0.201356
    },
    {
      "acc": 0.78636661,
      "epoch": 1.2579944297817116,
      "grad_norm": 5.4375,
      "learning_rate": 3.195439343676617e-06,
      "loss": 0.77123709,
      "memory(GiB)": 147.13,
      "step": 53920,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.78531208,
      "epoch": 1.2582277373540005,
      "grad_norm": 4.5,
      "learning_rate": 3.1936776873360947e-06,
      "loss": 0.78743391,
      "memory(GiB)": 147.13,
      "step": 53930,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.78387508,
      "epoch": 1.2584610449262894,
      "grad_norm": 9.875,
      "learning_rate": 3.1919162888475586e-06,
      "loss": 0.77252707,
      "memory(GiB)": 147.13,
      "step": 53940,
      "train_speed(iter/s)": 0.201416
    },
    {
      "acc": 0.79411707,
      "epoch": 1.2586943524985783,
      "grad_norm": 5.5,
      "learning_rate": 3.190155148462446e-06,
      "loss": 0.72856312,
      "memory(GiB)": 147.13,
      "step": 53950,
      "train_speed(iter/s)": 0.201436
    },
    {
      "acc": 0.76296749,
      "epoch": 1.2589276600708672,
      "grad_norm": 4.96875,
      "learning_rate": 3.188394266432162e-06,
      "loss": 0.8607048,
      "memory(GiB)": 147.13,
      "step": 53960,
      "train_speed(iter/s)": 0.201455
    },
    {
      "acc": 0.77531281,
      "epoch": 1.259160967643156,
      "grad_norm": 5.625,
      "learning_rate": 3.186633643008069e-06,
      "loss": 0.80578537,
      "memory(GiB)": 147.13,
      "step": 53970,
      "train_speed(iter/s)": 0.201475
    },
    {
      "acc": 0.78390398,
      "epoch": 1.259394275215445,
      "grad_norm": 4.46875,
      "learning_rate": 3.1848732784414965e-06,
      "loss": 0.75349741,
      "memory(GiB)": 147.13,
      "step": 53980,
      "train_speed(iter/s)": 0.201495
    },
    {
      "acc": 0.7925087,
      "epoch": 1.259627582787734,
      "grad_norm": 5.15625,
      "learning_rate": 3.183113172983736e-06,
      "loss": 0.75053062,
      "memory(GiB)": 147.13,
      "step": 53990,
      "train_speed(iter/s)": 0.201513
    },
    {
      "acc": 0.79274836,
      "epoch": 1.2598608903600228,
      "grad_norm": 6.0625,
      "learning_rate": 3.181353326886042e-06,
      "loss": 0.74251266,
      "memory(GiB)": 147.13,
      "step": 54000,
      "train_speed(iter/s)": 0.201533
    },
    {
      "epoch": 1.2598608903600228,
      "eval_acc": 0.7445457677002215,
      "eval_loss": 0.8053033947944641,
      "eval_runtime": 1272.3358,
      "eval_samples_per_second": 28.287,
      "eval_steps_per_second": 14.144,
      "step": 54000
    },
    {
      "acc": 0.78218327,
      "epoch": 1.2600941979323117,
      "grad_norm": 6.46875,
      "learning_rate": 3.1795937403996324e-06,
      "loss": 0.76867433,
      "memory(GiB)": 147.13,
      "step": 54010,
      "train_speed(iter/s)": 0.200585
    },
    {
      "acc": 0.78693171,
      "epoch": 1.2603275055046006,
      "grad_norm": 5.40625,
      "learning_rate": 3.1778344137756887e-06,
      "loss": 0.75604353,
      "memory(GiB)": 147.13,
      "step": 54020,
      "train_speed(iter/s)": 0.200605
    },
    {
      "acc": 0.79023509,
      "epoch": 1.2605608130768895,
      "grad_norm": 4.84375,
      "learning_rate": 3.176075347265352e-06,
      "loss": 0.7607203,
      "memory(GiB)": 147.13,
      "step": 54030,
      "train_speed(iter/s)": 0.200623
    },
    {
      "acc": 0.77073202,
      "epoch": 1.2607941206491784,
      "grad_norm": 4.75,
      "learning_rate": 3.17431654111973e-06,
      "loss": 0.82227421,
      "memory(GiB)": 147.13,
      "step": 54040,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.78074164,
      "epoch": 1.2610274282214673,
      "grad_norm": 6.25,
      "learning_rate": 3.1725579955898904e-06,
      "loss": 0.76902056,
      "memory(GiB)": 147.13,
      "step": 54050,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.77640028,
      "epoch": 1.2612607357937562,
      "grad_norm": 6.5625,
      "learning_rate": 3.170799710926867e-06,
      "loss": 0.80473099,
      "memory(GiB)": 147.13,
      "step": 54060,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.76053233,
      "epoch": 1.2614940433660449,
      "grad_norm": 5.53125,
      "learning_rate": 3.1690416873816533e-06,
      "loss": 0.87321119,
      "memory(GiB)": 147.13,
      "step": 54070,
      "train_speed(iter/s)": 0.200699
    },
    {
      "acc": 0.78000383,
      "epoch": 1.261727350938334,
      "grad_norm": 4.875,
      "learning_rate": 3.1672839252052083e-06,
      "loss": 0.77255969,
      "memory(GiB)": 147.13,
      "step": 54080,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.79675488,
      "epoch": 1.2619606585106227,
      "grad_norm": 5.125,
      "learning_rate": 3.165526424648449e-06,
      "loss": 0.71965032,
      "memory(GiB)": 147.13,
      "step": 54090,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.7872025,
      "epoch": 1.2621939660829118,
      "grad_norm": 5.4375,
      "learning_rate": 3.1637691859622612e-06,
      "loss": 0.78567123,
      "memory(GiB)": 147.13,
      "step": 54100,
      "train_speed(iter/s)": 0.200757
    },
    {
      "acc": 0.7730463,
      "epoch": 1.2624272736552005,
      "grad_norm": 5.46875,
      "learning_rate": 3.1620122093974864e-06,
      "loss": 0.83145771,
      "memory(GiB)": 147.13,
      "step": 54110,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.78709173,
      "epoch": 1.2626605812274896,
      "grad_norm": 4.8125,
      "learning_rate": 3.160255495204936e-06,
      "loss": 0.76123199,
      "memory(GiB)": 147.13,
      "step": 54120,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.77258062,
      "epoch": 1.2628938887997783,
      "grad_norm": 7.15625,
      "learning_rate": 3.158499043635378e-06,
      "loss": 0.81969166,
      "memory(GiB)": 147.13,
      "step": 54130,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.78251991,
      "epoch": 1.2631271963720674,
      "grad_norm": 7.46875,
      "learning_rate": 3.156742854939547e-06,
      "loss": 0.79453034,
      "memory(GiB)": 147.13,
      "step": 54140,
      "train_speed(iter/s)": 0.200835
    },
    {
      "acc": 0.77685394,
      "epoch": 1.263360503944356,
      "grad_norm": 7.5625,
      "learning_rate": 3.1549869293681385e-06,
      "loss": 0.81942472,
      "memory(GiB)": 147.13,
      "step": 54150,
      "train_speed(iter/s)": 0.200855
    },
    {
      "acc": 0.77604828,
      "epoch": 1.263593811516645,
      "grad_norm": 5.0,
      "learning_rate": 3.1532312671718102e-06,
      "loss": 0.79661341,
      "memory(GiB)": 147.13,
      "step": 54160,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.77102032,
      "epoch": 1.2638271190889339,
      "grad_norm": 4.90625,
      "learning_rate": 3.1514758686011816e-06,
      "loss": 0.84676743,
      "memory(GiB)": 147.13,
      "step": 54170,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.77398539,
      "epoch": 1.2640604266612228,
      "grad_norm": 6.25,
      "learning_rate": 3.149720733906836e-06,
      "loss": 0.82969208,
      "memory(GiB)": 147.13,
      "step": 54180,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.77549362,
      "epoch": 1.2642937342335117,
      "grad_norm": 5.96875,
      "learning_rate": 3.1479658633393194e-06,
      "loss": 0.82915554,
      "memory(GiB)": 147.13,
      "step": 54190,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.78671007,
      "epoch": 1.2645270418058006,
      "grad_norm": 5.71875,
      "learning_rate": 3.146211257149136e-06,
      "loss": 0.77449756,
      "memory(GiB)": 147.13,
      "step": 54200,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.79366622,
      "epoch": 1.2647603493780895,
      "grad_norm": 7.75,
      "learning_rate": 3.1444569155867573e-06,
      "loss": 0.7421545,
      "memory(GiB)": 147.13,
      "step": 54210,
      "train_speed(iter/s)": 0.200969
    },
    {
      "acc": 0.77127199,
      "epoch": 1.2649936569503784,
      "grad_norm": 5.28125,
      "learning_rate": 3.1427028389026147e-06,
      "loss": 0.83733711,
      "memory(GiB)": 147.13,
      "step": 54220,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.77401648,
      "epoch": 1.2652269645226673,
      "grad_norm": 5.03125,
      "learning_rate": 3.140949027347102e-06,
      "loss": 0.82600069,
      "memory(GiB)": 147.13,
      "step": 54230,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78014426,
      "epoch": 1.2654602720949562,
      "grad_norm": 7.84375,
      "learning_rate": 3.139195481170577e-06,
      "loss": 0.78453398,
      "memory(GiB)": 147.13,
      "step": 54240,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.79164677,
      "epoch": 1.265693579667245,
      "grad_norm": 5.6875,
      "learning_rate": 3.1374422006233553e-06,
      "loss": 0.72470675,
      "memory(GiB)": 147.13,
      "step": 54250,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.78896961,
      "epoch": 1.265926887239534,
      "grad_norm": 5.625,
      "learning_rate": 3.1356891859557187e-06,
      "loss": 0.76291442,
      "memory(GiB)": 147.13,
      "step": 54260,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.78453579,
      "epoch": 1.2661601948118228,
      "grad_norm": 5.46875,
      "learning_rate": 3.1339364374179092e-06,
      "loss": 0.76488409,
      "memory(GiB)": 147.13,
      "step": 54270,
      "train_speed(iter/s)": 0.201082
    },
    {
      "acc": 0.76442003,
      "epoch": 1.2663935023841117,
      "grad_norm": 7.03125,
      "learning_rate": 3.1321839552601308e-06,
      "loss": 0.83735924,
      "memory(GiB)": 147.13,
      "step": 54280,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.7594471,
      "epoch": 1.2666268099564006,
      "grad_norm": 5.5,
      "learning_rate": 3.1304317397325503e-06,
      "loss": 0.87646704,
      "memory(GiB)": 147.13,
      "step": 54290,
      "train_speed(iter/s)": 0.201121
    },
    {
      "acc": 0.78775072,
      "epoch": 1.2668601175286895,
      "grad_norm": 6.09375,
      "learning_rate": 3.128679791085297e-06,
      "loss": 0.74031334,
      "memory(GiB)": 147.13,
      "step": 54300,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.78536415,
      "epoch": 1.2670934251009784,
      "grad_norm": 4.5625,
      "learning_rate": 3.1269281095684594e-06,
      "loss": 0.78184915,
      "memory(GiB)": 147.13,
      "step": 54310,
      "train_speed(iter/s)": 0.20116
    },
    {
      "acc": 0.76834879,
      "epoch": 1.2673267326732673,
      "grad_norm": 5.09375,
      "learning_rate": 3.1251766954320906e-06,
      "loss": 0.83152952,
      "memory(GiB)": 147.13,
      "step": 54320,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.78719902,
      "epoch": 1.2675600402455562,
      "grad_norm": 8.5625,
      "learning_rate": 3.123425548926203e-06,
      "loss": 0.75794401,
      "memory(GiB)": 147.13,
      "step": 54330,
      "train_speed(iter/s)": 0.201198
    },
    {
      "acc": 0.77632055,
      "epoch": 1.2677933478178451,
      "grad_norm": 7.1875,
      "learning_rate": 3.121674670300773e-06,
      "loss": 0.81995049,
      "memory(GiB)": 147.13,
      "step": 54340,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.79015565,
      "epoch": 1.268026655390134,
      "grad_norm": 3.25,
      "learning_rate": 3.1199240598057377e-06,
      "loss": 0.7804615,
      "memory(GiB)": 147.13,
      "step": 54350,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.76458197,
      "epoch": 1.268259962962423,
      "grad_norm": 5.46875,
      "learning_rate": 3.1181737176909967e-06,
      "loss": 0.85183592,
      "memory(GiB)": 147.13,
      "step": 54360,
      "train_speed(iter/s)": 0.201259
    },
    {
      "acc": 0.76942658,
      "epoch": 1.2684932705347118,
      "grad_norm": 4.96875,
      "learning_rate": 3.116423644206411e-06,
      "loss": 0.83050613,
      "memory(GiB)": 147.13,
      "step": 54370,
      "train_speed(iter/s)": 0.201278
    },
    {
      "acc": 0.76355877,
      "epoch": 1.2687265781070007,
      "grad_norm": 7.28125,
      "learning_rate": 3.1146738396018043e-06,
      "loss": 0.86715822,
      "memory(GiB)": 147.13,
      "step": 54380,
      "train_speed(iter/s)": 0.201298
    },
    {
      "acc": 0.77561574,
      "epoch": 1.2689598856792896,
      "grad_norm": 5.5,
      "learning_rate": 3.112924304126958e-06,
      "loss": 0.81895523,
      "memory(GiB)": 147.13,
      "step": 54390,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.79888744,
      "epoch": 1.2691931932515785,
      "grad_norm": 4.125,
      "learning_rate": 3.111175038031619e-06,
      "loss": 0.70149899,
      "memory(GiB)": 147.13,
      "step": 54400,
      "train_speed(iter/s)": 0.201337
    },
    {
      "acc": 0.78107786,
      "epoch": 1.2694265008238674,
      "grad_norm": 4.96875,
      "learning_rate": 3.1094260415654955e-06,
      "loss": 0.79072466,
      "memory(GiB)": 147.13,
      "step": 54410,
      "train_speed(iter/s)": 0.201357
    },
    {
      "acc": 0.77758923,
      "epoch": 1.2696598083961563,
      "grad_norm": 5.75,
      "learning_rate": 3.1076773149782557e-06,
      "loss": 0.80385323,
      "memory(GiB)": 147.13,
      "step": 54420,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.79401674,
      "epoch": 1.2698931159684452,
      "grad_norm": 6.0,
      "learning_rate": 3.105928858519529e-06,
      "loss": 0.72925024,
      "memory(GiB)": 147.13,
      "step": 54430,
      "train_speed(iter/s)": 0.201397
    },
    {
      "acc": 0.79929819,
      "epoch": 1.270126423540734,
      "grad_norm": 4.5,
      "learning_rate": 3.1041806724389067e-06,
      "loss": 0.70583544,
      "memory(GiB)": 147.13,
      "step": 54440,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.78400726,
      "epoch": 1.270359731113023,
      "grad_norm": 6.46875,
      "learning_rate": 3.1024327569859425e-06,
      "loss": 0.79743586,
      "memory(GiB)": 147.13,
      "step": 54450,
      "train_speed(iter/s)": 0.201434
    },
    {
      "acc": 0.75207291,
      "epoch": 1.2705930386853117,
      "grad_norm": 5.96875,
      "learning_rate": 3.1006851124101524e-06,
      "loss": 0.90693626,
      "memory(GiB)": 147.13,
      "step": 54460,
      "train_speed(iter/s)": 0.201453
    },
    {
      "acc": 0.77595568,
      "epoch": 1.2708263462576008,
      "grad_norm": 18.375,
      "learning_rate": 3.0989377389610097e-06,
      "loss": 0.82628527,
      "memory(GiB)": 147.13,
      "step": 54470,
      "train_speed(iter/s)": 0.201473
    },
    {
      "acc": 0.77218075,
      "epoch": 1.2710596538298895,
      "grad_norm": 4.9375,
      "learning_rate": 3.0971906368879524e-06,
      "loss": 0.8140975,
      "memory(GiB)": 147.13,
      "step": 54480,
      "train_speed(iter/s)": 0.201491
    },
    {
      "acc": 0.80995159,
      "epoch": 1.2712929614021786,
      "grad_norm": 4.6875,
      "learning_rate": 3.095443806440379e-06,
      "loss": 0.68186216,
      "memory(GiB)": 147.13,
      "step": 54490,
      "train_speed(iter/s)": 0.20151
    },
    {
      "acc": 0.78228736,
      "epoch": 1.2715262689744673,
      "grad_norm": 5.65625,
      "learning_rate": 3.0936972478676493e-06,
      "loss": 0.77988653,
      "memory(GiB)": 147.13,
      "step": 54500,
      "train_speed(iter/s)": 0.201529
    },
    {
      "epoch": 1.2715262689744673,
      "eval_acc": 0.7445822805851863,
      "eval_loss": 0.80512934923172,
      "eval_runtime": 1270.2737,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 54500
    },
    {
      "acc": 0.78444986,
      "epoch": 1.2717595765467564,
      "grad_norm": 4.90625,
      "learning_rate": 3.0919509614190836e-06,
      "loss": 0.77315531,
      "memory(GiB)": 147.13,
      "step": 54510,
      "train_speed(iter/s)": 0.200592
    },
    {
      "acc": 0.77153635,
      "epoch": 1.271992884119045,
      "grad_norm": 4.625,
      "learning_rate": 3.0902049473439643e-06,
      "loss": 0.80885544,
      "memory(GiB)": 147.13,
      "step": 54520,
      "train_speed(iter/s)": 0.200612
    },
    {
      "acc": 0.7847683,
      "epoch": 1.2722261916913342,
      "grad_norm": 9.9375,
      "learning_rate": 3.0884592058915342e-06,
      "loss": 0.79745474,
      "memory(GiB)": 147.13,
      "step": 54530,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.76501255,
      "epoch": 1.272459499263623,
      "grad_norm": 5.28125,
      "learning_rate": 3.0867137373109972e-06,
      "loss": 0.86392994,
      "memory(GiB)": 147.13,
      "step": 54540,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78538918,
      "epoch": 1.2726928068359118,
      "grad_norm": 9.4375,
      "learning_rate": 3.0849685418515174e-06,
      "loss": 0.77651844,
      "memory(GiB)": 147.13,
      "step": 54550,
      "train_speed(iter/s)": 0.20067
    },
    {
      "acc": 0.76775522,
      "epoch": 1.2729261144082007,
      "grad_norm": 5.53125,
      "learning_rate": 3.0832236197622223e-06,
      "loss": 0.8431736,
      "memory(GiB)": 147.13,
      "step": 54560,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.79953661,
      "epoch": 1.2731594219804896,
      "grad_norm": 6.0625,
      "learning_rate": 3.0814789712921977e-06,
      "loss": 0.70614519,
      "memory(GiB)": 147.13,
      "step": 54570,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.78566427,
      "epoch": 1.2733927295527785,
      "grad_norm": 4.34375,
      "learning_rate": 3.0797345966904933e-06,
      "loss": 0.74611931,
      "memory(GiB)": 147.13,
      "step": 54580,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78306842,
      "epoch": 1.2736260371250674,
      "grad_norm": 5.40625,
      "learning_rate": 3.0779904962061173e-06,
      "loss": 0.75449877,
      "memory(GiB)": 147.13,
      "step": 54590,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.78306794,
      "epoch": 1.2738593446973563,
      "grad_norm": 5.625,
      "learning_rate": 3.076246670088041e-06,
      "loss": 0.77245879,
      "memory(GiB)": 147.13,
      "step": 54600,
      "train_speed(iter/s)": 0.200765
    },
    {
      "acc": 0.77822618,
      "epoch": 1.2740926522696452,
      "grad_norm": 6.59375,
      "learning_rate": 3.074503118585192e-06,
      "loss": 0.78750162,
      "memory(GiB)": 147.13,
      "step": 54610,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.76630201,
      "epoch": 1.274325959841934,
      "grad_norm": 5.90625,
      "learning_rate": 3.072759841946464e-06,
      "loss": 0.84698448,
      "memory(GiB)": 147.13,
      "step": 54620,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.757019,
      "epoch": 1.274559267414223,
      "grad_norm": 6.5,
      "learning_rate": 3.0710168404207086e-06,
      "loss": 0.88410797,
      "memory(GiB)": 147.13,
      "step": 54630,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.7815589,
      "epoch": 1.2747925749865119,
      "grad_norm": 3.984375,
      "learning_rate": 3.0692741142567385e-06,
      "loss": 0.78019571,
      "memory(GiB)": 147.13,
      "step": 54640,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.78222027,
      "epoch": 1.2750258825588008,
      "grad_norm": 6.3125,
      "learning_rate": 3.0675316637033296e-06,
      "loss": 0.80412626,
      "memory(GiB)": 147.13,
      "step": 54650,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.77893991,
      "epoch": 1.2752591901310897,
      "grad_norm": 4.125,
      "learning_rate": 3.0657894890092134e-06,
      "loss": 0.79963918,
      "memory(GiB)": 147.13,
      "step": 54660,
      "train_speed(iter/s)": 0.200879
    },
    {
      "acc": 0.79304705,
      "epoch": 1.2754924977033786,
      "grad_norm": 7.5,
      "learning_rate": 3.0640475904230848e-06,
      "loss": 0.75699606,
      "memory(GiB)": 147.13,
      "step": 54670,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.78414087,
      "epoch": 1.2757258052756675,
      "grad_norm": 4.375,
      "learning_rate": 3.062305968193601e-06,
      "loss": 0.77377367,
      "memory(GiB)": 147.13,
      "step": 54680,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.80083361,
      "epoch": 1.2759591128479564,
      "grad_norm": 4.09375,
      "learning_rate": 3.060564622569377e-06,
      "loss": 0.69888325,
      "memory(GiB)": 147.13,
      "step": 54690,
      "train_speed(iter/s)": 0.200935
    },
    {
      "acc": 0.77480173,
      "epoch": 1.2761924204202453,
      "grad_norm": 6.34375,
      "learning_rate": 3.0588235537989897e-06,
      "loss": 0.79366236,
      "memory(GiB)": 147.13,
      "step": 54700,
      "train_speed(iter/s)": 0.200954
    },
    {
      "acc": 0.75772133,
      "epoch": 1.2764257279925342,
      "grad_norm": 5.0,
      "learning_rate": 3.057082762130976e-06,
      "loss": 0.87921181,
      "memory(GiB)": 147.13,
      "step": 54710,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.80236292,
      "epoch": 1.276659035564823,
      "grad_norm": 5.875,
      "learning_rate": 3.0553422478138333e-06,
      "loss": 0.69562078,
      "memory(GiB)": 147.13,
      "step": 54720,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.76373925,
      "epoch": 1.276892343137112,
      "grad_norm": 6.59375,
      "learning_rate": 3.0536020110960214e-06,
      "loss": 0.86020823,
      "memory(GiB)": 147.13,
      "step": 54730,
      "train_speed(iter/s)": 0.201011
    },
    {
      "acc": 0.76700792,
      "epoch": 1.2771256507094009,
      "grad_norm": 4.4375,
      "learning_rate": 3.0518620522259557e-06,
      "loss": 0.83871212,
      "memory(GiB)": 147.13,
      "step": 54740,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.78003569,
      "epoch": 1.2773589582816898,
      "grad_norm": 7.4375,
      "learning_rate": 3.0501223714520155e-06,
      "loss": 0.78800702,
      "memory(GiB)": 147.13,
      "step": 54750,
      "train_speed(iter/s)": 0.201049
    },
    {
      "acc": 0.7942296,
      "epoch": 1.2775922658539787,
      "grad_norm": 6.28125,
      "learning_rate": 3.048382969022543e-06,
      "loss": 0.75022469,
      "memory(GiB)": 147.13,
      "step": 54760,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.77333212,
      "epoch": 1.2778255734262676,
      "grad_norm": 8.4375,
      "learning_rate": 3.0466438451858326e-06,
      "loss": 0.81570168,
      "memory(GiB)": 147.13,
      "step": 54770,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.7831358,
      "epoch": 1.2780588809985565,
      "grad_norm": 6.625,
      "learning_rate": 3.044905000190146e-06,
      "loss": 0.77250023,
      "memory(GiB)": 147.13,
      "step": 54780,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.78039427,
      "epoch": 1.2782921885708454,
      "grad_norm": 6.53125,
      "learning_rate": 3.043166434283703e-06,
      "loss": 0.78244643,
      "memory(GiB)": 147.13,
      "step": 54790,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.7836205,
      "epoch": 1.2785254961431343,
      "grad_norm": 5.96875,
      "learning_rate": 3.0414281477146823e-06,
      "loss": 0.79010315,
      "memory(GiB)": 147.13,
      "step": 54800,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.77984056,
      "epoch": 1.2787588037154232,
      "grad_norm": 5.625,
      "learning_rate": 3.0396901407312263e-06,
      "loss": 0.80035505,
      "memory(GiB)": 147.13,
      "step": 54810,
      "train_speed(iter/s)": 0.201166
    },
    {
      "acc": 0.78728938,
      "epoch": 1.278992111287712,
      "grad_norm": 5.15625,
      "learning_rate": 3.037952413581431e-06,
      "loss": 0.74682913,
      "memory(GiB)": 147.13,
      "step": 54820,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.7650311,
      "epoch": 1.2792254188600007,
      "grad_norm": 7.59375,
      "learning_rate": 3.03621496651336e-06,
      "loss": 0.84308777,
      "memory(GiB)": 147.13,
      "step": 54830,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.75912447,
      "epoch": 1.2794587264322899,
      "grad_norm": 6.75,
      "learning_rate": 3.0344777997750313e-06,
      "loss": 0.8658433,
      "memory(GiB)": 147.13,
      "step": 54840,
      "train_speed(iter/s)": 0.201224
    },
    {
      "acc": 0.7930964,
      "epoch": 1.2796920340045785,
      "grad_norm": 4.125,
      "learning_rate": 3.0327409136144257e-06,
      "loss": 0.73651738,
      "memory(GiB)": 147.13,
      "step": 54850,
      "train_speed(iter/s)": 0.201243
    },
    {
      "acc": 0.7795516,
      "epoch": 1.2799253415768677,
      "grad_norm": 5.09375,
      "learning_rate": 3.031004308279484e-06,
      "loss": 0.77885942,
      "memory(GiB)": 147.13,
      "step": 54860,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.7870728,
      "epoch": 1.2801586491491563,
      "grad_norm": 4.1875,
      "learning_rate": 3.0292679840181048e-06,
      "loss": 0.77090311,
      "memory(GiB)": 147.13,
      "step": 54870,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.79632759,
      "epoch": 1.2803919567214455,
      "grad_norm": 5.53125,
      "learning_rate": 3.02753194107815e-06,
      "loss": 0.7337678,
      "memory(GiB)": 147.13,
      "step": 54880,
      "train_speed(iter/s)": 0.201298
    },
    {
      "acc": 0.78915548,
      "epoch": 1.2806252642937341,
      "grad_norm": 4.65625,
      "learning_rate": 3.0257961797074353e-06,
      "loss": 0.75042048,
      "memory(GiB)": 147.13,
      "step": 54890,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.78317914,
      "epoch": 1.2808585718660233,
      "grad_norm": 6.96875,
      "learning_rate": 3.0240607001537442e-06,
      "loss": 0.78657742,
      "memory(GiB)": 147.13,
      "step": 54900,
      "train_speed(iter/s)": 0.201335
    },
    {
      "acc": 0.7978034,
      "epoch": 1.281091879438312,
      "grad_norm": 6.4375,
      "learning_rate": 3.022325502664813e-06,
      "loss": 0.71349802,
      "memory(GiB)": 147.13,
      "step": 54910,
      "train_speed(iter/s)": 0.201354
    },
    {
      "acc": 0.78730545,
      "epoch": 1.2813251870106008,
      "grad_norm": 8.125,
      "learning_rate": 3.020590587488342e-06,
      "loss": 0.7752285,
      "memory(GiB)": 147.13,
      "step": 54920,
      "train_speed(iter/s)": 0.201372
    },
    {
      "acc": 0.76834173,
      "epoch": 1.2815584945828897,
      "grad_norm": 5.875,
      "learning_rate": 3.0188559548719888e-06,
      "loss": 0.83232594,
      "memory(GiB)": 147.13,
      "step": 54930,
      "train_speed(iter/s)": 0.201391
    },
    {
      "acc": 0.78525715,
      "epoch": 1.2817918021551786,
      "grad_norm": 7.96875,
      "learning_rate": 3.0171216050633735e-06,
      "loss": 0.7697998,
      "memory(GiB)": 147.13,
      "step": 54940,
      "train_speed(iter/s)": 0.20141
    },
    {
      "acc": 0.7783596,
      "epoch": 1.2820251097274675,
      "grad_norm": 8.375,
      "learning_rate": 3.0153875383100732e-06,
      "loss": 0.78331108,
      "memory(GiB)": 147.13,
      "step": 54950,
      "train_speed(iter/s)": 0.201429
    },
    {
      "acc": 0.77114954,
      "epoch": 1.2822584172997564,
      "grad_norm": 6.8125,
      "learning_rate": 3.0136537548596247e-06,
      "loss": 0.83518686,
      "memory(GiB)": 147.13,
      "step": 54960,
      "train_speed(iter/s)": 0.201448
    },
    {
      "acc": 0.79357519,
      "epoch": 1.2824917248720453,
      "grad_norm": 3.984375,
      "learning_rate": 3.011920254959526e-06,
      "loss": 0.72738924,
      "memory(GiB)": 147.13,
      "step": 54970,
      "train_speed(iter/s)": 0.201468
    },
    {
      "acc": 0.7828373,
      "epoch": 1.2827250324443342,
      "grad_norm": 4.21875,
      "learning_rate": 3.010187038857233e-06,
      "loss": 0.78417239,
      "memory(GiB)": 147.13,
      "step": 54980,
      "train_speed(iter/s)": 0.201488
    },
    {
      "acc": 0.78416605,
      "epoch": 1.2829583400166231,
      "grad_norm": 5.59375,
      "learning_rate": 3.008454106800164e-06,
      "loss": 0.77895746,
      "memory(GiB)": 147.13,
      "step": 54990,
      "train_speed(iter/s)": 0.201507
    },
    {
      "acc": 0.77086582,
      "epoch": 1.283191647588912,
      "grad_norm": 5.3125,
      "learning_rate": 3.006721459035691e-06,
      "loss": 0.81353455,
      "memory(GiB)": 147.13,
      "step": 55000,
      "train_speed(iter/s)": 0.201526
    },
    {
      "epoch": 1.283191647588912,
      "eval_acc": 0.7444704999110399,
      "eval_loss": 0.8050407767295837,
      "eval_runtime": 1270.065,
      "eval_samples_per_second": 28.338,
      "eval_steps_per_second": 14.169,
      "step": 55000
    },
    {
      "acc": 0.76830349,
      "epoch": 1.283424955161201,
      "grad_norm": 4.90625,
      "learning_rate": 3.0049890958111505e-06,
      "loss": 0.82744799,
      "memory(GiB)": 147.13,
      "step": 55010,
      "train_speed(iter/s)": 0.200597
    },
    {
      "acc": 0.78124275,
      "epoch": 1.2836582627334898,
      "grad_norm": 5.25,
      "learning_rate": 3.0032570173738367e-06,
      "loss": 0.77643175,
      "memory(GiB)": 147.13,
      "step": 55020,
      "train_speed(iter/s)": 0.200615
    },
    {
      "acc": 0.77881341,
      "epoch": 1.2838915703057787,
      "grad_norm": 5.5,
      "learning_rate": 3.0015252239710052e-06,
      "loss": 0.8051384,
      "memory(GiB)": 147.13,
      "step": 55030,
      "train_speed(iter/s)": 0.200634
    },
    {
      "acc": 0.78495493,
      "epoch": 1.2841248778780676,
      "grad_norm": 6.09375,
      "learning_rate": 2.9997937158498657e-06,
      "loss": 0.77997541,
      "memory(GiB)": 147.13,
      "step": 55040,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.76635275,
      "epoch": 1.2843581854503565,
      "grad_norm": 5.15625,
      "learning_rate": 2.998062493257593e-06,
      "loss": 0.85216427,
      "memory(GiB)": 147.13,
      "step": 55050,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.77827249,
      "epoch": 1.2845914930226454,
      "grad_norm": 4.84375,
      "learning_rate": 2.9963315564413174e-06,
      "loss": 0.78949223,
      "memory(GiB)": 147.13,
      "step": 55060,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.7691534,
      "epoch": 1.2848248005949343,
      "grad_norm": 4.59375,
      "learning_rate": 2.994600905648131e-06,
      "loss": 0.83677711,
      "memory(GiB)": 147.13,
      "step": 55070,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.7861763,
      "epoch": 1.2850581081672232,
      "grad_norm": 6.5625,
      "learning_rate": 2.9928705411250813e-06,
      "loss": 0.77848387,
      "memory(GiB)": 147.13,
      "step": 55080,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.766677,
      "epoch": 1.285291415739512,
      "grad_norm": 4.625,
      "learning_rate": 2.9911404631191796e-06,
      "loss": 0.84276142,
      "memory(GiB)": 147.13,
      "step": 55090,
      "train_speed(iter/s)": 0.200745
    },
    {
      "acc": 0.77530727,
      "epoch": 1.285524723311801,
      "grad_norm": 5.21875,
      "learning_rate": 2.9894106718773936e-06,
      "loss": 0.82425938,
      "memory(GiB)": 147.13,
      "step": 55100,
      "train_speed(iter/s)": 0.200763
    },
    {
      "acc": 0.78108807,
      "epoch": 1.28575803088409,
      "grad_norm": 6.34375,
      "learning_rate": 2.987681167646652e-06,
      "loss": 0.78073077,
      "memory(GiB)": 147.13,
      "step": 55110,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.76942391,
      "epoch": 1.2859913384563788,
      "grad_norm": 6.5625,
      "learning_rate": 2.985951950673836e-06,
      "loss": 0.8330018,
      "memory(GiB)": 147.13,
      "step": 55120,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.75507021,
      "epoch": 1.2862246460286677,
      "grad_norm": 6.03125,
      "learning_rate": 2.984223021205795e-06,
      "loss": 0.89126616,
      "memory(GiB)": 147.13,
      "step": 55130,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.78265476,
      "epoch": 1.2864579536009566,
      "grad_norm": 4.75,
      "learning_rate": 2.9824943794893312e-06,
      "loss": 0.78735905,
      "memory(GiB)": 147.13,
      "step": 55140,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.78866234,
      "epoch": 1.2866912611732455,
      "grad_norm": 6.34375,
      "learning_rate": 2.9807660257712097e-06,
      "loss": 0.75422587,
      "memory(GiB)": 147.13,
      "step": 55150,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.78413725,
      "epoch": 1.2869245687455344,
      "grad_norm": 4.40625,
      "learning_rate": 2.9790379602981508e-06,
      "loss": 0.78166876,
      "memory(GiB)": 147.13,
      "step": 55160,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.77496414,
      "epoch": 1.2871578763178233,
      "grad_norm": 4.59375,
      "learning_rate": 2.9773101833168374e-06,
      "loss": 0.81109505,
      "memory(GiB)": 147.13,
      "step": 55170,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.77975092,
      "epoch": 1.2873911838901122,
      "grad_norm": 4.375,
      "learning_rate": 2.9755826950739057e-06,
      "loss": 0.79754815,
      "memory(GiB)": 147.13,
      "step": 55180,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.77278199,
      "epoch": 1.287624491462401,
      "grad_norm": 5.125,
      "learning_rate": 2.973855495815957e-06,
      "loss": 0.80997,
      "memory(GiB)": 147.13,
      "step": 55190,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.77235184,
      "epoch": 1.28785779903469,
      "grad_norm": 6.46875,
      "learning_rate": 2.9721285857895475e-06,
      "loss": 0.82727203,
      "memory(GiB)": 147.13,
      "step": 55200,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.78080778,
      "epoch": 1.288091106606979,
      "grad_norm": 5.6875,
      "learning_rate": 2.9704019652411933e-06,
      "loss": 0.80346479,
      "memory(GiB)": 147.13,
      "step": 55210,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.78474374,
      "epoch": 1.2883244141792676,
      "grad_norm": 5.96875,
      "learning_rate": 2.9686756344173712e-06,
      "loss": 0.77900343,
      "memory(GiB)": 147.13,
      "step": 55220,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.77824039,
      "epoch": 1.2885577217515567,
      "grad_norm": 25.75,
      "learning_rate": 2.96694959356451e-06,
      "loss": 0.81776686,
      "memory(GiB)": 147.13,
      "step": 55230,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.78159637,
      "epoch": 1.2887910293238454,
      "grad_norm": 6.09375,
      "learning_rate": 2.9652238429290036e-06,
      "loss": 0.79265308,
      "memory(GiB)": 147.13,
      "step": 55240,
      "train_speed(iter/s)": 0.20103
    },
    {
      "acc": 0.79291801,
      "epoch": 1.2890243368961345,
      "grad_norm": 6.40625,
      "learning_rate": 2.9634983827572038e-06,
      "loss": 0.73101878,
      "memory(GiB)": 147.13,
      "step": 55250,
      "train_speed(iter/s)": 0.20105
    },
    {
      "acc": 0.7721282,
      "epoch": 1.2892576444684232,
      "grad_norm": 4.53125,
      "learning_rate": 2.961773213295417e-06,
      "loss": 0.82999897,
      "memory(GiB)": 147.13,
      "step": 55260,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.79432802,
      "epoch": 1.2894909520407123,
      "grad_norm": 4.78125,
      "learning_rate": 2.960048334789912e-06,
      "loss": 0.73425961,
      "memory(GiB)": 147.13,
      "step": 55270,
      "train_speed(iter/s)": 0.20109
    },
    {
      "acc": 0.78572264,
      "epoch": 1.289724259613001,
      "grad_norm": 5.625,
      "learning_rate": 2.9583237474869143e-06,
      "loss": 0.76063013,
      "memory(GiB)": 147.13,
      "step": 55280,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.77698021,
      "epoch": 1.28995756718529,
      "grad_norm": 5.40625,
      "learning_rate": 2.956599451632609e-06,
      "loss": 0.79158425,
      "memory(GiB)": 147.13,
      "step": 55290,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.76085668,
      "epoch": 1.2901908747575788,
      "grad_norm": 5.375,
      "learning_rate": 2.9548754474731376e-06,
      "loss": 0.85578661,
      "memory(GiB)": 147.13,
      "step": 55300,
      "train_speed(iter/s)": 0.201145
    },
    {
      "acc": 0.78259788,
      "epoch": 1.2904241823298677,
      "grad_norm": 5.75,
      "learning_rate": 2.953151735254604e-06,
      "loss": 0.77954855,
      "memory(GiB)": 147.13,
      "step": 55310,
      "train_speed(iter/s)": 0.201164
    },
    {
      "acc": 0.78067522,
      "epoch": 1.2906574899021566,
      "grad_norm": 5.375,
      "learning_rate": 2.9514283152230637e-06,
      "loss": 0.78616638,
      "memory(GiB)": 147.13,
      "step": 55320,
      "train_speed(iter/s)": 0.201181
    },
    {
      "acc": 0.77021255,
      "epoch": 1.2908907974744455,
      "grad_norm": 8.5625,
      "learning_rate": 2.949705187624539e-06,
      "loss": 0.80933561,
      "memory(GiB)": 147.13,
      "step": 55330,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.777321,
      "epoch": 1.2911241050467344,
      "grad_norm": 5.53125,
      "learning_rate": 2.947982352705001e-06,
      "loss": 0.78500934,
      "memory(GiB)": 147.13,
      "step": 55340,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.78850975,
      "epoch": 1.2913574126190233,
      "grad_norm": 5.96875,
      "learning_rate": 2.9462598107103855e-06,
      "loss": 0.7612771,
      "memory(GiB)": 147.13,
      "step": 55350,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.77932158,
      "epoch": 1.2915907201913122,
      "grad_norm": 6.78125,
      "learning_rate": 2.9445375618865857e-06,
      "loss": 0.79323964,
      "memory(GiB)": 147.13,
      "step": 55360,
      "train_speed(iter/s)": 0.201256
    },
    {
      "acc": 0.75293274,
      "epoch": 1.291824027763601,
      "grad_norm": 6.15625,
      "learning_rate": 2.942815606479452e-06,
      "loss": 0.89559402,
      "memory(GiB)": 147.13,
      "step": 55370,
      "train_speed(iter/s)": 0.201275
    },
    {
      "acc": 0.79058542,
      "epoch": 1.29205733533589,
      "grad_norm": 4.8125,
      "learning_rate": 2.941093944734793e-06,
      "loss": 0.76939631,
      "memory(GiB)": 147.13,
      "step": 55380,
      "train_speed(iter/s)": 0.201294
    },
    {
      "acc": 0.78246756,
      "epoch": 1.2922906429081789,
      "grad_norm": 5.4375,
      "learning_rate": 2.939372576898376e-06,
      "loss": 0.78994184,
      "memory(GiB)": 147.13,
      "step": 55390,
      "train_speed(iter/s)": 0.201313
    },
    {
      "acc": 0.7745213,
      "epoch": 1.2925239504804678,
      "grad_norm": 5.84375,
      "learning_rate": 2.937651503215924e-06,
      "loss": 0.80446148,
      "memory(GiB)": 147.13,
      "step": 55400,
      "train_speed(iter/s)": 0.20133
    },
    {
      "acc": 0.77480874,
      "epoch": 1.2927572580527567,
      "grad_norm": 5.625,
      "learning_rate": 2.9359307239331214e-06,
      "loss": 0.8080328,
      "memory(GiB)": 147.13,
      "step": 55410,
      "train_speed(iter/s)": 0.20135
    },
    {
      "acc": 0.78159122,
      "epoch": 1.2929905656250456,
      "grad_norm": 5.1875,
      "learning_rate": 2.9342102392956075e-06,
      "loss": 0.77628255,
      "memory(GiB)": 147.13,
      "step": 55420,
      "train_speed(iter/s)": 0.201368
    },
    {
      "acc": 0.77849216,
      "epoch": 1.2932238731973345,
      "grad_norm": 5.59375,
      "learning_rate": 2.932490049548982e-06,
      "loss": 0.79256525,
      "memory(GiB)": 147.13,
      "step": 55430,
      "train_speed(iter/s)": 0.201385
    },
    {
      "acc": 0.80485229,
      "epoch": 1.2934571807696233,
      "grad_norm": 3.953125,
      "learning_rate": 2.9307701549388025e-06,
      "loss": 0.68343935,
      "memory(GiB)": 147.13,
      "step": 55440,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.77289133,
      "epoch": 1.2936904883419122,
      "grad_norm": 4.75,
      "learning_rate": 2.929050555710582e-06,
      "loss": 0.8112318,
      "memory(GiB)": 147.13,
      "step": 55450,
      "train_speed(iter/s)": 0.201421
    },
    {
      "acc": 0.7689023,
      "epoch": 1.2939237959142011,
      "grad_norm": 8.125,
      "learning_rate": 2.9273312521097926e-06,
      "loss": 0.82355614,
      "memory(GiB)": 147.13,
      "step": 55460,
      "train_speed(iter/s)": 0.201441
    },
    {
      "acc": 0.78700824,
      "epoch": 1.29415710348649,
      "grad_norm": 6.34375,
      "learning_rate": 2.9256122443818657e-06,
      "loss": 0.75792155,
      "memory(GiB)": 147.13,
      "step": 55470,
      "train_speed(iter/s)": 0.201459
    },
    {
      "acc": 0.78975463,
      "epoch": 1.294390411058779,
      "grad_norm": 5.90625,
      "learning_rate": 2.923893532772187e-06,
      "loss": 0.77099237,
      "memory(GiB)": 147.13,
      "step": 55480,
      "train_speed(iter/s)": 0.201477
    },
    {
      "acc": 0.77440877,
      "epoch": 1.2946237186310678,
      "grad_norm": 5.6875,
      "learning_rate": 2.9221751175261036e-06,
      "loss": 0.81504545,
      "memory(GiB)": 147.13,
      "step": 55490,
      "train_speed(iter/s)": 0.201496
    },
    {
      "acc": 0.78191586,
      "epoch": 1.2948570262033567,
      "grad_norm": 7.34375,
      "learning_rate": 2.9204569988889186e-06,
      "loss": 0.78614321,
      "memory(GiB)": 147.13,
      "step": 55500,
      "train_speed(iter/s)": 0.201513
    },
    {
      "epoch": 1.2948570262033567,
      "eval_acc": 0.7445686683254407,
      "eval_loss": 0.8049291968345642,
      "eval_runtime": 1270.1062,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 55500
    },
    {
      "acc": 0.79387369,
      "epoch": 1.2950903337756456,
      "grad_norm": 6.3125,
      "learning_rate": 2.9187391771058938e-06,
      "loss": 0.7472805,
      "memory(GiB)": 147.13,
      "step": 55510,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.77921963,
      "epoch": 1.2953236413479345,
      "grad_norm": 4.96875,
      "learning_rate": 2.9170216524222446e-06,
      "loss": 0.78324275,
      "memory(GiB)": 147.13,
      "step": 55520,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.76559649,
      "epoch": 1.2955569489202234,
      "grad_norm": 7.96875,
      "learning_rate": 2.9153044250831512e-06,
      "loss": 0.84542322,
      "memory(GiB)": 147.13,
      "step": 55530,
      "train_speed(iter/s)": 0.20063
    },
    {
      "acc": 0.76427712,
      "epoch": 1.2957902564925123,
      "grad_norm": 6.21875,
      "learning_rate": 2.913587495333744e-06,
      "loss": 0.83981133,
      "memory(GiB)": 147.13,
      "step": 55540,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.77819748,
      "epoch": 1.2960235640648012,
      "grad_norm": 6.46875,
      "learning_rate": 2.9118708634191177e-06,
      "loss": 0.78465223,
      "memory(GiB)": 147.13,
      "step": 55550,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.78325009,
      "epoch": 1.2962568716370901,
      "grad_norm": 5.03125,
      "learning_rate": 2.910154529584319e-06,
      "loss": 0.78194418,
      "memory(GiB)": 147.13,
      "step": 55560,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.79050236,
      "epoch": 1.296490179209379,
      "grad_norm": 4.09375,
      "learning_rate": 2.9084384940743543e-06,
      "loss": 0.76087713,
      "memory(GiB)": 147.13,
      "step": 55570,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.78712444,
      "epoch": 1.296723486781668,
      "grad_norm": 5.21875,
      "learning_rate": 2.9067227571341873e-06,
      "loss": 0.76762724,
      "memory(GiB)": 147.13,
      "step": 55580,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.776087,
      "epoch": 1.2969567943539568,
      "grad_norm": 4.9375,
      "learning_rate": 2.905007319008736e-06,
      "loss": 0.81573505,
      "memory(GiB)": 147.13,
      "step": 55590,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.77835026,
      "epoch": 1.2971901019262457,
      "grad_norm": 4.8125,
      "learning_rate": 2.903292179942883e-06,
      "loss": 0.79308224,
      "memory(GiB)": 147.13,
      "step": 55600,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.78184395,
      "epoch": 1.2974234094985344,
      "grad_norm": 5.0,
      "learning_rate": 2.9015773401814606e-06,
      "loss": 0.79529061,
      "memory(GiB)": 147.13,
      "step": 55610,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.78858995,
      "epoch": 1.2976567170708235,
      "grad_norm": 5.65625,
      "learning_rate": 2.899862799969265e-06,
      "loss": 0.77276478,
      "memory(GiB)": 147.13,
      "step": 55620,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.77310009,
      "epoch": 1.2978900246431122,
      "grad_norm": 5.15625,
      "learning_rate": 2.898148559551045e-06,
      "loss": 0.81262197,
      "memory(GiB)": 147.13,
      "step": 55630,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.79125342,
      "epoch": 1.2981233322154013,
      "grad_norm": 5.15625,
      "learning_rate": 2.8964346191715058e-06,
      "loss": 0.74343634,
      "memory(GiB)": 147.13,
      "step": 55640,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.78252659,
      "epoch": 1.29835663978769,
      "grad_norm": 4.5625,
      "learning_rate": 2.894720979075315e-06,
      "loss": 0.76678948,
      "memory(GiB)": 147.13,
      "step": 55650,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.77767839,
      "epoch": 1.2985899473599791,
      "grad_norm": 4.84375,
      "learning_rate": 2.8930076395070915e-06,
      "loss": 0.76314602,
      "memory(GiB)": 147.13,
      "step": 55660,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.76303263,
      "epoch": 1.2988232549322678,
      "grad_norm": 4.71875,
      "learning_rate": 2.8912946007114175e-06,
      "loss": 0.85604687,
      "memory(GiB)": 147.13,
      "step": 55670,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.79654713,
      "epoch": 1.299056562504557,
      "grad_norm": 4.65625,
      "learning_rate": 2.8895818629328254e-06,
      "loss": 0.71905622,
      "memory(GiB)": 147.13,
      "step": 55680,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.76574378,
      "epoch": 1.2992898700768456,
      "grad_norm": 11.75,
      "learning_rate": 2.8878694264158103e-06,
      "loss": 0.84654961,
      "memory(GiB)": 147.13,
      "step": 55690,
      "train_speed(iter/s)": 0.200924
    },
    {
      "acc": 0.78273115,
      "epoch": 1.2995231776491345,
      "grad_norm": 5.71875,
      "learning_rate": 2.8861572914048184e-06,
      "loss": 0.76711287,
      "memory(GiB)": 147.13,
      "step": 55700,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.77995882,
      "epoch": 1.2997564852214234,
      "grad_norm": 5.84375,
      "learning_rate": 2.8844454581442614e-06,
      "loss": 0.77336879,
      "memory(GiB)": 147.13,
      "step": 55710,
      "train_speed(iter/s)": 0.200962
    },
    {
      "acc": 0.80577364,
      "epoch": 1.2999897927937123,
      "grad_norm": 4.90625,
      "learning_rate": 2.8827339268785015e-06,
      "loss": 0.70034103,
      "memory(GiB)": 147.13,
      "step": 55720,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.76289902,
      "epoch": 1.3002231003660012,
      "grad_norm": 6.09375,
      "learning_rate": 2.881022697851855e-06,
      "loss": 0.84976912,
      "memory(GiB)": 147.13,
      "step": 55730,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.76065493,
      "epoch": 1.30045640793829,
      "grad_norm": 7.09375,
      "learning_rate": 2.879311771308606e-06,
      "loss": 0.87251511,
      "memory(GiB)": 147.13,
      "step": 55740,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.7798687,
      "epoch": 1.300689715510579,
      "grad_norm": 9.0,
      "learning_rate": 2.877601147492983e-06,
      "loss": 0.78704138,
      "memory(GiB)": 147.13,
      "step": 55750,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.76645966,
      "epoch": 1.300923023082868,
      "grad_norm": 4.5,
      "learning_rate": 2.8758908266491815e-06,
      "loss": 0.84115009,
      "memory(GiB)": 147.13,
      "step": 55760,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.77044306,
      "epoch": 1.3011563306551568,
      "grad_norm": 7.0,
      "learning_rate": 2.874180809021348e-06,
      "loss": 0.82898483,
      "memory(GiB)": 147.13,
      "step": 55770,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.78014908,
      "epoch": 1.3013896382274457,
      "grad_norm": 5.65625,
      "learning_rate": 2.872471094853584e-06,
      "loss": 0.78329229,
      "memory(GiB)": 147.13,
      "step": 55780,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.78297763,
      "epoch": 1.3016229457997346,
      "grad_norm": 6.15625,
      "learning_rate": 2.8707616843899554e-06,
      "loss": 0.76991262,
      "memory(GiB)": 147.13,
      "step": 55790,
      "train_speed(iter/s)": 0.201112
    },
    {
      "acc": 0.78463106,
      "epoch": 1.3018562533720235,
      "grad_norm": 7.28125,
      "learning_rate": 2.8690525778744777e-06,
      "loss": 0.81536741,
      "memory(GiB)": 147.13,
      "step": 55800,
      "train_speed(iter/s)": 0.201131
    },
    {
      "acc": 0.78778915,
      "epoch": 1.3020895609443124,
      "grad_norm": 5.1875,
      "learning_rate": 2.867343775551126e-06,
      "loss": 0.76017818,
      "memory(GiB)": 147.13,
      "step": 55810,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.76485157,
      "epoch": 1.3023228685166013,
      "grad_norm": 7.0625,
      "learning_rate": 2.8656352776638274e-06,
      "loss": 0.84671221,
      "memory(GiB)": 147.13,
      "step": 55820,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.78802285,
      "epoch": 1.3025561760888902,
      "grad_norm": 5.03125,
      "learning_rate": 2.863927084456476e-06,
      "loss": 0.74512553,
      "memory(GiB)": 147.13,
      "step": 55830,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.78229799,
      "epoch": 1.302789483661179,
      "grad_norm": 6.0,
      "learning_rate": 2.862219196172911e-06,
      "loss": 0.78259425,
      "memory(GiB)": 147.13,
      "step": 55840,
      "train_speed(iter/s)": 0.201207
    },
    {
      "acc": 0.78135662,
      "epoch": 1.303022791233468,
      "grad_norm": 5.1875,
      "learning_rate": 2.8605116130569355e-06,
      "loss": 0.78383131,
      "memory(GiB)": 147.13,
      "step": 55850,
      "train_speed(iter/s)": 0.201226
    },
    {
      "acc": 0.77223597,
      "epoch": 1.3032560988057569,
      "grad_norm": 7.78125,
      "learning_rate": 2.8588043353523066e-06,
      "loss": 0.83154545,
      "memory(GiB)": 147.13,
      "step": 55860,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.76503639,
      "epoch": 1.3034894063780458,
      "grad_norm": 6.0,
      "learning_rate": 2.8570973633027342e-06,
      "loss": 0.856847,
      "memory(GiB)": 147.13,
      "step": 55870,
      "train_speed(iter/s)": 0.201264
    },
    {
      "acc": 0.78612194,
      "epoch": 1.3037227139503347,
      "grad_norm": 4.59375,
      "learning_rate": 2.8553906971518936e-06,
      "loss": 0.76604691,
      "memory(GiB)": 147.13,
      "step": 55880,
      "train_speed(iter/s)": 0.201282
    },
    {
      "acc": 0.7799561,
      "epoch": 1.3039560215226236,
      "grad_norm": 8.3125,
      "learning_rate": 2.8536843371434054e-06,
      "loss": 0.78653412,
      "memory(GiB)": 147.13,
      "step": 55890,
      "train_speed(iter/s)": 0.201301
    },
    {
      "acc": 0.77653108,
      "epoch": 1.3041893290949125,
      "grad_norm": 6.0,
      "learning_rate": 2.851978283520859e-06,
      "loss": 0.79027066,
      "memory(GiB)": 147.13,
      "step": 55900,
      "train_speed(iter/s)": 0.20132
    },
    {
      "acc": 0.78584161,
      "epoch": 1.3044226366672014,
      "grad_norm": 6.0,
      "learning_rate": 2.850272536527784e-06,
      "loss": 0.78225594,
      "memory(GiB)": 147.13,
      "step": 55910,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.77251587,
      "epoch": 1.3046559442394903,
      "grad_norm": 8.75,
      "learning_rate": 2.848567096407682e-06,
      "loss": 0.83527536,
      "memory(GiB)": 147.13,
      "step": 55920,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.77410121,
      "epoch": 1.3048892518117792,
      "grad_norm": 6.3125,
      "learning_rate": 2.8468619634040017e-06,
      "loss": 0.820961,
      "memory(GiB)": 147.13,
      "step": 55930,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.76235828,
      "epoch": 1.305122559384068,
      "grad_norm": 4.78125,
      "learning_rate": 2.8451571377601495e-06,
      "loss": 0.85515633,
      "memory(GiB)": 147.13,
      "step": 55940,
      "train_speed(iter/s)": 0.201395
    },
    {
      "acc": 0.79035673,
      "epoch": 1.305355866956357,
      "grad_norm": 5.125,
      "learning_rate": 2.8434526197194915e-06,
      "loss": 0.73724895,
      "memory(GiB)": 147.13,
      "step": 55950,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.78144627,
      "epoch": 1.3055891745286459,
      "grad_norm": 5.4375,
      "learning_rate": 2.8417484095253434e-06,
      "loss": 0.78020916,
      "memory(GiB)": 147.13,
      "step": 55960,
      "train_speed(iter/s)": 0.201432
    },
    {
      "acc": 0.77665806,
      "epoch": 1.3058224821009348,
      "grad_norm": 5.4375,
      "learning_rate": 2.8400445074209852e-06,
      "loss": 0.78938322,
      "memory(GiB)": 147.13,
      "step": 55970,
      "train_speed(iter/s)": 0.201451
    },
    {
      "acc": 0.76400418,
      "epoch": 1.3060557896732234,
      "grad_norm": 6.1875,
      "learning_rate": 2.8383409136496443e-06,
      "loss": 0.86329784,
      "memory(GiB)": 147.13,
      "step": 55980,
      "train_speed(iter/s)": 0.20147
    },
    {
      "acc": 0.78050137,
      "epoch": 1.3062890972455126,
      "grad_norm": 4.59375,
      "learning_rate": 2.8366376284545117e-06,
      "loss": 0.77180653,
      "memory(GiB)": 147.13,
      "step": 55990,
      "train_speed(iter/s)": 0.201489
    },
    {
      "acc": 0.79701538,
      "epoch": 1.3065224048178012,
      "grad_norm": 9.75,
      "learning_rate": 2.8349346520787284e-06,
      "loss": 0.72172403,
      "memory(GiB)": 147.13,
      "step": 56000,
      "train_speed(iter/s)": 0.201506
    },
    {
      "epoch": 1.3065224048178012,
      "eval_acc": 0.7445838820275092,
      "eval_loss": 0.8049860000610352,
      "eval_runtime": 1271.539,
      "eval_samples_per_second": 28.305,
      "eval_steps_per_second": 14.153,
      "step": 56000
    },
    {
      "acc": 0.7685607,
      "epoch": 1.3067557123900904,
      "grad_norm": 4.6875,
      "learning_rate": 2.833231984765393e-06,
      "loss": 0.81922808,
      "memory(GiB)": 147.13,
      "step": 56010,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.79894352,
      "epoch": 1.306989019962379,
      "grad_norm": 5.125,
      "learning_rate": 2.8315296267575672e-06,
      "loss": 0.68040066,
      "memory(GiB)": 147.13,
      "step": 56020,
      "train_speed(iter/s)": 0.200612
    },
    {
      "acc": 0.77997193,
      "epoch": 1.3072223275346682,
      "grad_norm": 4.6875,
      "learning_rate": 2.8298275782982525e-06,
      "loss": 0.81127281,
      "memory(GiB)": 147.13,
      "step": 56030,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.77440147,
      "epoch": 1.3074556351069568,
      "grad_norm": 4.625,
      "learning_rate": 2.8281258396304224e-06,
      "loss": 0.8060461,
      "memory(GiB)": 147.13,
      "step": 56040,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.7795207,
      "epoch": 1.307688942679246,
      "grad_norm": 5.03125,
      "learning_rate": 2.8264244109969963e-06,
      "loss": 0.82625637,
      "memory(GiB)": 147.13,
      "step": 56050,
      "train_speed(iter/s)": 0.200669
    },
    {
      "acc": 0.77957468,
      "epoch": 1.3079222502515346,
      "grad_norm": 6.59375,
      "learning_rate": 2.824723292640856e-06,
      "loss": 0.77296095,
      "memory(GiB)": 147.13,
      "step": 56060,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.76619778,
      "epoch": 1.3081555578238238,
      "grad_norm": 5.8125,
      "learning_rate": 2.823022484804834e-06,
      "loss": 0.82944441,
      "memory(GiB)": 147.13,
      "step": 56070,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.77313166,
      "epoch": 1.3083888653961124,
      "grad_norm": 7.875,
      "learning_rate": 2.8213219877317164e-06,
      "loss": 0.82456112,
      "memory(GiB)": 147.13,
      "step": 56080,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.76039181,
      "epoch": 1.3086221729684013,
      "grad_norm": 5.71875,
      "learning_rate": 2.819621801664256e-06,
      "loss": 0.85574379,
      "memory(GiB)": 147.13,
      "step": 56090,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.78070183,
      "epoch": 1.3088554805406902,
      "grad_norm": 4.78125,
      "learning_rate": 2.817921926845147e-06,
      "loss": 0.81434116,
      "memory(GiB)": 147.13,
      "step": 56100,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.78396792,
      "epoch": 1.3090887881129791,
      "grad_norm": 5.1875,
      "learning_rate": 2.8162223635170515e-06,
      "loss": 0.76618247,
      "memory(GiB)": 147.13,
      "step": 56110,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.79208441,
      "epoch": 1.309322095685268,
      "grad_norm": 6.53125,
      "learning_rate": 2.814523111922577e-06,
      "loss": 0.74408183,
      "memory(GiB)": 147.13,
      "step": 56120,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.79604063,
      "epoch": 1.309555403257557,
      "grad_norm": 4.53125,
      "learning_rate": 2.812824172304297e-06,
      "loss": 0.75110483,
      "memory(GiB)": 147.13,
      "step": 56130,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.77951555,
      "epoch": 1.3097887108298458,
      "grad_norm": 5.75,
      "learning_rate": 2.8111255449047277e-06,
      "loss": 0.76333828,
      "memory(GiB)": 147.13,
      "step": 56140,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.79114752,
      "epoch": 1.3100220184021347,
      "grad_norm": 5.21875,
      "learning_rate": 2.809427229966353e-06,
      "loss": 0.73384008,
      "memory(GiB)": 147.13,
      "step": 56150,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.77196131,
      "epoch": 1.3102553259744236,
      "grad_norm": 8.3125,
      "learning_rate": 2.8077292277316036e-06,
      "loss": 0.80360489,
      "memory(GiB)": 147.13,
      "step": 56160,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.78124895,
      "epoch": 1.3104886335467125,
      "grad_norm": 5.375,
      "learning_rate": 2.8060315384428692e-06,
      "loss": 0.79064064,
      "memory(GiB)": 147.13,
      "step": 56170,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.77550502,
      "epoch": 1.3107219411190014,
      "grad_norm": 8.8125,
      "learning_rate": 2.8043341623424974e-06,
      "loss": 0.82532978,
      "memory(GiB)": 147.13,
      "step": 56180,
      "train_speed(iter/s)": 0.200907
    },
    {
      "acc": 0.80522442,
      "epoch": 1.3109552486912903,
      "grad_norm": 5.15625,
      "learning_rate": 2.8026370996727835e-06,
      "loss": 0.68403687,
      "memory(GiB)": 147.13,
      "step": 56190,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.79306288,
      "epoch": 1.3111885562635792,
      "grad_norm": 7.21875,
      "learning_rate": 2.800940350675988e-06,
      "loss": 0.7425447,
      "memory(GiB)": 147.13,
      "step": 56200,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.76478882,
      "epoch": 1.3114218638358681,
      "grad_norm": 5.40625,
      "learning_rate": 2.7992439155943185e-06,
      "loss": 0.87208843,
      "memory(GiB)": 147.13,
      "step": 56210,
      "train_speed(iter/s)": 0.200963
    },
    {
      "acc": 0.7814106,
      "epoch": 1.311655171408157,
      "grad_norm": 4.34375,
      "learning_rate": 2.797547794669938e-06,
      "loss": 0.7701395,
      "memory(GiB)": 147.13,
      "step": 56220,
      "train_speed(iter/s)": 0.200982
    },
    {
      "acc": 0.76465588,
      "epoch": 1.311888478980446,
      "grad_norm": 6.15625,
      "learning_rate": 2.7958519881449723e-06,
      "loss": 0.860077,
      "memory(GiB)": 147.13,
      "step": 56230,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.78399577,
      "epoch": 1.3121217865527348,
      "grad_norm": 4.40625,
      "learning_rate": 2.794156496261493e-06,
      "loss": 0.75324211,
      "memory(GiB)": 147.13,
      "step": 56240,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.77755909,
      "epoch": 1.3123550941250237,
      "grad_norm": 5.78125,
      "learning_rate": 2.792461319261538e-06,
      "loss": 0.79980249,
      "memory(GiB)": 147.13,
      "step": 56250,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.77361746,
      "epoch": 1.3125884016973126,
      "grad_norm": 4.96875,
      "learning_rate": 2.790766457387083e-06,
      "loss": 0.81588135,
      "memory(GiB)": 147.13,
      "step": 56260,
      "train_speed(iter/s)": 0.201056
    },
    {
      "acc": 0.78619328,
      "epoch": 1.3128217092696015,
      "grad_norm": 4.25,
      "learning_rate": 2.7890719108800766e-06,
      "loss": 0.75692482,
      "memory(GiB)": 147.13,
      "step": 56270,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.77736025,
      "epoch": 1.3130550168418904,
      "grad_norm": 6.0,
      "learning_rate": 2.7873776799824115e-06,
      "loss": 0.79954443,
      "memory(GiB)": 147.13,
      "step": 56280,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.8001195,
      "epoch": 1.3132883244141793,
      "grad_norm": 5.0,
      "learning_rate": 2.7856837649359416e-06,
      "loss": 0.7024128,
      "memory(GiB)": 147.13,
      "step": 56290,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.78101416,
      "epoch": 1.3135216319864682,
      "grad_norm": 6.75,
      "learning_rate": 2.7839901659824707e-06,
      "loss": 0.80548544,
      "memory(GiB)": 147.13,
      "step": 56300,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.7754415,
      "epoch": 1.313754939558757,
      "grad_norm": 5.25,
      "learning_rate": 2.7822968833637577e-06,
      "loss": 0.79988079,
      "memory(GiB)": 147.13,
      "step": 56310,
      "train_speed(iter/s)": 0.201147
    },
    {
      "acc": 0.7806459,
      "epoch": 1.313988247131046,
      "grad_norm": 4.78125,
      "learning_rate": 2.7806039173215225e-06,
      "loss": 0.78693638,
      "memory(GiB)": 147.13,
      "step": 56320,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.78040733,
      "epoch": 1.314221554703335,
      "grad_norm": 6.5625,
      "learning_rate": 2.7789112680974316e-06,
      "loss": 0.78004808,
      "memory(GiB)": 147.13,
      "step": 56330,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.80179815,
      "epoch": 1.3144548622756238,
      "grad_norm": 3.546875,
      "learning_rate": 2.7772189359331136e-06,
      "loss": 0.68882008,
      "memory(GiB)": 147.13,
      "step": 56340,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.7934073,
      "epoch": 1.3146881698479127,
      "grad_norm": 6.09375,
      "learning_rate": 2.7755269210701475e-06,
      "loss": 0.76322498,
      "memory(GiB)": 147.13,
      "step": 56350,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.78926401,
      "epoch": 1.3149214774202016,
      "grad_norm": 6.03125,
      "learning_rate": 2.7738352237500667e-06,
      "loss": 0.74684887,
      "memory(GiB)": 147.13,
      "step": 56360,
      "train_speed(iter/s)": 0.201235
    },
    {
      "acc": 0.76026993,
      "epoch": 1.3151547849924903,
      "grad_norm": 4.59375,
      "learning_rate": 2.7721438442143607e-06,
      "loss": 0.85534048,
      "memory(GiB)": 147.13,
      "step": 56370,
      "train_speed(iter/s)": 0.201253
    },
    {
      "acc": 0.77141409,
      "epoch": 1.3153880925647794,
      "grad_norm": 5.125,
      "learning_rate": 2.7704527827044714e-06,
      "loss": 0.81075706,
      "memory(GiB)": 147.13,
      "step": 56380,
      "train_speed(iter/s)": 0.201271
    },
    {
      "acc": 0.79955773,
      "epoch": 1.315621400137068,
      "grad_norm": 3.953125,
      "learning_rate": 2.7687620394618025e-06,
      "loss": 0.71365733,
      "memory(GiB)": 147.13,
      "step": 56390,
      "train_speed(iter/s)": 0.20129
    },
    {
      "acc": 0.78430367,
      "epoch": 1.3158547077093572,
      "grad_norm": 6.34375,
      "learning_rate": 2.767071614727702e-06,
      "loss": 0.75429764,
      "memory(GiB)": 147.13,
      "step": 56400,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.76308584,
      "epoch": 1.3160880152816459,
      "grad_norm": 6.65625,
      "learning_rate": 2.765381508743482e-06,
      "loss": 0.86523905,
      "memory(GiB)": 147.13,
      "step": 56410,
      "train_speed(iter/s)": 0.201326
    },
    {
      "acc": 0.78555737,
      "epoch": 1.316321322853935,
      "grad_norm": 5.65625,
      "learning_rate": 2.7636917217504007e-06,
      "loss": 0.77236929,
      "memory(GiB)": 147.13,
      "step": 56420,
      "train_speed(iter/s)": 0.201345
    },
    {
      "acc": 0.78901734,
      "epoch": 1.3165546304262237,
      "grad_norm": 4.21875,
      "learning_rate": 2.762002253989678e-06,
      "loss": 0.75951052,
      "memory(GiB)": 147.13,
      "step": 56430,
      "train_speed(iter/s)": 0.201364
    },
    {
      "acc": 0.76562176,
      "epoch": 1.3167879379985128,
      "grad_norm": 5.375,
      "learning_rate": 2.7603131057024835e-06,
      "loss": 0.84810133,
      "memory(GiB)": 147.13,
      "step": 56440,
      "train_speed(iter/s)": 0.201383
    },
    {
      "acc": 0.78627138,
      "epoch": 1.3170212455708015,
      "grad_norm": 5.125,
      "learning_rate": 2.7586242771299404e-06,
      "loss": 0.76413832,
      "memory(GiB)": 147.13,
      "step": 56450,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.78318043,
      "epoch": 1.3172545531430904,
      "grad_norm": 4.15625,
      "learning_rate": 2.7569357685131325e-06,
      "loss": 0.80174465,
      "memory(GiB)": 147.13,
      "step": 56460,
      "train_speed(iter/s)": 0.201419
    },
    {
      "acc": 0.79765291,
      "epoch": 1.3174878607153793,
      "grad_norm": 6.84375,
      "learning_rate": 2.7552475800930907e-06,
      "loss": 0.72690916,
      "memory(GiB)": 147.13,
      "step": 56470,
      "train_speed(iter/s)": 0.201438
    },
    {
      "acc": 0.79450531,
      "epoch": 1.3177211682876682,
      "grad_norm": 4.3125,
      "learning_rate": 2.753559712110808e-06,
      "loss": 0.73781071,
      "memory(GiB)": 147.13,
      "step": 56480,
      "train_speed(iter/s)": 0.201455
    },
    {
      "acc": 0.78455567,
      "epoch": 1.317954475859957,
      "grad_norm": 4.84375,
      "learning_rate": 2.75187216480722e-06,
      "loss": 0.7775054,
      "memory(GiB)": 147.13,
      "step": 56490,
      "train_speed(iter/s)": 0.201473
    },
    {
      "acc": 0.79019089,
      "epoch": 1.318187783432246,
      "grad_norm": 4.90625,
      "learning_rate": 2.75018493842323e-06,
      "loss": 0.75038633,
      "memory(GiB)": 147.13,
      "step": 56500,
      "train_speed(iter/s)": 0.201491
    },
    {
      "epoch": 1.318187783432246,
      "eval_acc": 0.7445172620268719,
      "eval_loss": 0.8049691915512085,
      "eval_runtime": 1270.3048,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 56500
    },
    {
      "acc": 0.77661104,
      "epoch": 1.3184210910045349,
      "grad_norm": 4.21875,
      "learning_rate": 2.748498033199686e-06,
      "loss": 0.79541645,
      "memory(GiB)": 147.13,
      "step": 56510,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.78313751,
      "epoch": 1.3186543985768238,
      "grad_norm": 4.90625,
      "learning_rate": 2.7468114493773913e-06,
      "loss": 0.80430899,
      "memory(GiB)": 147.13,
      "step": 56520,
      "train_speed(iter/s)": 0.200606
    },
    {
      "acc": 0.76499681,
      "epoch": 1.3188877061491127,
      "grad_norm": 5.71875,
      "learning_rate": 2.7451251871971103e-06,
      "loss": 0.85248184,
      "memory(GiB)": 147.13,
      "step": 56530,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.78148131,
      "epoch": 1.3191210137214016,
      "grad_norm": 5.09375,
      "learning_rate": 2.743439246899552e-06,
      "loss": 0.80053949,
      "memory(GiB)": 147.13,
      "step": 56540,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.78812108,
      "epoch": 1.3193543212936905,
      "grad_norm": 4.28125,
      "learning_rate": 2.7417536287253864e-06,
      "loss": 0.77857547,
      "memory(GiB)": 147.13,
      "step": 56550,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.78670983,
      "epoch": 1.3195876288659794,
      "grad_norm": 4.4375,
      "learning_rate": 2.7400683329152358e-06,
      "loss": 0.77823715,
      "memory(GiB)": 147.13,
      "step": 56560,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.78537912,
      "epoch": 1.3198209364382683,
      "grad_norm": 4.15625,
      "learning_rate": 2.738383359709671e-06,
      "loss": 0.76351776,
      "memory(GiB)": 147.13,
      "step": 56570,
      "train_speed(iter/s)": 0.200698
    },
    {
      "acc": 0.77593765,
      "epoch": 1.3200542440105572,
      "grad_norm": 9.25,
      "learning_rate": 2.736698709349227e-06,
      "loss": 0.80890398,
      "memory(GiB)": 147.13,
      "step": 56580,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.7956008,
      "epoch": 1.320287551582846,
      "grad_norm": 5.5,
      "learning_rate": 2.7350143820743847e-06,
      "loss": 0.7315671,
      "memory(GiB)": 147.13,
      "step": 56590,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.77545271,
      "epoch": 1.320520859155135,
      "grad_norm": 24.25,
      "learning_rate": 2.7333303781255816e-06,
      "loss": 0.78528709,
      "memory(GiB)": 147.13,
      "step": 56600,
      "train_speed(iter/s)": 0.20075
    },
    {
      "acc": 0.76972427,
      "epoch": 1.3207541667274239,
      "grad_norm": 9.625,
      "learning_rate": 2.7316466977432067e-06,
      "loss": 0.80668039,
      "memory(GiB)": 147.13,
      "step": 56610,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.77437797,
      "epoch": 1.3209874742997127,
      "grad_norm": 5.65625,
      "learning_rate": 2.729963341167608e-06,
      "loss": 0.82345848,
      "memory(GiB)": 147.13,
      "step": 56620,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.79564867,
      "epoch": 1.3212207818720016,
      "grad_norm": 4.9375,
      "learning_rate": 2.728280308639081e-06,
      "loss": 0.72023449,
      "memory(GiB)": 147.13,
      "step": 56630,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.77554989,
      "epoch": 1.3214540894442905,
      "grad_norm": 5.40625,
      "learning_rate": 2.7265976003978828e-06,
      "loss": 0.79516258,
      "memory(GiB)": 147.13,
      "step": 56640,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.76518278,
      "epoch": 1.3216873970165794,
      "grad_norm": 6.0625,
      "learning_rate": 2.7249152166842164e-06,
      "loss": 0.8493721,
      "memory(GiB)": 147.13,
      "step": 56650,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.78103828,
      "epoch": 1.3219207045888683,
      "grad_norm": 5.9375,
      "learning_rate": 2.72323315773824e-06,
      "loss": 0.78014269,
      "memory(GiB)": 147.13,
      "step": 56660,
      "train_speed(iter/s)": 0.200862
    },
    {
      "acc": 0.77679877,
      "epoch": 1.3221540121611572,
      "grad_norm": 4.03125,
      "learning_rate": 2.72155142380007e-06,
      "loss": 0.80789299,
      "memory(GiB)": 147.13,
      "step": 56670,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.79273858,
      "epoch": 1.3223873197334461,
      "grad_norm": 6.65625,
      "learning_rate": 2.7198700151097714e-06,
      "loss": 0.74300385,
      "memory(GiB)": 147.13,
      "step": 56680,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.76670465,
      "epoch": 1.322620627305735,
      "grad_norm": 4.90625,
      "learning_rate": 2.7181889319073674e-06,
      "loss": 0.83968229,
      "memory(GiB)": 147.13,
      "step": 56690,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.80488224,
      "epoch": 1.322853934878024,
      "grad_norm": 6.09375,
      "learning_rate": 2.7165081744328304e-06,
      "loss": 0.69187484,
      "memory(GiB)": 147.13,
      "step": 56700,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.78700209,
      "epoch": 1.3230872424503128,
      "grad_norm": 5.125,
      "learning_rate": 2.714827742926088e-06,
      "loss": 0.75780425,
      "memory(GiB)": 147.13,
      "step": 56710,
      "train_speed(iter/s)": 0.200954
    },
    {
      "acc": 0.80247622,
      "epoch": 1.3233205500226017,
      "grad_norm": 5.03125,
      "learning_rate": 2.7131476376270215e-06,
      "loss": 0.71583662,
      "memory(GiB)": 147.13,
      "step": 56720,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.78960152,
      "epoch": 1.3235538575948906,
      "grad_norm": 4.75,
      "learning_rate": 2.711467858775464e-06,
      "loss": 0.75800562,
      "memory(GiB)": 147.13,
      "step": 56730,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.77212892,
      "epoch": 1.3237871651671795,
      "grad_norm": 4.625,
      "learning_rate": 2.7097884066112062e-06,
      "loss": 0.81948633,
      "memory(GiB)": 147.13,
      "step": 56740,
      "train_speed(iter/s)": 0.201009
    },
    {
      "acc": 0.77226896,
      "epoch": 1.3240204727394684,
      "grad_norm": 4.96875,
      "learning_rate": 2.7081092813739863e-06,
      "loss": 0.81515799,
      "memory(GiB)": 147.13,
      "step": 56750,
      "train_speed(iter/s)": 0.201028
    },
    {
      "acc": 0.7703146,
      "epoch": 1.3242537803117571,
      "grad_norm": 9.4375,
      "learning_rate": 2.7064304833035027e-06,
      "loss": 0.82806835,
      "memory(GiB)": 147.13,
      "step": 56760,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.75851698,
      "epoch": 1.3244870878840462,
      "grad_norm": 9.625,
      "learning_rate": 2.704752012639399e-06,
      "loss": 0.87062492,
      "memory(GiB)": 147.13,
      "step": 56770,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.78109846,
      "epoch": 1.324720395456335,
      "grad_norm": 7.53125,
      "learning_rate": 2.703073869621281e-06,
      "loss": 0.75848188,
      "memory(GiB)": 147.13,
      "step": 56780,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.78177109,
      "epoch": 1.324953703028624,
      "grad_norm": 5.78125,
      "learning_rate": 2.7013960544887007e-06,
      "loss": 0.77262831,
      "memory(GiB)": 147.13,
      "step": 56790,
      "train_speed(iter/s)": 0.201101
    },
    {
      "acc": 0.79652271,
      "epoch": 1.3251870106009127,
      "grad_norm": 3.84375,
      "learning_rate": 2.699718567481164e-06,
      "loss": 0.72971768,
      "memory(GiB)": 147.13,
      "step": 56800,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.76938486,
      "epoch": 1.3254203181732018,
      "grad_norm": 5.4375,
      "learning_rate": 2.698041408838136e-06,
      "loss": 0.81897049,
      "memory(GiB)": 147.13,
      "step": 56810,
      "train_speed(iter/s)": 0.201139
    },
    {
      "acc": 0.796984,
      "epoch": 1.3256536257454905,
      "grad_norm": 4.21875,
      "learning_rate": 2.696364578799028e-06,
      "loss": 0.69904175,
      "memory(GiB)": 147.13,
      "step": 56820,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.79061108,
      "epoch": 1.3258869333177796,
      "grad_norm": 5.03125,
      "learning_rate": 2.694688077603207e-06,
      "loss": 0.75381279,
      "memory(GiB)": 147.13,
      "step": 56830,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.77822804,
      "epoch": 1.3261202408900683,
      "grad_norm": 4.75,
      "learning_rate": 2.6930119054899905e-06,
      "loss": 0.78712869,
      "memory(GiB)": 147.13,
      "step": 56840,
      "train_speed(iter/s)": 0.201195
    },
    {
      "acc": 0.79944668,
      "epoch": 1.3263535484623572,
      "grad_norm": 5.59375,
      "learning_rate": 2.6913360626986575e-06,
      "loss": 0.72601347,
      "memory(GiB)": 147.13,
      "step": 56850,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.78989477,
      "epoch": 1.326586856034646,
      "grad_norm": 5.34375,
      "learning_rate": 2.68966054946843e-06,
      "loss": 0.72735806,
      "memory(GiB)": 147.13,
      "step": 56860,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.78181467,
      "epoch": 1.326820163606935,
      "grad_norm": 4.96875,
      "learning_rate": 2.687985366038486e-06,
      "loss": 0.79019494,
      "memory(GiB)": 147.13,
      "step": 56870,
      "train_speed(iter/s)": 0.201251
    },
    {
      "acc": 0.76266632,
      "epoch": 1.327053471179224,
      "grad_norm": 9.0,
      "learning_rate": 2.6863105126479616e-06,
      "loss": 0.85104504,
      "memory(GiB)": 147.13,
      "step": 56880,
      "train_speed(iter/s)": 0.20127
    },
    {
      "acc": 0.77862854,
      "epoch": 1.3272867787515128,
      "grad_norm": 4.03125,
      "learning_rate": 2.6846359895359373e-06,
      "loss": 0.79643354,
      "memory(GiB)": 147.13,
      "step": 56890,
      "train_speed(iter/s)": 0.201288
    },
    {
      "acc": 0.77985401,
      "epoch": 1.3275200863238017,
      "grad_norm": 4.9375,
      "learning_rate": 2.682961796941456e-06,
      "loss": 0.78767076,
      "memory(GiB)": 147.13,
      "step": 56900,
      "train_speed(iter/s)": 0.201306
    },
    {
      "acc": 0.7925395,
      "epoch": 1.3277533938960906,
      "grad_norm": 5.0,
      "learning_rate": 2.6812879351035015e-06,
      "loss": 0.71750937,
      "memory(GiB)": 147.13,
      "step": 56910,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.80067816,
      "epoch": 1.3279867014683795,
      "grad_norm": 4.53125,
      "learning_rate": 2.679614404261023e-06,
      "loss": 0.70950232,
      "memory(GiB)": 147.13,
      "step": 56920,
      "train_speed(iter/s)": 0.201343
    },
    {
      "acc": 0.80294123,
      "epoch": 1.3282200090406684,
      "grad_norm": 9.4375,
      "learning_rate": 2.677941204652914e-06,
      "loss": 0.70604153,
      "memory(GiB)": 147.13,
      "step": 56930,
      "train_speed(iter/s)": 0.201362
    },
    {
      "acc": 0.78235593,
      "epoch": 1.3284533166129573,
      "grad_norm": 5.65625,
      "learning_rate": 2.676268336518024e-06,
      "loss": 0.77650847,
      "memory(GiB)": 147.13,
      "step": 56940,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.78522987,
      "epoch": 1.3286866241852462,
      "grad_norm": 4.34375,
      "learning_rate": 2.6745958000951546e-06,
      "loss": 0.76834965,
      "memory(GiB)": 147.13,
      "step": 56950,
      "train_speed(iter/s)": 0.2014
    },
    {
      "acc": 0.77212758,
      "epoch": 1.328919931757535,
      "grad_norm": 5.8125,
      "learning_rate": 2.672923595623056e-06,
      "loss": 0.78489156,
      "memory(GiB)": 147.13,
      "step": 56960,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.78244338,
      "epoch": 1.329153239329824,
      "grad_norm": 4.46875,
      "learning_rate": 2.67125172334044e-06,
      "loss": 0.77883863,
      "memory(GiB)": 147.13,
      "step": 56970,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.76625214,
      "epoch": 1.3293865469021129,
      "grad_norm": 5.3125,
      "learning_rate": 2.669580183485963e-06,
      "loss": 0.8450387,
      "memory(GiB)": 147.13,
      "step": 56980,
      "train_speed(iter/s)": 0.201453
    },
    {
      "acc": 0.76922102,
      "epoch": 1.3296198544744018,
      "grad_norm": 5.09375,
      "learning_rate": 2.667908976298239e-06,
      "loss": 0.81567163,
      "memory(GiB)": 147.13,
      "step": 56990,
      "train_speed(iter/s)": 0.201472
    },
    {
      "acc": 0.76809058,
      "epoch": 1.3298531620466907,
      "grad_norm": 9.375,
      "learning_rate": 2.666238102015832e-06,
      "loss": 0.83717842,
      "memory(GiB)": 147.13,
      "step": 57000,
      "train_speed(iter/s)": 0.201488
    },
    {
      "epoch": 1.3298531620466907,
      "eval_acc": 0.7444829911611593,
      "eval_loss": 0.8050174117088318,
      "eval_runtime": 1269.7274,
      "eval_samples_per_second": 28.345,
      "eval_steps_per_second": 14.173,
      "step": 57000
    },
    {
      "acc": 0.79596758,
      "epoch": 1.3300864696189796,
      "grad_norm": 6.75,
      "learning_rate": 2.6645675608772554e-06,
      "loss": 0.73099294,
      "memory(GiB)": 147.13,
      "step": 57010,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.76844263,
      "epoch": 1.3303197771912685,
      "grad_norm": 6.03125,
      "learning_rate": 2.662897353120983e-06,
      "loss": 0.82072315,
      "memory(GiB)": 147.13,
      "step": 57020,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.77861929,
      "epoch": 1.3305530847635574,
      "grad_norm": 8.375,
      "learning_rate": 2.6612274789854326e-06,
      "loss": 0.75851078,
      "memory(GiB)": 147.13,
      "step": 57030,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.77308245,
      "epoch": 1.3307863923358463,
      "grad_norm": 6.5,
      "learning_rate": 2.659557938708982e-06,
      "loss": 0.81746635,
      "memory(GiB)": 147.13,
      "step": 57040,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.79103942,
      "epoch": 1.3310196999081352,
      "grad_norm": 6.78125,
      "learning_rate": 2.657888732529956e-06,
      "loss": 0.75392189,
      "memory(GiB)": 147.13,
      "step": 57050,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.76411114,
      "epoch": 1.331253007480424,
      "grad_norm": 5.1875,
      "learning_rate": 2.656219860686633e-06,
      "loss": 0.86522112,
      "memory(GiB)": 147.13,
      "step": 57060,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.77968445,
      "epoch": 1.331486315052713,
      "grad_norm": 5.9375,
      "learning_rate": 2.6545513234172413e-06,
      "loss": 0.79840469,
      "memory(GiB)": 147.13,
      "step": 57070,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.79253349,
      "epoch": 1.3317196226250019,
      "grad_norm": 4.875,
      "learning_rate": 2.65288312095997e-06,
      "loss": 0.7571218,
      "memory(GiB)": 147.13,
      "step": 57080,
      "train_speed(iter/s)": 0.200724
    },
    {
      "acc": 0.78285084,
      "epoch": 1.3319529301972908,
      "grad_norm": 3.703125,
      "learning_rate": 2.651215253552951e-06,
      "loss": 0.77423506,
      "memory(GiB)": 147.13,
      "step": 57090,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.78827653,
      "epoch": 1.3321862377695797,
      "grad_norm": 5.8125,
      "learning_rate": 2.6495477214342704e-06,
      "loss": 0.74848604,
      "memory(GiB)": 147.13,
      "step": 57100,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.77849426,
      "epoch": 1.3324195453418686,
      "grad_norm": 5.21875,
      "learning_rate": 2.647880524841971e-06,
      "loss": 0.79774141,
      "memory(GiB)": 147.13,
      "step": 57110,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.78932242,
      "epoch": 1.3326528529141575,
      "grad_norm": 6.125,
      "learning_rate": 2.646213664014042e-06,
      "loss": 0.74397902,
      "memory(GiB)": 147.13,
      "step": 57120,
      "train_speed(iter/s)": 0.200797
    },
    {
      "acc": 0.78925648,
      "epoch": 1.3328861604864464,
      "grad_norm": 4.65625,
      "learning_rate": 2.6445471391884304e-06,
      "loss": 0.76678162,
      "memory(GiB)": 147.13,
      "step": 57130,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.78613939,
      "epoch": 1.3331194680587353,
      "grad_norm": 5.0625,
      "learning_rate": 2.6428809506030306e-06,
      "loss": 0.78517361,
      "memory(GiB)": 147.13,
      "step": 57140,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.77853289,
      "epoch": 1.333352775631024,
      "grad_norm": 5.75,
      "learning_rate": 2.641215098495688e-06,
      "loss": 0.80418797,
      "memory(GiB)": 147.13,
      "step": 57150,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.7710371,
      "epoch": 1.333586083203313,
      "grad_norm": 7.1875,
      "learning_rate": 2.639549583104209e-06,
      "loss": 0.82036028,
      "memory(GiB)": 147.13,
      "step": 57160,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.7829278,
      "epoch": 1.3338193907756017,
      "grad_norm": 6.40625,
      "learning_rate": 2.6378844046663375e-06,
      "loss": 0.78318968,
      "memory(GiB)": 147.13,
      "step": 57170,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.77414856,
      "epoch": 1.3340526983478909,
      "grad_norm": 4.59375,
      "learning_rate": 2.636219563419783e-06,
      "loss": 0.79386172,
      "memory(GiB)": 147.13,
      "step": 57180,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.78854723,
      "epoch": 1.3342860059201795,
      "grad_norm": 6.0625,
      "learning_rate": 2.6345550596021967e-06,
      "loss": 0.76046,
      "memory(GiB)": 147.13,
      "step": 57190,
      "train_speed(iter/s)": 0.200924
    },
    {
      "acc": 0.77748866,
      "epoch": 1.3345193134924687,
      "grad_norm": 5.9375,
      "learning_rate": 2.632890893451191e-06,
      "loss": 0.80022764,
      "memory(GiB)": 147.13,
      "step": 57200,
      "train_speed(iter/s)": 0.200942
    },
    {
      "acc": 0.78416834,
      "epoch": 1.3347526210647573,
      "grad_norm": 4.5,
      "learning_rate": 2.63122706520432e-06,
      "loss": 0.79278574,
      "memory(GiB)": 147.13,
      "step": 57210,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.7817687,
      "epoch": 1.3349859286370465,
      "grad_norm": 5.59375,
      "learning_rate": 2.6295635750990998e-06,
      "loss": 0.80557747,
      "memory(GiB)": 147.13,
      "step": 57220,
      "train_speed(iter/s)": 0.200979
    },
    {
      "acc": 0.7696763,
      "epoch": 1.3352192362093351,
      "grad_norm": 5.90625,
      "learning_rate": 2.627900423372991e-06,
      "loss": 0.81974831,
      "memory(GiB)": 147.13,
      "step": 57230,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.79114909,
      "epoch": 1.335452543781624,
      "grad_norm": 4.28125,
      "learning_rate": 2.626237610263406e-06,
      "loss": 0.74017973,
      "memory(GiB)": 147.13,
      "step": 57240,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.7918015,
      "epoch": 1.335685851353913,
      "grad_norm": 7.0625,
      "learning_rate": 2.6245751360077133e-06,
      "loss": 0.72015886,
      "memory(GiB)": 147.13,
      "step": 57250,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.775524,
      "epoch": 1.3359191589262018,
      "grad_norm": 7.34375,
      "learning_rate": 2.622913000843228e-06,
      "loss": 0.79100475,
      "memory(GiB)": 147.13,
      "step": 57260,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.7828311,
      "epoch": 1.3361524664984907,
      "grad_norm": 5.15625,
      "learning_rate": 2.6212512050072236e-06,
      "loss": 0.77825603,
      "memory(GiB)": 147.13,
      "step": 57270,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.79701376,
      "epoch": 1.3363857740707796,
      "grad_norm": 5.15625,
      "learning_rate": 2.6195897487369195e-06,
      "loss": 0.71078596,
      "memory(GiB)": 147.13,
      "step": 57280,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.77222824,
      "epoch": 1.3366190816430685,
      "grad_norm": 6.03125,
      "learning_rate": 2.6179286322694866e-06,
      "loss": 0.80330944,
      "memory(GiB)": 147.13,
      "step": 57290,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.7781415,
      "epoch": 1.3368523892153574,
      "grad_norm": 6.4375,
      "learning_rate": 2.6162678558420484e-06,
      "loss": 0.79362383,
      "memory(GiB)": 147.13,
      "step": 57300,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.76344051,
      "epoch": 1.3370856967876463,
      "grad_norm": 6.40625,
      "learning_rate": 2.6146074196916806e-06,
      "loss": 0.84555454,
      "memory(GiB)": 147.13,
      "step": 57310,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.79169607,
      "epoch": 1.3373190043599352,
      "grad_norm": 5.03125,
      "learning_rate": 2.6129473240554126e-06,
      "loss": 0.7247673,
      "memory(GiB)": 147.13,
      "step": 57320,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.77501116,
      "epoch": 1.3375523119322241,
      "grad_norm": 5.1875,
      "learning_rate": 2.6112875691702176e-06,
      "loss": 0.79619384,
      "memory(GiB)": 147.13,
      "step": 57330,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.79097958,
      "epoch": 1.337785619504513,
      "grad_norm": 10.5,
      "learning_rate": 2.609628155273032e-06,
      "loss": 0.75640821,
      "memory(GiB)": 147.13,
      "step": 57340,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.77276196,
      "epoch": 1.338018927076802,
      "grad_norm": 4.96875,
      "learning_rate": 2.6079690826007307e-06,
      "loss": 0.83116341,
      "memory(GiB)": 147.13,
      "step": 57350,
      "train_speed(iter/s)": 0.20122
    },
    {
      "acc": 0.77244496,
      "epoch": 1.3382522346490908,
      "grad_norm": 5.125,
      "learning_rate": 2.606310351390148e-06,
      "loss": 0.79508491,
      "memory(GiB)": 147.13,
      "step": 57360,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.78902702,
      "epoch": 1.3384855422213797,
      "grad_norm": 5.125,
      "learning_rate": 2.6046519618780673e-06,
      "loss": 0.73824391,
      "memory(GiB)": 147.13,
      "step": 57370,
      "train_speed(iter/s)": 0.201256
    },
    {
      "acc": 0.78787184,
      "epoch": 1.3387188497936686,
      "grad_norm": 6.4375,
      "learning_rate": 2.6029939143012228e-06,
      "loss": 0.75440736,
      "memory(GiB)": 147.13,
      "step": 57380,
      "train_speed(iter/s)": 0.201273
    },
    {
      "acc": 0.80366573,
      "epoch": 1.3389521573659575,
      "grad_norm": 8.125,
      "learning_rate": 2.601336208896304e-06,
      "loss": 0.69952521,
      "memory(GiB)": 147.13,
      "step": 57390,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.77663946,
      "epoch": 1.3391854649382464,
      "grad_norm": 5.5625,
      "learning_rate": 2.5996788458999404e-06,
      "loss": 0.80009689,
      "memory(GiB)": 147.13,
      "step": 57400,
      "train_speed(iter/s)": 0.20131
    },
    {
      "acc": 0.77313061,
      "epoch": 1.3394187725105353,
      "grad_norm": 5.3125,
      "learning_rate": 2.598021825548727e-06,
      "loss": 0.81944647,
      "memory(GiB)": 147.13,
      "step": 57410,
      "train_speed(iter/s)": 0.201328
    },
    {
      "acc": 0.78171992,
      "epoch": 1.3396520800828242,
      "grad_norm": 5.46875,
      "learning_rate": 2.596365148079197e-06,
      "loss": 0.77206979,
      "memory(GiB)": 147.13,
      "step": 57420,
      "train_speed(iter/s)": 0.201347
    },
    {
      "acc": 0.78008041,
      "epoch": 1.3398853876551131,
      "grad_norm": 4.46875,
      "learning_rate": 2.594708813727847e-06,
      "loss": 0.79095898,
      "memory(GiB)": 147.13,
      "step": 57430,
      "train_speed(iter/s)": 0.201365
    },
    {
      "acc": 0.79991326,
      "epoch": 1.340118695227402,
      "grad_norm": 5.78125,
      "learning_rate": 2.5930528227311148e-06,
      "loss": 0.69914985,
      "memory(GiB)": 147.13,
      "step": 57440,
      "train_speed(iter/s)": 0.201382
    },
    {
      "acc": 0.79570603,
      "epoch": 1.340352002799691,
      "grad_norm": 5.1875,
      "learning_rate": 2.591397175325391e-06,
      "loss": 0.71980028,
      "memory(GiB)": 147.13,
      "step": 57450,
      "train_speed(iter/s)": 0.201399
    },
    {
      "acc": 0.78314619,
      "epoch": 1.3405853103719798,
      "grad_norm": 5.53125,
      "learning_rate": 2.5897418717470224e-06,
      "loss": 0.77125511,
      "memory(GiB)": 147.13,
      "step": 57460,
      "train_speed(iter/s)": 0.201417
    },
    {
      "acc": 0.77202096,
      "epoch": 1.3408186179442687,
      "grad_norm": 6.9375,
      "learning_rate": 2.5880869122322994e-06,
      "loss": 0.82745008,
      "memory(GiB)": 147.13,
      "step": 57470,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.77319994,
      "epoch": 1.3410519255165576,
      "grad_norm": 5.0,
      "learning_rate": 2.5864322970174714e-06,
      "loss": 0.79846964,
      "memory(GiB)": 147.13,
      "step": 57480,
      "train_speed(iter/s)": 0.201454
    },
    {
      "acc": 0.79231257,
      "epoch": 1.3412852330888465,
      "grad_norm": 5.125,
      "learning_rate": 2.5847780263387314e-06,
      "loss": 0.73147917,
      "memory(GiB)": 147.13,
      "step": 57490,
      "train_speed(iter/s)": 0.201472
    },
    {
      "acc": 0.77735405,
      "epoch": 1.3415185406611354,
      "grad_norm": 6.53125,
      "learning_rate": 2.583124100432227e-06,
      "loss": 0.78960352,
      "memory(GiB)": 147.13,
      "step": 57500,
      "train_speed(iter/s)": 0.201491
    },
    {
      "epoch": 1.3415185406611354,
      "eval_acc": 0.7445643444311685,
      "eval_loss": 0.8049200177192688,
      "eval_runtime": 1269.9367,
      "eval_samples_per_second": 28.341,
      "eval_steps_per_second": 14.171,
      "step": 57500
    },
    {
      "acc": 0.77931428,
      "epoch": 1.3417518482334243,
      "grad_norm": 5.0625,
      "learning_rate": 2.5814705195340527e-06,
      "loss": 0.79522343,
      "memory(GiB)": 147.13,
      "step": 57510,
      "train_speed(iter/s)": 0.200603
    },
    {
      "acc": 0.77775626,
      "epoch": 1.341985155805713,
      "grad_norm": 5.3125,
      "learning_rate": 2.5798172838802616e-06,
      "loss": 0.79539862,
      "memory(GiB)": 147.13,
      "step": 57520,
      "train_speed(iter/s)": 0.20062
    },
    {
      "acc": 0.7882154,
      "epoch": 1.342218463378002,
      "grad_norm": 5.75,
      "learning_rate": 2.5781643937068495e-06,
      "loss": 0.75900679,
      "memory(GiB)": 147.13,
      "step": 57530,
      "train_speed(iter/s)": 0.200638
    },
    {
      "acc": 0.78802242,
      "epoch": 1.3424517709502908,
      "grad_norm": 4.3125,
      "learning_rate": 2.5765118492497654e-06,
      "loss": 0.75461683,
      "memory(GiB)": 147.13,
      "step": 57540,
      "train_speed(iter/s)": 0.200656
    },
    {
      "acc": 0.79044104,
      "epoch": 1.34268507852258,
      "grad_norm": 5.40625,
      "learning_rate": 2.5748596507449118e-06,
      "loss": 0.75172663,
      "memory(GiB)": 147.13,
      "step": 57550,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.78111725,
      "epoch": 1.3429183860948686,
      "grad_norm": 6.03125,
      "learning_rate": 2.5732077984281378e-06,
      "loss": 0.76782961,
      "memory(GiB)": 147.13,
      "step": 57560,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.78649011,
      "epoch": 1.3431516936671577,
      "grad_norm": 6.0,
      "learning_rate": 2.571556292535247e-06,
      "loss": 0.77026129,
      "memory(GiB)": 147.13,
      "step": 57570,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.79622579,
      "epoch": 1.3433850012394464,
      "grad_norm": 4.0625,
      "learning_rate": 2.5699051333019897e-06,
      "loss": 0.74895239,
      "memory(GiB)": 147.13,
      "step": 57580,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.77864823,
      "epoch": 1.3436183088117355,
      "grad_norm": 5.0,
      "learning_rate": 2.568254320964067e-06,
      "loss": 0.78803215,
      "memory(GiB)": 147.13,
      "step": 57590,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.776053,
      "epoch": 1.3438516163840242,
      "grad_norm": 6.46875,
      "learning_rate": 2.5666038557571355e-06,
      "loss": 0.81020985,
      "memory(GiB)": 147.13,
      "step": 57600,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.78326931,
      "epoch": 1.3440849239563133,
      "grad_norm": 3.53125,
      "learning_rate": 2.5649537379167944e-06,
      "loss": 0.78707523,
      "memory(GiB)": 147.13,
      "step": 57610,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.76431599,
      "epoch": 1.344318231528602,
      "grad_norm": 5.5,
      "learning_rate": 2.5633039676786044e-06,
      "loss": 0.8433094,
      "memory(GiB)": 147.13,
      "step": 57620,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.78314161,
      "epoch": 1.3445515391008909,
      "grad_norm": 4.71875,
      "learning_rate": 2.5616545452780607e-06,
      "loss": 0.76809273,
      "memory(GiB)": 147.13,
      "step": 57630,
      "train_speed(iter/s)": 0.200819
    },
    {
      "acc": 0.78058529,
      "epoch": 1.3447848466731798,
      "grad_norm": 5.28125,
      "learning_rate": 2.5600054709506244e-06,
      "loss": 0.78540258,
      "memory(GiB)": 147.13,
      "step": 57640,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.77517385,
      "epoch": 1.3450181542454687,
      "grad_norm": 6.75,
      "learning_rate": 2.5583567449316983e-06,
      "loss": 0.7901629,
      "memory(GiB)": 147.13,
      "step": 57650,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.75654125,
      "epoch": 1.3452514618177576,
      "grad_norm": 5.28125,
      "learning_rate": 2.5567083674566363e-06,
      "loss": 0.87824078,
      "memory(GiB)": 147.13,
      "step": 57660,
      "train_speed(iter/s)": 0.200873
    },
    {
      "acc": 0.79099903,
      "epoch": 1.3454847693900465,
      "grad_norm": 4.71875,
      "learning_rate": 2.555060338760746e-06,
      "loss": 0.74126825,
      "memory(GiB)": 147.13,
      "step": 57670,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.77133732,
      "epoch": 1.3457180769623354,
      "grad_norm": 5.5,
      "learning_rate": 2.553412659079281e-06,
      "loss": 0.83065491,
      "memory(GiB)": 147.13,
      "step": 57680,
      "train_speed(iter/s)": 0.200907
    },
    {
      "acc": 0.77394104,
      "epoch": 1.3459513845346243,
      "grad_norm": 4.5625,
      "learning_rate": 2.5517653286474486e-06,
      "loss": 0.7871973,
      "memory(GiB)": 147.13,
      "step": 57690,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.77455368,
      "epoch": 1.3461846921069132,
      "grad_norm": 7.375,
      "learning_rate": 2.5501183477004036e-06,
      "loss": 0.81267767,
      "memory(GiB)": 147.13,
      "step": 57700,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.78425789,
      "epoch": 1.346417999679202,
      "grad_norm": 6.65625,
      "learning_rate": 2.548471716473255e-06,
      "loss": 0.75778885,
      "memory(GiB)": 147.13,
      "step": 57710,
      "train_speed(iter/s)": 0.200963
    },
    {
      "acc": 0.77946496,
      "epoch": 1.346651307251491,
      "grad_norm": 5.5625,
      "learning_rate": 2.546825435201056e-06,
      "loss": 0.79724245,
      "memory(GiB)": 147.13,
      "step": 57720,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.76264691,
      "epoch": 1.3468846148237799,
      "grad_norm": 6.46875,
      "learning_rate": 2.5451795041188137e-06,
      "loss": 0.84872093,
      "memory(GiB)": 147.13,
      "step": 57730,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.78879805,
      "epoch": 1.3471179223960688,
      "grad_norm": 6.75,
      "learning_rate": 2.543533923461484e-06,
      "loss": 0.76141896,
      "memory(GiB)": 147.13,
      "step": 57740,
      "train_speed(iter/s)": 0.201016
    },
    {
      "acc": 0.78948154,
      "epoch": 1.3473512299683577,
      "grad_norm": 4.09375,
      "learning_rate": 2.541888693463971e-06,
      "loss": 0.73227172,
      "memory(GiB)": 147.13,
      "step": 57750,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.77823572,
      "epoch": 1.3475845375406466,
      "grad_norm": 4.9375,
      "learning_rate": 2.540243814361135e-06,
      "loss": 0.80870161,
      "memory(GiB)": 147.13,
      "step": 57760,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.77902465,
      "epoch": 1.3478178451129355,
      "grad_norm": 7.4375,
      "learning_rate": 2.5385992863877783e-06,
      "loss": 0.80094547,
      "memory(GiB)": 147.13,
      "step": 57770,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.77547317,
      "epoch": 1.3480511526852244,
      "grad_norm": 7.125,
      "learning_rate": 2.5369551097786606e-06,
      "loss": 0.79117107,
      "memory(GiB)": 147.13,
      "step": 57780,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.78636332,
      "epoch": 1.3482844602575133,
      "grad_norm": 4.75,
      "learning_rate": 2.5353112847684846e-06,
      "loss": 0.77924938,
      "memory(GiB)": 147.13,
      "step": 57790,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.79167924,
      "epoch": 1.3485177678298021,
      "grad_norm": 6.71875,
      "learning_rate": 2.5336678115919056e-06,
      "loss": 0.7614068,
      "memory(GiB)": 147.13,
      "step": 57800,
      "train_speed(iter/s)": 0.201122
    },
    {
      "acc": 0.773736,
      "epoch": 1.348751075402091,
      "grad_norm": 5.40625,
      "learning_rate": 2.532024690483531e-06,
      "loss": 0.81560125,
      "memory(GiB)": 147.13,
      "step": 57810,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.78321466,
      "epoch": 1.34898438297438,
      "grad_norm": 6.5,
      "learning_rate": 2.5303819216779134e-06,
      "loss": 0.76348276,
      "memory(GiB)": 147.13,
      "step": 57820,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.76863384,
      "epoch": 1.3492176905466688,
      "grad_norm": 5.15625,
      "learning_rate": 2.528739505409561e-06,
      "loss": 0.84750042,
      "memory(GiB)": 147.13,
      "step": 57830,
      "train_speed(iter/s)": 0.201177
    },
    {
      "acc": 0.79967451,
      "epoch": 1.3494509981189577,
      "grad_norm": 4.59375,
      "learning_rate": 2.5270974419129248e-06,
      "loss": 0.71103725,
      "memory(GiB)": 147.13,
      "step": 57840,
      "train_speed(iter/s)": 0.201194
    },
    {
      "acc": 0.77464585,
      "epoch": 1.3496843056912466,
      "grad_norm": 9.125,
      "learning_rate": 2.525455731422414e-06,
      "loss": 0.8018692,
      "memory(GiB)": 147.13,
      "step": 57850,
      "train_speed(iter/s)": 0.201213
    },
    {
      "acc": 0.79444895,
      "epoch": 1.3499176132635355,
      "grad_norm": 5.0,
      "learning_rate": 2.5238143741723743e-06,
      "loss": 0.72763662,
      "memory(GiB)": 147.13,
      "step": 57860,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.78565588,
      "epoch": 1.3501509208358244,
      "grad_norm": 6.0,
      "learning_rate": 2.5221733703971165e-06,
      "loss": 0.76235085,
      "memory(GiB)": 147.13,
      "step": 57870,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.78204556,
      "epoch": 1.3503842284081133,
      "grad_norm": 7.15625,
      "learning_rate": 2.5205327203308887e-06,
      "loss": 0.76153746,
      "memory(GiB)": 147.13,
      "step": 57880,
      "train_speed(iter/s)": 0.201268
    },
    {
      "acc": 0.78484373,
      "epoch": 1.3506175359804022,
      "grad_norm": 5.15625,
      "learning_rate": 2.518892424207894e-06,
      "loss": 0.77373028,
      "memory(GiB)": 147.13,
      "step": 57890,
      "train_speed(iter/s)": 0.201285
    },
    {
      "acc": 0.7609642,
      "epoch": 1.3508508435526911,
      "grad_norm": 6.65625,
      "learning_rate": 2.517252482262286e-06,
      "loss": 0.85898237,
      "memory(GiB)": 147.13,
      "step": 57900,
      "train_speed(iter/s)": 0.201304
    },
    {
      "acc": 0.78211231,
      "epoch": 1.3510841511249798,
      "grad_norm": 4.5625,
      "learning_rate": 2.515612894728164e-06,
      "loss": 0.78129263,
      "memory(GiB)": 147.13,
      "step": 57910,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.77921953,
      "epoch": 1.351317458697269,
      "grad_norm": 6.5,
      "learning_rate": 2.5139736618395804e-06,
      "loss": 0.7952693,
      "memory(GiB)": 147.13,
      "step": 57920,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.78653154,
      "epoch": 1.3515507662695576,
      "grad_norm": 6.09375,
      "learning_rate": 2.5123347838305354e-06,
      "loss": 0.75747843,
      "memory(GiB)": 147.13,
      "step": 57930,
      "train_speed(iter/s)": 0.201359
    },
    {
      "acc": 0.77528124,
      "epoch": 1.3517840738418467,
      "grad_norm": 6.96875,
      "learning_rate": 2.510696260934975e-06,
      "loss": 0.80986633,
      "memory(GiB)": 147.13,
      "step": 57940,
      "train_speed(iter/s)": 0.201375
    },
    {
      "acc": 0.77857499,
      "epoch": 1.3520173814141354,
      "grad_norm": 6.21875,
      "learning_rate": 2.509058093386802e-06,
      "loss": 0.79531021,
      "memory(GiB)": 147.13,
      "step": 57950,
      "train_speed(iter/s)": 0.201393
    },
    {
      "acc": 0.75173092,
      "epoch": 1.3522506889864245,
      "grad_norm": 5.0625,
      "learning_rate": 2.507420281419862e-06,
      "loss": 0.89647017,
      "memory(GiB)": 147.13,
      "step": 57960,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.77753296,
      "epoch": 1.3524839965587132,
      "grad_norm": 4.90625,
      "learning_rate": 2.505782825267954e-06,
      "loss": 0.79882941,
      "memory(GiB)": 147.13,
      "step": 57970,
      "train_speed(iter/s)": 0.201428
    },
    {
      "acc": 0.77016959,
      "epoch": 1.3527173041310023,
      "grad_norm": 4.5625,
      "learning_rate": 2.5041457251648204e-06,
      "loss": 0.84093132,
      "memory(GiB)": 147.13,
      "step": 57980,
      "train_speed(iter/s)": 0.201445
    },
    {
      "acc": 0.78002801,
      "epoch": 1.352950611703291,
      "grad_norm": 5.0,
      "learning_rate": 2.502508981344162e-06,
      "loss": 0.78813906,
      "memory(GiB)": 147.13,
      "step": 57990,
      "train_speed(iter/s)": 0.201463
    },
    {
      "acc": 0.78480358,
      "epoch": 1.35318391927558,
      "grad_norm": 5.65625,
      "learning_rate": 2.5008725940396182e-06,
      "loss": 0.76581554,
      "memory(GiB)": 147.13,
      "step": 58000,
      "train_speed(iter/s)": 0.20148
    },
    {
      "epoch": 1.35318391927558,
      "eval_acc": 0.7446276014029275,
      "eval_loss": 0.8048146963119507,
      "eval_runtime": 1270.2313,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "step": 58000
    },
    {
      "acc": 0.7872107,
      "epoch": 1.3534172268478688,
      "grad_norm": 6.875,
      "learning_rate": 2.499236563484788e-06,
      "loss": 0.75764198,
      "memory(GiB)": 147.13,
      "step": 58010,
      "train_speed(iter/s)": 0.2006
    },
    {
      "acc": 0.79398293,
      "epoch": 1.3536505344201577,
      "grad_norm": 4.78125,
      "learning_rate": 2.4976008899132122e-06,
      "loss": 0.73490696,
      "memory(GiB)": 147.13,
      "step": 58020,
      "train_speed(iter/s)": 0.200618
    },
    {
      "acc": 0.78950262,
      "epoch": 1.3538838419924466,
      "grad_norm": 4.96875,
      "learning_rate": 2.49596557355838e-06,
      "loss": 0.73174076,
      "memory(GiB)": 147.13,
      "step": 58030,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.78074455,
      "epoch": 1.3541171495647355,
      "grad_norm": 6.6875,
      "learning_rate": 2.4943306146537365e-06,
      "loss": 0.79538136,
      "memory(GiB)": 147.13,
      "step": 58040,
      "train_speed(iter/s)": 0.200655
    },
    {
      "acc": 0.77637463,
      "epoch": 1.3543504571370244,
      "grad_norm": 8.75,
      "learning_rate": 2.4926960134326684e-06,
      "loss": 0.79754944,
      "memory(GiB)": 147.13,
      "step": 58050,
      "train_speed(iter/s)": 0.200673
    },
    {
      "acc": 0.78122358,
      "epoch": 1.3545837647093133,
      "grad_norm": 6.0625,
      "learning_rate": 2.491061770128518e-06,
      "loss": 0.77927294,
      "memory(GiB)": 147.13,
      "step": 58060,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.77235141,
      "epoch": 1.3548170722816022,
      "grad_norm": 5.09375,
      "learning_rate": 2.4894278849745705e-06,
      "loss": 0.81898794,
      "memory(GiB)": 147.13,
      "step": 58070,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.77147989,
      "epoch": 1.355050379853891,
      "grad_norm": 7.71875,
      "learning_rate": 2.4877943582040636e-06,
      "loss": 0.81207123,
      "memory(GiB)": 147.13,
      "step": 58080,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.79436178,
      "epoch": 1.35528368742618,
      "grad_norm": 4.4375,
      "learning_rate": 2.486161190050182e-06,
      "loss": 0.73931255,
      "memory(GiB)": 147.13,
      "step": 58090,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.78587637,
      "epoch": 1.355516994998469,
      "grad_norm": 5.8125,
      "learning_rate": 2.4845283807460587e-06,
      "loss": 0.76380043,
      "memory(GiB)": 147.13,
      "step": 58100,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.79440575,
      "epoch": 1.3557503025707578,
      "grad_norm": 5.21875,
      "learning_rate": 2.4828959305247795e-06,
      "loss": 0.74594207,
      "memory(GiB)": 147.13,
      "step": 58110,
      "train_speed(iter/s)": 0.20078
    },
    {
      "acc": 0.79074926,
      "epoch": 1.3559836101430467,
      "grad_norm": 5.0625,
      "learning_rate": 2.4812638396193734e-06,
      "loss": 0.76517062,
      "memory(GiB)": 147.13,
      "step": 58120,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.7802947,
      "epoch": 1.3562169177153356,
      "grad_norm": 4.6875,
      "learning_rate": 2.479632108262825e-06,
      "loss": 0.77691612,
      "memory(GiB)": 147.13,
      "step": 58130,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.78983259,
      "epoch": 1.3564502252876245,
      "grad_norm": 4.90625,
      "learning_rate": 2.4780007366880584e-06,
      "loss": 0.75796638,
      "memory(GiB)": 147.13,
      "step": 58140,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.77917747,
      "epoch": 1.3566835328599134,
      "grad_norm": 3.953125,
      "learning_rate": 2.476369725127956e-06,
      "loss": 0.79284611,
      "memory(GiB)": 147.13,
      "step": 58150,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.76637774,
      "epoch": 1.3569168404322023,
      "grad_norm": 5.34375,
      "learning_rate": 2.474739073815342e-06,
      "loss": 0.84874744,
      "memory(GiB)": 147.13,
      "step": 58160,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.78275561,
      "epoch": 1.3571501480044912,
      "grad_norm": 4.59375,
      "learning_rate": 2.47310878298299e-06,
      "loss": 0.79091311,
      "memory(GiB)": 147.13,
      "step": 58170,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.78089781,
      "epoch": 1.35738345557678,
      "grad_norm": 5.5,
      "learning_rate": 2.4714788528636275e-06,
      "loss": 0.77343168,
      "memory(GiB)": 147.13,
      "step": 58180,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.77175679,
      "epoch": 1.357616763149069,
      "grad_norm": 4.625,
      "learning_rate": 2.4698492836899234e-06,
      "loss": 0.80777512,
      "memory(GiB)": 147.13,
      "step": 58190,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.78164663,
      "epoch": 1.3578500707213579,
      "grad_norm": 4.8125,
      "learning_rate": 2.4682200756944997e-06,
      "loss": 0.7826436,
      "memory(GiB)": 147.13,
      "step": 58200,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.79077911,
      "epoch": 1.3580833782936468,
      "grad_norm": 6.03125,
      "learning_rate": 2.4665912291099225e-06,
      "loss": 0.74198914,
      "memory(GiB)": 147.13,
      "step": 58210,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.78291769,
      "epoch": 1.3583166858659357,
      "grad_norm": 5.59375,
      "learning_rate": 2.4649627441687134e-06,
      "loss": 0.79139719,
      "memory(GiB)": 147.13,
      "step": 58220,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.77830348,
      "epoch": 1.3585499934382246,
      "grad_norm": 5.40625,
      "learning_rate": 2.463334621103336e-06,
      "loss": 0.7960639,
      "memory(GiB)": 147.13,
      "step": 58230,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.77357435,
      "epoch": 1.3587833010105135,
      "grad_norm": 6.53125,
      "learning_rate": 2.461706860146203e-06,
      "loss": 0.81305132,
      "memory(GiB)": 147.13,
      "step": 58240,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.76329145,
      "epoch": 1.3590166085828024,
      "grad_norm": 6.3125,
      "learning_rate": 2.4600794615296797e-06,
      "loss": 0.85375662,
      "memory(GiB)": 147.13,
      "step": 58250,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.78082247,
      "epoch": 1.3592499161550913,
      "grad_norm": 17.125,
      "learning_rate": 2.4584524254860736e-06,
      "loss": 0.78417535,
      "memory(GiB)": 147.13,
      "step": 58260,
      "train_speed(iter/s)": 0.201043
    },
    {
      "acc": 0.77918062,
      "epoch": 1.3594832237273802,
      "grad_norm": 5.0,
      "learning_rate": 2.4568257522476476e-06,
      "loss": 0.80585155,
      "memory(GiB)": 147.13,
      "step": 58270,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.76542234,
      "epoch": 1.359716531299669,
      "grad_norm": 4.34375,
      "learning_rate": 2.455199442046607e-06,
      "loss": 0.83563366,
      "memory(GiB)": 147.13,
      "step": 58280,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.80455227,
      "epoch": 1.359949838871958,
      "grad_norm": 5.25,
      "learning_rate": 2.453573495115104e-06,
      "loss": 0.68476171,
      "memory(GiB)": 147.13,
      "step": 58290,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.80062113,
      "epoch": 1.3601831464442466,
      "grad_norm": 4.21875,
      "learning_rate": 2.4519479116852476e-06,
      "loss": 0.7182889,
      "memory(GiB)": 147.13,
      "step": 58300,
      "train_speed(iter/s)": 0.201116
    },
    {
      "acc": 0.78591204,
      "epoch": 1.3604164540165358,
      "grad_norm": 5.71875,
      "learning_rate": 2.450322691989086e-06,
      "loss": 0.7478971,
      "memory(GiB)": 147.13,
      "step": 58310,
      "train_speed(iter/s)": 0.201134
    },
    {
      "acc": 0.78090954,
      "epoch": 1.3606497615888244,
      "grad_norm": 6.5625,
      "learning_rate": 2.4486978362586196e-06,
      "loss": 0.79978738,
      "memory(GiB)": 147.13,
      "step": 58320,
      "train_speed(iter/s)": 0.201151
    },
    {
      "acc": 0.77255669,
      "epoch": 1.3608830691611136,
      "grad_norm": 4.96875,
      "learning_rate": 2.447073344725794e-06,
      "loss": 0.80740366,
      "memory(GiB)": 147.13,
      "step": 58330,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.77630148,
      "epoch": 1.3611163767334022,
      "grad_norm": 6.34375,
      "learning_rate": 2.4454492176225087e-06,
      "loss": 0.79302568,
      "memory(GiB)": 147.13,
      "step": 58340,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.78025756,
      "epoch": 1.3613496843056914,
      "grad_norm": 5.34375,
      "learning_rate": 2.4438254551806034e-06,
      "loss": 0.77262883,
      "memory(GiB)": 147.13,
      "step": 58350,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.78731136,
      "epoch": 1.36158299187798,
      "grad_norm": 4.71875,
      "learning_rate": 2.4422020576318737e-06,
      "loss": 0.75996304,
      "memory(GiB)": 147.13,
      "step": 58360,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.75691671,
      "epoch": 1.3618162994502692,
      "grad_norm": 6.125,
      "learning_rate": 2.4405790252080576e-06,
      "loss": 0.90783367,
      "memory(GiB)": 147.13,
      "step": 58370,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.78348799,
      "epoch": 1.3620496070225578,
      "grad_norm": 6.125,
      "learning_rate": 2.4389563581408397e-06,
      "loss": 0.79257278,
      "memory(GiB)": 147.13,
      "step": 58380,
      "train_speed(iter/s)": 0.201255
    },
    {
      "acc": 0.76404047,
      "epoch": 1.3622829145948467,
      "grad_norm": 7.3125,
      "learning_rate": 2.4373340566618603e-06,
      "loss": 0.85965233,
      "memory(GiB)": 147.13,
      "step": 58390,
      "train_speed(iter/s)": 0.201273
    },
    {
      "acc": 0.77306118,
      "epoch": 1.3625162221671356,
      "grad_norm": 5.21875,
      "learning_rate": 2.435712121002698e-06,
      "loss": 0.82709122,
      "memory(GiB)": 147.13,
      "step": 58400,
      "train_speed(iter/s)": 0.20129
    },
    {
      "acc": 0.78636055,
      "epoch": 1.3627495297394245,
      "grad_norm": 4.59375,
      "learning_rate": 2.4340905513948866e-06,
      "loss": 0.76273184,
      "memory(GiB)": 147.13,
      "step": 58410,
      "train_speed(iter/s)": 0.201308
    },
    {
      "acc": 0.79748254,
      "epoch": 1.3629828373117134,
      "grad_norm": 4.34375,
      "learning_rate": 2.432469348069904e-06,
      "loss": 0.73208895,
      "memory(GiB)": 147.13,
      "step": 58420,
      "train_speed(iter/s)": 0.201323
    },
    {
      "acc": 0.79371405,
      "epoch": 1.3632161448840023,
      "grad_norm": 5.9375,
      "learning_rate": 2.4308485112591764e-06,
      "loss": 0.73580298,
      "memory(GiB)": 147.13,
      "step": 58430,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.79239931,
      "epoch": 1.3634494524562912,
      "grad_norm": 6.84375,
      "learning_rate": 2.429228041194077e-06,
      "loss": 0.74616742,
      "memory(GiB)": 147.13,
      "step": 58440,
      "train_speed(iter/s)": 0.201358
    },
    {
      "acc": 0.75693569,
      "epoch": 1.3636827600285801,
      "grad_norm": 4.5,
      "learning_rate": 2.4276079381059258e-06,
      "loss": 0.89039459,
      "memory(GiB)": 147.13,
      "step": 58450,
      "train_speed(iter/s)": 0.201376
    },
    {
      "acc": 0.76860232,
      "epoch": 1.363916067600869,
      "grad_norm": 7.15625,
      "learning_rate": 2.4259882022259968e-06,
      "loss": 0.82961311,
      "memory(GiB)": 147.13,
      "step": 58460,
      "train_speed(iter/s)": 0.201394
    },
    {
      "acc": 0.79684687,
      "epoch": 1.364149375173158,
      "grad_norm": 5.25,
      "learning_rate": 2.424368833785502e-06,
      "loss": 0.72636261,
      "memory(GiB)": 147.13,
      "step": 58470,
      "train_speed(iter/s)": 0.201411
    },
    {
      "acc": 0.77321219,
      "epoch": 1.3643826827454468,
      "grad_norm": 5.125,
      "learning_rate": 2.4227498330156095e-06,
      "loss": 0.82248058,
      "memory(GiB)": 147.13,
      "step": 58480,
      "train_speed(iter/s)": 0.20143
    },
    {
      "acc": 0.77851667,
      "epoch": 1.3646159903177357,
      "grad_norm": 5.4375,
      "learning_rate": 2.421131200147428e-06,
      "loss": 0.81334505,
      "memory(GiB)": 147.13,
      "step": 58490,
      "train_speed(iter/s)": 0.201447
    },
    {
      "acc": 0.78208332,
      "epoch": 1.3648492978900246,
      "grad_norm": 5.15625,
      "learning_rate": 2.4195129354120204e-06,
      "loss": 0.78902369,
      "memory(GiB)": 147.13,
      "step": 58500,
      "train_speed(iter/s)": 0.201465
    },
    {
      "epoch": 1.3648492978900246,
      "eval_acc": 0.7446186333259187,
      "eval_loss": 0.8047336339950562,
      "eval_runtime": 1268.6994,
      "eval_samples_per_second": 28.368,
      "eval_steps_per_second": 14.185,
      "step": 58500
    },
    {
      "acc": 0.78909907,
      "epoch": 1.3650826054623135,
      "grad_norm": 4.84375,
      "learning_rate": 2.4178950390403917e-06,
      "loss": 0.74475222,
      "memory(GiB)": 147.13,
      "step": 58510,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.75878892,
      "epoch": 1.3653159130346024,
      "grad_norm": 7.4375,
      "learning_rate": 2.416277511263494e-06,
      "loss": 0.86468029,
      "memory(GiB)": 147.13,
      "step": 58520,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.77081804,
      "epoch": 1.3655492206068913,
      "grad_norm": 4.6875,
      "learning_rate": 2.4146603523122347e-06,
      "loss": 0.83188496,
      "memory(GiB)": 147.13,
      "step": 58530,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.78759122,
      "epoch": 1.3657825281791802,
      "grad_norm": 6.4375,
      "learning_rate": 2.413043562417456e-06,
      "loss": 0.7472733,
      "memory(GiB)": 147.13,
      "step": 58540,
      "train_speed(iter/s)": 0.200647
    },
    {
      "acc": 0.79054499,
      "epoch": 1.3660158357514691,
      "grad_norm": 6.25,
      "learning_rate": 2.4114271418099583e-06,
      "loss": 0.77312899,
      "memory(GiB)": 147.13,
      "step": 58550,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.76001344,
      "epoch": 1.366249143323758,
      "grad_norm": 5.6875,
      "learning_rate": 2.4098110907204824e-06,
      "loss": 0.86084118,
      "memory(GiB)": 147.13,
      "step": 58560,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.7808567,
      "epoch": 1.366482450896047,
      "grad_norm": 5.0,
      "learning_rate": 2.4081954093797234e-06,
      "loss": 0.81140366,
      "memory(GiB)": 147.13,
      "step": 58570,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.78830757,
      "epoch": 1.3667157584683358,
      "grad_norm": 5.375,
      "learning_rate": 2.406580098018316e-06,
      "loss": 0.74808269,
      "memory(GiB)": 147.13,
      "step": 58580,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.78679028,
      "epoch": 1.3669490660406247,
      "grad_norm": 5.84375,
      "learning_rate": 2.4049651568668447e-06,
      "loss": 0.76290827,
      "memory(GiB)": 147.13,
      "step": 58590,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.77562418,
      "epoch": 1.3671823736129136,
      "grad_norm": 5.9375,
      "learning_rate": 2.403350586155845e-06,
      "loss": 0.82148495,
      "memory(GiB)": 147.13,
      "step": 58600,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.76451578,
      "epoch": 1.3674156811852025,
      "grad_norm": 5.46875,
      "learning_rate": 2.4017363861157927e-06,
      "loss": 0.85645256,
      "memory(GiB)": 147.13,
      "step": 58610,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.77976284,
      "epoch": 1.3676489887574914,
      "grad_norm": 5.96875,
      "learning_rate": 2.400122556977119e-06,
      "loss": 0.79148855,
      "memory(GiB)": 147.13,
      "step": 58620,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.77969866,
      "epoch": 1.3678822963297803,
      "grad_norm": 3.96875,
      "learning_rate": 2.398509098970193e-06,
      "loss": 0.77951374,
      "memory(GiB)": 147.13,
      "step": 58630,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.78442101,
      "epoch": 1.3681156039020692,
      "grad_norm": 4.0625,
      "learning_rate": 2.3968960123253392e-06,
      "loss": 0.76123886,
      "memory(GiB)": 147.13,
      "step": 58640,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.7682898,
      "epoch": 1.368348911474358,
      "grad_norm": 7.375,
      "learning_rate": 2.3952832972728234e-06,
      "loss": 0.83347187,
      "memory(GiB)": 147.13,
      "step": 58650,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.78368082,
      "epoch": 1.368582219046647,
      "grad_norm": 5.5625,
      "learning_rate": 2.39367095404286e-06,
      "loss": 0.78032255,
      "memory(GiB)": 147.13,
      "step": 58660,
      "train_speed(iter/s)": 0.200866
    },
    {
      "acc": 0.7843792,
      "epoch": 1.368815526618936,
      "grad_norm": 4.53125,
      "learning_rate": 2.392058982865611e-06,
      "loss": 0.76798601,
      "memory(GiB)": 147.13,
      "step": 58670,
      "train_speed(iter/s)": 0.200883
    },
    {
      "acc": 0.77215185,
      "epoch": 1.3690488341912248,
      "grad_norm": 7.34375,
      "learning_rate": 2.3904473839711826e-06,
      "loss": 0.81567993,
      "memory(GiB)": 147.13,
      "step": 58680,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.79290199,
      "epoch": 1.3692821417635135,
      "grad_norm": 5.96875,
      "learning_rate": 2.388836157589634e-06,
      "loss": 0.75185156,
      "memory(GiB)": 147.13,
      "step": 58690,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.78097115,
      "epoch": 1.3695154493358026,
      "grad_norm": 5.5,
      "learning_rate": 2.3872253039509637e-06,
      "loss": 0.78168497,
      "memory(GiB)": 147.13,
      "step": 58700,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.76524758,
      "epoch": 1.3697487569080913,
      "grad_norm": 6.3125,
      "learning_rate": 2.3856148232851237e-06,
      "loss": 0.90382605,
      "memory(GiB)": 147.13,
      "step": 58710,
      "train_speed(iter/s)": 0.200955
    },
    {
      "acc": 0.78163342,
      "epoch": 1.3699820644803804,
      "grad_norm": 5.53125,
      "learning_rate": 2.384004715822009e-06,
      "loss": 0.78530493,
      "memory(GiB)": 147.13,
      "step": 58720,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.78916397,
      "epoch": 1.370215372052669,
      "grad_norm": 9.8125,
      "learning_rate": 2.3823949817914584e-06,
      "loss": 0.75205698,
      "memory(GiB)": 147.13,
      "step": 58730,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.79455929,
      "epoch": 1.3704486796249582,
      "grad_norm": 6.84375,
      "learning_rate": 2.380785621423266e-06,
      "loss": 0.72274475,
      "memory(GiB)": 147.13,
      "step": 58740,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78989534,
      "epoch": 1.3706819871972469,
      "grad_norm": 6.0625,
      "learning_rate": 2.379176634947163e-06,
      "loss": 0.7328239,
      "memory(GiB)": 147.13,
      "step": 58750,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.77720261,
      "epoch": 1.370915294769536,
      "grad_norm": 5.75,
      "learning_rate": 2.377568022592838e-06,
      "loss": 0.79559016,
      "memory(GiB)": 147.13,
      "step": 58760,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.77461166,
      "epoch": 1.3711486023418247,
      "grad_norm": 4.15625,
      "learning_rate": 2.3759597845899123e-06,
      "loss": 0.79362354,
      "memory(GiB)": 147.13,
      "step": 58770,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.78339438,
      "epoch": 1.3713819099141136,
      "grad_norm": 9.4375,
      "learning_rate": 2.374351921167967e-06,
      "loss": 0.74716663,
      "memory(GiB)": 147.13,
      "step": 58780,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78279791,
      "epoch": 1.3716152174864025,
      "grad_norm": 5.34375,
      "learning_rate": 2.37274443255652e-06,
      "loss": 0.80144253,
      "memory(GiB)": 147.13,
      "step": 58790,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.78304019,
      "epoch": 1.3718485250586914,
      "grad_norm": 4.53125,
      "learning_rate": 2.3711373189850444e-06,
      "loss": 0.76427855,
      "memory(GiB)": 147.13,
      "step": 58800,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.76471519,
      "epoch": 1.3720818326309803,
      "grad_norm": 5.25,
      "learning_rate": 2.369530580682953e-06,
      "loss": 0.83371811,
      "memory(GiB)": 147.13,
      "step": 58810,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.7774107,
      "epoch": 1.3723151402032692,
      "grad_norm": 5.28125,
      "learning_rate": 2.367924217879604e-06,
      "loss": 0.80156479,
      "memory(GiB)": 147.13,
      "step": 58820,
      "train_speed(iter/s)": 0.201146
    },
    {
      "acc": 0.78052721,
      "epoch": 1.372548447775558,
      "grad_norm": 5.0625,
      "learning_rate": 2.3663182308043115e-06,
      "loss": 0.79353514,
      "memory(GiB)": 147.13,
      "step": 58830,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.77628222,
      "epoch": 1.372781755347847,
      "grad_norm": 6.34375,
      "learning_rate": 2.3647126196863234e-06,
      "loss": 0.78026848,
      "memory(GiB)": 147.13,
      "step": 58840,
      "train_speed(iter/s)": 0.20118
    },
    {
      "acc": 0.7791482,
      "epoch": 1.3730150629201359,
      "grad_norm": 6.0625,
      "learning_rate": 2.3631073847548457e-06,
      "loss": 0.80607452,
      "memory(GiB)": 147.13,
      "step": 58850,
      "train_speed(iter/s)": 0.201197
    },
    {
      "acc": 0.76585202,
      "epoch": 1.3732483704924248,
      "grad_norm": 4.46875,
      "learning_rate": 2.3615025262390228e-06,
      "loss": 0.8779665,
      "memory(GiB)": 147.13,
      "step": 58860,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.77455015,
      "epoch": 1.3734816780647137,
      "grad_norm": 5.59375,
      "learning_rate": 2.3598980443679483e-06,
      "loss": 0.80353298,
      "memory(GiB)": 147.13,
      "step": 58870,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.77604198,
      "epoch": 1.3737149856370026,
      "grad_norm": 5.625,
      "learning_rate": 2.3582939393706604e-06,
      "loss": 0.79859333,
      "memory(GiB)": 147.13,
      "step": 58880,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.79776068,
      "epoch": 1.3739482932092915,
      "grad_norm": 8.8125,
      "learning_rate": 2.3566902114761435e-06,
      "loss": 0.71753788,
      "memory(GiB)": 147.13,
      "step": 58890,
      "train_speed(iter/s)": 0.201267
    },
    {
      "acc": 0.78484306,
      "epoch": 1.3741816007815804,
      "grad_norm": 4.40625,
      "learning_rate": 2.3550868609133326e-06,
      "loss": 0.76917105,
      "memory(GiB)": 147.13,
      "step": 58900,
      "train_speed(iter/s)": 0.201285
    },
    {
      "acc": 0.78789654,
      "epoch": 1.3744149083538693,
      "grad_norm": 5.65625,
      "learning_rate": 2.3534838879111026e-06,
      "loss": 0.74949794,
      "memory(GiB)": 147.13,
      "step": 58910,
      "train_speed(iter/s)": 0.201303
    },
    {
      "acc": 0.7702579,
      "epoch": 1.3746482159261582,
      "grad_norm": 4.1875,
      "learning_rate": 2.35188129269828e-06,
      "loss": 0.82006226,
      "memory(GiB)": 147.13,
      "step": 58920,
      "train_speed(iter/s)": 0.201321
    },
    {
      "acc": 0.77914472,
      "epoch": 1.374881523498447,
      "grad_norm": 6.875,
      "learning_rate": 2.3502790755036324e-06,
      "loss": 0.80082273,
      "memory(GiB)": 147.13,
      "step": 58930,
      "train_speed(iter/s)": 0.201338
    },
    {
      "acc": 0.78248625,
      "epoch": 1.375114831070736,
      "grad_norm": 6.71875,
      "learning_rate": 2.3486772365558786e-06,
      "loss": 0.80548611,
      "memory(GiB)": 147.13,
      "step": 58940,
      "train_speed(iter/s)": 0.201356
    },
    {
      "acc": 0.7839232,
      "epoch": 1.3753481386430249,
      "grad_norm": 6.125,
      "learning_rate": 2.3470757760836794e-06,
      "loss": 0.75650473,
      "memory(GiB)": 147.13,
      "step": 58950,
      "train_speed(iter/s)": 0.201373
    },
    {
      "acc": 0.77035446,
      "epoch": 1.3755814462153138,
      "grad_norm": 4.40625,
      "learning_rate": 2.34547469431564e-06,
      "loss": 0.84554367,
      "memory(GiB)": 147.13,
      "step": 58960,
      "train_speed(iter/s)": 0.201391
    },
    {
      "acc": 0.77048821,
      "epoch": 1.3758147537876027,
      "grad_norm": 5.0,
      "learning_rate": 2.3438739914803193e-06,
      "loss": 0.82956161,
      "memory(GiB)": 147.13,
      "step": 58970,
      "train_speed(iter/s)": 0.201409
    },
    {
      "acc": 0.76621504,
      "epoch": 1.3760480613598916,
      "grad_norm": 5.0,
      "learning_rate": 2.3422736678062126e-06,
      "loss": 0.85359306,
      "memory(GiB)": 147.13,
      "step": 58980,
      "train_speed(iter/s)": 0.201427
    },
    {
      "acc": 0.8010005,
      "epoch": 1.3762813689321804,
      "grad_norm": 4.375,
      "learning_rate": 2.3406737235217714e-06,
      "loss": 0.69008727,
      "memory(GiB)": 147.13,
      "step": 58990,
      "train_speed(iter/s)": 0.201445
    },
    {
      "acc": 0.78054056,
      "epoch": 1.3765146765044693,
      "grad_norm": 5.65625,
      "learning_rate": 2.33907415885538e-06,
      "loss": 0.75995293,
      "memory(GiB)": 147.13,
      "step": 59000,
      "train_speed(iter/s)": 0.201462
    },
    {
      "epoch": 1.3765146765044693,
      "eval_acc": 0.7445920493833567,
      "eval_loss": 0.8047448992729187,
      "eval_runtime": 1269.8518,
      "eval_samples_per_second": 28.343,
      "eval_steps_per_second": 14.172,
      "step": 59000
    },
    {
      "acc": 0.78174009,
      "epoch": 1.3767479840767582,
      "grad_norm": 4.40625,
      "learning_rate": 2.3374749740353815e-06,
      "loss": 0.77599964,
      "memory(GiB)": 147.13,
      "step": 59010,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.78817472,
      "epoch": 1.3769812916490471,
      "grad_norm": 4.34375,
      "learning_rate": 2.335876169290056e-06,
      "loss": 0.76314001,
      "memory(GiB)": 147.13,
      "step": 59020,
      "train_speed(iter/s)": 0.200615
    },
    {
      "acc": 0.78163285,
      "epoch": 1.377214599221336,
      "grad_norm": 8.5625,
      "learning_rate": 2.3342777448476326e-06,
      "loss": 0.76157379,
      "memory(GiB)": 147.13,
      "step": 59030,
      "train_speed(iter/s)": 0.200632
    },
    {
      "acc": 0.7947587,
      "epoch": 1.377447906793625,
      "grad_norm": 5.71875,
      "learning_rate": 2.3326797009362884e-06,
      "loss": 0.72174067,
      "memory(GiB)": 147.13,
      "step": 59040,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78288536,
      "epoch": 1.3776812143659138,
      "grad_norm": 5.0625,
      "learning_rate": 2.33108203778414e-06,
      "loss": 0.78767552,
      "memory(GiB)": 147.13,
      "step": 59050,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.76853189,
      "epoch": 1.3779145219382025,
      "grad_norm": 4.5625,
      "learning_rate": 2.3294847556192575e-06,
      "loss": 0.83004322,
      "memory(GiB)": 147.13,
      "step": 59060,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.76078501,
      "epoch": 1.3781478295104916,
      "grad_norm": 6.40625,
      "learning_rate": 2.32788785466965e-06,
      "loss": 0.86728783,
      "memory(GiB)": 147.13,
      "step": 59070,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77539091,
      "epoch": 1.3783811370827803,
      "grad_norm": 5.625,
      "learning_rate": 2.3262913351632725e-06,
      "loss": 0.79250236,
      "memory(GiB)": 147.13,
      "step": 59080,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.77549553,
      "epoch": 1.3786144446550694,
      "grad_norm": 4.65625,
      "learning_rate": 2.3246951973280328e-06,
      "loss": 0.80941544,
      "memory(GiB)": 147.13,
      "step": 59090,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.78177032,
      "epoch": 1.3788477522273581,
      "grad_norm": 6.3125,
      "learning_rate": 2.3230994413917767e-06,
      "loss": 0.77314987,
      "memory(GiB)": 147.13,
      "step": 59100,
      "train_speed(iter/s)": 0.200757
    },
    {
      "acc": 0.77774806,
      "epoch": 1.3790810597996472,
      "grad_norm": 4.28125,
      "learning_rate": 2.3215040675822976e-06,
      "loss": 0.79984112,
      "memory(GiB)": 147.13,
      "step": 59110,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.78535266,
      "epoch": 1.379314367371936,
      "grad_norm": 5.0,
      "learning_rate": 2.319909076127333e-06,
      "loss": 0.76313772,
      "memory(GiB)": 147.13,
      "step": 59120,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.78793802,
      "epoch": 1.379547674944225,
      "grad_norm": 4.53125,
      "learning_rate": 2.3183144672545706e-06,
      "loss": 0.77101927,
      "memory(GiB)": 147.13,
      "step": 59130,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.79543657,
      "epoch": 1.3797809825165137,
      "grad_norm": 6.125,
      "learning_rate": 2.3167202411916372e-06,
      "loss": 0.71729631,
      "memory(GiB)": 147.13,
      "step": 59140,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.79009199,
      "epoch": 1.3800142900888028,
      "grad_norm": 7.875,
      "learning_rate": 2.315126398166112e-06,
      "loss": 0.75708284,
      "memory(GiB)": 147.13,
      "step": 59150,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.78260612,
      "epoch": 1.3802475976610915,
      "grad_norm": 6.625,
      "learning_rate": 2.3135329384055134e-06,
      "loss": 0.78136015,
      "memory(GiB)": 147.13,
      "step": 59160,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.75776453,
      "epoch": 1.3804809052333804,
      "grad_norm": 4.375,
      "learning_rate": 2.3119398621373055e-06,
      "loss": 0.87936306,
      "memory(GiB)": 147.13,
      "step": 59170,
      "train_speed(iter/s)": 0.200879
    },
    {
      "acc": 0.78566179,
      "epoch": 1.3807142128056693,
      "grad_norm": 6.46875,
      "learning_rate": 2.3103471695889035e-06,
      "loss": 0.75471678,
      "memory(GiB)": 147.13,
      "step": 59180,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.77722082,
      "epoch": 1.3809475203779582,
      "grad_norm": 3.84375,
      "learning_rate": 2.308754860987659e-06,
      "loss": 0.78902168,
      "memory(GiB)": 147.13,
      "step": 59190,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.78038301,
      "epoch": 1.381180827950247,
      "grad_norm": 5.34375,
      "learning_rate": 2.3071629365608793e-06,
      "loss": 0.7852129,
      "memory(GiB)": 147.13,
      "step": 59200,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.76663785,
      "epoch": 1.381414135522536,
      "grad_norm": 6.8125,
      "learning_rate": 2.305571396535807e-06,
      "loss": 0.84909973,
      "memory(GiB)": 147.13,
      "step": 59210,
      "train_speed(iter/s)": 0.200946
    },
    {
      "acc": 0.76717291,
      "epoch": 1.381647443094825,
      "grad_norm": 5.84375,
      "learning_rate": 2.303980241139636e-06,
      "loss": 0.84376945,
      "memory(GiB)": 147.13,
      "step": 59220,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.75790958,
      "epoch": 1.3818807506671138,
      "grad_norm": 9.0,
      "learning_rate": 2.3023894705995e-06,
      "loss": 0.85827541,
      "memory(GiB)": 147.13,
      "step": 59230,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.7841033,
      "epoch": 1.3821140582394027,
      "grad_norm": 10.6875,
      "learning_rate": 2.3007990851424862e-06,
      "loss": 0.79800854,
      "memory(GiB)": 147.13,
      "step": 59240,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.78057899,
      "epoch": 1.3823473658116916,
      "grad_norm": 4.15625,
      "learning_rate": 2.2992090849956176e-06,
      "loss": 0.8046237,
      "memory(GiB)": 147.13,
      "step": 59250,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.77489777,
      "epoch": 1.3825806733839805,
      "grad_norm": 6.46875,
      "learning_rate": 2.2976194703858666e-06,
      "loss": 0.80679083,
      "memory(GiB)": 147.13,
      "step": 59260,
      "train_speed(iter/s)": 0.201034
    },
    {
      "acc": 0.77184649,
      "epoch": 1.3828139809562694,
      "grad_norm": 6.0625,
      "learning_rate": 2.2960302415401525e-06,
      "loss": 0.84999313,
      "memory(GiB)": 147.13,
      "step": 59270,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.79643536,
      "epoch": 1.3830472885285583,
      "grad_norm": 6.625,
      "learning_rate": 2.2944413986853344e-06,
      "loss": 0.75381327,
      "memory(GiB)": 147.13,
      "step": 59280,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.77682009,
      "epoch": 1.3832805961008472,
      "grad_norm": 4.0,
      "learning_rate": 2.292852942048222e-06,
      "loss": 0.78831792,
      "memory(GiB)": 147.13,
      "step": 59290,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.78009706,
      "epoch": 1.383513903673136,
      "grad_norm": 5.375,
      "learning_rate": 2.2912648718555665e-06,
      "loss": 0.79728909,
      "memory(GiB)": 147.13,
      "step": 59300,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.77483168,
      "epoch": 1.383747211245425,
      "grad_norm": 5.40625,
      "learning_rate": 2.2896771883340614e-06,
      "loss": 0.81179886,
      "memory(GiB)": 147.13,
      "step": 59310,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.77034178,
      "epoch": 1.383980518817714,
      "grad_norm": 4.8125,
      "learning_rate": 2.2880898917103515e-06,
      "loss": 0.84435978,
      "memory(GiB)": 147.13,
      "step": 59320,
      "train_speed(iter/s)": 0.201142
    },
    {
      "acc": 0.78930626,
      "epoch": 1.3842138263900028,
      "grad_norm": 8.9375,
      "learning_rate": 2.2865029822110222e-06,
      "loss": 0.73878508,
      "memory(GiB)": 147.13,
      "step": 59330,
      "train_speed(iter/s)": 0.201159
    },
    {
      "acc": 0.78097978,
      "epoch": 1.3844471339622917,
      "grad_norm": 4.59375,
      "learning_rate": 2.2849164600626045e-06,
      "loss": 0.77706594,
      "memory(GiB)": 147.13,
      "step": 59340,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.77245169,
      "epoch": 1.3846804415345806,
      "grad_norm": 5.25,
      "learning_rate": 2.2833303254915713e-06,
      "loss": 0.80866556,
      "memory(GiB)": 147.13,
      "step": 59350,
      "train_speed(iter/s)": 0.201195
    },
    {
      "acc": 0.80767746,
      "epoch": 1.3849137491068695,
      "grad_norm": 5.71875,
      "learning_rate": 2.2817445787243464e-06,
      "loss": 0.6845746,
      "memory(GiB)": 147.13,
      "step": 59360,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.78989286,
      "epoch": 1.3851470566791584,
      "grad_norm": 7.21875,
      "learning_rate": 2.280159219987293e-06,
      "loss": 0.73846169,
      "memory(GiB)": 147.13,
      "step": 59370,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.76815262,
      "epoch": 1.3853803642514473,
      "grad_norm": 8.6875,
      "learning_rate": 2.27857424950672e-06,
      "loss": 0.82266207,
      "memory(GiB)": 147.13,
      "step": 59380,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.76027946,
      "epoch": 1.3856136718237362,
      "grad_norm": 6.03125,
      "learning_rate": 2.2769896675088833e-06,
      "loss": 0.87801542,
      "memory(GiB)": 147.13,
      "step": 59390,
      "train_speed(iter/s)": 0.201267
    },
    {
      "acc": 0.78674126,
      "epoch": 1.385846979396025,
      "grad_norm": 4.96875,
      "learning_rate": 2.2754054742199787e-06,
      "loss": 0.78030553,
      "memory(GiB)": 147.13,
      "step": 59400,
      "train_speed(iter/s)": 0.201285
    },
    {
      "acc": 0.78899231,
      "epoch": 1.386080286968314,
      "grad_norm": 4.625,
      "learning_rate": 2.273821669866153e-06,
      "loss": 0.73768158,
      "memory(GiB)": 147.13,
      "step": 59410,
      "train_speed(iter/s)": 0.201301
    },
    {
      "acc": 0.77118654,
      "epoch": 1.3863135945406029,
      "grad_norm": 6.15625,
      "learning_rate": 2.2722382546734904e-06,
      "loss": 0.83121977,
      "memory(GiB)": 147.13,
      "step": 59420,
      "train_speed(iter/s)": 0.201319
    },
    {
      "acc": 0.76945696,
      "epoch": 1.3865469021128918,
      "grad_norm": 5.78125,
      "learning_rate": 2.270655228868026e-06,
      "loss": 0.84560337,
      "memory(GiB)": 147.13,
      "step": 59430,
      "train_speed(iter/s)": 0.201336
    },
    {
      "acc": 0.79610276,
      "epoch": 1.3867802096851807,
      "grad_norm": 5.25,
      "learning_rate": 2.2690725926757355e-06,
      "loss": 0.73223934,
      "memory(GiB)": 147.13,
      "step": 59440,
      "train_speed(iter/s)": 0.201354
    },
    {
      "acc": 0.76559458,
      "epoch": 1.3870135172574694,
      "grad_norm": 5.875,
      "learning_rate": 2.267490346322539e-06,
      "loss": 0.86651058,
      "memory(GiB)": 147.13,
      "step": 59450,
      "train_speed(iter/s)": 0.20137
    },
    {
      "acc": 0.80258989,
      "epoch": 1.3872468248297585,
      "grad_norm": 10.0,
      "learning_rate": 2.265908490034301e-06,
      "loss": 0.71374941,
      "memory(GiB)": 147.13,
      "step": 59460,
      "train_speed(iter/s)": 0.201388
    },
    {
      "acc": 0.77058544,
      "epoch": 1.3874801324020472,
      "grad_norm": 5.21875,
      "learning_rate": 2.2643270240368305e-06,
      "loss": 0.83985233,
      "memory(GiB)": 147.13,
      "step": 59470,
      "train_speed(iter/s)": 0.201406
    },
    {
      "acc": 0.77997141,
      "epoch": 1.3877134399743363,
      "grad_norm": 5.53125,
      "learning_rate": 2.2627459485558846e-06,
      "loss": 0.77843761,
      "memory(GiB)": 147.13,
      "step": 59480,
      "train_speed(iter/s)": 0.201423
    },
    {
      "acc": 0.76730309,
      "epoch": 1.387946747546625,
      "grad_norm": 7.03125,
      "learning_rate": 2.2611652638171568e-06,
      "loss": 0.82932968,
      "memory(GiB)": 147.13,
      "step": 59490,
      "train_speed(iter/s)": 0.20144
    },
    {
      "acc": 0.78150969,
      "epoch": 1.388180055118914,
      "grad_norm": 5.5,
      "learning_rate": 2.259584970046294e-06,
      "loss": 0.7819231,
      "memory(GiB)": 147.13,
      "step": 59500,
      "train_speed(iter/s)": 0.201457
    },
    {
      "epoch": 1.388180055118914,
      "eval_acc": 0.7445906080852659,
      "eval_loss": 0.8047671914100647,
      "eval_runtime": 1269.924,
      "eval_samples_per_second": 28.341,
      "eval_steps_per_second": 14.171,
      "step": 59500
    },
    {
      "acc": 0.74836121,
      "epoch": 1.3884133626912027,
      "grad_norm": 6.53125,
      "learning_rate": 2.2580050674688815e-06,
      "loss": 0.9253665,
      "memory(GiB)": 147.13,
      "step": 59510,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.77755685,
      "epoch": 1.3886466702634919,
      "grad_norm": 5.25,
      "learning_rate": 2.2564255563104465e-06,
      "loss": 0.79394531,
      "memory(GiB)": 147.13,
      "step": 59520,
      "train_speed(iter/s)": 0.200617
    },
    {
      "acc": 0.77132187,
      "epoch": 1.3888799778357805,
      "grad_norm": 5.90625,
      "learning_rate": 2.254846436796468e-06,
      "loss": 0.82167749,
      "memory(GiB)": 147.13,
      "step": 59530,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.76772861,
      "epoch": 1.3891132854080694,
      "grad_norm": 5.75,
      "learning_rate": 2.2532677091523615e-06,
      "loss": 0.84193869,
      "memory(GiB)": 147.13,
      "step": 59540,
      "train_speed(iter/s)": 0.200654
    },
    {
      "acc": 0.79113283,
      "epoch": 1.3893465929803583,
      "grad_norm": 6.0,
      "learning_rate": 2.2516893736034935e-06,
      "loss": 0.74261956,
      "memory(GiB)": 147.13,
      "step": 59550,
      "train_speed(iter/s)": 0.20067
    },
    {
      "acc": 0.76318316,
      "epoch": 1.3895799005526472,
      "grad_norm": 5.46875,
      "learning_rate": 2.250111430375169e-06,
      "loss": 0.8669055,
      "memory(GiB)": 147.13,
      "step": 59560,
      "train_speed(iter/s)": 0.200688
    },
    {
      "acc": 0.780229,
      "epoch": 1.3898132081249361,
      "grad_norm": 6.5,
      "learning_rate": 2.248533879692639e-06,
      "loss": 0.77526116,
      "memory(GiB)": 147.13,
      "step": 59570,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.7953208,
      "epoch": 1.390046515697225,
      "grad_norm": 4.5625,
      "learning_rate": 2.246956721781097e-06,
      "loss": 0.74956465,
      "memory(GiB)": 147.13,
      "step": 59580,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.79553127,
      "epoch": 1.390279823269514,
      "grad_norm": 4.78125,
      "learning_rate": 2.245379956865684e-06,
      "loss": 0.74106674,
      "memory(GiB)": 147.13,
      "step": 59590,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.74994259,
      "epoch": 1.3905131308418028,
      "grad_norm": 5.28125,
      "learning_rate": 2.243803585171483e-06,
      "loss": 0.89359455,
      "memory(GiB)": 147.13,
      "step": 59600,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.76904812,
      "epoch": 1.3907464384140917,
      "grad_norm": 5.53125,
      "learning_rate": 2.2422276069235174e-06,
      "loss": 0.82504959,
      "memory(GiB)": 147.13,
      "step": 59610,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.77541676,
      "epoch": 1.3909797459863806,
      "grad_norm": 7.625,
      "learning_rate": 2.240652022346761e-06,
      "loss": 0.80968227,
      "memory(GiB)": 147.13,
      "step": 59620,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.7667109,
      "epoch": 1.3912130535586695,
      "grad_norm": 5.34375,
      "learning_rate": 2.2390768316661256e-06,
      "loss": 0.82808847,
      "memory(GiB)": 147.13,
      "step": 59630,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.79110146,
      "epoch": 1.3914463611309584,
      "grad_norm": 4.71875,
      "learning_rate": 2.237502035106472e-06,
      "loss": 0.75989141,
      "memory(GiB)": 147.13,
      "step": 59640,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.76692033,
      "epoch": 1.3916796687032473,
      "grad_norm": 16.375,
      "learning_rate": 2.2359276328926007e-06,
      "loss": 0.83675632,
      "memory(GiB)": 147.13,
      "step": 59650,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.77610579,
      "epoch": 1.3919129762755362,
      "grad_norm": 5.25,
      "learning_rate": 2.2343536252492542e-06,
      "loss": 0.80217104,
      "memory(GiB)": 147.13,
      "step": 59660,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.77709975,
      "epoch": 1.3921462838478251,
      "grad_norm": 5.96875,
      "learning_rate": 2.2327800124011285e-06,
      "loss": 0.81001654,
      "memory(GiB)": 147.13,
      "step": 59670,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.77643595,
      "epoch": 1.392379591420114,
      "grad_norm": 5.4375,
      "learning_rate": 2.231206794572848e-06,
      "loss": 0.80721416,
      "memory(GiB)": 147.13,
      "step": 59680,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.77637901,
      "epoch": 1.392612898992403,
      "grad_norm": 5.34375,
      "learning_rate": 2.229633971988996e-06,
      "loss": 0.80078964,
      "memory(GiB)": 147.13,
      "step": 59690,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.78074894,
      "epoch": 1.3928462065646918,
      "grad_norm": 8.0625,
      "learning_rate": 2.2280615448740873e-06,
      "loss": 0.79391489,
      "memory(GiB)": 147.13,
      "step": 59700,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.78368201,
      "epoch": 1.3930795141369807,
      "grad_norm": 9.0625,
      "learning_rate": 2.2264895134525898e-06,
      "loss": 0.77110548,
      "memory(GiB)": 147.13,
      "step": 59710,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.79938531,
      "epoch": 1.3933128217092696,
      "grad_norm": 4.40625,
      "learning_rate": 2.2249178779489065e-06,
      "loss": 0.71430826,
      "memory(GiB)": 147.13,
      "step": 59720,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.78940001,
      "epoch": 1.3935461292815585,
      "grad_norm": 4.625,
      "learning_rate": 2.223346638587392e-06,
      "loss": 0.76045151,
      "memory(GiB)": 147.13,
      "step": 59730,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.79823899,
      "epoch": 1.3937794368538474,
      "grad_norm": 6.46875,
      "learning_rate": 2.2217757955923386e-06,
      "loss": 0.72001886,
      "memory(GiB)": 147.13,
      "step": 59740,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.79044533,
      "epoch": 1.3940127444261363,
      "grad_norm": 4.5,
      "learning_rate": 2.220205349187981e-06,
      "loss": 0.74306707,
      "memory(GiB)": 147.13,
      "step": 59750,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.78812528,
      "epoch": 1.3942460519984252,
      "grad_norm": 8.375,
      "learning_rate": 2.218635299598504e-06,
      "loss": 0.76315918,
      "memory(GiB)": 147.13,
      "step": 59760,
      "train_speed(iter/s)": 0.201036
    },
    {
      "acc": 0.78087091,
      "epoch": 1.3944793595707141,
      "grad_norm": 5.9375,
      "learning_rate": 2.2170656470480284e-06,
      "loss": 0.79584026,
      "memory(GiB)": 147.13,
      "step": 59770,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.77213335,
      "epoch": 1.394712667143003,
      "grad_norm": 5.96875,
      "learning_rate": 2.215496391760625e-06,
      "loss": 0.83666573,
      "memory(GiB)": 147.13,
      "step": 59780,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.79270315,
      "epoch": 1.394945974715292,
      "grad_norm": 5.65625,
      "learning_rate": 2.2139275339603023e-06,
      "loss": 0.7484087,
      "memory(GiB)": 147.13,
      "step": 59790,
      "train_speed(iter/s)": 0.201088
    },
    {
      "acc": 0.80184422,
      "epoch": 1.3951792822875808,
      "grad_norm": 4.8125,
      "learning_rate": 2.2123590738710153e-06,
      "loss": 0.71304121,
      "memory(GiB)": 147.13,
      "step": 59800,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.793783,
      "epoch": 1.3954125898598697,
      "grad_norm": 5.03125,
      "learning_rate": 2.2107910117166608e-06,
      "loss": 0.73716478,
      "memory(GiB)": 147.13,
      "step": 59810,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.77772017,
      "epoch": 1.3956458974321586,
      "grad_norm": 4.75,
      "learning_rate": 2.2092233477210767e-06,
      "loss": 0.78334846,
      "memory(GiB)": 147.13,
      "step": 59820,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.78544092,
      "epoch": 1.3958792050044475,
      "grad_norm": 5.9375,
      "learning_rate": 2.2076560821080515e-06,
      "loss": 0.77470407,
      "memory(GiB)": 147.13,
      "step": 59830,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.77739944,
      "epoch": 1.3961125125767362,
      "grad_norm": 4.875,
      "learning_rate": 2.2060892151013067e-06,
      "loss": 0.807547,
      "memory(GiB)": 147.13,
      "step": 59840,
      "train_speed(iter/s)": 0.201175
    },
    {
      "acc": 0.7652626,
      "epoch": 1.3963458201490253,
      "grad_norm": 4.75,
      "learning_rate": 2.2045227469245178e-06,
      "loss": 0.83140392,
      "memory(GiB)": 147.13,
      "step": 59850,
      "train_speed(iter/s)": 0.201194
    },
    {
      "acc": 0.78730431,
      "epoch": 1.396579127721314,
      "grad_norm": 4.9375,
      "learning_rate": 2.202956677801292e-06,
      "loss": 0.74223843,
      "memory(GiB)": 147.13,
      "step": 59860,
      "train_speed(iter/s)": 0.201211
    },
    {
      "acc": 0.77190733,
      "epoch": 1.396812435293603,
      "grad_norm": 5.8125,
      "learning_rate": 2.2013910079551905e-06,
      "loss": 0.82516975,
      "memory(GiB)": 147.13,
      "step": 59870,
      "train_speed(iter/s)": 0.201227
    },
    {
      "acc": 0.77773724,
      "epoch": 1.3970457428658918,
      "grad_norm": 6.1875,
      "learning_rate": 2.199825737609709e-06,
      "loss": 0.78944416,
      "memory(GiB)": 147.13,
      "step": 59880,
      "train_speed(iter/s)": 0.201245
    },
    {
      "acc": 0.77316818,
      "epoch": 1.397279050438181,
      "grad_norm": 6.125,
      "learning_rate": 2.198260866988288e-06,
      "loss": 0.81869354,
      "memory(GiB)": 147.13,
      "step": 59890,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.79717088,
      "epoch": 1.3975123580104696,
      "grad_norm": 4.75,
      "learning_rate": 2.1966963963143184e-06,
      "loss": 0.72420158,
      "memory(GiB)": 147.13,
      "step": 59900,
      "train_speed(iter/s)": 0.201279
    },
    {
      "acc": 0.7967824,
      "epoch": 1.3977456655827587,
      "grad_norm": 4.21875,
      "learning_rate": 2.1951323258111194e-06,
      "loss": 0.71405821,
      "memory(GiB)": 147.13,
      "step": 59910,
      "train_speed(iter/s)": 0.201295
    },
    {
      "acc": 0.77692866,
      "epoch": 1.3979789731550474,
      "grad_norm": 5.125,
      "learning_rate": 2.193568655701969e-06,
      "loss": 0.80145493,
      "memory(GiB)": 147.13,
      "step": 59920,
      "train_speed(iter/s)": 0.201313
    },
    {
      "acc": 0.77152061,
      "epoch": 1.3982122807273363,
      "grad_norm": 4.125,
      "learning_rate": 2.1920053862100754e-06,
      "loss": 0.82826338,
      "memory(GiB)": 147.13,
      "step": 59930,
      "train_speed(iter/s)": 0.201331
    },
    {
      "acc": 0.78626838,
      "epoch": 1.3984455882996252,
      "grad_norm": 22.0,
      "learning_rate": 2.190442517558599e-06,
      "loss": 0.76881523,
      "memory(GiB)": 147.13,
      "step": 59940,
      "train_speed(iter/s)": 0.201348
    },
    {
      "acc": 0.77308369,
      "epoch": 1.398678895871914,
      "grad_norm": 5.4375,
      "learning_rate": 2.188880049970637e-06,
      "loss": 0.82151947,
      "memory(GiB)": 147.13,
      "step": 59950,
      "train_speed(iter/s)": 0.201365
    },
    {
      "acc": 0.76596837,
      "epoch": 1.398912203444203,
      "grad_norm": 5.5625,
      "learning_rate": 2.18731798366923e-06,
      "loss": 0.8408473,
      "memory(GiB)": 147.13,
      "step": 59960,
      "train_speed(iter/s)": 0.201384
    },
    {
      "acc": 0.7691186,
      "epoch": 1.3991455110164919,
      "grad_norm": 6.09375,
      "learning_rate": 2.1857563188773644e-06,
      "loss": 0.84743719,
      "memory(GiB)": 147.13,
      "step": 59970,
      "train_speed(iter/s)": 0.201402
    },
    {
      "acc": 0.77092209,
      "epoch": 1.3993788185887808,
      "grad_norm": 8.4375,
      "learning_rate": 2.184195055817966e-06,
      "loss": 0.84977627,
      "memory(GiB)": 147.13,
      "step": 59980,
      "train_speed(iter/s)": 0.201418
    },
    {
      "acc": 0.77478681,
      "epoch": 1.3996121261610697,
      "grad_norm": 6.90625,
      "learning_rate": 2.1826341947139067e-06,
      "loss": 0.80044651,
      "memory(GiB)": 147.13,
      "step": 59990,
      "train_speed(iter/s)": 0.201435
    },
    {
      "acc": 0.7726613,
      "epoch": 1.3998454337333586,
      "grad_norm": 3.703125,
      "learning_rate": 2.181073735787998e-06,
      "loss": 0.83067513,
      "memory(GiB)": 147.13,
      "step": 60000,
      "train_speed(iter/s)": 0.201452
    },
    {
      "epoch": 1.3998454337333586,
      "eval_acc": 0.7445992558738102,
      "eval_loss": 0.8046594858169556,
      "eval_runtime": 1271.0218,
      "eval_samples_per_second": 28.317,
      "eval_steps_per_second": 14.159,
      "step": 60000
    },
    {
      "acc": 0.78201704,
      "epoch": 1.4000787413056475,
      "grad_norm": 4.84375,
      "learning_rate": 2.179513679262992e-06,
      "loss": 0.78135967,
      "memory(GiB)": 147.13,
      "step": 60010,
      "train_speed(iter/s)": 0.200603
    },
    {
      "acc": 0.7830049,
      "epoch": 1.4003120488779364,
      "grad_norm": 5.03125,
      "learning_rate": 2.1779540253615917e-06,
      "loss": 0.78097987,
      "memory(GiB)": 147.13,
      "step": 60020,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.80212574,
      "epoch": 1.4005453564502253,
      "grad_norm": 5.46875,
      "learning_rate": 2.176394774306434e-06,
      "loss": 0.71014204,
      "memory(GiB)": 147.13,
      "step": 60030,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.79081697,
      "epoch": 1.4007786640225142,
      "grad_norm": 6.875,
      "learning_rate": 2.174835926320102e-06,
      "loss": 0.75322866,
      "memory(GiB)": 147.13,
      "step": 60040,
      "train_speed(iter/s)": 0.200655
    },
    {
      "acc": 0.77095056,
      "epoch": 1.401011971594803,
      "grad_norm": 5.375,
      "learning_rate": 2.173277481625119e-06,
      "loss": 0.83664217,
      "memory(GiB)": 147.13,
      "step": 60050,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.76814718,
      "epoch": 1.401245279167092,
      "grad_norm": 5.375,
      "learning_rate": 2.1717194404439563e-06,
      "loss": 0.84824924,
      "memory(GiB)": 147.13,
      "step": 60060,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.77537007,
      "epoch": 1.4014785867393809,
      "grad_norm": 4.8125,
      "learning_rate": 2.17016180299902e-06,
      "loss": 0.79466648,
      "memory(GiB)": 147.13,
      "step": 60070,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.78756018,
      "epoch": 1.4017118943116698,
      "grad_norm": 4.5,
      "learning_rate": 2.168604569512666e-06,
      "loss": 0.75703373,
      "memory(GiB)": 147.13,
      "step": 60080,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.76781497,
      "epoch": 1.4019452018839587,
      "grad_norm": 7.03125,
      "learning_rate": 2.167047740207187e-06,
      "loss": 0.84280319,
      "memory(GiB)": 147.13,
      "step": 60090,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.76985674,
      "epoch": 1.4021785094562476,
      "grad_norm": 5.3125,
      "learning_rate": 2.1654913153048186e-06,
      "loss": 0.8241436,
      "memory(GiB)": 147.13,
      "step": 60100,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.79031215,
      "epoch": 1.4024118170285365,
      "grad_norm": 3.828125,
      "learning_rate": 2.1639352950277433e-06,
      "loss": 0.74541531,
      "memory(GiB)": 147.13,
      "step": 60110,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.7930346,
      "epoch": 1.4026451246008254,
      "grad_norm": 7.75,
      "learning_rate": 2.162379679598079e-06,
      "loss": 0.73493719,
      "memory(GiB)": 147.13,
      "step": 60120,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.77408686,
      "epoch": 1.4028784321731143,
      "grad_norm": 4.75,
      "learning_rate": 2.1608244692378946e-06,
      "loss": 0.83344536,
      "memory(GiB)": 147.13,
      "step": 60130,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.78932433,
      "epoch": 1.4031117397454032,
      "grad_norm": 5.84375,
      "learning_rate": 2.1592696641691884e-06,
      "loss": 0.744315,
      "memory(GiB)": 147.13,
      "step": 60140,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.78552046,
      "epoch": 1.403345047317692,
      "grad_norm": 5.40625,
      "learning_rate": 2.157715264613915e-06,
      "loss": 0.78410473,
      "memory(GiB)": 147.13,
      "step": 60150,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.76167541,
      "epoch": 1.403578354889981,
      "grad_norm": 6.46875,
      "learning_rate": 2.156161270793961e-06,
      "loss": 0.8703126,
      "memory(GiB)": 147.13,
      "step": 60160,
      "train_speed(iter/s)": 0.200858
    },
    {
      "acc": 0.78156233,
      "epoch": 1.4038116624622698,
      "grad_norm": 4.875,
      "learning_rate": 2.1546076829311584e-06,
      "loss": 0.7798255,
      "memory(GiB)": 147.13,
      "step": 60170,
      "train_speed(iter/s)": 0.200876
    },
    {
      "acc": 0.79834514,
      "epoch": 1.4040449700345587,
      "grad_norm": 5.5625,
      "learning_rate": 2.153054501247284e-06,
      "loss": 0.73483548,
      "memory(GiB)": 147.13,
      "step": 60180,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.77490888,
      "epoch": 1.4042782776068476,
      "grad_norm": 5.4375,
      "learning_rate": 2.151501725964051e-06,
      "loss": 0.79429326,
      "memory(GiB)": 147.13,
      "step": 60190,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.79103308,
      "epoch": 1.4045115851791365,
      "grad_norm": 3.796875,
      "learning_rate": 2.14994935730312e-06,
      "loss": 0.74287324,
      "memory(GiB)": 147.13,
      "step": 60200,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.76605501,
      "epoch": 1.4047448927514254,
      "grad_norm": 8.1875,
      "learning_rate": 2.1483973954860894e-06,
      "loss": 0.86529417,
      "memory(GiB)": 147.13,
      "step": 60210,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.7755722,
      "epoch": 1.4049782003237143,
      "grad_norm": 4.9375,
      "learning_rate": 2.146845840734504e-06,
      "loss": 0.79309669,
      "memory(GiB)": 147.13,
      "step": 60220,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.7734189,
      "epoch": 1.405211507896003,
      "grad_norm": 8.75,
      "learning_rate": 2.1452946932698454e-06,
      "loss": 0.84152288,
      "memory(GiB)": 147.13,
      "step": 60230,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.78702106,
      "epoch": 1.4054448154682921,
      "grad_norm": 4.5625,
      "learning_rate": 2.1437439533135386e-06,
      "loss": 0.77411451,
      "memory(GiB)": 147.13,
      "step": 60240,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.78498821,
      "epoch": 1.4056781230405808,
      "grad_norm": 4.96875,
      "learning_rate": 2.142193621086956e-06,
      "loss": 0.75319977,
      "memory(GiB)": 147.13,
      "step": 60250,
      "train_speed(iter/s)": 0.201011
    },
    {
      "acc": 0.7821517,
      "epoch": 1.40591143061287,
      "grad_norm": 4.46875,
      "learning_rate": 2.140643696811401e-06,
      "loss": 0.80938702,
      "memory(GiB)": 147.13,
      "step": 60260,
      "train_speed(iter/s)": 0.201027
    },
    {
      "acc": 0.77948647,
      "epoch": 1.4061447381851586,
      "grad_norm": 7.25,
      "learning_rate": 2.1390941807081285e-06,
      "loss": 0.79588933,
      "memory(GiB)": 147.13,
      "step": 60270,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.79571724,
      "epoch": 1.4063780457574477,
      "grad_norm": 7.15625,
      "learning_rate": 2.1375450729983294e-06,
      "loss": 0.72755728,
      "memory(GiB)": 147.13,
      "step": 60280,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.79610963,
      "epoch": 1.4066113533297364,
      "grad_norm": 4.90625,
      "learning_rate": 2.1359963739031407e-06,
      "loss": 0.728936,
      "memory(GiB)": 147.13,
      "step": 60290,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.7886919,
      "epoch": 1.4068446609020255,
      "grad_norm": 5.21875,
      "learning_rate": 2.134448083643638e-06,
      "loss": 0.76339221,
      "memory(GiB)": 147.13,
      "step": 60300,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.76663504,
      "epoch": 1.4070779684743142,
      "grad_norm": 5.59375,
      "learning_rate": 2.1329002024408375e-06,
      "loss": 0.84437771,
      "memory(GiB)": 147.13,
      "step": 60310,
      "train_speed(iter/s)": 0.201113
    },
    {
      "acc": 0.77732272,
      "epoch": 1.407311276046603,
      "grad_norm": 5.125,
      "learning_rate": 2.1313527305157015e-06,
      "loss": 0.78522806,
      "memory(GiB)": 147.13,
      "step": 60320,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.78382611,
      "epoch": 1.407544583618892,
      "grad_norm": 4.0625,
      "learning_rate": 2.1298056680891288e-06,
      "loss": 0.78302422,
      "memory(GiB)": 147.13,
      "step": 60330,
      "train_speed(iter/s)": 0.201145
    },
    {
      "acc": 0.77463741,
      "epoch": 1.407777891191181,
      "grad_norm": 6.0,
      "learning_rate": 2.1282590153819645e-06,
      "loss": 0.80812149,
      "memory(GiB)": 147.13,
      "step": 60340,
      "train_speed(iter/s)": 0.201161
    },
    {
      "acc": 0.79031725,
      "epoch": 1.4080111987634698,
      "grad_norm": 4.03125,
      "learning_rate": 2.1267127726149896e-06,
      "loss": 0.76469936,
      "memory(GiB)": 147.13,
      "step": 60350,
      "train_speed(iter/s)": 0.201179
    },
    {
      "acc": 0.76453934,
      "epoch": 1.4082445063357587,
      "grad_norm": 5.78125,
      "learning_rate": 2.1251669400089353e-06,
      "loss": 0.84970903,
      "memory(GiB)": 147.13,
      "step": 60360,
      "train_speed(iter/s)": 0.201195
    },
    {
      "acc": 0.79993305,
      "epoch": 1.4084778139080476,
      "grad_norm": 4.84375,
      "learning_rate": 2.1236215177844617e-06,
      "loss": 0.70226107,
      "memory(GiB)": 147.13,
      "step": 60370,
      "train_speed(iter/s)": 0.201212
    },
    {
      "acc": 0.80778503,
      "epoch": 1.4087111214803365,
      "grad_norm": 7.0625,
      "learning_rate": 2.1220765061621828e-06,
      "loss": 0.67411156,
      "memory(GiB)": 147.13,
      "step": 60380,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.77227788,
      "epoch": 1.4089444290526254,
      "grad_norm": 4.15625,
      "learning_rate": 2.120531905362646e-06,
      "loss": 0.81891384,
      "memory(GiB)": 147.13,
      "step": 60390,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.79898548,
      "epoch": 1.4091777366249143,
      "grad_norm": 4.8125,
      "learning_rate": 2.118987715606342e-06,
      "loss": 0.72313986,
      "memory(GiB)": 147.13,
      "step": 60400,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.78145227,
      "epoch": 1.4094110441972032,
      "grad_norm": 5.71875,
      "learning_rate": 2.1174439371137064e-06,
      "loss": 0.78863721,
      "memory(GiB)": 147.13,
      "step": 60410,
      "train_speed(iter/s)": 0.201278
    },
    {
      "acc": 0.78333859,
      "epoch": 1.409644351769492,
      "grad_norm": 5.875,
      "learning_rate": 2.1159005701051093e-06,
      "loss": 0.77123041,
      "memory(GiB)": 147.13,
      "step": 60420,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.79827003,
      "epoch": 1.409877659341781,
      "grad_norm": 5.375,
      "learning_rate": 2.11435761480087e-06,
      "loss": 0.71006422,
      "memory(GiB)": 147.13,
      "step": 60430,
      "train_speed(iter/s)": 0.201312
    },
    {
      "acc": 0.78789725,
      "epoch": 1.41011096691407,
      "grad_norm": 4.40625,
      "learning_rate": 2.112815071421243e-06,
      "loss": 0.7574213,
      "memory(GiB)": 147.13,
      "step": 60440,
      "train_speed(iter/s)": 0.201329
    },
    {
      "acc": 0.77921619,
      "epoch": 1.4103442744863588,
      "grad_norm": 4.3125,
      "learning_rate": 2.111272940186424e-06,
      "loss": 0.78693404,
      "memory(GiB)": 147.13,
      "step": 60450,
      "train_speed(iter/s)": 0.201347
    },
    {
      "acc": 0.76266708,
      "epoch": 1.4105775820586477,
      "grad_norm": 5.84375,
      "learning_rate": 2.109731221316555e-06,
      "loss": 0.87478523,
      "memory(GiB)": 147.13,
      "step": 60460,
      "train_speed(iter/s)": 0.201363
    },
    {
      "acc": 0.76366029,
      "epoch": 1.4108108896309366,
      "grad_norm": 7.03125,
      "learning_rate": 2.108189915031715e-06,
      "loss": 0.84197912,
      "memory(GiB)": 147.13,
      "step": 60470,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.78174658,
      "epoch": 1.4110441972032255,
      "grad_norm": 5.65625,
      "learning_rate": 2.1066490215519243e-06,
      "loss": 0.78470864,
      "memory(GiB)": 147.13,
      "step": 60480,
      "train_speed(iter/s)": 0.201398
    },
    {
      "acc": 0.80206575,
      "epoch": 1.4112775047755144,
      "grad_norm": 6.53125,
      "learning_rate": 2.105108541097143e-06,
      "loss": 0.71396837,
      "memory(GiB)": 147.13,
      "step": 60490,
      "train_speed(iter/s)": 0.201416
    },
    {
      "acc": 0.77739525,
      "epoch": 1.4115108123478033,
      "grad_norm": 6.40625,
      "learning_rate": 2.1035684738872792e-06,
      "loss": 0.78567958,
      "memory(GiB)": 147.13,
      "step": 60500,
      "train_speed(iter/s)": 0.201433
    },
    {
      "epoch": 1.4115108123478033,
      "eval_acc": 0.7446263202490692,
      "eval_loss": 0.8046724200248718,
      "eval_runtime": 1269.8878,
      "eval_samples_per_second": 28.342,
      "eval_steps_per_second": 14.171,
      "step": 60500
    },
    {
      "acc": 0.78144083,
      "epoch": 1.4117441199200922,
      "grad_norm": 4.96875,
      "learning_rate": 2.1020288201421722e-06,
      "loss": 0.7848321,
      "memory(GiB)": 147.13,
      "step": 60510,
      "train_speed(iter/s)": 0.200589
    },
    {
      "acc": 0.78034506,
      "epoch": 1.411977427492381,
      "grad_norm": 4.96875,
      "learning_rate": 2.100489580081611e-06,
      "loss": 0.79630418,
      "memory(GiB)": 147.13,
      "step": 60520,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.76285105,
      "epoch": 1.41221073506467,
      "grad_norm": 6.125,
      "learning_rate": 2.09895075392532e-06,
      "loss": 0.84258804,
      "memory(GiB)": 147.13,
      "step": 60530,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.7814455,
      "epoch": 1.4124440426369589,
      "grad_norm": 5.15625,
      "learning_rate": 2.0974123418929644e-06,
      "loss": 0.78600588,
      "memory(GiB)": 147.13,
      "step": 60540,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.79303546,
      "epoch": 1.4126773502092478,
      "grad_norm": 5.15625,
      "learning_rate": 2.095874344204155e-06,
      "loss": 0.73297596,
      "memory(GiB)": 147.13,
      "step": 60550,
      "train_speed(iter/s)": 0.200662
    },
    {
      "acc": 0.79392943,
      "epoch": 1.4129106577815367,
      "grad_norm": 6.0,
      "learning_rate": 2.094336761078438e-06,
      "loss": 0.73787174,
      "memory(GiB)": 147.13,
      "step": 60560,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.78089685,
      "epoch": 1.4131439653538256,
      "grad_norm": 5.9375,
      "learning_rate": 2.0927995927353062e-06,
      "loss": 0.79184189,
      "memory(GiB)": 147.13,
      "step": 60570,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.78633385,
      "epoch": 1.4133772729261145,
      "grad_norm": 5.0,
      "learning_rate": 2.091262839394188e-06,
      "loss": 0.76935081,
      "memory(GiB)": 147.13,
      "step": 60580,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.7927001,
      "epoch": 1.4136105804984034,
      "grad_norm": 6.125,
      "learning_rate": 2.0897265012744543e-06,
      "loss": 0.75596743,
      "memory(GiB)": 147.13,
      "step": 60590,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.77544589,
      "epoch": 1.413843888070692,
      "grad_norm": 5.15625,
      "learning_rate": 2.0881905785954172e-06,
      "loss": 0.79783726,
      "memory(GiB)": 147.13,
      "step": 60600,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76404762,
      "epoch": 1.4140771956429812,
      "grad_norm": 6.25,
      "learning_rate": 2.086655071576327e-06,
      "loss": 0.83482218,
      "memory(GiB)": 147.13,
      "step": 60610,
      "train_speed(iter/s)": 0.200763
    },
    {
      "acc": 0.77434025,
      "epoch": 1.4143105032152699,
      "grad_norm": 5.15625,
      "learning_rate": 2.085119980436381e-06,
      "loss": 0.79602251,
      "memory(GiB)": 147.13,
      "step": 60620,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.78777227,
      "epoch": 1.414543810787559,
      "grad_norm": 4.03125,
      "learning_rate": 2.083585305394709e-06,
      "loss": 0.75320654,
      "memory(GiB)": 147.13,
      "step": 60630,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.77788224,
      "epoch": 1.4147771183598477,
      "grad_norm": 6.0,
      "learning_rate": 2.0820510466703898e-06,
      "loss": 0.77969913,
      "memory(GiB)": 147.13,
      "step": 60640,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.79586306,
      "epoch": 1.4150104259321368,
      "grad_norm": 5.21875,
      "learning_rate": 2.080517204482434e-06,
      "loss": 0.73031406,
      "memory(GiB)": 147.13,
      "step": 60650,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.79818316,
      "epoch": 1.4152437335044254,
      "grad_norm": 4.09375,
      "learning_rate": 2.078983779049801e-06,
      "loss": 0.73085876,
      "memory(GiB)": 147.13,
      "step": 60660,
      "train_speed(iter/s)": 0.200849
    },
    {
      "acc": 0.78288202,
      "epoch": 1.4154770410767146,
      "grad_norm": 5.40625,
      "learning_rate": 2.0774507705913844e-06,
      "loss": 0.77484465,
      "memory(GiB)": 147.13,
      "step": 60670,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.77833548,
      "epoch": 1.4157103486490032,
      "grad_norm": 5.40625,
      "learning_rate": 2.07591817932602e-06,
      "loss": 0.78327904,
      "memory(GiB)": 147.13,
      "step": 60680,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.78331342,
      "epoch": 1.4159436562212924,
      "grad_norm": 10.5625,
      "learning_rate": 2.074386005472488e-06,
      "loss": 0.77935891,
      "memory(GiB)": 147.13,
      "step": 60690,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.77536631,
      "epoch": 1.416176963793581,
      "grad_norm": 4.96875,
      "learning_rate": 2.072854249249503e-06,
      "loss": 0.79027624,
      "memory(GiB)": 147.13,
      "step": 60700,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.77584419,
      "epoch": 1.41641027136587,
      "grad_norm": 4.34375,
      "learning_rate": 2.0713229108757244e-06,
      "loss": 0.80542765,
      "memory(GiB)": 147.13,
      "step": 60710,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.78596239,
      "epoch": 1.4166435789381588,
      "grad_norm": 4.8125,
      "learning_rate": 2.0697919905697474e-06,
      "loss": 0.77891278,
      "memory(GiB)": 147.13,
      "step": 60720,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78150387,
      "epoch": 1.4168768865104477,
      "grad_norm": 3.5,
      "learning_rate": 2.0682614885501147e-06,
      "loss": 0.77592754,
      "memory(GiB)": 147.13,
      "step": 60730,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.7803638,
      "epoch": 1.4171101940827366,
      "grad_norm": 6.21875,
      "learning_rate": 2.066731405035302e-06,
      "loss": 0.78199749,
      "memory(GiB)": 147.13,
      "step": 60740,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.79469795,
      "epoch": 1.4173435016550255,
      "grad_norm": 5.28125,
      "learning_rate": 2.065201740243728e-06,
      "loss": 0.72167587,
      "memory(GiB)": 147.13,
      "step": 60750,
      "train_speed(iter/s)": 0.200997
    },
    {
      "acc": 0.76823006,
      "epoch": 1.4175768092273144,
      "grad_norm": 5.53125,
      "learning_rate": 2.063672494393755e-06,
      "loss": 0.84083357,
      "memory(GiB)": 147.13,
      "step": 60760,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.79289522,
      "epoch": 1.4178101167996033,
      "grad_norm": 6.125,
      "learning_rate": 2.0621436677036775e-06,
      "loss": 0.74563189,
      "memory(GiB)": 147.13,
      "step": 60770,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.79262247,
      "epoch": 1.4180434243718922,
      "grad_norm": 5.0625,
      "learning_rate": 2.0606152603917406e-06,
      "loss": 0.74184866,
      "memory(GiB)": 147.13,
      "step": 60780,
      "train_speed(iter/s)": 0.20105
    },
    {
      "acc": 0.78077393,
      "epoch": 1.4182767319441811,
      "grad_norm": 7.8125,
      "learning_rate": 2.0590872726761215e-06,
      "loss": 0.77461243,
      "memory(GiB)": 147.13,
      "step": 60790,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.78589611,
      "epoch": 1.41851003951647,
      "grad_norm": 5.40625,
      "learning_rate": 2.057559704774938e-06,
      "loss": 0.78779116,
      "memory(GiB)": 147.13,
      "step": 60800,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.77445693,
      "epoch": 1.418743347088759,
      "grad_norm": 6.875,
      "learning_rate": 2.0560325569062535e-06,
      "loss": 0.80334835,
      "memory(GiB)": 147.13,
      "step": 60810,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.77701745,
      "epoch": 1.4189766546610478,
      "grad_norm": 5.21875,
      "learning_rate": 2.054505829288066e-06,
      "loss": 0.81892672,
      "memory(GiB)": 147.13,
      "step": 60820,
      "train_speed(iter/s)": 0.201121
    },
    {
      "acc": 0.79528999,
      "epoch": 1.4192099622333367,
      "grad_norm": 5.1875,
      "learning_rate": 2.0529795221383164e-06,
      "loss": 0.72682176,
      "memory(GiB)": 147.13,
      "step": 60830,
      "train_speed(iter/s)": 0.201136
    },
    {
      "acc": 0.75750227,
      "epoch": 1.4194432698056256,
      "grad_norm": 5.96875,
      "learning_rate": 2.0514536356748814e-06,
      "loss": 0.86790705,
      "memory(GiB)": 147.13,
      "step": 60840,
      "train_speed(iter/s)": 0.201154
    },
    {
      "acc": 0.76532421,
      "epoch": 1.4196765773779145,
      "grad_norm": 4.40625,
      "learning_rate": 2.0499281701155852e-06,
      "loss": 0.84386234,
      "memory(GiB)": 147.13,
      "step": 60850,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.7653913,
      "epoch": 1.4199098849502034,
      "grad_norm": 9.375,
      "learning_rate": 2.0484031256781845e-06,
      "loss": 0.84208984,
      "memory(GiB)": 147.13,
      "step": 60860,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.7715621,
      "epoch": 1.4201431925224923,
      "grad_norm": 5.65625,
      "learning_rate": 2.046878502580382e-06,
      "loss": 0.8072238,
      "memory(GiB)": 147.13,
      "step": 60870,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.77248592,
      "epoch": 1.4203765000947812,
      "grad_norm": 4.625,
      "learning_rate": 2.045354301039815e-06,
      "loss": 0.81728535,
      "memory(GiB)": 147.13,
      "step": 60880,
      "train_speed(iter/s)": 0.201223
    },
    {
      "acc": 0.79007959,
      "epoch": 1.4206098076670701,
      "grad_norm": 7.90625,
      "learning_rate": 2.043830521274061e-06,
      "loss": 0.74873753,
      "memory(GiB)": 147.13,
      "step": 60890,
      "train_speed(iter/s)": 0.20124
    },
    {
      "acc": 0.7762063,
      "epoch": 1.420843115239359,
      "grad_norm": 5.0625,
      "learning_rate": 2.0423071635006436e-06,
      "loss": 0.80805073,
      "memory(GiB)": 147.13,
      "step": 60900,
      "train_speed(iter/s)": 0.201257
    },
    {
      "acc": 0.77343888,
      "epoch": 1.421076422811648,
      "grad_norm": 5.5,
      "learning_rate": 2.0407842279370176e-06,
      "loss": 0.82168198,
      "memory(GiB)": 147.13,
      "step": 60910,
      "train_speed(iter/s)": 0.201274
    },
    {
      "acc": 0.7828186,
      "epoch": 1.4213097303839368,
      "grad_norm": 4.5,
      "learning_rate": 2.039261714800585e-06,
      "loss": 0.7554111,
      "memory(GiB)": 147.13,
      "step": 60920,
      "train_speed(iter/s)": 0.201291
    },
    {
      "acc": 0.78235793,
      "epoch": 1.4215430379562257,
      "grad_norm": 5.90625,
      "learning_rate": 2.0377396243086827e-06,
      "loss": 0.7957407,
      "memory(GiB)": 147.13,
      "step": 60930,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.77107277,
      "epoch": 1.4217763455285146,
      "grad_norm": 27.75,
      "learning_rate": 2.036217956678588e-06,
      "loss": 0.83353214,
      "memory(GiB)": 147.13,
      "step": 60940,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.78861694,
      "epoch": 1.4220096531008035,
      "grad_norm": 4.34375,
      "learning_rate": 2.034696712127518e-06,
      "loss": 0.75524387,
      "memory(GiB)": 147.13,
      "step": 60950,
      "train_speed(iter/s)": 0.201343
    },
    {
      "acc": 0.76062737,
      "epoch": 1.4222429606730924,
      "grad_norm": 6.53125,
      "learning_rate": 2.0331758908726323e-06,
      "loss": 0.87450056,
      "memory(GiB)": 147.13,
      "step": 60960,
      "train_speed(iter/s)": 0.201359
    },
    {
      "acc": 0.77146816,
      "epoch": 1.4224762682453813,
      "grad_norm": 5.25,
      "learning_rate": 2.031655493131026e-06,
      "loss": 0.80269995,
      "memory(GiB)": 147.13,
      "step": 60970,
      "train_speed(iter/s)": 0.201377
    },
    {
      "acc": 0.76841688,
      "epoch": 1.4227095758176702,
      "grad_norm": 5.5625,
      "learning_rate": 2.030135519119735e-06,
      "loss": 0.83026876,
      "memory(GiB)": 147.13,
      "step": 60980,
      "train_speed(iter/s)": 0.201393
    },
    {
      "acc": 0.78153791,
      "epoch": 1.422942883389959,
      "grad_norm": 8.375,
      "learning_rate": 2.0286159690557366e-06,
      "loss": 0.8033679,
      "memory(GiB)": 147.13,
      "step": 60990,
      "train_speed(iter/s)": 0.20141
    },
    {
      "acc": 0.79216089,
      "epoch": 1.423176190962248,
      "grad_norm": 8.0,
      "learning_rate": 2.027096843155944e-06,
      "loss": 0.76074972,
      "memory(GiB)": 147.13,
      "step": 61000,
      "train_speed(iter/s)": 0.201427
    },
    {
      "epoch": 1.423176190962248,
      "eval_acc": 0.744670359912952,
      "eval_loss": 0.8046051859855652,
      "eval_runtime": 1270.2783,
      "eval_samples_per_second": 28.333,
      "eval_steps_per_second": 14.167,
      "step": 61000
    },
    {
      "acc": 0.79125977,
      "epoch": 1.4234094985345367,
      "grad_norm": 10.5625,
      "learning_rate": 2.025578141637215e-06,
      "loss": 0.74310904,
      "memory(GiB)": 147.13,
      "step": 61010,
      "train_speed(iter/s)": 0.20059
    },
    {
      "acc": 0.76769581,
      "epoch": 1.4236428061068258,
      "grad_norm": 4.875,
      "learning_rate": 2.024059864716343e-06,
      "loss": 0.83147717,
      "memory(GiB)": 147.13,
      "step": 61020,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.77622566,
      "epoch": 1.4238761136791145,
      "grad_norm": 7.03125,
      "learning_rate": 2.022542012610058e-06,
      "loss": 0.80299397,
      "memory(GiB)": 147.13,
      "step": 61030,
      "train_speed(iter/s)": 0.200626
    },
    {
      "acc": 0.78763971,
      "epoch": 1.4241094212514036,
      "grad_norm": 4.53125,
      "learning_rate": 2.0210245855350397e-06,
      "loss": 0.7463274,
      "memory(GiB)": 147.13,
      "step": 61040,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.78232255,
      "epoch": 1.4243427288236923,
      "grad_norm": 5.8125,
      "learning_rate": 2.019507583707893e-06,
      "loss": 0.80083017,
      "memory(GiB)": 147.13,
      "step": 61050,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.79297686,
      "epoch": 1.4245760363959814,
      "grad_norm": 4.71875,
      "learning_rate": 2.017991007345175e-06,
      "loss": 0.73181958,
      "memory(GiB)": 147.13,
      "step": 61060,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.78155422,
      "epoch": 1.42480934396827,
      "grad_norm": 5.375,
      "learning_rate": 2.016474856663372e-06,
      "loss": 0.79105949,
      "memory(GiB)": 147.13,
      "step": 61070,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.77549515,
      "epoch": 1.425042651540559,
      "grad_norm": 6.0,
      "learning_rate": 2.014959131878918e-06,
      "loss": 0.81070833,
      "memory(GiB)": 147.13,
      "step": 61080,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.78981018,
      "epoch": 1.4252759591128479,
      "grad_norm": 4.03125,
      "learning_rate": 2.0134438332081814e-06,
      "loss": 0.76355982,
      "memory(GiB)": 147.13,
      "step": 61090,
      "train_speed(iter/s)": 0.200727
    },
    {
      "acc": 0.78231525,
      "epoch": 1.4255092666851368,
      "grad_norm": 7.0625,
      "learning_rate": 2.0119289608674682e-06,
      "loss": 0.77640409,
      "memory(GiB)": 147.13,
      "step": 61100,
      "train_speed(iter/s)": 0.200745
    },
    {
      "acc": 0.77027392,
      "epoch": 1.4257425742574257,
      "grad_norm": 6.28125,
      "learning_rate": 2.010414515073029e-06,
      "loss": 0.8242136,
      "memory(GiB)": 147.13,
      "step": 61110,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.78219414,
      "epoch": 1.4259758818297146,
      "grad_norm": 5.03125,
      "learning_rate": 2.0089004960410485e-06,
      "loss": 0.78226018,
      "memory(GiB)": 147.13,
      "step": 61120,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.79695072,
      "epoch": 1.4262091894020035,
      "grad_norm": 4.46875,
      "learning_rate": 2.007386903987654e-06,
      "loss": 0.72442718,
      "memory(GiB)": 147.13,
      "step": 61130,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.78832622,
      "epoch": 1.4264424969742924,
      "grad_norm": 5.15625,
      "learning_rate": 2.0058737391289085e-06,
      "loss": 0.7365797,
      "memory(GiB)": 147.13,
      "step": 61140,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.79924631,
      "epoch": 1.4266758045465813,
      "grad_norm": 10.625,
      "learning_rate": 2.0043610016808185e-06,
      "loss": 0.71290426,
      "memory(GiB)": 147.13,
      "step": 61150,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.793398,
      "epoch": 1.4269091121188702,
      "grad_norm": 5.25,
      "learning_rate": 2.0028486918593253e-06,
      "loss": 0.74249353,
      "memory(GiB)": 147.13,
      "step": 61160,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.79817839,
      "epoch": 1.427142419691159,
      "grad_norm": 7.1875,
      "learning_rate": 2.001336809880311e-06,
      "loss": 0.71638279,
      "memory(GiB)": 147.13,
      "step": 61170,
      "train_speed(iter/s)": 0.200864
    },
    {
      "acc": 0.7944067,
      "epoch": 1.427375727263448,
      "grad_norm": 7.3125,
      "learning_rate": 1.9998253559595952e-06,
      "loss": 0.72443352,
      "memory(GiB)": 147.13,
      "step": 61180,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.77642808,
      "epoch": 1.4276090348357369,
      "grad_norm": 5.40625,
      "learning_rate": 1.9983143303129373e-06,
      "loss": 0.81988869,
      "memory(GiB)": 147.13,
      "step": 61190,
      "train_speed(iter/s)": 0.200899
    },
    {
      "acc": 0.79813347,
      "epoch": 1.4278423424080258,
      "grad_norm": 4.875,
      "learning_rate": 1.996803733156038e-06,
      "loss": 0.72240438,
      "memory(GiB)": 147.13,
      "step": 61200,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.76945052,
      "epoch": 1.4280756499803147,
      "grad_norm": 5.96875,
      "learning_rate": 1.9952935647045317e-06,
      "loss": 0.83810911,
      "memory(GiB)": 147.13,
      "step": 61210,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.79682202,
      "epoch": 1.4283089575526036,
      "grad_norm": 7.59375,
      "learning_rate": 1.9937838251739983e-06,
      "loss": 0.72564445,
      "memory(GiB)": 147.13,
      "step": 61220,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.79631333,
      "epoch": 1.4285422651248925,
      "grad_norm": 6.0,
      "learning_rate": 1.9922745147799505e-06,
      "loss": 0.74118524,
      "memory(GiB)": 147.13,
      "step": 61230,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.80053787,
      "epoch": 1.4287755726971814,
      "grad_norm": 3.890625,
      "learning_rate": 1.9907656337378396e-06,
      "loss": 0.69944177,
      "memory(GiB)": 147.13,
      "step": 61240,
      "train_speed(iter/s)": 0.200983
    },
    {
      "acc": 0.76880045,
      "epoch": 1.4290088802694703,
      "grad_norm": 5.3125,
      "learning_rate": 1.9892571822630622e-06,
      "loss": 0.8351284,
      "memory(GiB)": 147.13,
      "step": 61250,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.76350574,
      "epoch": 1.4292421878417592,
      "grad_norm": 4.46875,
      "learning_rate": 1.987749160570946e-06,
      "loss": 0.85821857,
      "memory(GiB)": 147.13,
      "step": 61260,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.7792346,
      "epoch": 1.429475495414048,
      "grad_norm": 5.5,
      "learning_rate": 1.9862415688767657e-06,
      "loss": 0.78018122,
      "memory(GiB)": 147.13,
      "step": 61270,
      "train_speed(iter/s)": 0.201036
    },
    {
      "acc": 0.7726048,
      "epoch": 1.429708802986337,
      "grad_norm": 5.375,
      "learning_rate": 1.984734407395722e-06,
      "loss": 0.82185307,
      "memory(GiB)": 147.13,
      "step": 61280,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.78452177,
      "epoch": 1.4299421105586259,
      "grad_norm": 4.0625,
      "learning_rate": 1.9832276763429674e-06,
      "loss": 0.75924959,
      "memory(GiB)": 147.13,
      "step": 61290,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.77438474,
      "epoch": 1.4301754181309148,
      "grad_norm": 4.65625,
      "learning_rate": 1.9817213759335846e-06,
      "loss": 0.84851189,
      "memory(GiB)": 147.13,
      "step": 61300,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.77766914,
      "epoch": 1.4304087257032037,
      "grad_norm": 5.15625,
      "learning_rate": 1.9802155063825995e-06,
      "loss": 0.7944005,
      "memory(GiB)": 147.13,
      "step": 61310,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.74958205,
      "epoch": 1.4306420332754926,
      "grad_norm": 5.84375,
      "learning_rate": 1.9787100679049742e-06,
      "loss": 0.91778793,
      "memory(GiB)": 147.13,
      "step": 61320,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.79310417,
      "epoch": 1.4308753408477815,
      "grad_norm": 9.375,
      "learning_rate": 1.977205060715607e-06,
      "loss": 0.74190741,
      "memory(GiB)": 147.13,
      "step": 61330,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.77827301,
      "epoch": 1.4311086484200704,
      "grad_norm": 5.5,
      "learning_rate": 1.975700485029341e-06,
      "loss": 0.80124111,
      "memory(GiB)": 147.13,
      "step": 61340,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.78497138,
      "epoch": 1.4313419559923592,
      "grad_norm": 6.21875,
      "learning_rate": 1.9741963410609506e-06,
      "loss": 0.75833988,
      "memory(GiB)": 147.13,
      "step": 61350,
      "train_speed(iter/s)": 0.201175
    },
    {
      "acc": 0.77764912,
      "epoch": 1.4315752635646481,
      "grad_norm": 3.65625,
      "learning_rate": 1.9726926290251548e-06,
      "loss": 0.7967144,
      "memory(GiB)": 147.13,
      "step": 61360,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.76858315,
      "epoch": 1.431808571136937,
      "grad_norm": 6.1875,
      "learning_rate": 1.971189349136607e-06,
      "loss": 0.82962227,
      "memory(GiB)": 147.13,
      "step": 61370,
      "train_speed(iter/s)": 0.201209
    },
    {
      "acc": 0.7532167,
      "epoch": 1.4320418787092257,
      "grad_norm": 4.6875,
      "learning_rate": 1.969686501609898e-06,
      "loss": 0.89655132,
      "memory(GiB)": 147.13,
      "step": 61380,
      "train_speed(iter/s)": 0.201227
    },
    {
      "acc": 0.79799242,
      "epoch": 1.4322751862815148,
      "grad_norm": 3.984375,
      "learning_rate": 1.9681840866595644e-06,
      "loss": 0.73923264,
      "memory(GiB)": 147.13,
      "step": 61390,
      "train_speed(iter/s)": 0.201245
    },
    {
      "acc": 0.77811699,
      "epoch": 1.4325084938538035,
      "grad_norm": 6.15625,
      "learning_rate": 1.966682104500068e-06,
      "loss": 0.79530067,
      "memory(GiB)": 147.13,
      "step": 61400,
      "train_speed(iter/s)": 0.201263
    },
    {
      "acc": 0.76311293,
      "epoch": 1.4327418014260926,
      "grad_norm": 4.8125,
      "learning_rate": 1.9651805553458212e-06,
      "loss": 0.83997898,
      "memory(GiB)": 147.13,
      "step": 61410,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.77644825,
      "epoch": 1.4329751089983813,
      "grad_norm": 5.65625,
      "learning_rate": 1.9636794394111676e-06,
      "loss": 0.81493549,
      "memory(GiB)": 147.13,
      "step": 61420,
      "train_speed(iter/s)": 0.201298
    },
    {
      "acc": 0.7841002,
      "epoch": 1.4332084165706704,
      "grad_norm": 5.625,
      "learning_rate": 1.962178756910393e-06,
      "loss": 0.77840614,
      "memory(GiB)": 147.13,
      "step": 61430,
      "train_speed(iter/s)": 0.201315
    },
    {
      "acc": 0.77601957,
      "epoch": 1.4334417241429591,
      "grad_norm": 6.15625,
      "learning_rate": 1.9606785080577173e-06,
      "loss": 0.80642538,
      "memory(GiB)": 147.13,
      "step": 61440,
      "train_speed(iter/s)": 0.201332
    },
    {
      "acc": 0.78321104,
      "epoch": 1.4336750317152482,
      "grad_norm": 4.4375,
      "learning_rate": 1.959178693067303e-06,
      "loss": 0.77348614,
      "memory(GiB)": 147.13,
      "step": 61450,
      "train_speed(iter/s)": 0.201349
    },
    {
      "acc": 0.77611961,
      "epoch": 1.433908339287537,
      "grad_norm": 5.53125,
      "learning_rate": 1.9576793121532467e-06,
      "loss": 0.81834526,
      "memory(GiB)": 147.13,
      "step": 61460,
      "train_speed(iter/s)": 0.201365
    },
    {
      "acc": 0.76458263,
      "epoch": 1.4341416468598258,
      "grad_norm": 5.40625,
      "learning_rate": 1.9561803655295835e-06,
      "loss": 0.86758423,
      "memory(GiB)": 147.13,
      "step": 61470,
      "train_speed(iter/s)": 0.201381
    },
    {
      "acc": 0.78354859,
      "epoch": 1.4343749544321147,
      "grad_norm": 5.28125,
      "learning_rate": 1.9546818534102903e-06,
      "loss": 0.77626381,
      "memory(GiB)": 147.13,
      "step": 61480,
      "train_speed(iter/s)": 0.201397
    },
    {
      "acc": 0.79747005,
      "epoch": 1.4346082620044036,
      "grad_norm": 4.6875,
      "learning_rate": 1.9531837760092765e-06,
      "loss": 0.72363586,
      "memory(GiB)": 147.13,
      "step": 61490,
      "train_speed(iter/s)": 0.201414
    },
    {
      "acc": 0.77367868,
      "epoch": 1.4348415695766925,
      "grad_norm": 4.65625,
      "learning_rate": 1.9516861335403963e-06,
      "loss": 0.80418158,
      "memory(GiB)": 147.13,
      "step": 61500,
      "train_speed(iter/s)": 0.201431
    },
    {
      "epoch": 1.4348415695766925,
      "eval_acc": 0.7446573882301356,
      "eval_loss": 0.8045687675476074,
      "eval_runtime": 1271.1131,
      "eval_samples_per_second": 28.315,
      "eval_steps_per_second": 14.158,
      "step": 61500
    },
    {
      "acc": 0.77232461,
      "epoch": 1.4350748771489814,
      "grad_norm": 5.09375,
      "learning_rate": 1.9501889262174323e-06,
      "loss": 0.82037907,
      "memory(GiB)": 147.13,
      "step": 61510,
      "train_speed(iter/s)": 0.2006
    },
    {
      "acc": 0.77945395,
      "epoch": 1.4353081847212703,
      "grad_norm": 5.4375,
      "learning_rate": 1.9486921542541147e-06,
      "loss": 0.79027767,
      "memory(GiB)": 147.13,
      "step": 61520,
      "train_speed(iter/s)": 0.200618
    },
    {
      "acc": 0.79523678,
      "epoch": 1.4355414922935592,
      "grad_norm": 4.375,
      "learning_rate": 1.9471958178641055e-06,
      "loss": 0.7177619,
      "memory(GiB)": 147.13,
      "step": 61530,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.79970589,
      "epoch": 1.435774799865848,
      "grad_norm": 4.25,
      "learning_rate": 1.9456999172610046e-06,
      "loss": 0.71671066,
      "memory(GiB)": 147.13,
      "step": 61540,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.76421814,
      "epoch": 1.436008107438137,
      "grad_norm": 5.625,
      "learning_rate": 1.9442044526583555e-06,
      "loss": 0.86059494,
      "memory(GiB)": 147.13,
      "step": 61550,
      "train_speed(iter/s)": 0.200669
    },
    {
      "acc": 0.77591553,
      "epoch": 1.436241415010426,
      "grad_norm": 7.0,
      "learning_rate": 1.9427094242696304e-06,
      "loss": 0.81485806,
      "memory(GiB)": 147.13,
      "step": 61560,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.78154783,
      "epoch": 1.4364747225827148,
      "grad_norm": 6.46875,
      "learning_rate": 1.941214832308249e-06,
      "loss": 0.78967419,
      "memory(GiB)": 147.13,
      "step": 61570,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77090359,
      "epoch": 1.4367080301550037,
      "grad_norm": 5.5,
      "learning_rate": 1.9397206769875602e-06,
      "loss": 0.83600082,
      "memory(GiB)": 147.13,
      "step": 61580,
      "train_speed(iter/s)": 0.20072
    },
    {
      "acc": 0.77267971,
      "epoch": 1.4369413377272926,
      "grad_norm": 5.46875,
      "learning_rate": 1.9382269585208576e-06,
      "loss": 0.84045687,
      "memory(GiB)": 147.13,
      "step": 61590,
      "train_speed(iter/s)": 0.200737
    },
    {
      "acc": 0.75382733,
      "epoch": 1.4371746452995815,
      "grad_norm": 4.5625,
      "learning_rate": 1.936733677121367e-06,
      "loss": 0.87408772,
      "memory(GiB)": 147.13,
      "step": 61600,
      "train_speed(iter/s)": 0.200754
    },
    {
      "acc": 0.78201303,
      "epoch": 1.4374079528718704,
      "grad_norm": 6.21875,
      "learning_rate": 1.935240833002252e-06,
      "loss": 0.78332329,
      "memory(GiB)": 147.13,
      "step": 61610,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.8092041,
      "epoch": 1.4376412604441593,
      "grad_norm": 4.9375,
      "learning_rate": 1.933748426376622e-06,
      "loss": 0.68223801,
      "memory(GiB)": 147.13,
      "step": 61620,
      "train_speed(iter/s)": 0.200788
    },
    {
      "acc": 0.79494247,
      "epoch": 1.4378745680164482,
      "grad_norm": 5.40625,
      "learning_rate": 1.932256457457509e-06,
      "loss": 0.72881784,
      "memory(GiB)": 147.13,
      "step": 61630,
      "train_speed(iter/s)": 0.200805
    },
    {
      "acc": 0.78494654,
      "epoch": 1.438107875588737,
      "grad_norm": 4.78125,
      "learning_rate": 1.9307649264578982e-06,
      "loss": 0.75482392,
      "memory(GiB)": 147.13,
      "step": 61640,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.7854641,
      "epoch": 1.438341183161026,
      "grad_norm": 5.9375,
      "learning_rate": 1.9292738335907e-06,
      "loss": 0.77157822,
      "memory(GiB)": 147.13,
      "step": 61650,
      "train_speed(iter/s)": 0.200839
    },
    {
      "acc": 0.7829257,
      "epoch": 1.438574490733315,
      "grad_norm": 6.53125,
      "learning_rate": 1.9277831790687724e-06,
      "loss": 0.77883649,
      "memory(GiB)": 147.13,
      "step": 61660,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.78623981,
      "epoch": 1.4388077983056038,
      "grad_norm": 5.0625,
      "learning_rate": 1.9262929631049034e-06,
      "loss": 0.76112614,
      "memory(GiB)": 147.13,
      "step": 61670,
      "train_speed(iter/s)": 0.20087
    },
    {
      "acc": 0.7791275,
      "epoch": 1.4390411058778927,
      "grad_norm": 4.5,
      "learning_rate": 1.924803185911819e-06,
      "loss": 0.79928355,
      "memory(GiB)": 147.13,
      "step": 61680,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.79443464,
      "epoch": 1.4392744134501816,
      "grad_norm": 5.125,
      "learning_rate": 1.923313847702188e-06,
      "loss": 0.72661982,
      "memory(GiB)": 147.13,
      "step": 61690,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.78530979,
      "epoch": 1.4395077210224705,
      "grad_norm": 5.1875,
      "learning_rate": 1.9218249486886097e-06,
      "loss": 0.76591215,
      "memory(GiB)": 147.13,
      "step": 61700,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.78451734,
      "epoch": 1.4397410285947594,
      "grad_norm": 6.0,
      "learning_rate": 1.9203364890836277e-06,
      "loss": 0.78146687,
      "memory(GiB)": 147.13,
      "step": 61710,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.77538376,
      "epoch": 1.4399743361670483,
      "grad_norm": 5.78125,
      "learning_rate": 1.918848469099718e-06,
      "loss": 0.79503469,
      "memory(GiB)": 147.13,
      "step": 61720,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.79319973,
      "epoch": 1.4402076437393372,
      "grad_norm": 4.5,
      "learning_rate": 1.9173608889492936e-06,
      "loss": 0.7500802,
      "memory(GiB)": 147.13,
      "step": 61730,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.77077246,
      "epoch": 1.440440951311626,
      "grad_norm": 5.75,
      "learning_rate": 1.915873748844705e-06,
      "loss": 0.82878227,
      "memory(GiB)": 147.13,
      "step": 61740,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.78738832,
      "epoch": 1.440674258883915,
      "grad_norm": 6.25,
      "learning_rate": 1.9143870489982443e-06,
      "loss": 0.73900366,
      "memory(GiB)": 147.13,
      "step": 61750,
      "train_speed(iter/s)": 0.201002
    },
    {
      "acc": 0.76408119,
      "epoch": 1.4409075664562039,
      "grad_norm": 5.625,
      "learning_rate": 1.9129007896221365e-06,
      "loss": 0.88208513,
      "memory(GiB)": 147.13,
      "step": 61760,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.77571197,
      "epoch": 1.4411408740284926,
      "grad_norm": 5.4375,
      "learning_rate": 1.9114149709285416e-06,
      "loss": 0.8385725,
      "memory(GiB)": 147.13,
      "step": 61770,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.78287396,
      "epoch": 1.4413741816007817,
      "grad_norm": 5.1875,
      "learning_rate": 1.909929593129565e-06,
      "loss": 0.76827679,
      "memory(GiB)": 147.13,
      "step": 61780,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.76898689,
      "epoch": 1.4416074891730704,
      "grad_norm": 8.125,
      "learning_rate": 1.9084446564372393e-06,
      "loss": 0.83712349,
      "memory(GiB)": 147.13,
      "step": 61790,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.79256206,
      "epoch": 1.4418407967453595,
      "grad_norm": 6.46875,
      "learning_rate": 1.9069601610635424e-06,
      "loss": 0.74981198,
      "memory(GiB)": 147.13,
      "step": 61800,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.77438807,
      "epoch": 1.4420741043176482,
      "grad_norm": 5.6875,
      "learning_rate": 1.9054761072203843e-06,
      "loss": 0.80913887,
      "memory(GiB)": 147.13,
      "step": 61810,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.78683462,
      "epoch": 1.4423074118899373,
      "grad_norm": 4.125,
      "learning_rate": 1.9039924951196109e-06,
      "loss": 0.74516659,
      "memory(GiB)": 147.13,
      "step": 61820,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.76727543,
      "epoch": 1.442540719462226,
      "grad_norm": 4.9375,
      "learning_rate": 1.9025093249730108e-06,
      "loss": 0.85720892,
      "memory(GiB)": 147.13,
      "step": 61830,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.78560305,
      "epoch": 1.442774027034515,
      "grad_norm": 4.03125,
      "learning_rate": 1.9010265969923052e-06,
      "loss": 0.7749969,
      "memory(GiB)": 147.13,
      "step": 61840,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.77236147,
      "epoch": 1.4430073346068037,
      "grad_norm": 7.28125,
      "learning_rate": 1.8995443113891527e-06,
      "loss": 0.83206615,
      "memory(GiB)": 147.13,
      "step": 61850,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.77366638,
      "epoch": 1.4432406421790926,
      "grad_norm": 5.03125,
      "learning_rate": 1.898062468375147e-06,
      "loss": 0.80375004,
      "memory(GiB)": 147.13,
      "step": 61860,
      "train_speed(iter/s)": 0.201187
    },
    {
      "acc": 0.7846086,
      "epoch": 1.4434739497513815,
      "grad_norm": 6.0625,
      "learning_rate": 1.8965810681618251e-06,
      "loss": 0.78282852,
      "memory(GiB)": 147.13,
      "step": 61870,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.76645479,
      "epoch": 1.4437072573236704,
      "grad_norm": 9.0625,
      "learning_rate": 1.8951001109606538e-06,
      "loss": 0.84201212,
      "memory(GiB)": 147.13,
      "step": 61880,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.77387953,
      "epoch": 1.4439405648959593,
      "grad_norm": 4.15625,
      "learning_rate": 1.893619596983038e-06,
      "loss": 0.82105818,
      "memory(GiB)": 147.13,
      "step": 61890,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.7874526,
      "epoch": 1.4441738724682482,
      "grad_norm": 3.734375,
      "learning_rate": 1.8921395264403236e-06,
      "loss": 0.76474705,
      "memory(GiB)": 147.13,
      "step": 61900,
      "train_speed(iter/s)": 0.201255
    },
    {
      "acc": 0.78465261,
      "epoch": 1.4444071800405371,
      "grad_norm": 4.3125,
      "learning_rate": 1.890659899543788e-06,
      "loss": 0.77885513,
      "memory(GiB)": 147.13,
      "step": 61910,
      "train_speed(iter/s)": 0.201272
    },
    {
      "acc": 0.77660131,
      "epoch": 1.444640487612826,
      "grad_norm": 5.78125,
      "learning_rate": 1.88918071650465e-06,
      "loss": 0.8049757,
      "memory(GiB)": 147.13,
      "step": 61920,
      "train_speed(iter/s)": 0.201289
    },
    {
      "acc": 0.77938328,
      "epoch": 1.444873795185115,
      "grad_norm": 5.125,
      "learning_rate": 1.8877019775340587e-06,
      "loss": 0.81291084,
      "memory(GiB)": 147.13,
      "step": 61930,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.78789864,
      "epoch": 1.4451071027574038,
      "grad_norm": 6.28125,
      "learning_rate": 1.8862236828431086e-06,
      "loss": 0.76462269,
      "memory(GiB)": 147.13,
      "step": 61940,
      "train_speed(iter/s)": 0.201324
    },
    {
      "acc": 0.77977004,
      "epoch": 1.4453404103296927,
      "grad_norm": 6.0,
      "learning_rate": 1.8847458326428226e-06,
      "loss": 0.79729452,
      "memory(GiB)": 147.13,
      "step": 61950,
      "train_speed(iter/s)": 0.20134
    },
    {
      "acc": 0.78940058,
      "epoch": 1.4455737179019816,
      "grad_norm": 6.34375,
      "learning_rate": 1.8832684271441643e-06,
      "loss": 0.74841819,
      "memory(GiB)": 147.13,
      "step": 61960,
      "train_speed(iter/s)": 0.201357
    },
    {
      "acc": 0.78960233,
      "epoch": 1.4458070254742705,
      "grad_norm": 3.703125,
      "learning_rate": 1.8817914665580322e-06,
      "loss": 0.7424305,
      "memory(GiB)": 147.13,
      "step": 61970,
      "train_speed(iter/s)": 0.201373
    },
    {
      "acc": 0.80876637,
      "epoch": 1.4460403330465594,
      "grad_norm": 7.53125,
      "learning_rate": 1.8803149510952613e-06,
      "loss": 0.71806173,
      "memory(GiB)": 147.13,
      "step": 61980,
      "train_speed(iter/s)": 0.20139
    },
    {
      "acc": 0.77671194,
      "epoch": 1.4462736406188483,
      "grad_norm": 4.875,
      "learning_rate": 1.8788388809666259e-06,
      "loss": 0.79790506,
      "memory(GiB)": 147.13,
      "step": 61990,
      "train_speed(iter/s)": 0.201407
    },
    {
      "acc": 0.78528218,
      "epoch": 1.4465069481911372,
      "grad_norm": 5.21875,
      "learning_rate": 1.877363256382832e-06,
      "loss": 0.7618494,
      "memory(GiB)": 147.13,
      "step": 62000,
      "train_speed(iter/s)": 0.201423
    },
    {
      "epoch": 1.4465069481911372,
      "eval_acc": 0.7446714809225781,
      "eval_loss": 0.8045513033866882,
      "eval_runtime": 1270.3592,
      "eval_samples_per_second": 28.331,
      "eval_steps_per_second": 14.166,
      "step": 62000
    },
    {
      "acc": 0.78110247,
      "epoch": 1.4467402557634261,
      "grad_norm": 7.09375,
      "learning_rate": 1.8758880775545279e-06,
      "loss": 0.77745328,
      "memory(GiB)": 147.13,
      "step": 62010,
      "train_speed(iter/s)": 0.200597
    },
    {
      "acc": 0.78310804,
      "epoch": 1.446973563335715,
      "grad_norm": 4.15625,
      "learning_rate": 1.8744133446922935e-06,
      "loss": 0.78575201,
      "memory(GiB)": 147.13,
      "step": 62020,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.77370396,
      "epoch": 1.447206870908004,
      "grad_norm": 4.6875,
      "learning_rate": 1.8729390580066442e-06,
      "loss": 0.82233162,
      "memory(GiB)": 147.13,
      "step": 62030,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.77510133,
      "epoch": 1.4474401784802928,
      "grad_norm": 6.09375,
      "learning_rate": 1.8714652177080377e-06,
      "loss": 0.80517254,
      "memory(GiB)": 147.13,
      "step": 62040,
      "train_speed(iter/s)": 0.200645
    },
    {
      "acc": 0.78512564,
      "epoch": 1.4476734860525817,
      "grad_norm": 5.9375,
      "learning_rate": 1.869991824006861e-06,
      "loss": 0.78300362,
      "memory(GiB)": 147.13,
      "step": 62050,
      "train_speed(iter/s)": 0.200662
    },
    {
      "acc": 0.78685694,
      "epoch": 1.4479067936248706,
      "grad_norm": 20.25,
      "learning_rate": 1.8685188771134433e-06,
      "loss": 0.77022228,
      "memory(GiB)": 147.13,
      "step": 62060,
      "train_speed(iter/s)": 0.200678
    },
    {
      "acc": 0.76881313,
      "epoch": 1.4481401011971595,
      "grad_norm": 5.09375,
      "learning_rate": 1.8670463772380464e-06,
      "loss": 0.82338123,
      "memory(GiB)": 147.13,
      "step": 62070,
      "train_speed(iter/s)": 0.200695
    },
    {
      "acc": 0.79330425,
      "epoch": 1.4483734087694484,
      "grad_norm": 6.59375,
      "learning_rate": 1.8655743245908692e-06,
      "loss": 0.73466892,
      "memory(GiB)": 147.13,
      "step": 62080,
      "train_speed(iter/s)": 0.200711
    },
    {
      "acc": 0.79285131,
      "epoch": 1.4486067163417373,
      "grad_norm": 6.3125,
      "learning_rate": 1.864102719382045e-06,
      "loss": 0.73613644,
      "memory(GiB)": 147.13,
      "step": 62090,
      "train_speed(iter/s)": 0.200727
    },
    {
      "acc": 0.79956055,
      "epoch": 1.4488400239140262,
      "grad_norm": 4.125,
      "learning_rate": 1.8626315618216484e-06,
      "loss": 0.72277913,
      "memory(GiB)": 147.13,
      "step": 62100,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.77584467,
      "epoch": 1.4490733314863151,
      "grad_norm": 4.75,
      "learning_rate": 1.8611608521196844e-06,
      "loss": 0.80201588,
      "memory(GiB)": 147.13,
      "step": 62110,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.78103304,
      "epoch": 1.449306639058604,
      "grad_norm": 8.125,
      "learning_rate": 1.8596905904860956e-06,
      "loss": 0.78463149,
      "memory(GiB)": 147.13,
      "step": 62120,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.78027878,
      "epoch": 1.449539946630893,
      "grad_norm": 6.65625,
      "learning_rate": 1.8582207771307647e-06,
      "loss": 0.78982868,
      "memory(GiB)": 147.13,
      "step": 62130,
      "train_speed(iter/s)": 0.200794
    },
    {
      "acc": 0.80753584,
      "epoch": 1.4497732542031816,
      "grad_norm": 6.125,
      "learning_rate": 1.8567514122635027e-06,
      "loss": 0.68397255,
      "memory(GiB)": 147.13,
      "step": 62140,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.80782089,
      "epoch": 1.4500065617754707,
      "grad_norm": 7.21875,
      "learning_rate": 1.8552824960940658e-06,
      "loss": 0.67650642,
      "memory(GiB)": 147.13,
      "step": 62150,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.77864218,
      "epoch": 1.4502398693477594,
      "grad_norm": 7.0,
      "learning_rate": 1.8538140288321387e-06,
      "loss": 0.80679083,
      "memory(GiB)": 147.13,
      "step": 62160,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.79958267,
      "epoch": 1.4504731769200485,
      "grad_norm": 6.15625,
      "learning_rate": 1.8523460106873436e-06,
      "loss": 0.7120677,
      "memory(GiB)": 147.13,
      "step": 62170,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.77506552,
      "epoch": 1.4507064844923372,
      "grad_norm": 6.71875,
      "learning_rate": 1.8508784418692428e-06,
      "loss": 0.83289452,
      "memory(GiB)": 147.13,
      "step": 62180,
      "train_speed(iter/s)": 0.200876
    },
    {
      "acc": 0.80013189,
      "epoch": 1.4509397920646263,
      "grad_norm": 5.0625,
      "learning_rate": 1.8494113225873295e-06,
      "loss": 0.70295601,
      "memory(GiB)": 147.13,
      "step": 62190,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.79650593,
      "epoch": 1.451173099636915,
      "grad_norm": 8.4375,
      "learning_rate": 1.8479446530510348e-06,
      "loss": 0.72814693,
      "memory(GiB)": 147.13,
      "step": 62200,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.77118607,
      "epoch": 1.451406407209204,
      "grad_norm": 4.09375,
      "learning_rate": 1.8464784334697234e-06,
      "loss": 0.84197454,
      "memory(GiB)": 147.13,
      "step": 62210,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.79284525,
      "epoch": 1.4516397147814928,
      "grad_norm": 4.78125,
      "learning_rate": 1.845012664052701e-06,
      "loss": 0.74894667,
      "memory(GiB)": 147.13,
      "step": 62220,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.77012434,
      "epoch": 1.451873022353782,
      "grad_norm": 5.1875,
      "learning_rate": 1.843547345009203e-06,
      "loss": 0.81883812,
      "memory(GiB)": 147.13,
      "step": 62230,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.77381506,
      "epoch": 1.4521063299260706,
      "grad_norm": 6.71875,
      "learning_rate": 1.8420824765484058e-06,
      "loss": 0.79767299,
      "memory(GiB)": 147.13,
      "step": 62240,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.77434549,
      "epoch": 1.4523396374983595,
      "grad_norm": 4.9375,
      "learning_rate": 1.8406180588794176e-06,
      "loss": 0.81015167,
      "memory(GiB)": 147.13,
      "step": 62250,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.79392915,
      "epoch": 1.4525729450706484,
      "grad_norm": 6.53125,
      "learning_rate": 1.8391540922112822e-06,
      "loss": 0.72809219,
      "memory(GiB)": 147.13,
      "step": 62260,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.77831402,
      "epoch": 1.4528062526429373,
      "grad_norm": 5.59375,
      "learning_rate": 1.8376905767529834e-06,
      "loss": 0.78499942,
      "memory(GiB)": 147.13,
      "step": 62270,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.79829054,
      "epoch": 1.4530395602152262,
      "grad_norm": 6.21875,
      "learning_rate": 1.8362275127134348e-06,
      "loss": 0.71653814,
      "memory(GiB)": 147.13,
      "step": 62280,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.75779715,
      "epoch": 1.453272867787515,
      "grad_norm": 5.25,
      "learning_rate": 1.8347649003014911e-06,
      "loss": 0.86373796,
      "memory(GiB)": 147.13,
      "step": 62290,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.76059661,
      "epoch": 1.453506175359804,
      "grad_norm": 8.1875,
      "learning_rate": 1.833302739725939e-06,
      "loss": 0.87852764,
      "memory(GiB)": 147.13,
      "step": 62300,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.77951355,
      "epoch": 1.4537394829320929,
      "grad_norm": 11.6875,
      "learning_rate": 1.8318410311955003e-06,
      "loss": 0.7952219,
      "memory(GiB)": 147.13,
      "step": 62310,
      "train_speed(iter/s)": 0.201087
    },
    {
      "acc": 0.78401728,
      "epoch": 1.4539727905043818,
      "grad_norm": 4.6875,
      "learning_rate": 1.830379774918834e-06,
      "loss": 0.77956657,
      "memory(GiB)": 147.13,
      "step": 62320,
      "train_speed(iter/s)": 0.201102
    },
    {
      "acc": 0.78924007,
      "epoch": 1.4542060980766707,
      "grad_norm": 5.59375,
      "learning_rate": 1.8289189711045324e-06,
      "loss": 0.75395088,
      "memory(GiB)": 147.13,
      "step": 62330,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.77698288,
      "epoch": 1.4544394056489596,
      "grad_norm": 4.71875,
      "learning_rate": 1.8274586199611283e-06,
      "loss": 0.78425674,
      "memory(GiB)": 147.13,
      "step": 62340,
      "train_speed(iter/s)": 0.201135
    },
    {
      "acc": 0.78032331,
      "epoch": 1.4546727132212485,
      "grad_norm": 6.59375,
      "learning_rate": 1.8259987216970826e-06,
      "loss": 0.79403481,
      "memory(GiB)": 147.13,
      "step": 62350,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.7868978,
      "epoch": 1.4549060207935374,
      "grad_norm": 4.9375,
      "learning_rate": 1.8245392765207993e-06,
      "loss": 0.75227661,
      "memory(GiB)": 147.13,
      "step": 62360,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.78523664,
      "epoch": 1.4551393283658263,
      "grad_norm": 5.40625,
      "learning_rate": 1.8230802846406104e-06,
      "loss": 0.76713495,
      "memory(GiB)": 147.13,
      "step": 62370,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.77572908,
      "epoch": 1.4553726359381152,
      "grad_norm": 5.84375,
      "learning_rate": 1.821621746264789e-06,
      "loss": 0.80009136,
      "memory(GiB)": 147.13,
      "step": 62380,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.78314152,
      "epoch": 1.455605943510404,
      "grad_norm": 7.15625,
      "learning_rate": 1.8201636616015405e-06,
      "loss": 0.77063742,
      "memory(GiB)": 147.13,
      "step": 62390,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.77914481,
      "epoch": 1.455839251082693,
      "grad_norm": 4.53125,
      "learning_rate": 1.8187060308590038e-06,
      "loss": 0.77920589,
      "memory(GiB)": 147.13,
      "step": 62400,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.79587164,
      "epoch": 1.4560725586549819,
      "grad_norm": 5.65625,
      "learning_rate": 1.8172488542452583e-06,
      "loss": 0.73282957,
      "memory(GiB)": 147.13,
      "step": 62410,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.79226098,
      "epoch": 1.4563058662272708,
      "grad_norm": 5.65625,
      "learning_rate": 1.8157921319683147e-06,
      "loss": 0.73642225,
      "memory(GiB)": 147.13,
      "step": 62420,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.76984286,
      "epoch": 1.4565391737995597,
      "grad_norm": 5.0625,
      "learning_rate": 1.8143358642361191e-06,
      "loss": 0.83457851,
      "memory(GiB)": 147.13,
      "step": 62430,
      "train_speed(iter/s)": 0.201282
    },
    {
      "acc": 0.78718529,
      "epoch": 1.4567724813718486,
      "grad_norm": 5.3125,
      "learning_rate": 1.8128800512565514e-06,
      "loss": 0.75155096,
      "memory(GiB)": 147.13,
      "step": 62440,
      "train_speed(iter/s)": 0.201299
    },
    {
      "acc": 0.78033142,
      "epoch": 1.4570057889441375,
      "grad_norm": 4.03125,
      "learning_rate": 1.811424693237433e-06,
      "loss": 0.77643776,
      "memory(GiB)": 147.13,
      "step": 62450,
      "train_speed(iter/s)": 0.201317
    },
    {
      "acc": 0.80350819,
      "epoch": 1.4572390965164264,
      "grad_norm": 4.40625,
      "learning_rate": 1.8099697903865127e-06,
      "loss": 0.69405346,
      "memory(GiB)": 147.13,
      "step": 62460,
      "train_speed(iter/s)": 0.201333
    },
    {
      "acc": 0.7767561,
      "epoch": 1.4574724040887153,
      "grad_norm": 4.09375,
      "learning_rate": 1.8085153429114766e-06,
      "loss": 0.80634718,
      "memory(GiB)": 147.13,
      "step": 62470,
      "train_speed(iter/s)": 0.201351
    },
    {
      "acc": 0.79806414,
      "epoch": 1.4577057116610042,
      "grad_norm": 4.71875,
      "learning_rate": 1.8070613510199497e-06,
      "loss": 0.72472191,
      "memory(GiB)": 147.13,
      "step": 62480,
      "train_speed(iter/s)": 0.201368
    },
    {
      "acc": 0.7791873,
      "epoch": 1.457939019233293,
      "grad_norm": 4.75,
      "learning_rate": 1.8056078149194861e-06,
      "loss": 0.78082762,
      "memory(GiB)": 147.13,
      "step": 62490,
      "train_speed(iter/s)": 0.201385
    },
    {
      "acc": 0.78747339,
      "epoch": 1.458172326805582,
      "grad_norm": 7.03125,
      "learning_rate": 1.8041547348175803e-06,
      "loss": 0.76401696,
      "memory(GiB)": 147.13,
      "step": 62500,
      "train_speed(iter/s)": 0.201402
    },
    {
      "epoch": 1.458172326805582,
      "eval_acc": 0.7446378506337948,
      "eval_loss": 0.8045361042022705,
      "eval_runtime": 1270.7937,
      "eval_samples_per_second": 28.322,
      "eval_steps_per_second": 14.161,
      "step": 62500
    },
    {
      "acc": 0.76177197,
      "epoch": 1.4584056343778709,
      "grad_norm": 4.9375,
      "learning_rate": 1.802702110921658e-06,
      "loss": 0.85588417,
      "memory(GiB)": 147.13,
      "step": 62510,
      "train_speed(iter/s)": 0.200583
    },
    {
      "acc": 0.7588037,
      "epoch": 1.4586389419501598,
      "grad_norm": 6.59375,
      "learning_rate": 1.8012499434390784e-06,
      "loss": 0.8837925,
      "memory(GiB)": 147.13,
      "step": 62520,
      "train_speed(iter/s)": 0.200598
    },
    {
      "acc": 0.76614499,
      "epoch": 1.4588722495224484,
      "grad_norm": 5.90625,
      "learning_rate": 1.7997982325771425e-06,
      "loss": 0.83949127,
      "memory(GiB)": 147.13,
      "step": 62530,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.78617134,
      "epoch": 1.4591055570947375,
      "grad_norm": 6.21875,
      "learning_rate": 1.7983469785430785e-06,
      "loss": 0.79134507,
      "memory(GiB)": 147.13,
      "step": 62540,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.79341078,
      "epoch": 1.4593388646670262,
      "grad_norm": 4.3125,
      "learning_rate": 1.7968961815440534e-06,
      "loss": 0.76289606,
      "memory(GiB)": 147.13,
      "step": 62550,
      "train_speed(iter/s)": 0.200646
    },
    {
      "acc": 0.77591524,
      "epoch": 1.4595721722393153,
      "grad_norm": 7.96875,
      "learning_rate": 1.7954458417871667e-06,
      "loss": 0.80238419,
      "memory(GiB)": 147.13,
      "step": 62560,
      "train_speed(iter/s)": 0.200663
    },
    {
      "acc": 0.79817991,
      "epoch": 1.459805479811604,
      "grad_norm": 5.1875,
      "learning_rate": 1.7939959594794564e-06,
      "loss": 0.71526041,
      "memory(GiB)": 147.13,
      "step": 62570,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.77496409,
      "epoch": 1.4600387873838931,
      "grad_norm": 7.46875,
      "learning_rate": 1.7925465348278898e-06,
      "loss": 0.7962131,
      "memory(GiB)": 147.13,
      "step": 62580,
      "train_speed(iter/s)": 0.200697
    },
    {
      "acc": 0.78668814,
      "epoch": 1.4602720949561818,
      "grad_norm": 6.4375,
      "learning_rate": 1.7910975680393756e-06,
      "loss": 0.76099,
      "memory(GiB)": 147.13,
      "step": 62590,
      "train_speed(iter/s)": 0.200714
    },
    {
      "acc": 0.77080307,
      "epoch": 1.460505402528471,
      "grad_norm": 5.15625,
      "learning_rate": 1.789649059320751e-06,
      "loss": 0.82710533,
      "memory(GiB)": 147.13,
      "step": 62600,
      "train_speed(iter/s)": 0.20073
    },
    {
      "acc": 0.77813931,
      "epoch": 1.4607387101007596,
      "grad_norm": 5.8125,
      "learning_rate": 1.7882010088787888e-06,
      "loss": 0.801579,
      "memory(GiB)": 147.13,
      "step": 62610,
      "train_speed(iter/s)": 0.200749
    },
    {
      "acc": 0.77716465,
      "epoch": 1.4609720176730487,
      "grad_norm": 5.9375,
      "learning_rate": 1.7867534169202018e-06,
      "loss": 0.80519314,
      "memory(GiB)": 147.13,
      "step": 62620,
      "train_speed(iter/s)": 0.200766
    },
    {
      "acc": 0.75499105,
      "epoch": 1.4612053252453374,
      "grad_norm": 4.65625,
      "learning_rate": 1.785306283651629e-06,
      "loss": 0.88495131,
      "memory(GiB)": 147.13,
      "step": 62630,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.77682557,
      "epoch": 1.4614386328176263,
      "grad_norm": 6.34375,
      "learning_rate": 1.783859609279654e-06,
      "loss": 0.80335407,
      "memory(GiB)": 147.13,
      "step": 62640,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.78338518,
      "epoch": 1.4616719403899152,
      "grad_norm": 5.4375,
      "learning_rate": 1.7824133940107818e-06,
      "loss": 0.77647243,
      "memory(GiB)": 147.13,
      "step": 62650,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.77094107,
      "epoch": 1.4619052479622041,
      "grad_norm": 4.375,
      "learning_rate": 1.7809676380514646e-06,
      "loss": 0.84018555,
      "memory(GiB)": 147.13,
      "step": 62660,
      "train_speed(iter/s)": 0.200835
    },
    {
      "acc": 0.78348899,
      "epoch": 1.462138555534493,
      "grad_norm": 5.78125,
      "learning_rate": 1.7795223416080804e-06,
      "loss": 0.7956295,
      "memory(GiB)": 147.13,
      "step": 62670,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.77315598,
      "epoch": 1.462371863106782,
      "grad_norm": 5.59375,
      "learning_rate": 1.778077504886948e-06,
      "loss": 0.81796093,
      "memory(GiB)": 147.13,
      "step": 62680,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.7750021,
      "epoch": 1.4626051706790708,
      "grad_norm": 7.0625,
      "learning_rate": 1.7766331280943156e-06,
      "loss": 0.80663986,
      "memory(GiB)": 147.13,
      "step": 62690,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.78340206,
      "epoch": 1.4628384782513597,
      "grad_norm": 5.28125,
      "learning_rate": 1.775189211436366e-06,
      "loss": 0.78596182,
      "memory(GiB)": 147.13,
      "step": 62700,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.77824869,
      "epoch": 1.4630717858236486,
      "grad_norm": 5.25,
      "learning_rate": 1.7737457551192221e-06,
      "loss": 0.80983734,
      "memory(GiB)": 147.13,
      "step": 62710,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.77051554,
      "epoch": 1.4633050933959375,
      "grad_norm": 6.8125,
      "learning_rate": 1.7723027593489322e-06,
      "loss": 0.82581806,
      "memory(GiB)": 147.13,
      "step": 62720,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.76511116,
      "epoch": 1.4635384009682264,
      "grad_norm": 7.53125,
      "learning_rate": 1.7708602243314876e-06,
      "loss": 0.84343872,
      "memory(GiB)": 147.13,
      "step": 62730,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.78928823,
      "epoch": 1.4637717085405153,
      "grad_norm": 4.3125,
      "learning_rate": 1.7694181502728074e-06,
      "loss": 0.74881258,
      "memory(GiB)": 147.13,
      "step": 62740,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.77590837,
      "epoch": 1.4640050161128042,
      "grad_norm": 6.125,
      "learning_rate": 1.7679765373787467e-06,
      "loss": 0.78589211,
      "memory(GiB)": 147.13,
      "step": 62750,
      "train_speed(iter/s)": 0.200987
    },
    {
      "acc": 0.78272462,
      "epoch": 1.464238323685093,
      "grad_norm": 6.59375,
      "learning_rate": 1.7665353858550993e-06,
      "loss": 0.75262532,
      "memory(GiB)": 147.13,
      "step": 62760,
      "train_speed(iter/s)": 0.201003
    },
    {
      "acc": 0.77953529,
      "epoch": 1.464471631257382,
      "grad_norm": 7.4375,
      "learning_rate": 1.7650946959075833e-06,
      "loss": 0.81623087,
      "memory(GiB)": 147.13,
      "step": 62770,
      "train_speed(iter/s)": 0.20102
    },
    {
      "acc": 0.77969527,
      "epoch": 1.464704938829671,
      "grad_norm": 5.53125,
      "learning_rate": 1.763654467741861e-06,
      "loss": 0.78417301,
      "memory(GiB)": 147.13,
      "step": 62780,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.79427376,
      "epoch": 1.4649382464019598,
      "grad_norm": 6.53125,
      "learning_rate": 1.7622147015635222e-06,
      "loss": 0.75227299,
      "memory(GiB)": 147.13,
      "step": 62790,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.77105227,
      "epoch": 1.4651715539742487,
      "grad_norm": 4.75,
      "learning_rate": 1.760775397578095e-06,
      "loss": 0.83062248,
      "memory(GiB)": 147.13,
      "step": 62800,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.7859479,
      "epoch": 1.4654048615465376,
      "grad_norm": 5.375,
      "learning_rate": 1.7593365559910397e-06,
      "loss": 0.76885614,
      "memory(GiB)": 147.13,
      "step": 62810,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.76209989,
      "epoch": 1.4656381691188265,
      "grad_norm": 6.4375,
      "learning_rate": 1.7578981770077474e-06,
      "loss": 0.84865894,
      "memory(GiB)": 147.13,
      "step": 62820,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.78125958,
      "epoch": 1.4658714766911154,
      "grad_norm": 6.0625,
      "learning_rate": 1.7564602608335502e-06,
      "loss": 0.77670856,
      "memory(GiB)": 147.13,
      "step": 62830,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.79948587,
      "epoch": 1.4661047842634043,
      "grad_norm": 7.4375,
      "learning_rate": 1.7550228076737069e-06,
      "loss": 0.72959423,
      "memory(GiB)": 147.13,
      "step": 62840,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.75941429,
      "epoch": 1.4663380918356932,
      "grad_norm": 5.75,
      "learning_rate": 1.7535858177334163e-06,
      "loss": 0.86305752,
      "memory(GiB)": 147.13,
      "step": 62850,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.78896446,
      "epoch": 1.466571399407982,
      "grad_norm": 4.28125,
      "learning_rate": 1.7521492912178062e-06,
      "loss": 0.75683656,
      "memory(GiB)": 147.13,
      "step": 62860,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.79165177,
      "epoch": 1.466804706980271,
      "grad_norm": 5.40625,
      "learning_rate": 1.7507132283319445e-06,
      "loss": 0.73763084,
      "memory(GiB)": 147.13,
      "step": 62870,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.76576929,
      "epoch": 1.4670380145525599,
      "grad_norm": 4.1875,
      "learning_rate": 1.7492776292808217e-06,
      "loss": 0.84900112,
      "memory(GiB)": 147.13,
      "step": 62880,
      "train_speed(iter/s)": 0.201201
    },
    {
      "acc": 0.77859445,
      "epoch": 1.4672713221248488,
      "grad_norm": 4.78125,
      "learning_rate": 1.7478424942693751e-06,
      "loss": 0.79272537,
      "memory(GiB)": 147.13,
      "step": 62890,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.7813118,
      "epoch": 1.4675046296971377,
      "grad_norm": 6.9375,
      "learning_rate": 1.7464078235024678e-06,
      "loss": 0.78525038,
      "memory(GiB)": 147.13,
      "step": 62900,
      "train_speed(iter/s)": 0.201233
    },
    {
      "acc": 0.78674507,
      "epoch": 1.4677379372694266,
      "grad_norm": 7.4375,
      "learning_rate": 1.7449736171848964e-06,
      "loss": 0.76398363,
      "memory(GiB)": 147.13,
      "step": 62910,
      "train_speed(iter/s)": 0.201248
    },
    {
      "acc": 0.77175941,
      "epoch": 1.4679712448417153,
      "grad_norm": 9.0625,
      "learning_rate": 1.7435398755213977e-06,
      "loss": 0.8071166,
      "memory(GiB)": 147.13,
      "step": 62920,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.75375805,
      "epoch": 1.4682045524140044,
      "grad_norm": 6.59375,
      "learning_rate": 1.7421065987166335e-06,
      "loss": 0.87107162,
      "memory(GiB)": 147.13,
      "step": 62930,
      "train_speed(iter/s)": 0.201281
    },
    {
      "acc": 0.80961304,
      "epoch": 1.468437859986293,
      "grad_norm": 5.34375,
      "learning_rate": 1.7406737869752082e-06,
      "loss": 0.67591333,
      "memory(GiB)": 147.13,
      "step": 62940,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.77118597,
      "epoch": 1.4686711675585822,
      "grad_norm": 5.28125,
      "learning_rate": 1.7392414405016527e-06,
      "loss": 0.83532896,
      "memory(GiB)": 147.13,
      "step": 62950,
      "train_speed(iter/s)": 0.201312
    },
    {
      "acc": 0.76841412,
      "epoch": 1.4689044751308709,
      "grad_norm": 6.3125,
      "learning_rate": 1.7378095595004323e-06,
      "loss": 0.84016171,
      "memory(GiB)": 147.13,
      "step": 62960,
      "train_speed(iter/s)": 0.201329
    },
    {
      "acc": 0.77326422,
      "epoch": 1.46913778270316,
      "grad_norm": 6.40625,
      "learning_rate": 1.736378144175952e-06,
      "loss": 0.83091669,
      "memory(GiB)": 147.13,
      "step": 62970,
      "train_speed(iter/s)": 0.201346
    },
    {
      "acc": 0.7910758,
      "epoch": 1.4693710902754487,
      "grad_norm": 6.0625,
      "learning_rate": 1.7349471947325414e-06,
      "loss": 0.72511082,
      "memory(GiB)": 147.13,
      "step": 62980,
      "train_speed(iter/s)": 0.201363
    },
    {
      "acc": 0.78480339,
      "epoch": 1.4696043978477378,
      "grad_norm": 4.28125,
      "learning_rate": 1.7335167113744732e-06,
      "loss": 0.76047554,
      "memory(GiB)": 147.13,
      "step": 62990,
      "train_speed(iter/s)": 0.20138
    },
    {
      "acc": 0.78779182,
      "epoch": 1.4698377054200265,
      "grad_norm": 4.3125,
      "learning_rate": 1.7320866943059427e-06,
      "loss": 0.75346622,
      "memory(GiB)": 147.13,
      "step": 63000,
      "train_speed(iter/s)": 0.201397
    },
    {
      "epoch": 1.4698377054200265,
      "eval_acc": 0.7446705200571843,
      "eval_loss": 0.8045252561569214,
      "eval_runtime": 1270.0191,
      "eval_samples_per_second": 28.339,
      "eval_steps_per_second": 14.17,
      "step": 63000
    },
    {
      "acc": 0.79027567,
      "epoch": 1.4700710129923154,
      "grad_norm": 5.84375,
      "learning_rate": 1.7306571437310893e-06,
      "loss": 0.75496302,
      "memory(GiB)": 147.13,
      "step": 63010,
      "train_speed(iter/s)": 0.200586
    },
    {
      "acc": 0.77779145,
      "epoch": 1.4703043205646043,
      "grad_norm": 5.15625,
      "learning_rate": 1.7292280598539769e-06,
      "loss": 0.80010624,
      "memory(GiB)": 147.13,
      "step": 63020,
      "train_speed(iter/s)": 0.200603
    },
    {
      "acc": 0.78809614,
      "epoch": 1.4705376281368931,
      "grad_norm": 5.375,
      "learning_rate": 1.72779944287861e-06,
      "loss": 0.7661644,
      "memory(GiB)": 147.13,
      "step": 63030,
      "train_speed(iter/s)": 0.20062
    },
    {
      "acc": 0.78059311,
      "epoch": 1.470770935709182,
      "grad_norm": 4.875,
      "learning_rate": 1.7263712930089227e-06,
      "loss": 0.78118305,
      "memory(GiB)": 147.13,
      "step": 63040,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.78940554,
      "epoch": 1.471004243281471,
      "grad_norm": 9.75,
      "learning_rate": 1.7249436104487805e-06,
      "loss": 0.74943719,
      "memory(GiB)": 147.13,
      "step": 63050,
      "train_speed(iter/s)": 0.200651
    },
    {
      "acc": 0.77817397,
      "epoch": 1.4712375508537598,
      "grad_norm": 3.9375,
      "learning_rate": 1.7235163954019878e-06,
      "loss": 0.83013039,
      "memory(GiB)": 147.13,
      "step": 63060,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.7942832,
      "epoch": 1.4714708584260487,
      "grad_norm": 4.71875,
      "learning_rate": 1.7220896480722766e-06,
      "loss": 0.71083398,
      "memory(GiB)": 147.13,
      "step": 63070,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.77893252,
      "epoch": 1.4717041659983376,
      "grad_norm": 5.28125,
      "learning_rate": 1.7206633686633172e-06,
      "loss": 0.7933465,
      "memory(GiB)": 147.13,
      "step": 63080,
      "train_speed(iter/s)": 0.200699
    },
    {
      "acc": 0.77962651,
      "epoch": 1.4719374735706265,
      "grad_norm": 3.625,
      "learning_rate": 1.719237557378709e-06,
      "loss": 0.79189463,
      "memory(GiB)": 147.13,
      "step": 63090,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.7757308,
      "epoch": 1.4721707811429154,
      "grad_norm": 4.59375,
      "learning_rate": 1.7178122144219873e-06,
      "loss": 0.79125972,
      "memory(GiB)": 147.13,
      "step": 63100,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.78209968,
      "epoch": 1.4724040887152043,
      "grad_norm": 4.25,
      "learning_rate": 1.716387339996618e-06,
      "loss": 0.78565693,
      "memory(GiB)": 147.13,
      "step": 63110,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.76956949,
      "epoch": 1.4726373962874932,
      "grad_norm": 7.1875,
      "learning_rate": 1.7149629343060003e-06,
      "loss": 0.82966146,
      "memory(GiB)": 147.13,
      "step": 63120,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.78773918,
      "epoch": 1.4728707038597821,
      "grad_norm": 4.8125,
      "learning_rate": 1.7135389975534711e-06,
      "loss": 0.76110039,
      "memory(GiB)": 147.13,
      "step": 63130,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.7758707,
      "epoch": 1.473104011432071,
      "grad_norm": 4.84375,
      "learning_rate": 1.7121155299422936e-06,
      "loss": 0.81253977,
      "memory(GiB)": 147.13,
      "step": 63140,
      "train_speed(iter/s)": 0.200795
    },
    {
      "acc": 0.75602198,
      "epoch": 1.47333731900436,
      "grad_norm": 5.1875,
      "learning_rate": 1.710692531675671e-06,
      "loss": 0.90366888,
      "memory(GiB)": 147.13,
      "step": 63150,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.77753906,
      "epoch": 1.4735706265766488,
      "grad_norm": 5.40625,
      "learning_rate": 1.709270002956732e-06,
      "loss": 0.77708693,
      "memory(GiB)": 147.13,
      "step": 63160,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.77087703,
      "epoch": 1.4738039341489377,
      "grad_norm": 5.6875,
      "learning_rate": 1.7078479439885458e-06,
      "loss": 0.82326393,
      "memory(GiB)": 147.13,
      "step": 63170,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.78759995,
      "epoch": 1.4740372417212266,
      "grad_norm": 7.46875,
      "learning_rate": 1.7064263549741095e-06,
      "loss": 0.76839981,
      "memory(GiB)": 147.13,
      "step": 63180,
      "train_speed(iter/s)": 0.200862
    },
    {
      "acc": 0.79642954,
      "epoch": 1.4742705492935155,
      "grad_norm": 5.34375,
      "learning_rate": 1.7050052361163522e-06,
      "loss": 0.7275032,
      "memory(GiB)": 147.13,
      "step": 63190,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.79400454,
      "epoch": 1.4745038568658044,
      "grad_norm": 5.375,
      "learning_rate": 1.7035845876181422e-06,
      "loss": 0.74391875,
      "memory(GiB)": 147.13,
      "step": 63200,
      "train_speed(iter/s)": 0.200895
    },
    {
      "acc": 0.7892807,
      "epoch": 1.4747371644380933,
      "grad_norm": 12.75,
      "learning_rate": 1.7021644096822748e-06,
      "loss": 0.76604948,
      "memory(GiB)": 147.13,
      "step": 63210,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.76153727,
      "epoch": 1.4749704720103822,
      "grad_norm": 4.125,
      "learning_rate": 1.7007447025114798e-06,
      "loss": 0.86720562,
      "memory(GiB)": 147.13,
      "step": 63220,
      "train_speed(iter/s)": 0.200928
    },
    {
      "acc": 0.77364707,
      "epoch": 1.4752037795826711,
      "grad_norm": 5.8125,
      "learning_rate": 1.699325466308418e-06,
      "loss": 0.81828022,
      "memory(GiB)": 147.13,
      "step": 63230,
      "train_speed(iter/s)": 0.200946
    },
    {
      "acc": 0.77053213,
      "epoch": 1.47543708715496,
      "grad_norm": 4.21875,
      "learning_rate": 1.6979067012756888e-06,
      "loss": 0.83672915,
      "memory(GiB)": 147.13,
      "step": 63240,
      "train_speed(iter/s)": 0.200962
    },
    {
      "acc": 0.78983154,
      "epoch": 1.475670394727249,
      "grad_norm": 5.1875,
      "learning_rate": 1.6964884076158194e-06,
      "loss": 0.76191545,
      "memory(GiB)": 147.13,
      "step": 63250,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.75502505,
      "epoch": 1.4759037022995378,
      "grad_norm": 5.3125,
      "learning_rate": 1.6950705855312677e-06,
      "loss": 0.89188004,
      "memory(GiB)": 147.13,
      "step": 63260,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.7857728,
      "epoch": 1.4761370098718267,
      "grad_norm": 5.21875,
      "learning_rate": 1.6936532352244316e-06,
      "loss": 0.75499687,
      "memory(GiB)": 147.13,
      "step": 63270,
      "train_speed(iter/s)": 0.201011
    },
    {
      "acc": 0.78429489,
      "epoch": 1.4763703174441156,
      "grad_norm": 5.625,
      "learning_rate": 1.6922363568976347e-06,
      "loss": 0.78085718,
      "memory(GiB)": 147.13,
      "step": 63280,
      "train_speed(iter/s)": 0.201027
    },
    {
      "acc": 0.78143988,
      "epoch": 1.4766036250164045,
      "grad_norm": 5.75,
      "learning_rate": 1.690819950753138e-06,
      "loss": 0.77364426,
      "memory(GiB)": 147.13,
      "step": 63290,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.7872983,
      "epoch": 1.4768369325886934,
      "grad_norm": 4.6875,
      "learning_rate": 1.6894040169931303e-06,
      "loss": 0.75831628,
      "memory(GiB)": 147.13,
      "step": 63300,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.7711751,
      "epoch": 1.477070240160982,
      "grad_norm": 4.875,
      "learning_rate": 1.6879885558197395e-06,
      "loss": 0.80848656,
      "memory(GiB)": 147.13,
      "step": 63310,
      "train_speed(iter/s)": 0.201076
    },
    {
      "acc": 0.78801699,
      "epoch": 1.4773035477332712,
      "grad_norm": 4.125,
      "learning_rate": 1.6865735674350198e-06,
      "loss": 0.74979692,
      "memory(GiB)": 147.13,
      "step": 63320,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.8035552,
      "epoch": 1.47753685530556,
      "grad_norm": 4.4375,
      "learning_rate": 1.6851590520409611e-06,
      "loss": 0.69476547,
      "memory(GiB)": 147.13,
      "step": 63330,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.78817482,
      "epoch": 1.477770162877849,
      "grad_norm": 6.46875,
      "learning_rate": 1.6837450098394848e-06,
      "loss": 0.76696267,
      "memory(GiB)": 147.13,
      "step": 63340,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.79101725,
      "epoch": 1.4780034704501377,
      "grad_norm": 5.46875,
      "learning_rate": 1.6823314410324426e-06,
      "loss": 0.74547434,
      "memory(GiB)": 147.13,
      "step": 63350,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.78399458,
      "epoch": 1.4782367780224268,
      "grad_norm": 9.375,
      "learning_rate": 1.680918345821626e-06,
      "loss": 0.78179579,
      "memory(GiB)": 147.13,
      "step": 63360,
      "train_speed(iter/s)": 0.201158
    },
    {
      "acc": 0.80156345,
      "epoch": 1.4784700855947155,
      "grad_norm": 5.4375,
      "learning_rate": 1.6795057244087493e-06,
      "loss": 0.70179181,
      "memory(GiB)": 147.13,
      "step": 63370,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.79969625,
      "epoch": 1.4787033931670046,
      "grad_norm": 7.09375,
      "learning_rate": 1.678093576995467e-06,
      "loss": 0.71199322,
      "memory(GiB)": 147.13,
      "step": 63380,
      "train_speed(iter/s)": 0.201189
    },
    {
      "acc": 0.77803707,
      "epoch": 1.4789367007392933,
      "grad_norm": 5.8125,
      "learning_rate": 1.676681903783362e-06,
      "loss": 0.77450371,
      "memory(GiB)": 147.13,
      "step": 63390,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.79645801,
      "epoch": 1.4791700083115822,
      "grad_norm": 5.0,
      "learning_rate": 1.6752707049739487e-06,
      "loss": 0.72130475,
      "memory(GiB)": 147.13,
      "step": 63400,
      "train_speed(iter/s)": 0.201221
    },
    {
      "acc": 0.78377075,
      "epoch": 1.479403315883871,
      "grad_norm": 5.9375,
      "learning_rate": 1.6738599807686774e-06,
      "loss": 0.78707418,
      "memory(GiB)": 147.13,
      "step": 63410,
      "train_speed(iter/s)": 0.201237
    },
    {
      "acc": 0.7851202,
      "epoch": 1.47963662345616,
      "grad_norm": 5.53125,
      "learning_rate": 1.6724497313689258e-06,
      "loss": 0.76640143,
      "memory(GiB)": 147.13,
      "step": 63420,
      "train_speed(iter/s)": 0.201253
    },
    {
      "acc": 0.77378364,
      "epoch": 1.4798699310284489,
      "grad_norm": 4.5,
      "learning_rate": 1.6710399569760105e-06,
      "loss": 0.81367893,
      "memory(GiB)": 147.13,
      "step": 63430,
      "train_speed(iter/s)": 0.201268
    },
    {
      "acc": 0.77930741,
      "epoch": 1.4801032386007378,
      "grad_norm": 5.65625,
      "learning_rate": 1.669630657791174e-06,
      "loss": 0.77642231,
      "memory(GiB)": 147.13,
      "step": 63440,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.773598,
      "epoch": 1.4803365461730267,
      "grad_norm": 5.5,
      "learning_rate": 1.6682218340155936e-06,
      "loss": 0.80465155,
      "memory(GiB)": 147.13,
      "step": 63450,
      "train_speed(iter/s)": 0.201302
    },
    {
      "acc": 0.76938334,
      "epoch": 1.4805698537453156,
      "grad_norm": 6.1875,
      "learning_rate": 1.666813485850377e-06,
      "loss": 0.82184029,
      "memory(GiB)": 147.13,
      "step": 63460,
      "train_speed(iter/s)": 0.201318
    },
    {
      "acc": 0.77726431,
      "epoch": 1.4808031613176045,
      "grad_norm": 6.25,
      "learning_rate": 1.665405613496569e-06,
      "loss": 0.81066351,
      "memory(GiB)": 147.13,
      "step": 63470,
      "train_speed(iter/s)": 0.201334
    },
    {
      "acc": 0.75288639,
      "epoch": 1.4810364688898934,
      "grad_norm": 6.03125,
      "learning_rate": 1.6639982171551405e-06,
      "loss": 0.89472895,
      "memory(GiB)": 147.13,
      "step": 63480,
      "train_speed(iter/s)": 0.20135
    },
    {
      "acc": 0.77361879,
      "epoch": 1.4812697764621823,
      "grad_norm": 7.03125,
      "learning_rate": 1.6625912970269958e-06,
      "loss": 0.80337019,
      "memory(GiB)": 147.13,
      "step": 63490,
      "train_speed(iter/s)": 0.201367
    },
    {
      "acc": 0.77668934,
      "epoch": 1.4815030840344712,
      "grad_norm": 4.65625,
      "learning_rate": 1.6611848533129754e-06,
      "loss": 0.79210997,
      "memory(GiB)": 147.13,
      "step": 63500,
      "train_speed(iter/s)": 0.201385
    },
    {
      "epoch": 1.4815030840344712,
      "eval_acc": 0.7447049510671291,
      "eval_loss": 0.8045029044151306,
      "eval_runtime": 1268.8732,
      "eval_samples_per_second": 28.365,
      "eval_steps_per_second": 14.183,
      "step": 63500
    },
    {
      "acc": 0.77970753,
      "epoch": 1.48173639160676,
      "grad_norm": 3.75,
      "learning_rate": 1.6597788862138458e-06,
      "loss": 0.78724799,
      "memory(GiB)": 147.13,
      "step": 63510,
      "train_speed(iter/s)": 0.200581
    },
    {
      "acc": 0.78490286,
      "epoch": 1.481969699179049,
      "grad_norm": 4.5625,
      "learning_rate": 1.6583733959303116e-06,
      "loss": 0.7604619,
      "memory(GiB)": 147.13,
      "step": 63520,
      "train_speed(iter/s)": 0.200596
    },
    {
      "acc": 0.77267179,
      "epoch": 1.4822030067513379,
      "grad_norm": 4.78125,
      "learning_rate": 1.6569683826630045e-06,
      "loss": 0.80196037,
      "memory(GiB)": 147.13,
      "step": 63530,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.78124847,
      "epoch": 1.4824363143236268,
      "grad_norm": 4.84375,
      "learning_rate": 1.6555638466124878e-06,
      "loss": 0.80149517,
      "memory(GiB)": 147.13,
      "step": 63540,
      "train_speed(iter/s)": 0.200628
    },
    {
      "acc": 0.77615232,
      "epoch": 1.4826696218959157,
      "grad_norm": 5.9375,
      "learning_rate": 1.654159787979262e-06,
      "loss": 0.81325369,
      "memory(GiB)": 147.13,
      "step": 63550,
      "train_speed(iter/s)": 0.200645
    },
    {
      "acc": 0.78834553,
      "epoch": 1.4829029294682046,
      "grad_norm": 6.03125,
      "learning_rate": 1.6527562069637543e-06,
      "loss": 0.77705812,
      "memory(GiB)": 147.13,
      "step": 63560,
      "train_speed(iter/s)": 0.200662
    },
    {
      "acc": 0.76742916,
      "epoch": 1.4831362370404935,
      "grad_norm": 7.125,
      "learning_rate": 1.6513531037663262e-06,
      "loss": 0.83876429,
      "memory(GiB)": 147.13,
      "step": 63570,
      "train_speed(iter/s)": 0.200678
    },
    {
      "acc": 0.79001141,
      "epoch": 1.4833695446127824,
      "grad_norm": 5.1875,
      "learning_rate": 1.6499504785872679e-06,
      "loss": 0.74909034,
      "memory(GiB)": 147.13,
      "step": 63580,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.79263821,
      "epoch": 1.4836028521850713,
      "grad_norm": 6.125,
      "learning_rate": 1.648548331626807e-06,
      "loss": 0.76054592,
      "memory(GiB)": 147.13,
      "step": 63590,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.78314829,
      "epoch": 1.4838361597573602,
      "grad_norm": 6.28125,
      "learning_rate": 1.6471466630850985e-06,
      "loss": 0.78806133,
      "memory(GiB)": 147.13,
      "step": 63600,
      "train_speed(iter/s)": 0.200724
    },
    {
      "acc": 0.7892365,
      "epoch": 1.484069467329649,
      "grad_norm": 4.6875,
      "learning_rate": 1.645745473162228e-06,
      "loss": 0.74208422,
      "memory(GiB)": 147.13,
      "step": 63610,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.80071468,
      "epoch": 1.484302774901938,
      "grad_norm": 4.84375,
      "learning_rate": 1.644344762058218e-06,
      "loss": 0.71267734,
      "memory(GiB)": 147.13,
      "step": 63620,
      "train_speed(iter/s)": 0.200757
    },
    {
      "acc": 0.77488637,
      "epoch": 1.4845360824742269,
      "grad_norm": 10.1875,
      "learning_rate": 1.6429445299730173e-06,
      "loss": 0.82684231,
      "memory(GiB)": 147.13,
      "step": 63630,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.75984926,
      "epoch": 1.4847693900465158,
      "grad_norm": 5.3125,
      "learning_rate": 1.6415447771065112e-06,
      "loss": 0.88017397,
      "memory(GiB)": 147.13,
      "step": 63640,
      "train_speed(iter/s)": 0.200789
    },
    {
      "acc": 0.76972971,
      "epoch": 1.4850026976188047,
      "grad_norm": 5.75,
      "learning_rate": 1.6401455036585111e-06,
      "loss": 0.82038412,
      "memory(GiB)": 147.13,
      "step": 63650,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.77063775,
      "epoch": 1.4852360051910936,
      "grad_norm": 6.125,
      "learning_rate": 1.6387467098287656e-06,
      "loss": 0.82443304,
      "memory(GiB)": 147.13,
      "step": 63660,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.79164023,
      "epoch": 1.4854693127633825,
      "grad_norm": 4.4375,
      "learning_rate": 1.637348395816951e-06,
      "loss": 0.76277428,
      "memory(GiB)": 147.13,
      "step": 63670,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.80538321,
      "epoch": 1.4857026203356711,
      "grad_norm": 4.9375,
      "learning_rate": 1.635950561822676e-06,
      "loss": 0.68404865,
      "memory(GiB)": 147.13,
      "step": 63680,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.77760887,
      "epoch": 1.4859359279079603,
      "grad_norm": 6.625,
      "learning_rate": 1.6345532080454813e-06,
      "loss": 0.80281401,
      "memory(GiB)": 147.13,
      "step": 63690,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.7884223,
      "epoch": 1.486169235480249,
      "grad_norm": 7.5625,
      "learning_rate": 1.6331563346848366e-06,
      "loss": 0.77055836,
      "memory(GiB)": 147.13,
      "step": 63700,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78238873,
      "epoch": 1.486402543052538,
      "grad_norm": 4.75,
      "learning_rate": 1.6317599419401486e-06,
      "loss": 0.78170815,
      "memory(GiB)": 147.13,
      "step": 63710,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.77347994,
      "epoch": 1.4866358506248267,
      "grad_norm": 8.0,
      "learning_rate": 1.6303640300107493e-06,
      "loss": 0.79208412,
      "memory(GiB)": 147.13,
      "step": 63720,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.7714962,
      "epoch": 1.4868691581971158,
      "grad_norm": 5.5,
      "learning_rate": 1.628968599095907e-06,
      "loss": 0.81774483,
      "memory(GiB)": 147.13,
      "step": 63730,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.78385878,
      "epoch": 1.4871024657694045,
      "grad_norm": 5.875,
      "learning_rate": 1.6275736493948174e-06,
      "loss": 0.77382607,
      "memory(GiB)": 147.13,
      "step": 63740,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.77496548,
      "epoch": 1.4873357733416936,
      "grad_norm": 4.40625,
      "learning_rate": 1.626179181106609e-06,
      "loss": 0.82866879,
      "memory(GiB)": 147.13,
      "step": 63750,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.78694777,
      "epoch": 1.4875690809139823,
      "grad_norm": 7.8125,
      "learning_rate": 1.6247851944303433e-06,
      "loss": 0.7435267,
      "memory(GiB)": 147.13,
      "step": 63760,
      "train_speed(iter/s)": 0.200983
    },
    {
      "acc": 0.7818573,
      "epoch": 1.4878023884862714,
      "grad_norm": 4.875,
      "learning_rate": 1.6233916895650093e-06,
      "loss": 0.77421417,
      "memory(GiB)": 147.13,
      "step": 63770,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.77949824,
      "epoch": 1.4880356960585601,
      "grad_norm": 5.84375,
      "learning_rate": 1.6219986667095323e-06,
      "loss": 0.81707954,
      "memory(GiB)": 147.13,
      "step": 63780,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.77718635,
      "epoch": 1.488269003630849,
      "grad_norm": 5.8125,
      "learning_rate": 1.6206061260627643e-06,
      "loss": 0.80593338,
      "memory(GiB)": 147.13,
      "step": 63790,
      "train_speed(iter/s)": 0.201034
    },
    {
      "acc": 0.79007106,
      "epoch": 1.488502311203138,
      "grad_norm": 6.40625,
      "learning_rate": 1.6192140678234903e-06,
      "loss": 0.75823188,
      "memory(GiB)": 147.13,
      "step": 63800,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.76932173,
      "epoch": 1.4887356187754268,
      "grad_norm": 6.09375,
      "learning_rate": 1.617822492190424e-06,
      "loss": 0.82226,
      "memory(GiB)": 147.13,
      "step": 63810,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.77304816,
      "epoch": 1.4889689263477157,
      "grad_norm": 4.78125,
      "learning_rate": 1.616431399362216e-06,
      "loss": 0.80309229,
      "memory(GiB)": 147.13,
      "step": 63820,
      "train_speed(iter/s)": 0.201084
    },
    {
      "acc": 0.79105062,
      "epoch": 1.4892022339200046,
      "grad_norm": 4.8125,
      "learning_rate": 1.615040789537443e-06,
      "loss": 0.73891435,
      "memory(GiB)": 147.13,
      "step": 63830,
      "train_speed(iter/s)": 0.201101
    },
    {
      "acc": 0.78510084,
      "epoch": 1.4894355414922935,
      "grad_norm": 6.0625,
      "learning_rate": 1.6136506629146125e-06,
      "loss": 0.76382818,
      "memory(GiB)": 147.13,
      "step": 63840,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.78627157,
      "epoch": 1.4896688490645824,
      "grad_norm": 4.84375,
      "learning_rate": 1.6122610196921673e-06,
      "loss": 0.75550165,
      "memory(GiB)": 147.13,
      "step": 63850,
      "train_speed(iter/s)": 0.201135
    },
    {
      "acc": 0.78514376,
      "epoch": 1.4899021566368713,
      "grad_norm": 6.1875,
      "learning_rate": 1.6108718600684764e-06,
      "loss": 0.78006501,
      "memory(GiB)": 147.13,
      "step": 63860,
      "train_speed(iter/s)": 0.201151
    },
    {
      "acc": 0.76863089,
      "epoch": 1.4901354642091602,
      "grad_norm": 4.71875,
      "learning_rate": 1.609483184241844e-06,
      "loss": 0.81679535,
      "memory(GiB)": 147.13,
      "step": 63870,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.79790592,
      "epoch": 1.490368771781449,
      "grad_norm": 5.5,
      "learning_rate": 1.6080949924105022e-06,
      "loss": 0.71203775,
      "memory(GiB)": 147.13,
      "step": 63880,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.77048893,
      "epoch": 1.490602079353738,
      "grad_norm": 6.25,
      "learning_rate": 1.6067072847726134e-06,
      "loss": 0.81707821,
      "memory(GiB)": 147.13,
      "step": 63890,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.75077691,
      "epoch": 1.490835386926027,
      "grad_norm": 4.84375,
      "learning_rate": 1.605320061526277e-06,
      "loss": 0.89881153,
      "memory(GiB)": 147.13,
      "step": 63900,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.81723709,
      "epoch": 1.4910686944983158,
      "grad_norm": 4.75,
      "learning_rate": 1.6039333228695132e-06,
      "loss": 0.65049648,
      "memory(GiB)": 147.13,
      "step": 63910,
      "train_speed(iter/s)": 0.201232
    },
    {
      "acc": 0.78162031,
      "epoch": 1.4913020020706047,
      "grad_norm": 6.34375,
      "learning_rate": 1.6025470690002815e-06,
      "loss": 0.77352018,
      "memory(GiB)": 147.13,
      "step": 63920,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.76750431,
      "epoch": 1.4915353096428936,
      "grad_norm": 4.40625,
      "learning_rate": 1.6011613001164677e-06,
      "loss": 0.84007797,
      "memory(GiB)": 147.13,
      "step": 63930,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.78385334,
      "epoch": 1.4917686172151825,
      "grad_norm": 6.71875,
      "learning_rate": 1.5997760164158927e-06,
      "loss": 0.77498088,
      "memory(GiB)": 147.13,
      "step": 63940,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.78162289,
      "epoch": 1.4920019247874714,
      "grad_norm": 4.8125,
      "learning_rate": 1.5983912180963012e-06,
      "loss": 0.76992779,
      "memory(GiB)": 147.13,
      "step": 63950,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.78982821,
      "epoch": 1.4922352323597603,
      "grad_norm": 5.03125,
      "learning_rate": 1.5970069053553776e-06,
      "loss": 0.74374228,
      "memory(GiB)": 147.13,
      "step": 63960,
      "train_speed(iter/s)": 0.201313
    },
    {
      "acc": 0.79276485,
      "epoch": 1.4924685399320492,
      "grad_norm": 5.46875,
      "learning_rate": 1.5956230783907294e-06,
      "loss": 0.72949905,
      "memory(GiB)": 147.13,
      "step": 63970,
      "train_speed(iter/s)": 0.20133
    },
    {
      "acc": 0.77740602,
      "epoch": 1.492701847504338,
      "grad_norm": 5.75,
      "learning_rate": 1.5942397373998959e-06,
      "loss": 0.79231834,
      "memory(GiB)": 147.13,
      "step": 63980,
      "train_speed(iter/s)": 0.201346
    },
    {
      "acc": 0.78017025,
      "epoch": 1.492935155076627,
      "grad_norm": 5.4375,
      "learning_rate": 1.5928568825803526e-06,
      "loss": 0.78005562,
      "memory(GiB)": 147.13,
      "step": 63990,
      "train_speed(iter/s)": 0.201364
    },
    {
      "acc": 0.78059478,
      "epoch": 1.493168462648916,
      "grad_norm": 7.40625,
      "learning_rate": 1.5914745141294974e-06,
      "loss": 0.76288352,
      "memory(GiB)": 147.13,
      "step": 64000,
      "train_speed(iter/s)": 0.20138
    },
    {
      "epoch": 1.493168462648916,
      "eval_acc": 0.7446312847202705,
      "eval_loss": 0.8045158982276917,
      "eval_runtime": 1269.109,
      "eval_samples_per_second": 28.359,
      "eval_steps_per_second": 14.18,
      "step": 64000
    },
    {
      "acc": 0.80024071,
      "epoch": 1.4934017702212048,
      "grad_norm": 6.28125,
      "learning_rate": 1.5900926322446686e-06,
      "loss": 0.74026098,
      "memory(GiB)": 147.13,
      "step": 64010,
      "train_speed(iter/s)": 0.200581
    },
    {
      "acc": 0.78730555,
      "epoch": 1.4936350777934937,
      "grad_norm": 5.5,
      "learning_rate": 1.5887112371231227e-06,
      "loss": 0.75264711,
      "memory(GiB)": 147.13,
      "step": 64020,
      "train_speed(iter/s)": 0.200598
    },
    {
      "acc": 0.76471243,
      "epoch": 1.4938683853657826,
      "grad_norm": 5.875,
      "learning_rate": 1.5873303289620585e-06,
      "loss": 0.85071449,
      "memory(GiB)": 147.13,
      "step": 64030,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.77108183,
      "epoch": 1.4941016929380715,
      "grad_norm": 8.1875,
      "learning_rate": 1.5859499079585982e-06,
      "loss": 0.81153488,
      "memory(GiB)": 147.13,
      "step": 64040,
      "train_speed(iter/s)": 0.20063
    },
    {
      "acc": 0.78847384,
      "epoch": 1.4943350005103604,
      "grad_norm": 5.84375,
      "learning_rate": 1.5845699743097953e-06,
      "loss": 0.75993419,
      "memory(GiB)": 147.13,
      "step": 64050,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.78438349,
      "epoch": 1.4945683080826493,
      "grad_norm": 5.28125,
      "learning_rate": 1.583190528212638e-06,
      "loss": 0.78886342,
      "memory(GiB)": 147.13,
      "step": 64060,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.78701124,
      "epoch": 1.494801615654938,
      "grad_norm": 5.8125,
      "learning_rate": 1.5818115698640386e-06,
      "loss": 0.77774706,
      "memory(GiB)": 147.13,
      "step": 64070,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.76535654,
      "epoch": 1.495034923227227,
      "grad_norm": 6.75,
      "learning_rate": 1.5804330994608463e-06,
      "loss": 0.84333496,
      "memory(GiB)": 147.13,
      "step": 64080,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.79128575,
      "epoch": 1.4952682307995158,
      "grad_norm": 5.46875,
      "learning_rate": 1.5790551171998337e-06,
      "loss": 0.75358858,
      "memory(GiB)": 147.13,
      "step": 64090,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.77005005,
      "epoch": 1.4955015383718049,
      "grad_norm": 7.6875,
      "learning_rate": 1.5776776232777114e-06,
      "loss": 0.84556427,
      "memory(GiB)": 147.13,
      "step": 64100,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.77952299,
      "epoch": 1.4957348459440936,
      "grad_norm": 9.6875,
      "learning_rate": 1.5763006178911139e-06,
      "loss": 0.80671349,
      "memory(GiB)": 147.13,
      "step": 64110,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.77896752,
      "epoch": 1.4959681535163827,
      "grad_norm": 5.4375,
      "learning_rate": 1.5749241012366068e-06,
      "loss": 0.78669481,
      "memory(GiB)": 147.13,
      "step": 64120,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.7698513,
      "epoch": 1.4962014610886714,
      "grad_norm": 5.8125,
      "learning_rate": 1.5735480735106927e-06,
      "loss": 0.82209358,
      "memory(GiB)": 147.13,
      "step": 64130,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.78666344,
      "epoch": 1.4964347686609605,
      "grad_norm": 5.03125,
      "learning_rate": 1.5721725349097926e-06,
      "loss": 0.74558525,
      "memory(GiB)": 147.13,
      "step": 64140,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.79377465,
      "epoch": 1.4966680762332492,
      "grad_norm": 4.40625,
      "learning_rate": 1.570797485630269e-06,
      "loss": 0.75561595,
      "memory(GiB)": 147.13,
      "step": 64150,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.77909813,
      "epoch": 1.4969013838055383,
      "grad_norm": 6.9375,
      "learning_rate": 1.5694229258684063e-06,
      "loss": 0.77388506,
      "memory(GiB)": 147.13,
      "step": 64160,
      "train_speed(iter/s)": 0.200827
    },
    {
      "acc": 0.77309704,
      "epoch": 1.497134691377827,
      "grad_norm": 4.5625,
      "learning_rate": 1.5680488558204259e-06,
      "loss": 0.82009621,
      "memory(GiB)": 147.13,
      "step": 64170,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.77360954,
      "epoch": 1.4973679989501159,
      "grad_norm": 5.03125,
      "learning_rate": 1.566675275682475e-06,
      "loss": 0.80817747,
      "memory(GiB)": 147.13,
      "step": 64180,
      "train_speed(iter/s)": 0.200858
    },
    {
      "acc": 0.78091269,
      "epoch": 1.4976013065224048,
      "grad_norm": 4.25,
      "learning_rate": 1.565302185650629e-06,
      "loss": 0.78140159,
      "memory(GiB)": 147.13,
      "step": 64190,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.79408712,
      "epoch": 1.4978346140946937,
      "grad_norm": 5.4375,
      "learning_rate": 1.5639295859208998e-06,
      "loss": 0.74546127,
      "memory(GiB)": 147.13,
      "step": 64200,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.79185085,
      "epoch": 1.4980679216669825,
      "grad_norm": 6.96875,
      "learning_rate": 1.562557476689222e-06,
      "loss": 0.7497364,
      "memory(GiB)": 147.13,
      "step": 64210,
      "train_speed(iter/s)": 0.200906
    },
    {
      "acc": 0.79052382,
      "epoch": 1.4983012292392714,
      "grad_norm": 4.75,
      "learning_rate": 1.5611858581514683e-06,
      "loss": 0.75888486,
      "memory(GiB)": 147.13,
      "step": 64220,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.78845158,
      "epoch": 1.4985345368115603,
      "grad_norm": 4.125,
      "learning_rate": 1.559814730503434e-06,
      "loss": 0.7904933,
      "memory(GiB)": 147.13,
      "step": 64230,
      "train_speed(iter/s)": 0.200939
    },
    {
      "acc": 0.77261376,
      "epoch": 1.4987678443838492,
      "grad_norm": 5.4375,
      "learning_rate": 1.5584440939408473e-06,
      "loss": 0.82468948,
      "memory(GiB)": 147.13,
      "step": 64240,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.76977367,
      "epoch": 1.4990011519561381,
      "grad_norm": 5.84375,
      "learning_rate": 1.557073948659365e-06,
      "loss": 0.83275337,
      "memory(GiB)": 147.13,
      "step": 64250,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.79217887,
      "epoch": 1.499234459528427,
      "grad_norm": 5.9375,
      "learning_rate": 1.555704294854578e-06,
      "loss": 0.74504166,
      "memory(GiB)": 147.13,
      "step": 64260,
      "train_speed(iter/s)": 0.200988
    },
    {
      "acc": 0.7652029,
      "epoch": 1.499467767100716,
      "grad_norm": 5.9375,
      "learning_rate": 1.5543351327220025e-06,
      "loss": 0.84907303,
      "memory(GiB)": 147.13,
      "step": 64270,
      "train_speed(iter/s)": 0.201003
    },
    {
      "acc": 0.79375896,
      "epoch": 1.4997010746730048,
      "grad_norm": 5.5,
      "learning_rate": 1.5529664624570839e-06,
      "loss": 0.72161312,
      "memory(GiB)": 147.13,
      "step": 64280,
      "train_speed(iter/s)": 0.20102
    },
    {
      "acc": 0.78259649,
      "epoch": 1.4999343822452937,
      "grad_norm": 6.53125,
      "learning_rate": 1.551598284255203e-06,
      "loss": 0.79077225,
      "memory(GiB)": 147.13,
      "step": 64290,
      "train_speed(iter/s)": 0.201036
    },
    {
      "acc": 0.78028545,
      "epoch": 1.5001676898175826,
      "grad_norm": 5.375,
      "learning_rate": 1.550230598311664e-06,
      "loss": 0.81074438,
      "memory(GiB)": 147.13,
      "step": 64300,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.76743765,
      "epoch": 1.5004009973898715,
      "grad_norm": 4.21875,
      "learning_rate": 1.548863404821706e-06,
      "loss": 0.83184929,
      "memory(GiB)": 147.13,
      "step": 64310,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.80110559,
      "epoch": 1.5006343049621604,
      "grad_norm": 11.0,
      "learning_rate": 1.547496703980495e-06,
      "loss": 0.71932917,
      "memory(GiB)": 147.13,
      "step": 64320,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.77408819,
      "epoch": 1.5008676125344493,
      "grad_norm": 5.0,
      "learning_rate": 1.5461304959831248e-06,
      "loss": 0.81594849,
      "memory(GiB)": 147.13,
      "step": 64330,
      "train_speed(iter/s)": 0.201101
    },
    {
      "acc": 0.80996132,
      "epoch": 1.5011009201067382,
      "grad_norm": 5.53125,
      "learning_rate": 1.5447647810246241e-06,
      "loss": 0.68073368,
      "memory(GiB)": 147.13,
      "step": 64340,
      "train_speed(iter/s)": 0.201116
    },
    {
      "acc": 0.77236247,
      "epoch": 1.5013342276790271,
      "grad_norm": 7.9375,
      "learning_rate": 1.5433995592999457e-06,
      "loss": 0.79192848,
      "memory(GiB)": 147.13,
      "step": 64350,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.78151517,
      "epoch": 1.501567535251316,
      "grad_norm": 4.4375,
      "learning_rate": 1.5420348310039796e-06,
      "loss": 0.79086199,
      "memory(GiB)": 147.13,
      "step": 64360,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.77429748,
      "epoch": 1.501800842823605,
      "grad_norm": 5.78125,
      "learning_rate": 1.5406705963315333e-06,
      "loss": 0.80469446,
      "memory(GiB)": 147.13,
      "step": 64370,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.78121309,
      "epoch": 1.5020341503958938,
      "grad_norm": 5.125,
      "learning_rate": 1.539306855477356e-06,
      "loss": 0.7872479,
      "memory(GiB)": 147.13,
      "step": 64380,
      "train_speed(iter/s)": 0.201181
    },
    {
      "acc": 0.78482914,
      "epoch": 1.5022674579681827,
      "grad_norm": 4.875,
      "learning_rate": 1.5379436086361187e-06,
      "loss": 0.77998705,
      "memory(GiB)": 147.13,
      "step": 64390,
      "train_speed(iter/s)": 0.201197
    },
    {
      "acc": 0.76260514,
      "epoch": 1.5025007655404716,
      "grad_norm": 4.03125,
      "learning_rate": 1.5365808560024264e-06,
      "loss": 0.84974766,
      "memory(GiB)": 147.13,
      "step": 64400,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.79643979,
      "epoch": 1.5027340731127605,
      "grad_norm": 8.0625,
      "learning_rate": 1.5352185977708112e-06,
      "loss": 0.72695284,
      "memory(GiB)": 147.13,
      "step": 64410,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.80471954,
      "epoch": 1.5029673806850492,
      "grad_norm": 4.28125,
      "learning_rate": 1.533856834135733e-06,
      "loss": 0.71714687,
      "memory(GiB)": 147.13,
      "step": 64420,
      "train_speed(iter/s)": 0.201244
    },
    {
      "acc": 0.77157187,
      "epoch": 1.5032006882573383,
      "grad_norm": 6.53125,
      "learning_rate": 1.532495565291587e-06,
      "loss": 0.81608639,
      "memory(GiB)": 147.13,
      "step": 64430,
      "train_speed(iter/s)": 0.20126
    },
    {
      "acc": 0.79674616,
      "epoch": 1.503433995829627,
      "grad_norm": 6.09375,
      "learning_rate": 1.5311347914326891e-06,
      "loss": 0.72117443,
      "memory(GiB)": 147.13,
      "step": 64440,
      "train_speed(iter/s)": 0.201276
    },
    {
      "acc": 0.77385588,
      "epoch": 1.5036673034019161,
      "grad_norm": 3.9375,
      "learning_rate": 1.5297745127532942e-06,
      "loss": 0.83259687,
      "memory(GiB)": 147.13,
      "step": 64450,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.76449928,
      "epoch": 1.5039006109742048,
      "grad_norm": 4.71875,
      "learning_rate": 1.5284147294475792e-06,
      "loss": 0.84181862,
      "memory(GiB)": 147.13,
      "step": 64460,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.79227133,
      "epoch": 1.504133918546494,
      "grad_norm": 5.03125,
      "learning_rate": 1.5270554417096533e-06,
      "loss": 0.75533457,
      "memory(GiB)": 147.13,
      "step": 64470,
      "train_speed(iter/s)": 0.201322
    },
    {
      "acc": 0.77802868,
      "epoch": 1.5043672261187826,
      "grad_norm": 5.6875,
      "learning_rate": 1.5256966497335541e-06,
      "loss": 0.78939829,
      "memory(GiB)": 147.13,
      "step": 64480,
      "train_speed(iter/s)": 0.201339
    },
    {
      "acc": 0.77074051,
      "epoch": 1.5046005336910717,
      "grad_norm": 5.875,
      "learning_rate": 1.5243383537132473e-06,
      "loss": 0.83445225,
      "memory(GiB)": 147.13,
      "step": 64490,
      "train_speed(iter/s)": 0.201355
    },
    {
      "acc": 0.78632889,
      "epoch": 1.5048338412633604,
      "grad_norm": 4.84375,
      "learning_rate": 1.5229805538426323e-06,
      "loss": 0.76857004,
      "memory(GiB)": 147.13,
      "step": 64500,
      "train_speed(iter/s)": 0.201371
    },
    {
      "epoch": 1.5048338412633604,
      "eval_acc": 0.7446565875089741,
      "eval_loss": 0.8044771552085876,
      "eval_runtime": 1269.4786,
      "eval_samples_per_second": 28.351,
      "eval_steps_per_second": 14.176,
      "step": 64500
    },
    {
      "acc": 0.79818277,
      "epoch": 1.5050671488356495,
      "grad_norm": 6.90625,
      "learning_rate": 1.5216232503155314e-06,
      "loss": 0.71413069,
      "memory(GiB)": 147.13,
      "step": 64510,
      "train_speed(iter/s)": 0.200578
    },
    {
      "acc": 0.79765759,
      "epoch": 1.5053004564079382,
      "grad_norm": 4.875,
      "learning_rate": 1.520266443325703e-06,
      "loss": 0.70702105,
      "memory(GiB)": 147.13,
      "step": 64520,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.77393379,
      "epoch": 1.5055337639802273,
      "grad_norm": 6.03125,
      "learning_rate": 1.5189101330668288e-06,
      "loss": 0.82264185,
      "memory(GiB)": 147.13,
      "step": 64530,
      "train_speed(iter/s)": 0.200609
    },
    {
      "acc": 0.78522334,
      "epoch": 1.505767071552516,
      "grad_norm": 5.59375,
      "learning_rate": 1.5175543197325205e-06,
      "loss": 0.76188078,
      "memory(GiB)": 147.13,
      "step": 64540,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.78678837,
      "epoch": 1.506000379124805,
      "grad_norm": 6.25,
      "learning_rate": 1.5161990035163226e-06,
      "loss": 0.77836008,
      "memory(GiB)": 147.13,
      "step": 64550,
      "train_speed(iter/s)": 0.20064
    },
    {
      "acc": 0.76740236,
      "epoch": 1.5062336866970938,
      "grad_norm": 5.71875,
      "learning_rate": 1.5148441846117035e-06,
      "loss": 0.8500349,
      "memory(GiB)": 147.13,
      "step": 64560,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.77085843,
      "epoch": 1.506466994269383,
      "grad_norm": 6.15625,
      "learning_rate": 1.5134898632120659e-06,
      "loss": 0.80457802,
      "memory(GiB)": 147.13,
      "step": 64570,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.78143206,
      "epoch": 1.5067003018416716,
      "grad_norm": 6.34375,
      "learning_rate": 1.5121360395107366e-06,
      "loss": 0.80713491,
      "memory(GiB)": 147.13,
      "step": 64580,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.7703723,
      "epoch": 1.5069336094139607,
      "grad_norm": 4.34375,
      "learning_rate": 1.5107827137009772e-06,
      "loss": 0.83621893,
      "memory(GiB)": 147.13,
      "step": 64590,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.78048077,
      "epoch": 1.5071669169862494,
      "grad_norm": 5.03125,
      "learning_rate": 1.509429885975968e-06,
      "loss": 0.77270036,
      "memory(GiB)": 147.13,
      "step": 64600,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.78544827,
      "epoch": 1.5074002245585383,
      "grad_norm": 5.78125,
      "learning_rate": 1.5080775565288314e-06,
      "loss": 0.76025243,
      "memory(GiB)": 147.13,
      "step": 64610,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.77681675,
      "epoch": 1.5076335321308272,
      "grad_norm": 5.21875,
      "learning_rate": 1.5067257255526085e-06,
      "loss": 0.79332089,
      "memory(GiB)": 147.13,
      "step": 64620,
      "train_speed(iter/s)": 0.200753
    },
    {
      "acc": 0.79631433,
      "epoch": 1.507866839703116,
      "grad_norm": 7.15625,
      "learning_rate": 1.505374393240272e-06,
      "loss": 0.71731257,
      "memory(GiB)": 147.13,
      "step": 64630,
      "train_speed(iter/s)": 0.200769
    },
    {
      "acc": 0.7795043,
      "epoch": 1.508100147275405,
      "grad_norm": 6.6875,
      "learning_rate": 1.5040235597847268e-06,
      "loss": 0.79805608,
      "memory(GiB)": 147.13,
      "step": 64640,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.77450428,
      "epoch": 1.5083334548476939,
      "grad_norm": 6.15625,
      "learning_rate": 1.5026732253788018e-06,
      "loss": 0.82841396,
      "memory(GiB)": 147.13,
      "step": 64650,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.78522072,
      "epoch": 1.5085667624199828,
      "grad_norm": 5.3125,
      "learning_rate": 1.501323390215259e-06,
      "loss": 0.77809391,
      "memory(GiB)": 147.13,
      "step": 64660,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.77315788,
      "epoch": 1.5088000699922717,
      "grad_norm": 6.84375,
      "learning_rate": 1.4999740544867864e-06,
      "loss": 0.82334805,
      "memory(GiB)": 147.13,
      "step": 64670,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.77878227,
      "epoch": 1.5090333775645606,
      "grad_norm": 5.03125,
      "learning_rate": 1.498625218385999e-06,
      "loss": 0.80296698,
      "memory(GiB)": 147.13,
      "step": 64680,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.76942329,
      "epoch": 1.5092666851368495,
      "grad_norm": 6.125,
      "learning_rate": 1.4972768821054461e-06,
      "loss": 0.82304773,
      "memory(GiB)": 147.13,
      "step": 64690,
      "train_speed(iter/s)": 0.200864
    },
    {
      "acc": 0.78032999,
      "epoch": 1.5094999927091384,
      "grad_norm": 4.96875,
      "learning_rate": 1.4959290458376008e-06,
      "loss": 0.78106308,
      "memory(GiB)": 147.13,
      "step": 64700,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.77347794,
      "epoch": 1.5097333002814273,
      "grad_norm": 6.5,
      "learning_rate": 1.494581709774866e-06,
      "loss": 0.82325249,
      "memory(GiB)": 147.13,
      "step": 64710,
      "train_speed(iter/s)": 0.200895
    },
    {
      "acc": 0.78451128,
      "epoch": 1.5099666078537162,
      "grad_norm": 5.65625,
      "learning_rate": 1.4932348741095726e-06,
      "loss": 0.76474838,
      "memory(GiB)": 147.13,
      "step": 64720,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.78300848,
      "epoch": 1.510199915426005,
      "grad_norm": 6.125,
      "learning_rate": 1.4918885390339837e-06,
      "loss": 0.77643466,
      "memory(GiB)": 147.13,
      "step": 64730,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.79654002,
      "epoch": 1.510433222998294,
      "grad_norm": 5.3125,
      "learning_rate": 1.4905427047402848e-06,
      "loss": 0.72235746,
      "memory(GiB)": 147.13,
      "step": 64740,
      "train_speed(iter/s)": 0.200942
    },
    {
      "acc": 0.77989998,
      "epoch": 1.5106665305705829,
      "grad_norm": 5.75,
      "learning_rate": 1.4891973714205971e-06,
      "loss": 0.78111591,
      "memory(GiB)": 147.13,
      "step": 64750,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.75031438,
      "epoch": 1.5108998381428718,
      "grad_norm": 6.78125,
      "learning_rate": 1.4878525392669652e-06,
      "loss": 0.91278858,
      "memory(GiB)": 147.13,
      "step": 64760,
      "train_speed(iter/s)": 0.200976
    },
    {
      "acc": 0.79479699,
      "epoch": 1.5111331457151607,
      "grad_norm": 4.46875,
      "learning_rate": 1.4865082084713605e-06,
      "loss": 0.73966794,
      "memory(GiB)": 147.13,
      "step": 64770,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.79615707,
      "epoch": 1.5113664532874496,
      "grad_norm": 5.34375,
      "learning_rate": 1.485164379225691e-06,
      "loss": 0.72713776,
      "memory(GiB)": 147.13,
      "step": 64780,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.79181237,
      "epoch": 1.5115997608597382,
      "grad_norm": 6.25,
      "learning_rate": 1.4838210517217827e-06,
      "loss": 0.75146961,
      "memory(GiB)": 147.13,
      "step": 64790,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.7745163,
      "epoch": 1.5118330684320274,
      "grad_norm": 5.5,
      "learning_rate": 1.4824782261513997e-06,
      "loss": 0.81358576,
      "memory(GiB)": 147.13,
      "step": 64800,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.77751856,
      "epoch": 1.512066376004316,
      "grad_norm": 4.71875,
      "learning_rate": 1.4811359027062282e-06,
      "loss": 0.79387336,
      "memory(GiB)": 147.13,
      "step": 64810,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.78064179,
      "epoch": 1.5122996835766052,
      "grad_norm": 7.125,
      "learning_rate": 1.4797940815778849e-06,
      "loss": 0.78270473,
      "memory(GiB)": 147.13,
      "step": 64820,
      "train_speed(iter/s)": 0.201069
    },
    {
      "acc": 0.77359581,
      "epoch": 1.5125329911488938,
      "grad_norm": 6.25,
      "learning_rate": 1.4784527629579132e-06,
      "loss": 0.8067338,
      "memory(GiB)": 147.13,
      "step": 64830,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.77926979,
      "epoch": 1.512766298721183,
      "grad_norm": 6.59375,
      "learning_rate": 1.4771119470377853e-06,
      "loss": 0.78685246,
      "memory(GiB)": 147.13,
      "step": 64840,
      "train_speed(iter/s)": 0.201101
    },
    {
      "acc": 0.77433491,
      "epoch": 1.5129996062934716,
      "grad_norm": 6.75,
      "learning_rate": 1.4757716340089046e-06,
      "loss": 0.81569548,
      "memory(GiB)": 147.13,
      "step": 64850,
      "train_speed(iter/s)": 0.201116
    },
    {
      "acc": 0.77438469,
      "epoch": 1.5132329138657608,
      "grad_norm": 7.1875,
      "learning_rate": 1.4744318240625981e-06,
      "loss": 0.79644213,
      "memory(GiB)": 147.13,
      "step": 64860,
      "train_speed(iter/s)": 0.201133
    },
    {
      "acc": 0.78981161,
      "epoch": 1.5134662214380494,
      "grad_norm": 4.375,
      "learning_rate": 1.4730925173901262e-06,
      "loss": 0.74776039,
      "memory(GiB)": 147.13,
      "step": 64870,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.76372499,
      "epoch": 1.5136995290103386,
      "grad_norm": 4.75,
      "learning_rate": 1.4717537141826716e-06,
      "loss": 0.84560843,
      "memory(GiB)": 147.13,
      "step": 64880,
      "train_speed(iter/s)": 0.201164
    },
    {
      "acc": 0.77633152,
      "epoch": 1.5139328365826272,
      "grad_norm": 4.125,
      "learning_rate": 1.4704154146313503e-06,
      "loss": 0.80207596,
      "memory(GiB)": 147.13,
      "step": 64890,
      "train_speed(iter/s)": 0.20118
    },
    {
      "acc": 0.78325543,
      "epoch": 1.5141661441549163,
      "grad_norm": 7.15625,
      "learning_rate": 1.4690776189272033e-06,
      "loss": 0.78658934,
      "memory(GiB)": 147.13,
      "step": 64900,
      "train_speed(iter/s)": 0.201197
    },
    {
      "acc": 0.78068142,
      "epoch": 1.514399451727205,
      "grad_norm": 5.0625,
      "learning_rate": 1.467740327261199e-06,
      "loss": 0.79103022,
      "memory(GiB)": 147.13,
      "step": 64910,
      "train_speed(iter/s)": 0.201213
    },
    {
      "acc": 0.79627781,
      "epoch": 1.5146327592994941,
      "grad_norm": 5.6875,
      "learning_rate": 1.4664035398242387e-06,
      "loss": 0.72616043,
      "memory(GiB)": 147.13,
      "step": 64920,
      "train_speed(iter/s)": 0.201228
    },
    {
      "acc": 0.77021713,
      "epoch": 1.5148660668717828,
      "grad_norm": 4.875,
      "learning_rate": 1.4650672568071461e-06,
      "loss": 0.81304054,
      "memory(GiB)": 147.13,
      "step": 64930,
      "train_speed(iter/s)": 0.201244
    },
    {
      "acc": 0.78109665,
      "epoch": 1.515099374444072,
      "grad_norm": 5.40625,
      "learning_rate": 1.4637314784006761e-06,
      "loss": 0.77189884,
      "memory(GiB)": 147.13,
      "step": 64940,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.76999512,
      "epoch": 1.5153326820163606,
      "grad_norm": 6.84375,
      "learning_rate": 1.4623962047955087e-06,
      "loss": 0.82561445,
      "memory(GiB)": 147.13,
      "step": 64950,
      "train_speed(iter/s)": 0.201276
    },
    {
      "acc": 0.77311716,
      "epoch": 1.5155659895886497,
      "grad_norm": 5.34375,
      "learning_rate": 1.4610614361822567e-06,
      "loss": 0.817449,
      "memory(GiB)": 147.13,
      "step": 64960,
      "train_speed(iter/s)": 0.201293
    },
    {
      "acc": 0.79791188,
      "epoch": 1.5157992971609384,
      "grad_norm": 5.0625,
      "learning_rate": 1.4597271727514568e-06,
      "loss": 0.7242517,
      "memory(GiB)": 147.13,
      "step": 64970,
      "train_speed(iter/s)": 0.201307
    },
    {
      "acc": 0.78501654,
      "epoch": 1.5160326047332273,
      "grad_norm": 5.53125,
      "learning_rate": 1.4583934146935725e-06,
      "loss": 0.75701132,
      "memory(GiB)": 147.13,
      "step": 64980,
      "train_speed(iter/s)": 0.201323
    },
    {
      "acc": 0.76452885,
      "epoch": 1.5162659123055162,
      "grad_norm": 5.84375,
      "learning_rate": 1.4570601621990016e-06,
      "loss": 0.86551952,
      "memory(GiB)": 147.13,
      "step": 64990,
      "train_speed(iter/s)": 0.201339
    },
    {
      "acc": 0.7722949,
      "epoch": 1.5164992198778051,
      "grad_norm": 5.625,
      "learning_rate": 1.4557274154580614e-06,
      "loss": 0.81062641,
      "memory(GiB)": 147.13,
      "step": 65000,
      "train_speed(iter/s)": 0.201353
    },
    {
      "epoch": 1.5164992198778051,
      "eval_acc": 0.7446660360186799,
      "eval_loss": 0.8044557571411133,
      "eval_runtime": 1269.9569,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.171,
      "step": 65000
    },
    {
      "acc": 0.79000292,
      "epoch": 1.516732527450094,
      "grad_norm": 5.25,
      "learning_rate": 1.4543951746610047e-06,
      "loss": 0.73190441,
      "memory(GiB)": 147.13,
      "step": 65010,
      "train_speed(iter/s)": 0.200566
    },
    {
      "acc": 0.7913847,
      "epoch": 1.516965835022383,
      "grad_norm": 4.34375,
      "learning_rate": 1.4530634399980049e-06,
      "loss": 0.75735092,
      "memory(GiB)": 147.13,
      "step": 65020,
      "train_speed(iter/s)": 0.200581
    },
    {
      "acc": 0.78307304,
      "epoch": 1.5171991425946718,
      "grad_norm": 4.25,
      "learning_rate": 1.4517322116591709e-06,
      "loss": 0.77868781,
      "memory(GiB)": 147.13,
      "step": 65030,
      "train_speed(iter/s)": 0.200596
    },
    {
      "acc": 0.79364333,
      "epoch": 1.5174324501669607,
      "grad_norm": 6.65625,
      "learning_rate": 1.4504014898345326e-06,
      "loss": 0.73498554,
      "memory(GiB)": 147.13,
      "step": 65040,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.78174887,
      "epoch": 1.5176657577392496,
      "grad_norm": 6.53125,
      "learning_rate": 1.44907127471405e-06,
      "loss": 0.78112001,
      "memory(GiB)": 147.13,
      "step": 65050,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.79200649,
      "epoch": 1.5178990653115385,
      "grad_norm": 6.53125,
      "learning_rate": 1.447741566487612e-06,
      "loss": 0.73226194,
      "memory(GiB)": 147.13,
      "step": 65060,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.77733145,
      "epoch": 1.5181323728838274,
      "grad_norm": 5.5625,
      "learning_rate": 1.4464123653450318e-06,
      "loss": 0.7879797,
      "memory(GiB)": 147.13,
      "step": 65070,
      "train_speed(iter/s)": 0.200658
    },
    {
      "acc": 0.76886935,
      "epoch": 1.5183656804561163,
      "grad_norm": 5.0,
      "learning_rate": 1.4450836714760553e-06,
      "loss": 0.81756496,
      "memory(GiB)": 147.13,
      "step": 65080,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.79252605,
      "epoch": 1.5185989880284052,
      "grad_norm": 7.21875,
      "learning_rate": 1.4437554850703506e-06,
      "loss": 0.73690176,
      "memory(GiB)": 147.13,
      "step": 65090,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.79318552,
      "epoch": 1.518832295600694,
      "grad_norm": 4.625,
      "learning_rate": 1.442427806317519e-06,
      "loss": 0.73230796,
      "memory(GiB)": 147.13,
      "step": 65100,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.78320851,
      "epoch": 1.519065603172983,
      "grad_norm": 6.375,
      "learning_rate": 1.4411006354070844e-06,
      "loss": 0.76603732,
      "memory(GiB)": 147.13,
      "step": 65110,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.77989206,
      "epoch": 1.519298910745272,
      "grad_norm": 4.90625,
      "learning_rate": 1.4397739725284988e-06,
      "loss": 0.77163534,
      "memory(GiB)": 147.13,
      "step": 65120,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.77356272,
      "epoch": 1.5195322183175608,
      "grad_norm": 5.40625,
      "learning_rate": 1.4384478178711458e-06,
      "loss": 0.7995605,
      "memory(GiB)": 147.13,
      "step": 65130,
      "train_speed(iter/s)": 0.200754
    },
    {
      "acc": 0.79077578,
      "epoch": 1.5197655258898497,
      "grad_norm": 5.46875,
      "learning_rate": 1.437122171624331e-06,
      "loss": 0.73075666,
      "memory(GiB)": 147.13,
      "step": 65140,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.77648382,
      "epoch": 1.5199988334621386,
      "grad_norm": 4.9375,
      "learning_rate": 1.4357970339772924e-06,
      "loss": 0.79962687,
      "memory(GiB)": 147.13,
      "step": 65150,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.78098822,
      "epoch": 1.5202321410344275,
      "grad_norm": 6.6875,
      "learning_rate": 1.4344724051191917e-06,
      "loss": 0.80490036,
      "memory(GiB)": 147.13,
      "step": 65160,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.78283033,
      "epoch": 1.5204654486067164,
      "grad_norm": 6.1875,
      "learning_rate": 1.43314828523912e-06,
      "loss": 0.77967253,
      "memory(GiB)": 147.13,
      "step": 65170,
      "train_speed(iter/s)": 0.200819
    },
    {
      "acc": 0.78426676,
      "epoch": 1.520698756179005,
      "grad_norm": 10.1875,
      "learning_rate": 1.431824674526092e-06,
      "loss": 0.7804903,
      "memory(GiB)": 147.13,
      "step": 65180,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.77625027,
      "epoch": 1.5209320637512942,
      "grad_norm": 5.59375,
      "learning_rate": 1.430501573169057e-06,
      "loss": 0.79841747,
      "memory(GiB)": 147.13,
      "step": 65190,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.79016428,
      "epoch": 1.5211653713235829,
      "grad_norm": 4.75,
      "learning_rate": 1.4291789813568858e-06,
      "loss": 0.75621462,
      "memory(GiB)": 147.13,
      "step": 65200,
      "train_speed(iter/s)": 0.200864
    },
    {
      "acc": 0.78785486,
      "epoch": 1.521398678895872,
      "grad_norm": 6.53125,
      "learning_rate": 1.4278568992783758e-06,
      "loss": 0.76169577,
      "memory(GiB)": 147.13,
      "step": 65210,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.78738542,
      "epoch": 1.5216319864681607,
      "grad_norm": 4.625,
      "learning_rate": 1.4265353271222577e-06,
      "loss": 0.75058613,
      "memory(GiB)": 147.13,
      "step": 65220,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.78331351,
      "epoch": 1.5218652940404498,
      "grad_norm": 5.8125,
      "learning_rate": 1.4252142650771811e-06,
      "loss": 0.78213634,
      "memory(GiB)": 147.13,
      "step": 65230,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.768677,
      "epoch": 1.5220986016127385,
      "grad_norm": 4.375,
      "learning_rate": 1.4238937133317322e-06,
      "loss": 0.81841431,
      "memory(GiB)": 147.13,
      "step": 65240,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.78544569,
      "epoch": 1.5223319091850276,
      "grad_norm": 5.4375,
      "learning_rate": 1.4225736720744164e-06,
      "loss": 0.77090549,
      "memory(GiB)": 147.13,
      "step": 65250,
      "train_speed(iter/s)": 0.200946
    },
    {
      "acc": 0.77120905,
      "epoch": 1.5225652167573163,
      "grad_norm": 5.53125,
      "learning_rate": 1.4212541414936682e-06,
      "loss": 0.81031246,
      "memory(GiB)": 147.13,
      "step": 65260,
      "train_speed(iter/s)": 0.200962
    },
    {
      "acc": 0.77259464,
      "epoch": 1.5227985243296054,
      "grad_norm": 5.0625,
      "learning_rate": 1.4199351217778545e-06,
      "loss": 0.80552759,
      "memory(GiB)": 147.13,
      "step": 65270,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.78281403,
      "epoch": 1.523031831901894,
      "grad_norm": 6.375,
      "learning_rate": 1.4186166131152595e-06,
      "loss": 0.76773696,
      "memory(GiB)": 147.13,
      "step": 65280,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.75774269,
      "epoch": 1.5232651394741832,
      "grad_norm": 5.59375,
      "learning_rate": 1.4172986156941038e-06,
      "loss": 0.87673063,
      "memory(GiB)": 147.13,
      "step": 65290,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.76485977,
      "epoch": 1.5234984470464719,
      "grad_norm": 4.53125,
      "learning_rate": 1.4159811297025284e-06,
      "loss": 0.86806583,
      "memory(GiB)": 147.13,
      "step": 65300,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.80621672,
      "epoch": 1.523731754618761,
      "grad_norm": 6.125,
      "learning_rate": 1.4146641553286071e-06,
      "loss": 0.68341751,
      "memory(GiB)": 147.13,
      "step": 65310,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.78828001,
      "epoch": 1.5239650621910497,
      "grad_norm": 6.4375,
      "learning_rate": 1.4133476927603362e-06,
      "loss": 0.74977875,
      "memory(GiB)": 147.13,
      "step": 65320,
      "train_speed(iter/s)": 0.201058
    },
    {
      "acc": 0.76758041,
      "epoch": 1.5241983697633388,
      "grad_norm": 6.1875,
      "learning_rate": 1.4120317421856389e-06,
      "loss": 0.81543188,
      "memory(GiB)": 147.13,
      "step": 65330,
      "train_speed(iter/s)": 0.201074
    },
    {
      "acc": 0.76192322,
      "epoch": 1.5244316773356275,
      "grad_norm": 5.4375,
      "learning_rate": 1.4107163037923693e-06,
      "loss": 0.84419556,
      "memory(GiB)": 147.13,
      "step": 65340,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.77184172,
      "epoch": 1.5246649849079166,
      "grad_norm": 5.6875,
      "learning_rate": 1.409401377768303e-06,
      "loss": 0.82547474,
      "memory(GiB)": 147.13,
      "step": 65350,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.77231297,
      "epoch": 1.5248982924802053,
      "grad_norm": 4.5625,
      "learning_rate": 1.408086964301149e-06,
      "loss": 0.83197184,
      "memory(GiB)": 147.13,
      "step": 65360,
      "train_speed(iter/s)": 0.201122
    },
    {
      "acc": 0.78483472,
      "epoch": 1.5251316000524942,
      "grad_norm": 6.21875,
      "learning_rate": 1.4067730635785354e-06,
      "loss": 0.7758481,
      "memory(GiB)": 147.13,
      "step": 65370,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.77703638,
      "epoch": 1.525364907624783,
      "grad_norm": 9.125,
      "learning_rate": 1.4054596757880262e-06,
      "loss": 0.81458502,
      "memory(GiB)": 147.13,
      "step": 65380,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.77877493,
      "epoch": 1.525598215197072,
      "grad_norm": 6.21875,
      "learning_rate": 1.4041468011171017e-06,
      "loss": 0.81068916,
      "memory(GiB)": 147.13,
      "step": 65390,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.77528429,
      "epoch": 1.5258315227693608,
      "grad_norm": 4.8125,
      "learning_rate": 1.4028344397531773e-06,
      "loss": 0.80191021,
      "memory(GiB)": 147.13,
      "step": 65400,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.7893209,
      "epoch": 1.5260648303416497,
      "grad_norm": 9.5625,
      "learning_rate": 1.4015225918835923e-06,
      "loss": 0.7487329,
      "memory(GiB)": 147.13,
      "step": 65410,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.78093977,
      "epoch": 1.5262981379139386,
      "grad_norm": 5.96875,
      "learning_rate": 1.4002112576956102e-06,
      "loss": 0.78119683,
      "memory(GiB)": 147.13,
      "step": 65420,
      "train_speed(iter/s)": 0.201215
    },
    {
      "acc": 0.7891861,
      "epoch": 1.5265314454862275,
      "grad_norm": 8.75,
      "learning_rate": 1.3989004373764264e-06,
      "loss": 0.73629584,
      "memory(GiB)": 147.13,
      "step": 65430,
      "train_speed(iter/s)": 0.20123
    },
    {
      "acc": 0.76216254,
      "epoch": 1.5267647530585164,
      "grad_norm": 5.5625,
      "learning_rate": 1.3975901311131584e-06,
      "loss": 0.85819883,
      "memory(GiB)": 147.13,
      "step": 65440,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.77317934,
      "epoch": 1.5269980606308053,
      "grad_norm": 5.75,
      "learning_rate": 1.3962803390928537e-06,
      "loss": 0.80435085,
      "memory(GiB)": 147.13,
      "step": 65450,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.78658466,
      "epoch": 1.5272313682030942,
      "grad_norm": 6.8125,
      "learning_rate": 1.3949710615024836e-06,
      "loss": 0.76053772,
      "memory(GiB)": 147.13,
      "step": 65460,
      "train_speed(iter/s)": 0.201277
    },
    {
      "acc": 0.79117346,
      "epoch": 1.5274646757753831,
      "grad_norm": 5.125,
      "learning_rate": 1.3936622985289462e-06,
      "loss": 0.75075483,
      "memory(GiB)": 147.13,
      "step": 65470,
      "train_speed(iter/s)": 0.201292
    },
    {
      "acc": 0.77793059,
      "epoch": 1.527697983347672,
      "grad_norm": 5.34375,
      "learning_rate": 1.3923540503590689e-06,
      "loss": 0.79492455,
      "memory(GiB)": 147.13,
      "step": 65480,
      "train_speed(iter/s)": 0.201309
    },
    {
      "acc": 0.79856424,
      "epoch": 1.527931290919961,
      "grad_norm": 4.25,
      "learning_rate": 1.3910463171796018e-06,
      "loss": 0.71239781,
      "memory(GiB)": 147.13,
      "step": 65490,
      "train_speed(iter/s)": 0.201325
    },
    {
      "acc": 0.77887478,
      "epoch": 1.5281645984922498,
      "grad_norm": 8.3125,
      "learning_rate": 1.3897390991772269e-06,
      "loss": 0.78372169,
      "memory(GiB)": 147.13,
      "step": 65500,
      "train_speed(iter/s)": 0.201341
    },
    {
      "epoch": 1.5281645984922498,
      "eval_acc": 0.7447174423172486,
      "eval_loss": 0.8044983148574829,
      "eval_runtime": 1270.2082,
      "eval_samples_per_second": 28.335,
      "eval_steps_per_second": 14.168,
      "step": 65500
    },
    {
      "acc": 0.77596331,
      "epoch": 1.5283979060645387,
      "grad_norm": 5.75,
      "learning_rate": 1.3884323965385443e-06,
      "loss": 0.80863295,
      "memory(GiB)": 147.13,
      "step": 65510,
      "train_speed(iter/s)": 0.20056
    },
    {
      "acc": 0.79642587,
      "epoch": 1.5286312136368276,
      "grad_norm": 4.53125,
      "learning_rate": 1.3871262094500897e-06,
      "loss": 0.72295847,
      "memory(GiB)": 147.13,
      "step": 65520,
      "train_speed(iter/s)": 0.200576
    },
    {
      "acc": 0.78992147,
      "epoch": 1.5288645212091165,
      "grad_norm": 4.25,
      "learning_rate": 1.3858205380983175e-06,
      "loss": 0.7313539,
      "memory(GiB)": 147.13,
      "step": 65530,
      "train_speed(iter/s)": 0.200591
    },
    {
      "acc": 0.78537593,
      "epoch": 1.5290978287814054,
      "grad_norm": 7.5,
      "learning_rate": 1.3845153826696144e-06,
      "loss": 0.76099243,
      "memory(GiB)": 147.13,
      "step": 65540,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.76554108,
      "epoch": 1.5293311363536943,
      "grad_norm": 5.6875,
      "learning_rate": 1.3832107433502912e-06,
      "loss": 0.85681934,
      "memory(GiB)": 147.13,
      "step": 65550,
      "train_speed(iter/s)": 0.200623
    },
    {
      "acc": 0.77308493,
      "epoch": 1.5295644439259832,
      "grad_norm": 14.5,
      "learning_rate": 1.3819066203265813e-06,
      "loss": 0.82602463,
      "memory(GiB)": 147.13,
      "step": 65560,
      "train_speed(iter/s)": 0.200639
    },
    {
      "acc": 0.76433573,
      "epoch": 1.529797751498272,
      "grad_norm": 4.375,
      "learning_rate": 1.3806030137846521e-06,
      "loss": 0.84086971,
      "memory(GiB)": 147.13,
      "step": 65570,
      "train_speed(iter/s)": 0.200654
    },
    {
      "acc": 0.79202843,
      "epoch": 1.530031059070561,
      "grad_norm": 5.09375,
      "learning_rate": 1.3792999239105898e-06,
      "loss": 0.74395723,
      "memory(GiB)": 147.13,
      "step": 65580,
      "train_speed(iter/s)": 0.20067
    },
    {
      "acc": 0.79133291,
      "epoch": 1.5302643666428497,
      "grad_norm": 6.25,
      "learning_rate": 1.3779973508904132e-06,
      "loss": 0.74699202,
      "memory(GiB)": 147.13,
      "step": 65590,
      "train_speed(iter/s)": 0.200684
    },
    {
      "acc": 0.77008934,
      "epoch": 1.5304976742151388,
      "grad_norm": 5.5,
      "learning_rate": 1.3766952949100625e-06,
      "loss": 0.83408842,
      "memory(GiB)": 147.13,
      "step": 65600,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.77300768,
      "epoch": 1.5307309817874275,
      "grad_norm": 6.65625,
      "learning_rate": 1.3753937561554053e-06,
      "loss": 0.81059189,
      "memory(GiB)": 147.13,
      "step": 65610,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.76012034,
      "epoch": 1.5309642893597166,
      "grad_norm": 6.8125,
      "learning_rate": 1.3740927348122373e-06,
      "loss": 0.87639198,
      "memory(GiB)": 147.13,
      "step": 65620,
      "train_speed(iter/s)": 0.200731
    },
    {
      "acc": 0.77505989,
      "epoch": 1.5311975969320053,
      "grad_norm": 6.65625,
      "learning_rate": 1.3727922310662762e-06,
      "loss": 0.79479666,
      "memory(GiB)": 147.13,
      "step": 65630,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.78588977,
      "epoch": 1.5314309045042944,
      "grad_norm": 4.75,
      "learning_rate": 1.371492245103172e-06,
      "loss": 0.79581938,
      "memory(GiB)": 147.13,
      "step": 65640,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.78697853,
      "epoch": 1.531664212076583,
      "grad_norm": 7.75,
      "learning_rate": 1.3701927771084939e-06,
      "loss": 0.77796259,
      "memory(GiB)": 147.13,
      "step": 65650,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.7804883,
      "epoch": 1.5318975196488722,
      "grad_norm": 4.9375,
      "learning_rate": 1.3688938272677442e-06,
      "loss": 0.78984337,
      "memory(GiB)": 147.13,
      "step": 65660,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.7731575,
      "epoch": 1.532130827221161,
      "grad_norm": 7.46875,
      "learning_rate": 1.3675953957663441e-06,
      "loss": 0.8263773,
      "memory(GiB)": 147.13,
      "step": 65670,
      "train_speed(iter/s)": 0.200808
    },
    {
      "acc": 0.81379738,
      "epoch": 1.53236413479345,
      "grad_norm": 5.09375,
      "learning_rate": 1.3662974827896474e-06,
      "loss": 0.65503149,
      "memory(GiB)": 147.13,
      "step": 65680,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.77931385,
      "epoch": 1.5325974423657387,
      "grad_norm": 5.125,
      "learning_rate": 1.3650000885229297e-06,
      "loss": 0.80769215,
      "memory(GiB)": 147.13,
      "step": 65690,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.77855949,
      "epoch": 1.5328307499380278,
      "grad_norm": 5.6875,
      "learning_rate": 1.3637032131513922e-06,
      "loss": 0.78796816,
      "memory(GiB)": 147.13,
      "step": 65700,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.77442112,
      "epoch": 1.5330640575103165,
      "grad_norm": 6.3125,
      "learning_rate": 1.3624068568601655e-06,
      "loss": 0.79497447,
      "memory(GiB)": 147.13,
      "step": 65710,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.77679834,
      "epoch": 1.5332973650826056,
      "grad_norm": 8.5625,
      "learning_rate": 1.3611110198343025e-06,
      "loss": 0.81760941,
      "memory(GiB)": 147.13,
      "step": 65720,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78379855,
      "epoch": 1.5335306726548943,
      "grad_norm": 5.09375,
      "learning_rate": 1.359815702258787e-06,
      "loss": 0.77201395,
      "memory(GiB)": 147.13,
      "step": 65730,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.78287749,
      "epoch": 1.5337639802271834,
      "grad_norm": 6.8125,
      "learning_rate": 1.3585209043185193e-06,
      "loss": 0.76966333,
      "memory(GiB)": 147.13,
      "step": 65740,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.78400383,
      "epoch": 1.533997287799472,
      "grad_norm": 6.375,
      "learning_rate": 1.3572266261983363e-06,
      "loss": 0.77230244,
      "memory(GiB)": 147.13,
      "step": 65750,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.78760571,
      "epoch": 1.534230595371761,
      "grad_norm": 5.59375,
      "learning_rate": 1.3559328680829942e-06,
      "loss": 0.76585751,
      "memory(GiB)": 147.13,
      "step": 65760,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.79107695,
      "epoch": 1.5344639029440499,
      "grad_norm": 4.75,
      "learning_rate": 1.3546396301571751e-06,
      "loss": 0.74846153,
      "memory(GiB)": 147.13,
      "step": 65770,
      "train_speed(iter/s)": 0.200965
    },
    {
      "acc": 0.78852139,
      "epoch": 1.5346972105163388,
      "grad_norm": 5.53125,
      "learning_rate": 1.353346912605491e-06,
      "loss": 0.77424564,
      "memory(GiB)": 147.13,
      "step": 65780,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.79274092,
      "epoch": 1.5349305180886277,
      "grad_norm": 7.1875,
      "learning_rate": 1.3520547156124748e-06,
      "loss": 0.74635301,
      "memory(GiB)": 147.13,
      "step": 65790,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.77839308,
      "epoch": 1.5351638256609166,
      "grad_norm": 4.875,
      "learning_rate": 1.3507630393625893e-06,
      "loss": 0.80371504,
      "memory(GiB)": 147.13,
      "step": 65800,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.77611036,
      "epoch": 1.5353971332332055,
      "grad_norm": 5.25,
      "learning_rate": 1.3494718840402188e-06,
      "loss": 0.78693495,
      "memory(GiB)": 147.13,
      "step": 65810,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.77409744,
      "epoch": 1.5356304408054944,
      "grad_norm": 6.59375,
      "learning_rate": 1.348181249829677e-06,
      "loss": 0.82686939,
      "memory(GiB)": 147.13,
      "step": 65820,
      "train_speed(iter/s)": 0.201047
    },
    {
      "acc": 0.76908383,
      "epoch": 1.5358637483777833,
      "grad_norm": 4.625,
      "learning_rate": 1.3468911369152015e-06,
      "loss": 0.83825998,
      "memory(GiB)": 147.13,
      "step": 65830,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.77742887,
      "epoch": 1.5360970559500722,
      "grad_norm": 4.6875,
      "learning_rate": 1.3456015454809551e-06,
      "loss": 0.79348636,
      "memory(GiB)": 147.13,
      "step": 65840,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.77535996,
      "epoch": 1.536330363522361,
      "grad_norm": 7.0,
      "learning_rate": 1.3443124757110266e-06,
      "loss": 0.81135139,
      "memory(GiB)": 147.13,
      "step": 65850,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.77499065,
      "epoch": 1.53656367109465,
      "grad_norm": 6.96875,
      "learning_rate": 1.3430239277894281e-06,
      "loss": 0.82661295,
      "memory(GiB)": 147.13,
      "step": 65860,
      "train_speed(iter/s)": 0.201109
    },
    {
      "acc": 0.78962278,
      "epoch": 1.5367969786669389,
      "grad_norm": 5.09375,
      "learning_rate": 1.3417359019001037e-06,
      "loss": 0.74881072,
      "memory(GiB)": 147.13,
      "step": 65870,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.79519157,
      "epoch": 1.5370302862392278,
      "grad_norm": 9.3125,
      "learning_rate": 1.3404483982269145e-06,
      "loss": 0.74976549,
      "memory(GiB)": 147.13,
      "step": 65880,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.7780304,
      "epoch": 1.5372635938115167,
      "grad_norm": 6.15625,
      "learning_rate": 1.339161416953655e-06,
      "loss": 0.78960228,
      "memory(GiB)": 147.13,
      "step": 65890,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.74467463,
      "epoch": 1.5374969013838056,
      "grad_norm": 6.0,
      "learning_rate": 1.3378749582640398e-06,
      "loss": 0.95666943,
      "memory(GiB)": 147.13,
      "step": 65900,
      "train_speed(iter/s)": 0.201174
    },
    {
      "acc": 0.78630295,
      "epoch": 1.5377302089560945,
      "grad_norm": 4.125,
      "learning_rate": 1.3365890223417088e-06,
      "loss": 0.74238262,
      "memory(GiB)": 147.13,
      "step": 65910,
      "train_speed(iter/s)": 0.20119
    },
    {
      "acc": 0.7730278,
      "epoch": 1.5379635165283834,
      "grad_norm": 5.15625,
      "learning_rate": 1.3353036093702326e-06,
      "loss": 0.7966548,
      "memory(GiB)": 147.13,
      "step": 65920,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.78292246,
      "epoch": 1.5381968241006723,
      "grad_norm": 4.625,
      "learning_rate": 1.3340187195331e-06,
      "loss": 0.7709209,
      "memory(GiB)": 147.13,
      "step": 65930,
      "train_speed(iter/s)": 0.201223
    },
    {
      "acc": 0.78873224,
      "epoch": 1.5384301316729612,
      "grad_norm": 5.96875,
      "learning_rate": 1.3327343530137316e-06,
      "loss": 0.76129274,
      "memory(GiB)": 147.13,
      "step": 65940,
      "train_speed(iter/s)": 0.201238
    },
    {
      "acc": 0.77018852,
      "epoch": 1.53866343924525,
      "grad_norm": 3.5,
      "learning_rate": 1.3314505099954683e-06,
      "loss": 0.82236538,
      "memory(GiB)": 147.13,
      "step": 65950,
      "train_speed(iter/s)": 0.201254
    },
    {
      "acc": 0.77579556,
      "epoch": 1.5388967468175387,
      "grad_norm": 6.65625,
      "learning_rate": 1.33016719066158e-06,
      "loss": 0.7895669,
      "memory(GiB)": 147.13,
      "step": 65960,
      "train_speed(iter/s)": 0.201269
    },
    {
      "acc": 0.78516469,
      "epoch": 1.5391300543898279,
      "grad_norm": 4.78125,
      "learning_rate": 1.328884395195257e-06,
      "loss": 0.77448907,
      "memory(GiB)": 147.13,
      "step": 65970,
      "train_speed(iter/s)": 0.201284
    },
    {
      "acc": 0.77121625,
      "epoch": 1.5393633619621165,
      "grad_norm": 6.0,
      "learning_rate": 1.3276021237796216e-06,
      "loss": 0.83493958,
      "memory(GiB)": 147.13,
      "step": 65980,
      "train_speed(iter/s)": 0.201298
    },
    {
      "acc": 0.79430099,
      "epoch": 1.5395966695344057,
      "grad_norm": 4.28125,
      "learning_rate": 1.3263203765977168e-06,
      "loss": 0.73301487,
      "memory(GiB)": 147.13,
      "step": 65990,
      "train_speed(iter/s)": 0.201314
    },
    {
      "acc": 0.78175182,
      "epoch": 1.5398299771066943,
      "grad_norm": 5.03125,
      "learning_rate": 1.3250391538325085e-06,
      "loss": 0.75961099,
      "memory(GiB)": 147.13,
      "step": 66000,
      "train_speed(iter/s)": 0.20133
    },
    {
      "epoch": 1.5398299771066943,
      "eval_acc": 0.7446849330380915,
      "eval_loss": 0.8045194149017334,
      "eval_runtime": 1270.9323,
      "eval_samples_per_second": 28.319,
      "eval_steps_per_second": 14.16,
      "step": 66000
    },
    {
      "acc": 0.79298277,
      "epoch": 1.5400632846789835,
      "grad_norm": 5.65625,
      "learning_rate": 1.3237584556668958e-06,
      "loss": 0.74688911,
      "memory(GiB)": 147.13,
      "step": 66010,
      "train_speed(iter/s)": 0.200553
    },
    {
      "acc": 0.78268442,
      "epoch": 1.5402965922512721,
      "grad_norm": 4.0,
      "learning_rate": 1.322478282283694e-06,
      "loss": 0.77756462,
      "memory(GiB)": 147.13,
      "step": 66020,
      "train_speed(iter/s)": 0.200569
    },
    {
      "acc": 0.77548428,
      "epoch": 1.5405298998235613,
      "grad_norm": 6.875,
      "learning_rate": 1.3211986338656503e-06,
      "loss": 0.80724964,
      "memory(GiB)": 147.13,
      "step": 66030,
      "train_speed(iter/s)": 0.200586
    },
    {
      "acc": 0.77102299,
      "epoch": 1.54076320739585,
      "grad_norm": 5.03125,
      "learning_rate": 1.3199195105954331e-06,
      "loss": 0.82279663,
      "memory(GiB)": 147.13,
      "step": 66040,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.77918787,
      "epoch": 1.540996514968139,
      "grad_norm": 5.46875,
      "learning_rate": 1.318640912655635e-06,
      "loss": 0.79191918,
      "memory(GiB)": 147.13,
      "step": 66050,
      "train_speed(iter/s)": 0.200616
    },
    {
      "acc": 0.80649509,
      "epoch": 1.5412298225404277,
      "grad_norm": 4.40625,
      "learning_rate": 1.3173628402287785e-06,
      "loss": 0.68745565,
      "memory(GiB)": 147.13,
      "step": 66060,
      "train_speed(iter/s)": 0.200632
    },
    {
      "acc": 0.77775507,
      "epoch": 1.5414631301127169,
      "grad_norm": 7.46875,
      "learning_rate": 1.3160852934973073e-06,
      "loss": 0.79977551,
      "memory(GiB)": 147.13,
      "step": 66070,
      "train_speed(iter/s)": 0.200647
    },
    {
      "acc": 0.76136436,
      "epoch": 1.5416964376850055,
      "grad_norm": 11.75,
      "learning_rate": 1.31480827264359e-06,
      "loss": 0.86434231,
      "memory(GiB)": 147.13,
      "step": 66080,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.77416496,
      "epoch": 1.5419297452572946,
      "grad_norm": 7.0,
      "learning_rate": 1.3135317778499196e-06,
      "loss": 0.78873339,
      "memory(GiB)": 147.13,
      "step": 66090,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.76674585,
      "epoch": 1.5421630528295833,
      "grad_norm": 5.8125,
      "learning_rate": 1.3122558092985188e-06,
      "loss": 0.83854027,
      "memory(GiB)": 147.13,
      "step": 66100,
      "train_speed(iter/s)": 0.200697
    },
    {
      "acc": 0.77215676,
      "epoch": 1.5423963604018724,
      "grad_norm": 4.375,
      "learning_rate": 1.3109803671715283e-06,
      "loss": 0.81484003,
      "memory(GiB)": 147.13,
      "step": 66110,
      "train_speed(iter/s)": 0.200713
    },
    {
      "acc": 0.79621944,
      "epoch": 1.5426296679741611,
      "grad_norm": 5.78125,
      "learning_rate": 1.30970545165102e-06,
      "loss": 0.72149839,
      "memory(GiB)": 147.13,
      "step": 66120,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.79580064,
      "epoch": 1.5428629755464502,
      "grad_norm": 12.9375,
      "learning_rate": 1.3084310629189868e-06,
      "loss": 0.73481159,
      "memory(GiB)": 147.13,
      "step": 66130,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.77218895,
      "epoch": 1.543096283118739,
      "grad_norm": 5.0625,
      "learning_rate": 1.3071572011573453e-06,
      "loss": 0.83505878,
      "memory(GiB)": 147.13,
      "step": 66140,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.78478584,
      "epoch": 1.5433295906910278,
      "grad_norm": 3.75,
      "learning_rate": 1.3058838665479418e-06,
      "loss": 0.79148293,
      "memory(GiB)": 147.13,
      "step": 66150,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.7856843,
      "epoch": 1.5435628982633167,
      "grad_norm": 4.21875,
      "learning_rate": 1.304611059272542e-06,
      "loss": 0.77415447,
      "memory(GiB)": 147.13,
      "step": 66160,
      "train_speed(iter/s)": 0.200788
    },
    {
      "acc": 0.77835255,
      "epoch": 1.5437962058356056,
      "grad_norm": 6.25,
      "learning_rate": 1.3033387795128416e-06,
      "loss": 0.79931955,
      "memory(GiB)": 147.13,
      "step": 66170,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.76517811,
      "epoch": 1.5440295134078945,
      "grad_norm": 5.3125,
      "learning_rate": 1.302067027450456e-06,
      "loss": 0.84638748,
      "memory(GiB)": 147.13,
      "step": 66180,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.78675337,
      "epoch": 1.5442628209801834,
      "grad_norm": 8.125,
      "learning_rate": 1.3007958032669283e-06,
      "loss": 0.76565952,
      "memory(GiB)": 147.13,
      "step": 66190,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.78021317,
      "epoch": 1.5444961285524723,
      "grad_norm": 5.0625,
      "learning_rate": 1.2995251071437253e-06,
      "loss": 0.78182669,
      "memory(GiB)": 147.13,
      "step": 66200,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.78736315,
      "epoch": 1.5447294361247612,
      "grad_norm": 5.75,
      "learning_rate": 1.2982549392622362e-06,
      "loss": 0.75720816,
      "memory(GiB)": 147.13,
      "step": 66210,
      "train_speed(iter/s)": 0.200866
    },
    {
      "acc": 0.76521492,
      "epoch": 1.5449627436970501,
      "grad_norm": 5.3125,
      "learning_rate": 1.2969852998037813e-06,
      "loss": 0.8522131,
      "memory(GiB)": 147.13,
      "step": 66220,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.78474207,
      "epoch": 1.545196051269339,
      "grad_norm": 4.21875,
      "learning_rate": 1.2957161889495972e-06,
      "loss": 0.77432103,
      "memory(GiB)": 147.13,
      "step": 66230,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.77806253,
      "epoch": 1.545429358841628,
      "grad_norm": 4.59375,
      "learning_rate": 1.2944476068808526e-06,
      "loss": 0.80501852,
      "memory(GiB)": 147.13,
      "step": 66240,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.7818716,
      "epoch": 1.5456626664139168,
      "grad_norm": 4.6875,
      "learning_rate": 1.2931795537786357e-06,
      "loss": 0.77636614,
      "memory(GiB)": 147.13,
      "step": 66250,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.77749758,
      "epoch": 1.5458959739862057,
      "grad_norm": 5.40625,
      "learning_rate": 1.2919120298239591e-06,
      "loss": 0.80647802,
      "memory(GiB)": 147.13,
      "step": 66260,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.77590075,
      "epoch": 1.5461292815584946,
      "grad_norm": 5.75,
      "learning_rate": 1.2906450351977646e-06,
      "loss": 0.81553898,
      "memory(GiB)": 147.13,
      "step": 66270,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.78080235,
      "epoch": 1.5463625891307835,
      "grad_norm": 6.0,
      "learning_rate": 1.2893785700809118e-06,
      "loss": 0.79015241,
      "memory(GiB)": 147.13,
      "step": 66280,
      "train_speed(iter/s)": 0.200975
    },
    {
      "acc": 0.78328552,
      "epoch": 1.5465958967030724,
      "grad_norm": 4.8125,
      "learning_rate": 1.2881126346541922e-06,
      "loss": 0.76925888,
      "memory(GiB)": 147.13,
      "step": 66290,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.77661171,
      "epoch": 1.5468292042753613,
      "grad_norm": 4.46875,
      "learning_rate": 1.2868472290983158e-06,
      "loss": 0.80561304,
      "memory(GiB)": 147.13,
      "step": 66300,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.79191103,
      "epoch": 1.5470625118476502,
      "grad_norm": 3.734375,
      "learning_rate": 1.2855823535939188e-06,
      "loss": 0.74329491,
      "memory(GiB)": 147.13,
      "step": 66310,
      "train_speed(iter/s)": 0.20102
    },
    {
      "acc": 0.77668848,
      "epoch": 1.547295819419939,
      "grad_norm": 4.9375,
      "learning_rate": 1.28431800832156e-06,
      "loss": 0.79781227,
      "memory(GiB)": 147.13,
      "step": 66320,
      "train_speed(iter/s)": 0.201036
    },
    {
      "acc": 0.77691069,
      "epoch": 1.5475291269922278,
      "grad_norm": 5.03125,
      "learning_rate": 1.2830541934617274e-06,
      "loss": 0.79807272,
      "memory(GiB)": 147.13,
      "step": 66330,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.78547478,
      "epoch": 1.547762434564517,
      "grad_norm": 5.1875,
      "learning_rate": 1.2817909091948293e-06,
      "loss": 0.76715317,
      "memory(GiB)": 147.13,
      "step": 66340,
      "train_speed(iter/s)": 0.201066
    },
    {
      "acc": 0.78639536,
      "epoch": 1.5479957421368056,
      "grad_norm": 5.0625,
      "learning_rate": 1.2805281557011972e-06,
      "loss": 0.75163832,
      "memory(GiB)": 147.13,
      "step": 66350,
      "train_speed(iter/s)": 0.201082
    },
    {
      "acc": 0.79573755,
      "epoch": 1.5482290497090947,
      "grad_norm": 7.09375,
      "learning_rate": 1.2792659331610919e-06,
      "loss": 0.72670212,
      "memory(GiB)": 147.13,
      "step": 66360,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.77826233,
      "epoch": 1.5484623572813834,
      "grad_norm": 6.21875,
      "learning_rate": 1.2780042417546917e-06,
      "loss": 0.80692101,
      "memory(GiB)": 147.13,
      "step": 66370,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.76512957,
      "epoch": 1.5486956648536725,
      "grad_norm": 5.71875,
      "learning_rate": 1.276743081662107e-06,
      "loss": 0.85662422,
      "memory(GiB)": 147.13,
      "step": 66380,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.78918648,
      "epoch": 1.5489289724259612,
      "grad_norm": 4.78125,
      "learning_rate": 1.2754824530633654e-06,
      "loss": 0.77457085,
      "memory(GiB)": 147.13,
      "step": 66390,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.76402283,
      "epoch": 1.5491622799982503,
      "grad_norm": 4.71875,
      "learning_rate": 1.2742223561384204e-06,
      "loss": 0.86086817,
      "memory(GiB)": 147.13,
      "step": 66400,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.77913628,
      "epoch": 1.549395587570539,
      "grad_norm": 4.9375,
      "learning_rate": 1.2729627910671544e-06,
      "loss": 0.79207058,
      "memory(GiB)": 147.13,
      "step": 66410,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.80253363,
      "epoch": 1.549628895142828,
      "grad_norm": 5.9375,
      "learning_rate": 1.271703758029364e-06,
      "loss": 0.71166973,
      "memory(GiB)": 147.13,
      "step": 66420,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.77526503,
      "epoch": 1.5498622027151168,
      "grad_norm": 5.53125,
      "learning_rate": 1.270445257204781e-06,
      "loss": 0.78889947,
      "memory(GiB)": 147.13,
      "step": 66430,
      "train_speed(iter/s)": 0.201205
    },
    {
      "acc": 0.78784337,
      "epoch": 1.5500955102874059,
      "grad_norm": 6.625,
      "learning_rate": 1.2691872887730528e-06,
      "loss": 0.74168453,
      "memory(GiB)": 147.13,
      "step": 66440,
      "train_speed(iter/s)": 0.20122
    },
    {
      "acc": 0.79704084,
      "epoch": 1.5503288178596946,
      "grad_norm": 7.65625,
      "learning_rate": 1.2679298529137563e-06,
      "loss": 0.75291963,
      "memory(GiB)": 147.13,
      "step": 66450,
      "train_speed(iter/s)": 0.201236
    },
    {
      "acc": 0.78185368,
      "epoch": 1.5505621254319837,
      "grad_norm": 4.8125,
      "learning_rate": 1.266672949806388e-06,
      "loss": 0.78402071,
      "memory(GiB)": 147.13,
      "step": 66460,
      "train_speed(iter/s)": 0.201252
    },
    {
      "acc": 0.78436918,
      "epoch": 1.5507954330042724,
      "grad_norm": 5.875,
      "learning_rate": 1.265416579630373e-06,
      "loss": 0.76729245,
      "memory(GiB)": 147.13,
      "step": 66470,
      "train_speed(iter/s)": 0.201267
    },
    {
      "acc": 0.77176619,
      "epoch": 1.5510287405765615,
      "grad_norm": 5.6875,
      "learning_rate": 1.2641607425650565e-06,
      "loss": 0.8262764,
      "memory(GiB)": 147.13,
      "step": 66480,
      "train_speed(iter/s)": 0.201283
    },
    {
      "acc": 0.79059591,
      "epoch": 1.5512620481488502,
      "grad_norm": 4.125,
      "learning_rate": 1.2629054387897066e-06,
      "loss": 0.75890532,
      "memory(GiB)": 147.13,
      "step": 66490,
      "train_speed(iter/s)": 0.201299
    },
    {
      "acc": 0.78669443,
      "epoch": 1.5514953557211393,
      "grad_norm": 6.75,
      "learning_rate": 1.2616506684835217e-06,
      "loss": 0.77961187,
      "memory(GiB)": 147.13,
      "step": 66500,
      "train_speed(iter/s)": 0.201314
    },
    {
      "epoch": 1.5514953557211393,
      "eval_acc": 0.7446834917400007,
      "eval_loss": 0.8045360445976257,
      "eval_runtime": 1270.6085,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "step": 66500
    },
    {
      "acc": 0.7779171,
      "epoch": 1.551728663293428,
      "grad_norm": 3.71875,
      "learning_rate": 1.2603964318256167e-06,
      "loss": 0.82731237,
      "memory(GiB)": 147.13,
      "step": 66510,
      "train_speed(iter/s)": 0.200544
    },
    {
      "acc": 0.76602592,
      "epoch": 1.5519619708657169,
      "grad_norm": 5.5625,
      "learning_rate": 1.2591427289950358e-06,
      "loss": 0.85228939,
      "memory(GiB)": 147.13,
      "step": 66520,
      "train_speed(iter/s)": 0.20056
    },
    {
      "acc": 0.79903426,
      "epoch": 1.5521952784380058,
      "grad_norm": 4.40625,
      "learning_rate": 1.2578895601707435e-06,
      "loss": 0.74350095,
      "memory(GiB)": 147.13,
      "step": 66530,
      "train_speed(iter/s)": 0.200576
    },
    {
      "acc": 0.77694921,
      "epoch": 1.5524285860102947,
      "grad_norm": 5.6875,
      "learning_rate": 1.2566369255316296e-06,
      "loss": 0.81319304,
      "memory(GiB)": 147.13,
      "step": 66540,
      "train_speed(iter/s)": 0.200592
    },
    {
      "acc": 0.77841153,
      "epoch": 1.5526618935825836,
      "grad_norm": 5.46875,
      "learning_rate": 1.255384825256507e-06,
      "loss": 0.79920754,
      "memory(GiB)": 147.13,
      "step": 66550,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.77843218,
      "epoch": 1.5528952011548725,
      "grad_norm": 4.9375,
      "learning_rate": 1.2541332595241112e-06,
      "loss": 0.8021946,
      "memory(GiB)": 147.13,
      "step": 66560,
      "train_speed(iter/s)": 0.200622
    },
    {
      "acc": 0.77249784,
      "epoch": 1.5531285087271613,
      "grad_norm": 7.15625,
      "learning_rate": 1.2528822285131059e-06,
      "loss": 0.8305788,
      "memory(GiB)": 147.13,
      "step": 66570,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.77560177,
      "epoch": 1.5533618162994502,
      "grad_norm": 5.5625,
      "learning_rate": 1.2516317324020727e-06,
      "loss": 0.79391565,
      "memory(GiB)": 147.13,
      "step": 66580,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.78499255,
      "epoch": 1.5535951238717391,
      "grad_norm": 4.21875,
      "learning_rate": 1.2503817713695221e-06,
      "loss": 0.76304283,
      "memory(GiB)": 147.13,
      "step": 66590,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.79315147,
      "epoch": 1.553828431444028,
      "grad_norm": 5.0625,
      "learning_rate": 1.2491323455938831e-06,
      "loss": 0.75981688,
      "memory(GiB)": 147.13,
      "step": 66600,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.79385643,
      "epoch": 1.554061739016317,
      "grad_norm": 3.984375,
      "learning_rate": 1.2478834552535135e-06,
      "loss": 0.72674108,
      "memory(GiB)": 147.13,
      "step": 66610,
      "train_speed(iter/s)": 0.200699
    },
    {
      "acc": 0.77791653,
      "epoch": 1.5542950465886058,
      "grad_norm": 5.71875,
      "learning_rate": 1.2466351005266903e-06,
      "loss": 0.78302855,
      "memory(GiB)": 147.13,
      "step": 66620,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.80931139,
      "epoch": 1.5545283541608947,
      "grad_norm": 3.828125,
      "learning_rate": 1.2453872815916151e-06,
      "loss": 0.67272167,
      "memory(GiB)": 147.13,
      "step": 66630,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.76477842,
      "epoch": 1.5547616617331836,
      "grad_norm": 5.25,
      "learning_rate": 1.2441399986264174e-06,
      "loss": 0.8524189,
      "memory(GiB)": 147.13,
      "step": 66640,
      "train_speed(iter/s)": 0.200745
    },
    {
      "acc": 0.78072309,
      "epoch": 1.5549949693054725,
      "grad_norm": 6.5,
      "learning_rate": 1.2428932518091413e-06,
      "loss": 0.79402523,
      "memory(GiB)": 147.13,
      "step": 66650,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.77371817,
      "epoch": 1.5552282768777614,
      "grad_norm": 30.25,
      "learning_rate": 1.2416470413177633e-06,
      "loss": 0.82794876,
      "memory(GiB)": 147.13,
      "step": 66660,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.7927494,
      "epoch": 1.5554615844500503,
      "grad_norm": 5.90625,
      "learning_rate": 1.2404013673301768e-06,
      "loss": 0.73954892,
      "memory(GiB)": 147.13,
      "step": 66670,
      "train_speed(iter/s)": 0.200794
    },
    {
      "acc": 0.78622789,
      "epoch": 1.5556948920223392,
      "grad_norm": 7.71875,
      "learning_rate": 1.2391562300242044e-06,
      "loss": 0.79023104,
      "memory(GiB)": 147.13,
      "step": 66680,
      "train_speed(iter/s)": 0.200809
    },
    {
      "acc": 0.77736959,
      "epoch": 1.5559281995946281,
      "grad_norm": 4.21875,
      "learning_rate": 1.2379116295775877e-06,
      "loss": 0.79935203,
      "memory(GiB)": 147.13,
      "step": 66690,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.77949438,
      "epoch": 1.556161507166917,
      "grad_norm": 6.78125,
      "learning_rate": 1.236667566167991e-06,
      "loss": 0.791997,
      "memory(GiB)": 147.13,
      "step": 66700,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.78041005,
      "epoch": 1.556394814739206,
      "grad_norm": 4.53125,
      "learning_rate": 1.2354240399730083e-06,
      "loss": 0.78324399,
      "memory(GiB)": 147.13,
      "step": 66710,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.77618508,
      "epoch": 1.5566281223114946,
      "grad_norm": 6.875,
      "learning_rate": 1.2341810511701485e-06,
      "loss": 0.80764313,
      "memory(GiB)": 147.13,
      "step": 66720,
      "train_speed(iter/s)": 0.20087
    },
    {
      "acc": 0.78065491,
      "epoch": 1.5568614298837837,
      "grad_norm": 5.125,
      "learning_rate": 1.2329385999368509e-06,
      "loss": 0.80852127,
      "memory(GiB)": 147.13,
      "step": 66730,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.78683767,
      "epoch": 1.5570947374560724,
      "grad_norm": 4.65625,
      "learning_rate": 1.2316966864504725e-06,
      "loss": 0.76271248,
      "memory(GiB)": 147.13,
      "step": 66740,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.76667438,
      "epoch": 1.5573280450283615,
      "grad_norm": 6.0625,
      "learning_rate": 1.230455310888299e-06,
      "loss": 0.85402393,
      "memory(GiB)": 147.13,
      "step": 66750,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.80291309,
      "epoch": 1.5575613526006502,
      "grad_norm": 5.21875,
      "learning_rate": 1.2292144734275347e-06,
      "loss": 0.70809097,
      "memory(GiB)": 147.13,
      "step": 66760,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.79057379,
      "epoch": 1.5577946601729393,
      "grad_norm": 4.46875,
      "learning_rate": 1.2279741742453088e-06,
      "loss": 0.75554724,
      "memory(GiB)": 147.13,
      "step": 66770,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.79762211,
      "epoch": 1.558027967745228,
      "grad_norm": 4.1875,
      "learning_rate": 1.2267344135186743e-06,
      "loss": 0.72075491,
      "memory(GiB)": 147.13,
      "step": 66780,
      "train_speed(iter/s)": 0.200965
    },
    {
      "acc": 0.78778682,
      "epoch": 1.5582612753175171,
      "grad_norm": 6.0,
      "learning_rate": 1.2254951914246038e-06,
      "loss": 0.75706754,
      "memory(GiB)": 147.13,
      "step": 66790,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.77799454,
      "epoch": 1.5584945828898058,
      "grad_norm": 4.5,
      "learning_rate": 1.2242565081400004e-06,
      "loss": 0.80059433,
      "memory(GiB)": 147.13,
      "step": 66800,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.78043733,
      "epoch": 1.558727890462095,
      "grad_norm": 7.78125,
      "learning_rate": 1.223018363841682e-06,
      "loss": 0.78616896,
      "memory(GiB)": 147.13,
      "step": 66810,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.78293238,
      "epoch": 1.5589611980343836,
      "grad_norm": 6.84375,
      "learning_rate": 1.2217807587063962e-06,
      "loss": 0.77822418,
      "memory(GiB)": 147.13,
      "step": 66820,
      "train_speed(iter/s)": 0.201029
    },
    {
      "acc": 0.75709181,
      "epoch": 1.5591945056066727,
      "grad_norm": 5.3125,
      "learning_rate": 1.2205436929108093e-06,
      "loss": 0.87782154,
      "memory(GiB)": 147.13,
      "step": 66830,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.78140383,
      "epoch": 1.5594278131789614,
      "grad_norm": 5.4375,
      "learning_rate": 1.2193071666315114e-06,
      "loss": 0.77247772,
      "memory(GiB)": 147.13,
      "step": 66840,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.77636423,
      "epoch": 1.5596611207512505,
      "grad_norm": 4.6875,
      "learning_rate": 1.2180711800450184e-06,
      "loss": 0.79241691,
      "memory(GiB)": 147.13,
      "step": 66850,
      "train_speed(iter/s)": 0.201076
    },
    {
      "acc": 0.76072507,
      "epoch": 1.5598944283235392,
      "grad_norm": 6.34375,
      "learning_rate": 1.2168357333277641e-06,
      "loss": 0.85695286,
      "memory(GiB)": 147.13,
      "step": 66860,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.80281935,
      "epoch": 1.5601277358958283,
      "grad_norm": 4.09375,
      "learning_rate": 1.215600826656113e-06,
      "loss": 0.72951822,
      "memory(GiB)": 147.13,
      "step": 66870,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.77838449,
      "epoch": 1.560361043468117,
      "grad_norm": 4.65625,
      "learning_rate": 1.2143664602063415e-06,
      "loss": 0.79508176,
      "memory(GiB)": 147.13,
      "step": 66880,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.78288064,
      "epoch": 1.5605943510404061,
      "grad_norm": 6.15625,
      "learning_rate": 1.2131326341546596e-06,
      "loss": 0.78689842,
      "memory(GiB)": 147.13,
      "step": 66890,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.7747613,
      "epoch": 1.5608276586126948,
      "grad_norm": 5.5,
      "learning_rate": 1.2118993486771924e-06,
      "loss": 0.80352983,
      "memory(GiB)": 147.13,
      "step": 66900,
      "train_speed(iter/s)": 0.201153
    },
    {
      "acc": 0.78817844,
      "epoch": 1.5610609661849837,
      "grad_norm": 6.28125,
      "learning_rate": 1.2106666039499942e-06,
      "loss": 0.77384362,
      "memory(GiB)": 147.13,
      "step": 66910,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.75901203,
      "epoch": 1.5612942737572726,
      "grad_norm": 5.40625,
      "learning_rate": 1.2094344001490383e-06,
      "loss": 0.88302488,
      "memory(GiB)": 147.13,
      "step": 66920,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.78414526,
      "epoch": 1.5615275813295615,
      "grad_norm": 6.0625,
      "learning_rate": 1.2082027374502181e-06,
      "loss": 0.79911737,
      "memory(GiB)": 147.13,
      "step": 66930,
      "train_speed(iter/s)": 0.201201
    },
    {
      "acc": 0.77624192,
      "epoch": 1.5617608889018504,
      "grad_norm": 5.9375,
      "learning_rate": 1.2069716160293577e-06,
      "loss": 0.80140715,
      "memory(GiB)": 147.13,
      "step": 66940,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.7758502,
      "epoch": 1.5619941964741393,
      "grad_norm": 5.0625,
      "learning_rate": 1.2057410360621952e-06,
      "loss": 0.80642414,
      "memory(GiB)": 147.13,
      "step": 66950,
      "train_speed(iter/s)": 0.201233
    },
    {
      "acc": 0.76256318,
      "epoch": 1.5622275040464282,
      "grad_norm": 10.5625,
      "learning_rate": 1.2045109977243996e-06,
      "loss": 0.84569578,
      "memory(GiB)": 147.13,
      "step": 66960,
      "train_speed(iter/s)": 0.201249
    },
    {
      "acc": 0.79698315,
      "epoch": 1.562460811618717,
      "grad_norm": 5.59375,
      "learning_rate": 1.203281501191556e-06,
      "loss": 0.72654285,
      "memory(GiB)": 147.13,
      "step": 66970,
      "train_speed(iter/s)": 0.201265
    },
    {
      "acc": 0.78373365,
      "epoch": 1.562694119191006,
      "grad_norm": 7.03125,
      "learning_rate": 1.2020525466391758e-06,
      "loss": 0.76392503,
      "memory(GiB)": 147.13,
      "step": 66980,
      "train_speed(iter/s)": 0.20128
    },
    {
      "acc": 0.79392319,
      "epoch": 1.5629274267632949,
      "grad_norm": 7.875,
      "learning_rate": 1.2008241342426907e-06,
      "loss": 0.71769762,
      "memory(GiB)": 147.13,
      "step": 66990,
      "train_speed(iter/s)": 0.201296
    },
    {
      "acc": 0.79626703,
      "epoch": 1.5631607343355838,
      "grad_norm": 6.46875,
      "learning_rate": 1.1995962641774556e-06,
      "loss": 0.71886544,
      "memory(GiB)": 147.13,
      "step": 67000,
      "train_speed(iter/s)": 0.201311
    },
    {
      "epoch": 1.5631607343355838,
      "eval_acc": 0.7446565875089741,
      "eval_loss": 0.80445796251297,
      "eval_runtime": 1269.6632,
      "eval_samples_per_second": 28.347,
      "eval_steps_per_second": 14.174,
      "step": 67000
    },
    {
      "acc": 0.7860218,
      "epoch": 1.5633940419078727,
      "grad_norm": 5.0,
      "learning_rate": 1.1983689366187512e-06,
      "loss": 0.77764053,
      "memory(GiB)": 147.13,
      "step": 67010,
      "train_speed(iter/s)": 0.200547
    },
    {
      "acc": 0.77029128,
      "epoch": 1.5636273494801616,
      "grad_norm": 5.8125,
      "learning_rate": 1.1971421517417748e-06,
      "loss": 0.85604916,
      "memory(GiB)": 147.13,
      "step": 67020,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.78226252,
      "epoch": 1.5638606570524505,
      "grad_norm": 5.09375,
      "learning_rate": 1.1959159097216533e-06,
      "loss": 0.78034925,
      "memory(GiB)": 147.13,
      "step": 67030,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.78731737,
      "epoch": 1.5640939646247394,
      "grad_norm": 6.03125,
      "learning_rate": 1.1946902107334308e-06,
      "loss": 0.7751092,
      "memory(GiB)": 147.13,
      "step": 67040,
      "train_speed(iter/s)": 0.200595
    },
    {
      "acc": 0.77418489,
      "epoch": 1.5643272721970283,
      "grad_norm": 6.65625,
      "learning_rate": 1.1934650549520737e-06,
      "loss": 0.81517506,
      "memory(GiB)": 147.13,
      "step": 67050,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.7982223,
      "epoch": 1.5645605797693172,
      "grad_norm": 5.6875,
      "learning_rate": 1.1922404425524753e-06,
      "loss": 0.72336531,
      "memory(GiB)": 147.13,
      "step": 67060,
      "train_speed(iter/s)": 0.200626
    },
    {
      "acc": 0.78747759,
      "epoch": 1.564793887341606,
      "grad_norm": 5.03125,
      "learning_rate": 1.1910163737094465e-06,
      "loss": 0.75430288,
      "memory(GiB)": 147.13,
      "step": 67070,
      "train_speed(iter/s)": 0.200641
    },
    {
      "acc": 0.78929777,
      "epoch": 1.565027194913895,
      "grad_norm": 5.21875,
      "learning_rate": 1.189792848597725e-06,
      "loss": 0.77023878,
      "memory(GiB)": 147.13,
      "step": 67080,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.75839481,
      "epoch": 1.5652605024861839,
      "grad_norm": 6.5,
      "learning_rate": 1.1885698673919666e-06,
      "loss": 0.86226034,
      "memory(GiB)": 147.13,
      "step": 67090,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.77864447,
      "epoch": 1.5654938100584728,
      "grad_norm": 5.59375,
      "learning_rate": 1.1873474302667548e-06,
      "loss": 0.81831455,
      "memory(GiB)": 147.13,
      "step": 67100,
      "train_speed(iter/s)": 0.200688
    },
    {
      "acc": 0.77476091,
      "epoch": 1.5657271176307614,
      "grad_norm": 5.625,
      "learning_rate": 1.186125537396587e-06,
      "loss": 0.80706577,
      "memory(GiB)": 147.13,
      "step": 67110,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.76268578,
      "epoch": 1.5659604252030506,
      "grad_norm": 4.84375,
      "learning_rate": 1.1849041889558922e-06,
      "loss": 0.85323725,
      "memory(GiB)": 147.13,
      "step": 67120,
      "train_speed(iter/s)": 0.200718
    },
    {
      "acc": 0.78476624,
      "epoch": 1.5661937327753392,
      "grad_norm": 4.375,
      "learning_rate": 1.1836833851190161e-06,
      "loss": 0.78540063,
      "memory(GiB)": 147.13,
      "step": 67130,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.79616303,
      "epoch": 1.5664270403476284,
      "grad_norm": 6.25,
      "learning_rate": 1.1824631260602266e-06,
      "loss": 0.71438951,
      "memory(GiB)": 147.13,
      "step": 67140,
      "train_speed(iter/s)": 0.200749
    },
    {
      "acc": 0.79416599,
      "epoch": 1.566660347919917,
      "grad_norm": 4.59375,
      "learning_rate": 1.1812434119537187e-06,
      "loss": 0.74181514,
      "memory(GiB)": 147.13,
      "step": 67150,
      "train_speed(iter/s)": 0.200765
    },
    {
      "acc": 0.78638306,
      "epoch": 1.5668936554922062,
      "grad_norm": 8.25,
      "learning_rate": 1.1800242429736025e-06,
      "loss": 0.75746689,
      "memory(GiB)": 147.13,
      "step": 67160,
      "train_speed(iter/s)": 0.200781
    },
    {
      "acc": 0.76691604,
      "epoch": 1.5671269630644948,
      "grad_norm": 4.84375,
      "learning_rate": 1.1788056192939173e-06,
      "loss": 0.85675087,
      "memory(GiB)": 147.13,
      "step": 67170,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.78023224,
      "epoch": 1.567360270636784,
      "grad_norm": 5.75,
      "learning_rate": 1.1775875410886206e-06,
      "loss": 0.80645647,
      "memory(GiB)": 147.13,
      "step": 67180,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.80352058,
      "epoch": 1.5675935782090726,
      "grad_norm": 3.640625,
      "learning_rate": 1.17637000853159e-06,
      "loss": 0.68619394,
      "memory(GiB)": 147.13,
      "step": 67190,
      "train_speed(iter/s)": 0.200826
    },
    {
      "acc": 0.77900271,
      "epoch": 1.5678268857813618,
      "grad_norm": 4.90625,
      "learning_rate": 1.1751530217966312e-06,
      "loss": 0.79527159,
      "memory(GiB)": 147.13,
      "step": 67200,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.79193454,
      "epoch": 1.5680601933536504,
      "grad_norm": 9.1875,
      "learning_rate": 1.1739365810574677e-06,
      "loss": 0.74616985,
      "memory(GiB)": 147.13,
      "step": 67210,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.79721117,
      "epoch": 1.5682935009259396,
      "grad_norm": 9.0,
      "learning_rate": 1.1727206864877456e-06,
      "loss": 0.72338243,
      "memory(GiB)": 147.13,
      "step": 67220,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.77702656,
      "epoch": 1.5685268084982282,
      "grad_norm": 4.8125,
      "learning_rate": 1.1715053382610325e-06,
      "loss": 0.79841452,
      "memory(GiB)": 147.13,
      "step": 67230,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.74713612,
      "epoch": 1.5687601160705174,
      "grad_norm": 4.5625,
      "learning_rate": 1.170290536550821e-06,
      "loss": 0.89974098,
      "memory(GiB)": 147.13,
      "step": 67240,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.77886939,
      "epoch": 1.568993423642806,
      "grad_norm": 3.78125,
      "learning_rate": 1.1690762815305224e-06,
      "loss": 0.80204468,
      "memory(GiB)": 147.13,
      "step": 67250,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.77801399,
      "epoch": 1.5692267312150951,
      "grad_norm": 5.78125,
      "learning_rate": 1.1678625733734722e-06,
      "loss": 0.79122267,
      "memory(GiB)": 147.13,
      "step": 67260,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.77463903,
      "epoch": 1.5694600387873838,
      "grad_norm": 5.875,
      "learning_rate": 1.166649412252927e-06,
      "loss": 0.80886192,
      "memory(GiB)": 147.13,
      "step": 67270,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.81241455,
      "epoch": 1.569693346359673,
      "grad_norm": 4.8125,
      "learning_rate": 1.1654367983420628e-06,
      "loss": 0.65604534,
      "memory(GiB)": 147.13,
      "step": 67280,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.78449936,
      "epoch": 1.5699266539319616,
      "grad_norm": 5.0,
      "learning_rate": 1.1642247318139837e-06,
      "loss": 0.77122612,
      "memory(GiB)": 147.13,
      "step": 67290,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.77493362,
      "epoch": 1.5701599615042505,
      "grad_norm": 5.46875,
      "learning_rate": 1.1630132128417083e-06,
      "loss": 0.80600281,
      "memory(GiB)": 147.13,
      "step": 67300,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.7923779,
      "epoch": 1.5703932690765394,
      "grad_norm": 6.875,
      "learning_rate": 1.1618022415981827e-06,
      "loss": 0.7523242,
      "memory(GiB)": 147.13,
      "step": 67310,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.7808217,
      "epoch": 1.5706265766488283,
      "grad_norm": 5.9375,
      "learning_rate": 1.1605918182562731e-06,
      "loss": 0.784408,
      "memory(GiB)": 147.13,
      "step": 67320,
      "train_speed(iter/s)": 0.20103
    },
    {
      "acc": 0.77961884,
      "epoch": 1.5708598842211172,
      "grad_norm": 5.96875,
      "learning_rate": 1.1593819429887655e-06,
      "loss": 0.78548775,
      "memory(GiB)": 147.13,
      "step": 67330,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.79387894,
      "epoch": 1.5710931917934061,
      "grad_norm": 5.25,
      "learning_rate": 1.1581726159683698e-06,
      "loss": 0.73550348,
      "memory(GiB)": 147.13,
      "step": 67340,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.77411294,
      "epoch": 1.571326499365695,
      "grad_norm": 17.125,
      "learning_rate": 1.1569638373677162e-06,
      "loss": 0.80191069,
      "memory(GiB)": 147.13,
      "step": 67350,
      "train_speed(iter/s)": 0.201076
    },
    {
      "acc": 0.77730083,
      "epoch": 1.571559806937984,
      "grad_norm": 5.03125,
      "learning_rate": 1.1557556073593595e-06,
      "loss": 0.79704728,
      "memory(GiB)": 147.13,
      "step": 67360,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.77600565,
      "epoch": 1.5717931145102728,
      "grad_norm": 5.875,
      "learning_rate": 1.1545479261157715e-06,
      "loss": 0.80014782,
      "memory(GiB)": 147.13,
      "step": 67370,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.76335773,
      "epoch": 1.5720264220825617,
      "grad_norm": 5.9375,
      "learning_rate": 1.1533407938093515e-06,
      "loss": 0.8461832,
      "memory(GiB)": 147.13,
      "step": 67380,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.7748148,
      "epoch": 1.5722597296548506,
      "grad_norm": 5.34375,
      "learning_rate": 1.1521342106124145e-06,
      "loss": 0.80657759,
      "memory(GiB)": 147.13,
      "step": 67390,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.80296421,
      "epoch": 1.5724930372271395,
      "grad_norm": 4.21875,
      "learning_rate": 1.1509281766972026e-06,
      "loss": 0.6991787,
      "memory(GiB)": 147.13,
      "step": 67400,
      "train_speed(iter/s)": 0.201153
    },
    {
      "acc": 0.78165808,
      "epoch": 1.5727263447994284,
      "grad_norm": 6.0,
      "learning_rate": 1.149722692235875e-06,
      "loss": 0.79229851,
      "memory(GiB)": 147.13,
      "step": 67410,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.77584095,
      "epoch": 1.5729596523717173,
      "grad_norm": 5.46875,
      "learning_rate": 1.1485177574005134e-06,
      "loss": 0.79919806,
      "memory(GiB)": 147.13,
      "step": 67420,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.77663813,
      "epoch": 1.5731929599440062,
      "grad_norm": 6.40625,
      "learning_rate": 1.1473133723631241e-06,
      "loss": 0.83092327,
      "memory(GiB)": 147.13,
      "step": 67430,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.761483,
      "epoch": 1.573426267516295,
      "grad_norm": 4.53125,
      "learning_rate": 1.1461095372956322e-06,
      "loss": 0.86643829,
      "memory(GiB)": 147.13,
      "step": 67440,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.76875048,
      "epoch": 1.573659575088584,
      "grad_norm": 3.765625,
      "learning_rate": 1.1449062523698839e-06,
      "loss": 0.82976036,
      "memory(GiB)": 147.13,
      "step": 67450,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.77960482,
      "epoch": 1.573892882660873,
      "grad_norm": 5.375,
      "learning_rate": 1.1437035177576467e-06,
      "loss": 0.79715147,
      "memory(GiB)": 147.13,
      "step": 67460,
      "train_speed(iter/s)": 0.201245
    },
    {
      "acc": 0.77590551,
      "epoch": 1.5741261902331618,
      "grad_norm": 4.9375,
      "learning_rate": 1.142501333630614e-06,
      "loss": 0.79287214,
      "memory(GiB)": 147.13,
      "step": 67470,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.7959466,
      "epoch": 1.5743594978054507,
      "grad_norm": 3.984375,
      "learning_rate": 1.1412997001603947e-06,
      "loss": 0.74863319,
      "memory(GiB)": 147.13,
      "step": 67480,
      "train_speed(iter/s)": 0.201277
    },
    {
      "acc": 0.80535631,
      "epoch": 1.5745928053777396,
      "grad_norm": 35.5,
      "learning_rate": 1.1400986175185214e-06,
      "loss": 0.70440435,
      "memory(GiB)": 147.13,
      "step": 67490,
      "train_speed(iter/s)": 0.201291
    },
    {
      "acc": 0.78224182,
      "epoch": 1.5748261129500283,
      "grad_norm": 5.3125,
      "learning_rate": 1.1388980858764504e-06,
      "loss": 0.76127481,
      "memory(GiB)": 147.13,
      "step": 67500,
      "train_speed(iter/s)": 0.201306
    },
    {
      "epoch": 1.5748261129500283,
      "eval_acc": 0.74470014674016,
      "eval_loss": 0.804486095905304,
      "eval_runtime": 1270.1666,
      "eval_samples_per_second": 28.336,
      "eval_steps_per_second": 14.168,
      "step": 67500
    },
    {
      "acc": 0.78972077,
      "epoch": 1.5750594205223174,
      "grad_norm": 5.1875,
      "learning_rate": 1.1376981054055542e-06,
      "loss": 0.75714273,
      "memory(GiB)": 147.13,
      "step": 67510,
      "train_speed(iter/s)": 0.200548
    },
    {
      "acc": 0.75635891,
      "epoch": 1.575292728094606,
      "grad_norm": 5.65625,
      "learning_rate": 1.136498676277133e-06,
      "loss": 0.8773056,
      "memory(GiB)": 147.13,
      "step": 67520,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.78398147,
      "epoch": 1.5755260356668952,
      "grad_norm": 4.59375,
      "learning_rate": 1.135299798662402e-06,
      "loss": 0.76907773,
      "memory(GiB)": 147.13,
      "step": 67530,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.77541685,
      "epoch": 1.5757593432391839,
      "grad_norm": 5.28125,
      "learning_rate": 1.1341014727325038e-06,
      "loss": 0.83951645,
      "memory(GiB)": 147.13,
      "step": 67540,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.77365732,
      "epoch": 1.575992650811473,
      "grad_norm": 4.9375,
      "learning_rate": 1.1329036986584968e-06,
      "loss": 0.80199337,
      "memory(GiB)": 147.13,
      "step": 67550,
      "train_speed(iter/s)": 0.20061
    },
    {
      "acc": 0.77182055,
      "epoch": 1.5762259583837617,
      "grad_norm": 10.0,
      "learning_rate": 1.131706476611364e-06,
      "loss": 0.84103651,
      "memory(GiB)": 147.13,
      "step": 67560,
      "train_speed(iter/s)": 0.200624
    },
    {
      "acc": 0.78937941,
      "epoch": 1.5764592659560508,
      "grad_norm": 5.375,
      "learning_rate": 1.1305098067620074e-06,
      "loss": 0.74375906,
      "memory(GiB)": 147.13,
      "step": 67570,
      "train_speed(iter/s)": 0.200638
    },
    {
      "acc": 0.78117437,
      "epoch": 1.5766925735283395,
      "grad_norm": 5.3125,
      "learning_rate": 1.1293136892812507e-06,
      "loss": 0.79897671,
      "memory(GiB)": 147.13,
      "step": 67580,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.78500509,
      "epoch": 1.5769258811006286,
      "grad_norm": 4.625,
      "learning_rate": 1.1281181243398414e-06,
      "loss": 0.76446457,
      "memory(GiB)": 147.13,
      "step": 67590,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.76707001,
      "epoch": 1.5771591886729173,
      "grad_norm": 4.9375,
      "learning_rate": 1.1269231121084439e-06,
      "loss": 0.83135767,
      "memory(GiB)": 147.13,
      "step": 67600,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.78357925,
      "epoch": 1.5773924962452064,
      "grad_norm": 5.6875,
      "learning_rate": 1.1257286527576488e-06,
      "loss": 0.77646589,
      "memory(GiB)": 147.13,
      "step": 67610,
      "train_speed(iter/s)": 0.200698
    },
    {
      "acc": 0.76686993,
      "epoch": 1.577625803817495,
      "grad_norm": 6.46875,
      "learning_rate": 1.1245347464579626e-06,
      "loss": 0.83330984,
      "memory(GiB)": 147.13,
      "step": 67620,
      "train_speed(iter/s)": 0.200714
    },
    {
      "acc": 0.77705078,
      "epoch": 1.5778591113897842,
      "grad_norm": 5.1875,
      "learning_rate": 1.1233413933798143e-06,
      "loss": 0.80012417,
      "memory(GiB)": 147.13,
      "step": 67630,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78694525,
      "epoch": 1.5780924189620729,
      "grad_norm": 6.53125,
      "learning_rate": 1.1221485936935571e-06,
      "loss": 0.75556831,
      "memory(GiB)": 147.13,
      "step": 67640,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.7862113,
      "epoch": 1.578325726534362,
      "grad_norm": 4.59375,
      "learning_rate": 1.1209563475694608e-06,
      "loss": 0.77506971,
      "memory(GiB)": 147.13,
      "step": 67650,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.76903219,
      "epoch": 1.5785590341066507,
      "grad_norm": 6.15625,
      "learning_rate": 1.1197646551777196e-06,
      "loss": 0.83937769,
      "memory(GiB)": 147.13,
      "step": 67660,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.78912687,
      "epoch": 1.5787923416789398,
      "grad_norm": 4.53125,
      "learning_rate": 1.118573516688447e-06,
      "loss": 0.75033379,
      "memory(GiB)": 147.13,
      "step": 67670,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.7752059,
      "epoch": 1.5790256492512285,
      "grad_norm": 4.625,
      "learning_rate": 1.1173829322716774e-06,
      "loss": 0.80030231,
      "memory(GiB)": 147.13,
      "step": 67680,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.77555189,
      "epoch": 1.5792589568235174,
      "grad_norm": 4.8125,
      "learning_rate": 1.116192902097365e-06,
      "loss": 0.79320483,
      "memory(GiB)": 147.13,
      "step": 67690,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.77827835,
      "epoch": 1.5794922643958063,
      "grad_norm": 8.25,
      "learning_rate": 1.1150034263353887e-06,
      "loss": 0.78702092,
      "memory(GiB)": 147.13,
      "step": 67700,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.78876238,
      "epoch": 1.5797255719680952,
      "grad_norm": 4.40625,
      "learning_rate": 1.113814505155545e-06,
      "loss": 0.75948009,
      "memory(GiB)": 147.13,
      "step": 67710,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.78561726,
      "epoch": 1.579958879540384,
      "grad_norm": 4.3125,
      "learning_rate": 1.11262613872755e-06,
      "loss": 0.78061628,
      "memory(GiB)": 147.13,
      "step": 67720,
      "train_speed(iter/s)": 0.200863
    },
    {
      "acc": 0.76869831,
      "epoch": 1.580192187112673,
      "grad_norm": 4.6875,
      "learning_rate": 1.111438327221046e-06,
      "loss": 0.84357386,
      "memory(GiB)": 147.13,
      "step": 67730,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.79987097,
      "epoch": 1.5804254946849619,
      "grad_norm": 4.78125,
      "learning_rate": 1.1102510708055897e-06,
      "loss": 0.71598945,
      "memory(GiB)": 147.13,
      "step": 67740,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.76315088,
      "epoch": 1.5806588022572508,
      "grad_norm": 6.78125,
      "learning_rate": 1.1090643696506648e-06,
      "loss": 0.87976942,
      "memory(GiB)": 147.13,
      "step": 67750,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.79254284,
      "epoch": 1.5808921098295396,
      "grad_norm": 4.34375,
      "learning_rate": 1.1078782239256707e-06,
      "loss": 0.71899223,
      "memory(GiB)": 147.13,
      "step": 67760,
      "train_speed(iter/s)": 0.200924
    },
    {
      "acc": 0.78601794,
      "epoch": 1.5811254174018285,
      "grad_norm": 4.875,
      "learning_rate": 1.106692633799928e-06,
      "loss": 0.75329552,
      "memory(GiB)": 147.13,
      "step": 67770,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.77427597,
      "epoch": 1.5813587249741174,
      "grad_norm": 6.0625,
      "learning_rate": 1.1055075994426833e-06,
      "loss": 0.818151,
      "memory(GiB)": 147.13,
      "step": 67780,
      "train_speed(iter/s)": 0.200955
    },
    {
      "acc": 0.79199877,
      "epoch": 1.5815920325464063,
      "grad_norm": 6.875,
      "learning_rate": 1.1043231210230949e-06,
      "loss": 0.73157892,
      "memory(GiB)": 147.13,
      "step": 67790,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.77825437,
      "epoch": 1.5818253401186952,
      "grad_norm": 6.09375,
      "learning_rate": 1.1031391987102502e-06,
      "loss": 0.78899622,
      "memory(GiB)": 147.13,
      "step": 67800,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.7796669,
      "epoch": 1.5820586476909841,
      "grad_norm": 7.0625,
      "learning_rate": 1.1019558326731522e-06,
      "loss": 0.78667364,
      "memory(GiB)": 147.13,
      "step": 67810,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.76001234,
      "epoch": 1.582291955263273,
      "grad_norm": 6.125,
      "learning_rate": 1.100773023080728e-06,
      "loss": 0.8563343,
      "memory(GiB)": 147.13,
      "step": 67820,
      "train_speed(iter/s)": 0.201016
    },
    {
      "acc": 0.76715779,
      "epoch": 1.582525262835562,
      "grad_norm": 6.3125,
      "learning_rate": 1.09959077010182e-06,
      "loss": 0.83880901,
      "memory(GiB)": 147.13,
      "step": 67830,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.77683773,
      "epoch": 1.5827585704078508,
      "grad_norm": 4.09375,
      "learning_rate": 1.0984090739051984e-06,
      "loss": 0.79647036,
      "memory(GiB)": 147.13,
      "step": 67840,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.77974472,
      "epoch": 1.5829918779801397,
      "grad_norm": 4.6875,
      "learning_rate": 1.0972279346595477e-06,
      "loss": 0.78166647,
      "memory(GiB)": 147.13,
      "step": 67850,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.78405037,
      "epoch": 1.5832251855524286,
      "grad_norm": 5.03125,
      "learning_rate": 1.0960473525334747e-06,
      "loss": 0.76982818,
      "memory(GiB)": 147.13,
      "step": 67860,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.76590328,
      "epoch": 1.5834584931247173,
      "grad_norm": 4.3125,
      "learning_rate": 1.094867327695509e-06,
      "loss": 0.85223007,
      "memory(GiB)": 147.13,
      "step": 67870,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.79253726,
      "epoch": 1.5836918006970064,
      "grad_norm": 4.71875,
      "learning_rate": 1.0936878603140966e-06,
      "loss": 0.72532487,
      "memory(GiB)": 147.13,
      "step": 67880,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.79152641,
      "epoch": 1.5839251082692951,
      "grad_norm": 5.0625,
      "learning_rate": 1.0925089505576085e-06,
      "loss": 0.75220585,
      "memory(GiB)": 147.13,
      "step": 67890,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.77057228,
      "epoch": 1.5841584158415842,
      "grad_norm": 5.9375,
      "learning_rate": 1.0913305985943328e-06,
      "loss": 0.83439999,
      "memory(GiB)": 147.13,
      "step": 67900,
      "train_speed(iter/s)": 0.201138
    },
    {
      "acc": 0.78468051,
      "epoch": 1.584391723413873,
      "grad_norm": 7.0,
      "learning_rate": 1.0901528045924786e-06,
      "loss": 0.75254793,
      "memory(GiB)": 147.13,
      "step": 67910,
      "train_speed(iter/s)": 0.201152
    },
    {
      "acc": 0.77283545,
      "epoch": 1.584625030986162,
      "grad_norm": 4.5,
      "learning_rate": 1.0889755687201758e-06,
      "loss": 0.82278538,
      "memory(GiB)": 147.13,
      "step": 67920,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.78898163,
      "epoch": 1.5848583385584507,
      "grad_norm": 5.46875,
      "learning_rate": 1.087798891145473e-06,
      "loss": 0.75745392,
      "memory(GiB)": 147.13,
      "step": 67930,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.75619049,
      "epoch": 1.5850916461307398,
      "grad_norm": 6.5,
      "learning_rate": 1.0866227720363431e-06,
      "loss": 0.86945744,
      "memory(GiB)": 147.13,
      "step": 67940,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.78170376,
      "epoch": 1.5853249537030285,
      "grad_norm": 5.8125,
      "learning_rate": 1.0854472115606745e-06,
      "loss": 0.79862347,
      "memory(GiB)": 147.13,
      "step": 67950,
      "train_speed(iter/s)": 0.201216
    },
    {
      "acc": 0.80022554,
      "epoch": 1.5855582612753176,
      "grad_norm": 5.4375,
      "learning_rate": 1.0842722098862813e-06,
      "loss": 0.70132508,
      "memory(GiB)": 147.13,
      "step": 67960,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.79483395,
      "epoch": 1.5857915688476063,
      "grad_norm": 4.8125,
      "learning_rate": 1.0830977671808918e-06,
      "loss": 0.74464898,
      "memory(GiB)": 147.13,
      "step": 67970,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.7618938,
      "epoch": 1.5860248764198954,
      "grad_norm": 5.9375,
      "learning_rate": 1.081923883612157e-06,
      "loss": 0.85374565,
      "memory(GiB)": 147.13,
      "step": 67980,
      "train_speed(iter/s)": 0.201261
    },
    {
      "acc": 0.77546587,
      "epoch": 1.586258183992184,
      "grad_norm": 5.1875,
      "learning_rate": 1.080750559347651e-06,
      "loss": 0.79738269,
      "memory(GiB)": 147.13,
      "step": 67990,
      "train_speed(iter/s)": 0.201277
    },
    {
      "acc": 0.78763037,
      "epoch": 1.5864914915644732,
      "grad_norm": 4.5625,
      "learning_rate": 1.0795777945548624e-06,
      "loss": 0.78351851,
      "memory(GiB)": 147.13,
      "step": 68000,
      "train_speed(iter/s)": 0.201292
    },
    {
      "epoch": 1.5864914915644732,
      "eval_acc": 0.7446918192400804,
      "eval_loss": 0.8044730424880981,
      "eval_runtime": 1269.9877,
      "eval_samples_per_second": 28.34,
      "eval_steps_per_second": 14.17,
      "step": 68000
    },
    {
      "acc": 0.78458452,
      "epoch": 1.586724799136762,
      "grad_norm": 5.96875,
      "learning_rate": 1.078405589401208e-06,
      "loss": 0.77951698,
      "memory(GiB)": 147.13,
      "step": 68010,
      "train_speed(iter/s)": 0.200538
    },
    {
      "acc": 0.77816787,
      "epoch": 1.586958106709051,
      "grad_norm": 7.375,
      "learning_rate": 1.0772339440540135e-06,
      "loss": 0.77889547,
      "memory(GiB)": 147.13,
      "step": 68020,
      "train_speed(iter/s)": 0.200553
    },
    {
      "acc": 0.77970071,
      "epoch": 1.5871914142813397,
      "grad_norm": 5.6875,
      "learning_rate": 1.076062858680535e-06,
      "loss": 0.78750424,
      "memory(GiB)": 147.13,
      "step": 68030,
      "train_speed(iter/s)": 0.200567
    },
    {
      "acc": 0.79411736,
      "epoch": 1.5874247218536288,
      "grad_norm": 3.90625,
      "learning_rate": 1.0748923334479427e-06,
      "loss": 0.73421063,
      "memory(GiB)": 147.13,
      "step": 68040,
      "train_speed(iter/s)": 0.200582
    },
    {
      "acc": 0.78528967,
      "epoch": 1.5876580294259175,
      "grad_norm": 5.4375,
      "learning_rate": 1.0737223685233306e-06,
      "loss": 0.76053562,
      "memory(GiB)": 147.13,
      "step": 68050,
      "train_speed(iter/s)": 0.200597
    },
    {
      "acc": 0.77646742,
      "epoch": 1.5878913369982064,
      "grad_norm": 4.75,
      "learning_rate": 1.0725529640737098e-06,
      "loss": 0.80411501,
      "memory(GiB)": 147.13,
      "step": 68060,
      "train_speed(iter/s)": 0.200612
    },
    {
      "acc": 0.76218376,
      "epoch": 1.5881246445704953,
      "grad_norm": 5.15625,
      "learning_rate": 1.0713841202660114e-06,
      "loss": 0.86731377,
      "memory(GiB)": 147.13,
      "step": 68070,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.77366896,
      "epoch": 1.5883579521427842,
      "grad_norm": 5.0,
      "learning_rate": 1.0702158372670895e-06,
      "loss": 0.80369129,
      "memory(GiB)": 147.13,
      "step": 68080,
      "train_speed(iter/s)": 0.200641
    },
    {
      "acc": 0.77165804,
      "epoch": 1.588591259715073,
      "grad_norm": 5.0,
      "learning_rate": 1.0690481152437138e-06,
      "loss": 0.80997429,
      "memory(GiB)": 147.13,
      "step": 68090,
      "train_speed(iter/s)": 0.200655
    },
    {
      "acc": 0.7878973,
      "epoch": 1.588824567287362,
      "grad_norm": 6.15625,
      "learning_rate": 1.0678809543625796e-06,
      "loss": 0.76152925,
      "memory(GiB)": 147.13,
      "step": 68100,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.77761049,
      "epoch": 1.5890578748596509,
      "grad_norm": 6.40625,
      "learning_rate": 1.0667143547902964e-06,
      "loss": 0.80845604,
      "memory(GiB)": 147.13,
      "step": 68110,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.78582354,
      "epoch": 1.5892911824319398,
      "grad_norm": 4.0,
      "learning_rate": 1.065548316693395e-06,
      "loss": 0.75975571,
      "memory(GiB)": 147.13,
      "step": 68120,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.78570542,
      "epoch": 1.5895244900042287,
      "grad_norm": 4.6875,
      "learning_rate": 1.0643828402383317e-06,
      "loss": 0.7738452,
      "memory(GiB)": 147.13,
      "step": 68130,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.78605022,
      "epoch": 1.5897577975765176,
      "grad_norm": 5.3125,
      "learning_rate": 1.0632179255914716e-06,
      "loss": 0.76639547,
      "memory(GiB)": 147.13,
      "step": 68140,
      "train_speed(iter/s)": 0.200731
    },
    {
      "acc": 0.77231016,
      "epoch": 1.5899911051488065,
      "grad_norm": 5.5625,
      "learning_rate": 1.06205357291911e-06,
      "loss": 0.79767914,
      "memory(GiB)": 147.13,
      "step": 68150,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.78238726,
      "epoch": 1.5902244127210954,
      "grad_norm": 8.875,
      "learning_rate": 1.0608897823874565e-06,
      "loss": 0.76666489,
      "memory(GiB)": 147.13,
      "step": 68160,
      "train_speed(iter/s)": 0.200759
    },
    {
      "acc": 0.78485689,
      "epoch": 1.5904577202933843,
      "grad_norm": 5.875,
      "learning_rate": 1.0597265541626428e-06,
      "loss": 0.77945495,
      "memory(GiB)": 147.13,
      "step": 68170,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.789116,
      "epoch": 1.5906910278656732,
      "grad_norm": 4.625,
      "learning_rate": 1.0585638884107174e-06,
      "loss": 0.76332469,
      "memory(GiB)": 147.13,
      "step": 68180,
      "train_speed(iter/s)": 0.200789
    },
    {
      "acc": 0.77295685,
      "epoch": 1.590924335437962,
      "grad_norm": 5.96875,
      "learning_rate": 1.0574017852976538e-06,
      "loss": 0.80912657,
      "memory(GiB)": 147.13,
      "step": 68190,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.78879251,
      "epoch": 1.591157643010251,
      "grad_norm": 6.28125,
      "learning_rate": 1.0562402449893394e-06,
      "loss": 0.75802283,
      "memory(GiB)": 147.13,
      "step": 68200,
      "train_speed(iter/s)": 0.20082
    },
    {
      "acc": 0.77317877,
      "epoch": 1.5913909505825399,
      "grad_norm": 6.1875,
      "learning_rate": 1.0550792676515836e-06,
      "loss": 0.81081858,
      "memory(GiB)": 147.13,
      "step": 68210,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.79010024,
      "epoch": 1.5916242581548288,
      "grad_norm": 7.875,
      "learning_rate": 1.0539188534501176e-06,
      "loss": 0.75586977,
      "memory(GiB)": 147.13,
      "step": 68220,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.80472164,
      "epoch": 1.5918575657271177,
      "grad_norm": 7.15625,
      "learning_rate": 1.0527590025505873e-06,
      "loss": 0.71593409,
      "memory(GiB)": 147.13,
      "step": 68230,
      "train_speed(iter/s)": 0.200866
    },
    {
      "acc": 0.78382373,
      "epoch": 1.5920908732994066,
      "grad_norm": 5.28125,
      "learning_rate": 1.051599715118566e-06,
      "loss": 0.78045077,
      "memory(GiB)": 147.13,
      "step": 68240,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.79803543,
      "epoch": 1.5923241808716955,
      "grad_norm": 6.09375,
      "learning_rate": 1.0504409913195346e-06,
      "loss": 0.71405649,
      "memory(GiB)": 147.13,
      "step": 68250,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.79116182,
      "epoch": 1.5925574884439841,
      "grad_norm": 5.09375,
      "learning_rate": 1.0492828313189064e-06,
      "loss": 0.73879242,
      "memory(GiB)": 147.13,
      "step": 68260,
      "train_speed(iter/s)": 0.200911
    },
    {
      "acc": 0.76173649,
      "epoch": 1.5927907960162733,
      "grad_norm": 7.75,
      "learning_rate": 1.0481252352820064e-06,
      "loss": 0.85937166,
      "memory(GiB)": 147.13,
      "step": 68270,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.78311167,
      "epoch": 1.593024103588562,
      "grad_norm": 4.4375,
      "learning_rate": 1.046968203374079e-06,
      "loss": 0.78482008,
      "memory(GiB)": 147.13,
      "step": 68280,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.75329967,
      "epoch": 1.593257411160851,
      "grad_norm": 7.40625,
      "learning_rate": 1.0458117357602944e-06,
      "loss": 0.89326878,
      "memory(GiB)": 147.13,
      "step": 68290,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.77799082,
      "epoch": 1.5934907187331397,
      "grad_norm": 4.40625,
      "learning_rate": 1.0446558326057342e-06,
      "loss": 0.79334536,
      "memory(GiB)": 147.13,
      "step": 68300,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.75523701,
      "epoch": 1.5937240263054289,
      "grad_norm": 6.28125,
      "learning_rate": 1.0435004940754062e-06,
      "loss": 0.90748978,
      "memory(GiB)": 147.13,
      "step": 68310,
      "train_speed(iter/s)": 0.200987
    },
    {
      "acc": 0.80369549,
      "epoch": 1.5939573338777175,
      "grad_norm": 6.0,
      "learning_rate": 1.0423457203342318e-06,
      "loss": 0.68501949,
      "memory(GiB)": 147.13,
      "step": 68320,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.78252659,
      "epoch": 1.5941906414500067,
      "grad_norm": 5.8125,
      "learning_rate": 1.0411915115470578e-06,
      "loss": 0.772118,
      "memory(GiB)": 147.13,
      "step": 68330,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.78786469,
      "epoch": 1.5944239490222953,
      "grad_norm": 6.15625,
      "learning_rate": 1.0400378678786449e-06,
      "loss": 0.75693922,
      "memory(GiB)": 147.13,
      "step": 68340,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.77552619,
      "epoch": 1.5946572565945845,
      "grad_norm": 4.3125,
      "learning_rate": 1.0388847894936765e-06,
      "loss": 0.80412102,
      "memory(GiB)": 147.13,
      "step": 68350,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.7692029,
      "epoch": 1.5948905641668731,
      "grad_norm": 7.34375,
      "learning_rate": 1.0377322765567533e-06,
      "loss": 0.85158634,
      "memory(GiB)": 147.13,
      "step": 68360,
      "train_speed(iter/s)": 0.201064
    },
    {
      "acc": 0.79861155,
      "epoch": 1.5951238717391623,
      "grad_norm": 5.21875,
      "learning_rate": 1.0365803292323956e-06,
      "loss": 0.68969359,
      "memory(GiB)": 147.13,
      "step": 68370,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.80761919,
      "epoch": 1.595357179311451,
      "grad_norm": 4.5,
      "learning_rate": 1.0354289476850459e-06,
      "loss": 0.67954183,
      "memory(GiB)": 147.13,
      "step": 68380,
      "train_speed(iter/s)": 0.201094
    },
    {
      "acc": 0.79081321,
      "epoch": 1.59559048688374,
      "grad_norm": 9.625,
      "learning_rate": 1.0342781320790606e-06,
      "loss": 0.74034171,
      "memory(GiB)": 147.13,
      "step": 68390,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.79297962,
      "epoch": 1.5958237944560287,
      "grad_norm": 4.03125,
      "learning_rate": 1.0331278825787211e-06,
      "loss": 0.75901136,
      "memory(GiB)": 147.13,
      "step": 68400,
      "train_speed(iter/s)": 0.201124
    },
    {
      "acc": 0.78131299,
      "epoch": 1.5960571020283179,
      "grad_norm": 6.625,
      "learning_rate": 1.0319781993482242e-06,
      "loss": 0.77496767,
      "memory(GiB)": 147.13,
      "step": 68410,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.77413187,
      "epoch": 1.5962904096006065,
      "grad_norm": 7.40625,
      "learning_rate": 1.0308290825516852e-06,
      "loss": 0.80257893,
      "memory(GiB)": 147.13,
      "step": 68420,
      "train_speed(iter/s)": 0.201154
    },
    {
      "acc": 0.78656511,
      "epoch": 1.5965237171728957,
      "grad_norm": 5.96875,
      "learning_rate": 1.0296805323531435e-06,
      "loss": 0.79238644,
      "memory(GiB)": 147.13,
      "step": 68430,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.7843667,
      "epoch": 1.5967570247451843,
      "grad_norm": 5.9375,
      "learning_rate": 1.0285325489165503e-06,
      "loss": 0.78407121,
      "memory(GiB)": 147.13,
      "step": 68440,
      "train_speed(iter/s)": 0.201187
    },
    {
      "acc": 0.81157207,
      "epoch": 1.5969903323174732,
      "grad_norm": 5.90625,
      "learning_rate": 1.0273851324057838e-06,
      "loss": 0.67407074,
      "memory(GiB)": 147.13,
      "step": 68450,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.79256735,
      "epoch": 1.5972236398897621,
      "grad_norm": 6.03125,
      "learning_rate": 1.026238282984634e-06,
      "loss": 0.74679794,
      "memory(GiB)": 147.13,
      "step": 68460,
      "train_speed(iter/s)": 0.201217
    },
    {
      "acc": 0.79176388,
      "epoch": 1.597456947462051,
      "grad_norm": 9.5625,
      "learning_rate": 1.025092000816818e-06,
      "loss": 0.74075713,
      "memory(GiB)": 147.13,
      "step": 68470,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.78975224,
      "epoch": 1.59769025503434,
      "grad_norm": 5.5625,
      "learning_rate": 1.023946286065961e-06,
      "loss": 0.74442635,
      "memory(GiB)": 147.13,
      "step": 68480,
      "train_speed(iter/s)": 0.201247
    },
    {
      "acc": 0.77964411,
      "epoch": 1.5979235626066288,
      "grad_norm": 4.78125,
      "learning_rate": 1.0228011388956182e-06,
      "loss": 0.7978488,
      "memory(GiB)": 147.13,
      "step": 68490,
      "train_speed(iter/s)": 0.201262
    },
    {
      "acc": 0.78166089,
      "epoch": 1.5981568701789177,
      "grad_norm": 5.09375,
      "learning_rate": 1.0216565594692573e-06,
      "loss": 0.78926048,
      "memory(GiB)": 147.13,
      "step": 68500,
      "train_speed(iter/s)": 0.201277
    },
    {
      "epoch": 1.5981568701789177,
      "eval_acc": 0.7447232075096114,
      "eval_loss": 0.8043951988220215,
      "eval_runtime": 1270.2348,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "step": 68500
    },
    {
      "acc": 0.79227681,
      "epoch": 1.5983901777512066,
      "grad_norm": 6.90625,
      "learning_rate": 1.0205125479502658e-06,
      "loss": 0.74688244,
      "memory(GiB)": 147.13,
      "step": 68510,
      "train_speed(iter/s)": 0.200531
    },
    {
      "acc": 0.78717947,
      "epoch": 1.5986234853234955,
      "grad_norm": 5.5625,
      "learning_rate": 1.0193691045019533e-06,
      "loss": 0.7674963,
      "memory(GiB)": 147.13,
      "step": 68520,
      "train_speed(iter/s)": 0.200547
    },
    {
      "acc": 0.77596292,
      "epoch": 1.5988567928957844,
      "grad_norm": 4.5,
      "learning_rate": 1.0182262292875427e-06,
      "loss": 0.80983753,
      "memory(GiB)": 147.13,
      "step": 68530,
      "train_speed(iter/s)": 0.200563
    },
    {
      "acc": 0.77494078,
      "epoch": 1.5990901004680733,
      "grad_norm": 5.6875,
      "learning_rate": 1.0170839224701834e-06,
      "loss": 0.81508121,
      "memory(GiB)": 147.13,
      "step": 68540,
      "train_speed(iter/s)": 0.200578
    },
    {
      "acc": 0.78155966,
      "epoch": 1.5993234080403622,
      "grad_norm": 7.4375,
      "learning_rate": 1.015942184212937e-06,
      "loss": 0.81253033,
      "memory(GiB)": 147.13,
      "step": 68550,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.79896936,
      "epoch": 1.5995567156126511,
      "grad_norm": 7.65625,
      "learning_rate": 1.0148010146787845e-06,
      "loss": 0.70341702,
      "memory(GiB)": 148.85,
      "step": 68560,
      "train_speed(iter/s)": 0.200606
    },
    {
      "acc": 0.77949715,
      "epoch": 1.59979002318494,
      "grad_norm": 3.375,
      "learning_rate": 1.0136604140306312e-06,
      "loss": 0.7729764,
      "memory(GiB)": 138.1,
      "step": 68570,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.7776124,
      "epoch": 1.600023330757229,
      "grad_norm": 5.875,
      "learning_rate": 1.0125203824312957e-06,
      "loss": 0.78736258,
      "memory(GiB)": 138.1,
      "step": 68580,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.77759681,
      "epoch": 1.6002566383295178,
      "grad_norm": 5.875,
      "learning_rate": 1.0113809200435176e-06,
      "loss": 0.81777782,
      "memory(GiB)": 138.1,
      "step": 68590,
      "train_speed(iter/s)": 0.200651
    },
    {
      "acc": 0.78296614,
      "epoch": 1.6004899459018067,
      "grad_norm": 5.5,
      "learning_rate": 1.010242027029953e-06,
      "loss": 0.77656908,
      "memory(GiB)": 138.1,
      "step": 68600,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.77765617,
      "epoch": 1.6007232534740956,
      "grad_norm": 4.5,
      "learning_rate": 1.009103703553181e-06,
      "loss": 0.81307631,
      "memory(GiB)": 138.1,
      "step": 68610,
      "train_speed(iter/s)": 0.200682
    },
    {
      "acc": 0.78029776,
      "epoch": 1.6009565610463845,
      "grad_norm": 5.15625,
      "learning_rate": 1.0079659497756943e-06,
      "loss": 0.77876968,
      "memory(GiB)": 138.1,
      "step": 68620,
      "train_speed(iter/s)": 0.200697
    },
    {
      "acc": 0.77350893,
      "epoch": 1.6011898686186734,
      "grad_norm": 4.28125,
      "learning_rate": 1.0068287658599107e-06,
      "loss": 0.81584587,
      "memory(GiB)": 138.1,
      "step": 68630,
      "train_speed(iter/s)": 0.200711
    },
    {
      "acc": 0.77707477,
      "epoch": 1.6014231761909623,
      "grad_norm": 7.65625,
      "learning_rate": 1.0056921519681605e-06,
      "loss": 0.79458685,
      "memory(GiB)": 138.1,
      "step": 68640,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.80541954,
      "epoch": 1.601656483763251,
      "grad_norm": 6.09375,
      "learning_rate": 1.0045561082626936e-06,
      "loss": 0.68734684,
      "memory(GiB)": 138.1,
      "step": 68650,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.7977459,
      "epoch": 1.60188979133554,
      "grad_norm": 4.1875,
      "learning_rate": 1.0034206349056829e-06,
      "loss": 0.723417,
      "memory(GiB)": 138.1,
      "step": 68660,
      "train_speed(iter/s)": 0.200757
    },
    {
      "acc": 0.78180084,
      "epoch": 1.6021230989078288,
      "grad_norm": 6.5,
      "learning_rate": 1.002285732059215e-06,
      "loss": 0.79331684,
      "memory(GiB)": 138.1,
      "step": 68670,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.79122467,
      "epoch": 1.602356406480118,
      "grad_norm": 5.84375,
      "learning_rate": 1.001151399885298e-06,
      "loss": 0.74690819,
      "memory(GiB)": 138.1,
      "step": 68680,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.78143425,
      "epoch": 1.6025897140524066,
      "grad_norm": 6.125,
      "learning_rate": 1.000017638545857e-06,
      "loss": 0.77264366,
      "memory(GiB)": 138.1,
      "step": 68690,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.79719687,
      "epoch": 1.6028230216246957,
      "grad_norm": 5.90625,
      "learning_rate": 9.988844482027365e-07,
      "loss": 0.71532197,
      "memory(GiB)": 138.1,
      "step": 68700,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.77920446,
      "epoch": 1.6030563291969844,
      "grad_norm": 5.6875,
      "learning_rate": 9.97751829017699e-07,
      "loss": 0.77345262,
      "memory(GiB)": 138.1,
      "step": 68710,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.76342015,
      "epoch": 1.6032896367692735,
      "grad_norm": 6.75,
      "learning_rate": 9.966197811524231e-07,
      "loss": 0.85860195,
      "memory(GiB)": 138.1,
      "step": 68720,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.77934628,
      "epoch": 1.6035229443415622,
      "grad_norm": 6.21875,
      "learning_rate": 9.954883047685121e-07,
      "loss": 0.79521642,
      "memory(GiB)": 138.1,
      "step": 68730,
      "train_speed(iter/s)": 0.200862
    },
    {
      "acc": 0.781073,
      "epoch": 1.6037562519138513,
      "grad_norm": 5.1875,
      "learning_rate": 9.943574000274814e-07,
      "loss": 0.78763037,
      "memory(GiB)": 138.1,
      "step": 68740,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.77881775,
      "epoch": 1.60398955948614,
      "grad_norm": 5.59375,
      "learning_rate": 9.93227067090769e-07,
      "loss": 0.81242599,
      "memory(GiB)": 138.1,
      "step": 68750,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.77240114,
      "epoch": 1.604222867058429,
      "grad_norm": 5.75,
      "learning_rate": 9.920973061197291e-07,
      "loss": 0.80986681,
      "memory(GiB)": 138.1,
      "step": 68760,
      "train_speed(iter/s)": 0.200908
    },
    {
      "acc": 0.77040758,
      "epoch": 1.6044561746307178,
      "grad_norm": 5.53125,
      "learning_rate": 9.90968117275633e-07,
      "loss": 0.82993565,
      "memory(GiB)": 138.1,
      "step": 68770,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.76394424,
      "epoch": 1.604689482203007,
      "grad_norm": 5.625,
      "learning_rate": 9.898395007196747e-07,
      "loss": 0.84758568,
      "memory(GiB)": 138.1,
      "step": 68780,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.78739591,
      "epoch": 1.6049227897752956,
      "grad_norm": 6.0,
      "learning_rate": 9.887114566129613e-07,
      "loss": 0.75285759,
      "memory(GiB)": 138.1,
      "step": 68790,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.7844842,
      "epoch": 1.6051560973475847,
      "grad_norm": 8.5625,
      "learning_rate": 9.875839851165237e-07,
      "loss": 0.77806282,
      "memory(GiB)": 138.1,
      "step": 68800,
      "train_speed(iter/s)": 0.200966
    },
    {
      "acc": 0.77970557,
      "epoch": 1.6053894049198734,
      "grad_norm": 7.4375,
      "learning_rate": 9.86457086391307e-07,
      "loss": 0.8352437,
      "memory(GiB)": 138.1,
      "step": 68810,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.77098522,
      "epoch": 1.6056227124921625,
      "grad_norm": 7.4375,
      "learning_rate": 9.85330760598175e-07,
      "loss": 0.82055664,
      "memory(GiB)": 138.1,
      "step": 68820,
      "train_speed(iter/s)": 0.200995
    },
    {
      "acc": 0.77886381,
      "epoch": 1.6058560200644512,
      "grad_norm": 6.5,
      "learning_rate": 9.842050078979088e-07,
      "loss": 0.78484612,
      "memory(GiB)": 138.1,
      "step": 68830,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.79053154,
      "epoch": 1.60608932763674,
      "grad_norm": 3.765625,
      "learning_rate": 9.830798284512132e-07,
      "loss": 0.74266052,
      "memory(GiB)": 138.1,
      "step": 68840,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.79033446,
      "epoch": 1.606322635209029,
      "grad_norm": 5.15625,
      "learning_rate": 9.819552224187046e-07,
      "loss": 0.7502409,
      "memory(GiB)": 138.1,
      "step": 68850,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.78153863,
      "epoch": 1.6065559427813179,
      "grad_norm": 5.4375,
      "learning_rate": 9.808311899609197e-07,
      "loss": 0.76911235,
      "memory(GiB)": 138.1,
      "step": 68860,
      "train_speed(iter/s)": 0.201055
    },
    {
      "acc": 0.79210978,
      "epoch": 1.6067892503536068,
      "grad_norm": 5.03125,
      "learning_rate": 9.797077312383162e-07,
      "loss": 0.74023113,
      "memory(GiB)": 138.1,
      "step": 68870,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.79266319,
      "epoch": 1.6070225579258957,
      "grad_norm": 6.46875,
      "learning_rate": 9.785848464112647e-07,
      "loss": 0.76112881,
      "memory(GiB)": 138.1,
      "step": 68880,
      "train_speed(iter/s)": 0.201084
    },
    {
      "acc": 0.79358058,
      "epoch": 1.6072558654981846,
      "grad_norm": 3.65625,
      "learning_rate": 9.774625356400597e-07,
      "loss": 0.7407238,
      "memory(GiB)": 138.1,
      "step": 68890,
      "train_speed(iter/s)": 0.2011
    },
    {
      "acc": 0.78830051,
      "epoch": 1.6074891730704735,
      "grad_norm": 4.5625,
      "learning_rate": 9.763407990849089e-07,
      "loss": 0.75743866,
      "memory(GiB)": 138.1,
      "step": 68900,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.76708155,
      "epoch": 1.6077224806427624,
      "grad_norm": 5.34375,
      "learning_rate": 9.75219636905939e-07,
      "loss": 0.81055851,
      "memory(GiB)": 138.1,
      "step": 68910,
      "train_speed(iter/s)": 0.20113
    },
    {
      "acc": 0.78732371,
      "epoch": 1.6079557882150513,
      "grad_norm": 7.53125,
      "learning_rate": 9.74099049263198e-07,
      "loss": 0.76581669,
      "memory(GiB)": 138.1,
      "step": 68920,
      "train_speed(iter/s)": 0.201144
    },
    {
      "acc": 0.77004409,
      "epoch": 1.6081890957873402,
      "grad_norm": 4.6875,
      "learning_rate": 9.729790363166487e-07,
      "loss": 0.82656031,
      "memory(GiB)": 138.1,
      "step": 68930,
      "train_speed(iter/s)": 0.201159
    },
    {
      "acc": 0.79495058,
      "epoch": 1.608422403359629,
      "grad_norm": 4.1875,
      "learning_rate": 9.718595982261713e-07,
      "loss": 0.74895706,
      "memory(GiB)": 138.1,
      "step": 68940,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.78100176,
      "epoch": 1.608655710931918,
      "grad_norm": 6.09375,
      "learning_rate": 9.707407351515653e-07,
      "loss": 0.78046789,
      "memory(GiB)": 138.1,
      "step": 68950,
      "train_speed(iter/s)": 0.201188
    },
    {
      "acc": 0.78246336,
      "epoch": 1.6088890185042068,
      "grad_norm": 5.28125,
      "learning_rate": 9.696224472525494e-07,
      "loss": 0.78533907,
      "memory(GiB)": 138.1,
      "step": 68960,
      "train_speed(iter/s)": 0.201204
    },
    {
      "acc": 0.78976364,
      "epoch": 1.6091223260764957,
      "grad_norm": 7.0625,
      "learning_rate": 9.685047346887578e-07,
      "loss": 0.75304751,
      "memory(GiB)": 138.1,
      "step": 68970,
      "train_speed(iter/s)": 0.201219
    },
    {
      "acc": 0.78010139,
      "epoch": 1.6093556336487846,
      "grad_norm": 5.3125,
      "learning_rate": 9.673875976197455e-07,
      "loss": 0.79901323,
      "memory(GiB)": 138.1,
      "step": 68980,
      "train_speed(iter/s)": 0.201234
    },
    {
      "acc": 0.7747858,
      "epoch": 1.6095889412210735,
      "grad_norm": 5.03125,
      "learning_rate": 9.662710362049815e-07,
      "loss": 0.82582674,
      "memory(GiB)": 138.1,
      "step": 68990,
      "train_speed(iter/s)": 0.201248
    },
    {
      "acc": 0.78229184,
      "epoch": 1.6098222487933624,
      "grad_norm": 6.09375,
      "learning_rate": 9.651550506038543e-07,
      "loss": 0.78168783,
      "memory(GiB)": 138.1,
      "step": 69000,
      "train_speed(iter/s)": 0.201264
    },
    {
      "epoch": 1.6098222487933624,
      "eval_acc": 0.7446647548648214,
      "eval_loss": 0.8044345378875732,
      "eval_runtime": 1269.3193,
      "eval_samples_per_second": 28.355,
      "eval_steps_per_second": 14.178,
      "step": 69000
    },
    {
      "acc": 0.77631769,
      "epoch": 1.6100555563656513,
      "grad_norm": 4.40625,
      "learning_rate": 9.640396409756731e-07,
      "loss": 0.83473434,
      "memory(GiB)": 138.1,
      "step": 69010,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.79392805,
      "epoch": 1.6102888639379402,
      "grad_norm": 5.75,
      "learning_rate": 9.629248074796593e-07,
      "loss": 0.7308218,
      "memory(GiB)": 138.1,
      "step": 69020,
      "train_speed(iter/s)": 0.20054
    },
    {
      "acc": 0.78255825,
      "epoch": 1.6105221715102291,
      "grad_norm": 7.625,
      "learning_rate": 9.618105502749575e-07,
      "loss": 0.78719893,
      "memory(GiB)": 138.1,
      "step": 69030,
      "train_speed(iter/s)": 0.200555
    },
    {
      "acc": 0.80137396,
      "epoch": 1.6107554790825178,
      "grad_norm": 10.5625,
      "learning_rate": 9.606968695206264e-07,
      "loss": 0.69361744,
      "memory(GiB)": 138.1,
      "step": 69040,
      "train_speed(iter/s)": 0.200571
    },
    {
      "acc": 0.7979476,
      "epoch": 1.610988786654807,
      "grad_norm": 6.5625,
      "learning_rate": 9.59583765375644e-07,
      "loss": 0.72113686,
      "memory(GiB)": 138.1,
      "step": 69050,
      "train_speed(iter/s)": 0.200587
    },
    {
      "acc": 0.7878149,
      "epoch": 1.6112220942270956,
      "grad_norm": 8.5625,
      "learning_rate": 9.58471237998906e-07,
      "loss": 0.77528825,
      "memory(GiB)": 138.1,
      "step": 69060,
      "train_speed(iter/s)": 0.200602
    },
    {
      "acc": 0.77063704,
      "epoch": 1.6114554017993847,
      "grad_norm": 5.71875,
      "learning_rate": 9.57359287549222e-07,
      "loss": 0.84115477,
      "memory(GiB)": 138.1,
      "step": 69070,
      "train_speed(iter/s)": 0.200617
    },
    {
      "acc": 0.77148275,
      "epoch": 1.6116887093716734,
      "grad_norm": 5.3125,
      "learning_rate": 9.562479141853276e-07,
      "loss": 0.80791435,
      "memory(GiB)": 138.1,
      "step": 69080,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.77265015,
      "epoch": 1.6119220169439625,
      "grad_norm": 5.3125,
      "learning_rate": 9.551371180658675e-07,
      "loss": 0.79873414,
      "memory(GiB)": 138.1,
      "step": 69090,
      "train_speed(iter/s)": 0.200646
    },
    {
      "acc": 0.79981709,
      "epoch": 1.6121553245162512,
      "grad_norm": 4.8125,
      "learning_rate": 9.540268993494095e-07,
      "loss": 0.70313296,
      "memory(GiB)": 138.1,
      "step": 69100,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.77894669,
      "epoch": 1.6123886320885403,
      "grad_norm": 6.6875,
      "learning_rate": 9.529172581944352e-07,
      "loss": 0.81366968,
      "memory(GiB)": 138.1,
      "step": 69110,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.77654457,
      "epoch": 1.612621939660829,
      "grad_norm": 7.15625,
      "learning_rate": 9.518081947593477e-07,
      "loss": 0.81639051,
      "memory(GiB)": 138.1,
      "step": 69120,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.76891427,
      "epoch": 1.6128552472331181,
      "grad_norm": 4.8125,
      "learning_rate": 9.50699709202465e-07,
      "loss": 0.85439548,
      "memory(GiB)": 138.1,
      "step": 69130,
      "train_speed(iter/s)": 0.200706
    },
    {
      "acc": 0.78814936,
      "epoch": 1.6130885548054068,
      "grad_norm": 5.71875,
      "learning_rate": 9.495918016820204e-07,
      "loss": 0.7638484,
      "memory(GiB)": 138.1,
      "step": 69140,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.79379234,
      "epoch": 1.613321862377696,
      "grad_norm": 5.1875,
      "learning_rate": 9.484844723561726e-07,
      "loss": 0.7099072,
      "memory(GiB)": 138.1,
      "step": 69150,
      "train_speed(iter/s)": 0.200737
    },
    {
      "acc": 0.77234783,
      "epoch": 1.6135551699499846,
      "grad_norm": 6.90625,
      "learning_rate": 9.473777213829866e-07,
      "loss": 0.81089096,
      "memory(GiB)": 138.1,
      "step": 69160,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.77259874,
      "epoch": 1.6137884775222737,
      "grad_norm": 4.90625,
      "learning_rate": 9.462715489204549e-07,
      "loss": 0.81169949,
      "memory(GiB)": 138.1,
      "step": 69170,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.78828964,
      "epoch": 1.6140217850945624,
      "grad_norm": 7.78125,
      "learning_rate": 9.451659551264808e-07,
      "loss": 0.75277605,
      "memory(GiB)": 138.1,
      "step": 69180,
      "train_speed(iter/s)": 0.200783
    },
    {
      "acc": 0.79725246,
      "epoch": 1.6142550926668515,
      "grad_norm": 6.59375,
      "learning_rate": 9.440609401588901e-07,
      "loss": 0.72743855,
      "memory(GiB)": 138.1,
      "step": 69190,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.79967861,
      "epoch": 1.6144884002391402,
      "grad_norm": 4.21875,
      "learning_rate": 9.429565041754218e-07,
      "loss": 0.70557318,
      "memory(GiB)": 138.1,
      "step": 69200,
      "train_speed(iter/s)": 0.200813
    },
    {
      "acc": 0.77489271,
      "epoch": 1.6147217078114293,
      "grad_norm": 3.90625,
      "learning_rate": 9.418526473337325e-07,
      "loss": 0.81767092,
      "memory(GiB)": 138.1,
      "step": 69210,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.77260404,
      "epoch": 1.614955015383718,
      "grad_norm": 4.9375,
      "learning_rate": 9.407493697913999e-07,
      "loss": 0.82012691,
      "memory(GiB)": 138.1,
      "step": 69220,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.79442325,
      "epoch": 1.615188322956007,
      "grad_norm": 8.5,
      "learning_rate": 9.396466717059149e-07,
      "loss": 0.72898545,
      "memory(GiB)": 138.1,
      "step": 69230,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.78504844,
      "epoch": 1.6154216305282958,
      "grad_norm": 5.4375,
      "learning_rate": 9.385445532346887e-07,
      "loss": 0.76235442,
      "memory(GiB)": 138.1,
      "step": 69240,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.79455013,
      "epoch": 1.6156549381005847,
      "grad_norm": 4.40625,
      "learning_rate": 9.374430145350466e-07,
      "loss": 0.74329395,
      "memory(GiB)": 138.1,
      "step": 69250,
      "train_speed(iter/s)": 0.200888
    },
    {
      "acc": 0.7682219,
      "epoch": 1.6158882456728736,
      "grad_norm": 5.46875,
      "learning_rate": 9.363420557642355e-07,
      "loss": 0.84918575,
      "memory(GiB)": 138.1,
      "step": 69260,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.78665934,
      "epoch": 1.6161215532451625,
      "grad_norm": 5.25,
      "learning_rate": 9.352416770794154e-07,
      "loss": 0.76795254,
      "memory(GiB)": 138.1,
      "step": 69270,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.7946744,
      "epoch": 1.6163548608174514,
      "grad_norm": 4.71875,
      "learning_rate": 9.341418786376649e-07,
      "loss": 0.71400213,
      "memory(GiB)": 138.1,
      "step": 69280,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.77530904,
      "epoch": 1.6165881683897403,
      "grad_norm": 4.625,
      "learning_rate": 9.330426605959803e-07,
      "loss": 0.82718925,
      "memory(GiB)": 138.1,
      "step": 69290,
      "train_speed(iter/s)": 0.200947
    },
    {
      "acc": 0.7725615,
      "epoch": 1.6168214759620292,
      "grad_norm": 5.78125,
      "learning_rate": 9.319440231112725e-07,
      "loss": 0.81718159,
      "memory(GiB)": 138.1,
      "step": 69300,
      "train_speed(iter/s)": 0.200963
    },
    {
      "acc": 0.79436731,
      "epoch": 1.617054783534318,
      "grad_norm": 5.15625,
      "learning_rate": 9.308459663403757e-07,
      "loss": 0.75845556,
      "memory(GiB)": 138.1,
      "step": 69310,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.78687463,
      "epoch": 1.617288091106607,
      "grad_norm": 4.71875,
      "learning_rate": 9.297484904400333e-07,
      "loss": 0.7621346,
      "memory(GiB)": 138.1,
      "step": 69320,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.78693705,
      "epoch": 1.6175213986788959,
      "grad_norm": 4.90625,
      "learning_rate": 9.286515955669134e-07,
      "loss": 0.74903417,
      "memory(GiB)": 138.1,
      "step": 69330,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78143864,
      "epoch": 1.6177547062511848,
      "grad_norm": 5.53125,
      "learning_rate": 9.275552818775945e-07,
      "loss": 0.79808311,
      "memory(GiB)": 138.1,
      "step": 69340,
      "train_speed(iter/s)": 0.201023
    },
    {
      "acc": 0.7710865,
      "epoch": 1.6179880138234737,
      "grad_norm": 6.0625,
      "learning_rate": 9.264595495285755e-07,
      "loss": 0.83482437,
      "memory(GiB)": 138.1,
      "step": 69350,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.76026926,
      "epoch": 1.6182213213957626,
      "grad_norm": 8.1875,
      "learning_rate": 9.25364398676274e-07,
      "loss": 0.85732269,
      "memory(GiB)": 138.1,
      "step": 69360,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.78260975,
      "epoch": 1.6184546289680515,
      "grad_norm": 7.125,
      "learning_rate": 9.242698294770191e-07,
      "loss": 0.77640896,
      "memory(GiB)": 138.1,
      "step": 69370,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.79006805,
      "epoch": 1.6186879365403404,
      "grad_norm": 3.921875,
      "learning_rate": 9.231758420870645e-07,
      "loss": 0.75500803,
      "memory(GiB)": 138.1,
      "step": 69380,
      "train_speed(iter/s)": 0.201082
    },
    {
      "acc": 0.7738595,
      "epoch": 1.6189212441126293,
      "grad_norm": 4.4375,
      "learning_rate": 9.220824366625719e-07,
      "loss": 0.8050436,
      "memory(GiB)": 138.1,
      "step": 69390,
      "train_speed(iter/s)": 0.201097
    },
    {
      "acc": 0.77326818,
      "epoch": 1.6191545516849182,
      "grad_norm": 4.84375,
      "learning_rate": 9.20989613359628e-07,
      "loss": 0.83086958,
      "memory(GiB)": 138.1,
      "step": 69400,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.77323809,
      "epoch": 1.6193878592572069,
      "grad_norm": 5.90625,
      "learning_rate": 9.198973723342303e-07,
      "loss": 0.81775436,
      "memory(GiB)": 138.1,
      "step": 69410,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.76896219,
      "epoch": 1.619621166829496,
      "grad_norm": 4.84375,
      "learning_rate": 9.18805713742299e-07,
      "loss": 0.82727232,
      "memory(GiB)": 138.1,
      "step": 69420,
      "train_speed(iter/s)": 0.20114
    },
    {
      "acc": 0.77858291,
      "epoch": 1.6198544744017846,
      "grad_norm": 5.5,
      "learning_rate": 9.177146377396662e-07,
      "loss": 0.80608435,
      "memory(GiB)": 138.1,
      "step": 69430,
      "train_speed(iter/s)": 0.201156
    },
    {
      "acc": 0.78011804,
      "epoch": 1.6200877819740738,
      "grad_norm": 6.125,
      "learning_rate": 9.166241444820817e-07,
      "loss": 0.76758566,
      "memory(GiB)": 138.1,
      "step": 69440,
      "train_speed(iter/s)": 0.201171
    },
    {
      "acc": 0.78035154,
      "epoch": 1.6203210895463624,
      "grad_norm": 10.625,
      "learning_rate": 9.15534234125216e-07,
      "loss": 0.78598452,
      "memory(GiB)": 138.1,
      "step": 69450,
      "train_speed(iter/s)": 0.201186
    },
    {
      "acc": 0.80147018,
      "epoch": 1.6205543971186516,
      "grad_norm": 5.25,
      "learning_rate": 9.144449068246502e-07,
      "loss": 0.70722756,
      "memory(GiB)": 138.1,
      "step": 69460,
      "train_speed(iter/s)": 0.201202
    },
    {
      "acc": 0.75699992,
      "epoch": 1.6207877046909402,
      "grad_norm": 4.59375,
      "learning_rate": 9.133561627358884e-07,
      "loss": 0.86237135,
      "memory(GiB)": 138.1,
      "step": 69470,
      "train_speed(iter/s)": 0.201218
    },
    {
      "acc": 0.79911742,
      "epoch": 1.6210210122632294,
      "grad_norm": 8.875,
      "learning_rate": 9.122680020143476e-07,
      "loss": 0.70699673,
      "memory(GiB)": 138.1,
      "step": 69480,
      "train_speed(iter/s)": 0.201231
    },
    {
      "acc": 0.80146561,
      "epoch": 1.621254319835518,
      "grad_norm": 4.90625,
      "learning_rate": 9.111804248153605e-07,
      "loss": 0.70860691,
      "memory(GiB)": 138.1,
      "step": 69490,
      "train_speed(iter/s)": 0.201246
    },
    {
      "acc": 0.78607807,
      "epoch": 1.6214876274078072,
      "grad_norm": 5.40625,
      "learning_rate": 9.100934312941822e-07,
      "loss": 0.76731787,
      "memory(GiB)": 138.1,
      "step": 69500,
      "train_speed(iter/s)": 0.20126
    },
    {
      "epoch": 1.6214876274078072,
      "eval_acc": 0.744685733759253,
      "eval_loss": 0.8044191002845764,
      "eval_runtime": 1269.1743,
      "eval_samples_per_second": 28.358,
      "eval_steps_per_second": 14.179,
      "step": 69500
    },
    {
      "acc": 0.77421312,
      "epoch": 1.6217209349800958,
      "grad_norm": 6.46875,
      "learning_rate": 9.09007021605976e-07,
      "loss": 0.79150124,
      "memory(GiB)": 138.1,
      "step": 69510,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.76822271,
      "epoch": 1.621954242552385,
      "grad_norm": 10.875,
      "learning_rate": 9.079211959058304e-07,
      "loss": 0.8320261,
      "memory(GiB)": 138.1,
      "step": 69520,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.79691601,
      "epoch": 1.6221875501246736,
      "grad_norm": 8.0,
      "learning_rate": 9.068359543487442e-07,
      "loss": 0.73356771,
      "memory(GiB)": 138.1,
      "step": 69530,
      "train_speed(iter/s)": 0.200553
    },
    {
      "acc": 0.76938868,
      "epoch": 1.6224208576969628,
      "grad_norm": 4.40625,
      "learning_rate": 9.057512970896376e-07,
      "loss": 0.81400595,
      "memory(GiB)": 138.1,
      "step": 69540,
      "train_speed(iter/s)": 0.200568
    },
    {
      "acc": 0.77159843,
      "epoch": 1.6226541652692514,
      "grad_norm": 6.1875,
      "learning_rate": 9.046672242833427e-07,
      "loss": 0.82162743,
      "memory(GiB)": 138.1,
      "step": 69550,
      "train_speed(iter/s)": 0.200583
    },
    {
      "acc": 0.79073334,
      "epoch": 1.6228874728415406,
      "grad_norm": 5.71875,
      "learning_rate": 9.035837360846134e-07,
      "loss": 0.75625534,
      "memory(GiB)": 138.1,
      "step": 69560,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.77172422,
      "epoch": 1.6231207804138292,
      "grad_norm": 6.71875,
      "learning_rate": 9.02500832648115e-07,
      "loss": 0.82089863,
      "memory(GiB)": 138.1,
      "step": 69570,
      "train_speed(iter/s)": 0.200614
    },
    {
      "acc": 0.75787716,
      "epoch": 1.6233540879861184,
      "grad_norm": 5.8125,
      "learning_rate": 9.014185141284315e-07,
      "loss": 0.87999249,
      "memory(GiB)": 138.1,
      "step": 69580,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.78990684,
      "epoch": 1.623587395558407,
      "grad_norm": 3.5,
      "learning_rate": 9.003367806800661e-07,
      "loss": 0.76870031,
      "memory(GiB)": 138.1,
      "step": 69590,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.77759619,
      "epoch": 1.623820703130696,
      "grad_norm": 5.0,
      "learning_rate": 8.992556324574325e-07,
      "loss": 0.77327042,
      "memory(GiB)": 138.1,
      "step": 69600,
      "train_speed(iter/s)": 0.200658
    },
    {
      "acc": 0.77967892,
      "epoch": 1.6240540107029848,
      "grad_norm": 5.28125,
      "learning_rate": 8.981750696148689e-07,
      "loss": 0.78824787,
      "memory(GiB)": 138.1,
      "step": 69610,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.77583332,
      "epoch": 1.6242873182752737,
      "grad_norm": 5.21875,
      "learning_rate": 8.970950923066201e-07,
      "loss": 0.81284409,
      "memory(GiB)": 138.1,
      "step": 69620,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.79084225,
      "epoch": 1.6245206258475626,
      "grad_norm": 6.9375,
      "learning_rate": 8.960157006868564e-07,
      "loss": 0.76385317,
      "memory(GiB)": 138.1,
      "step": 69630,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77975607,
      "epoch": 1.6247539334198515,
      "grad_norm": 5.0,
      "learning_rate": 8.949368949096588e-07,
      "loss": 0.79590893,
      "memory(GiB)": 138.1,
      "step": 69640,
      "train_speed(iter/s)": 0.200719
    },
    {
      "acc": 0.76001158,
      "epoch": 1.6249872409921404,
      "grad_norm": 6.125,
      "learning_rate": 8.938586751290257e-07,
      "loss": 0.85949221,
      "memory(GiB)": 138.1,
      "step": 69650,
      "train_speed(iter/s)": 0.200734
    },
    {
      "acc": 0.78397055,
      "epoch": 1.6252205485644293,
      "grad_norm": 8.75,
      "learning_rate": 8.927810414988752e-07,
      "loss": 0.78344545,
      "memory(GiB)": 138.1,
      "step": 69660,
      "train_speed(iter/s)": 0.200749
    },
    {
      "acc": 0.7823205,
      "epoch": 1.6254538561367182,
      "grad_norm": 5.8125,
      "learning_rate": 8.917039941730365e-07,
      "loss": 0.77132821,
      "memory(GiB)": 138.1,
      "step": 69670,
      "train_speed(iter/s)": 0.200763
    },
    {
      "acc": 0.77989521,
      "epoch": 1.6256871637090071,
      "grad_norm": 5.21875,
      "learning_rate": 8.906275333052605e-07,
      "loss": 0.76878123,
      "memory(GiB)": 138.1,
      "step": 69680,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.78908534,
      "epoch": 1.625920471281296,
      "grad_norm": 4.59375,
      "learning_rate": 8.895516590492104e-07,
      "loss": 0.74415193,
      "memory(GiB)": 138.1,
      "step": 69690,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.80243616,
      "epoch": 1.626153778853585,
      "grad_norm": 4.84375,
      "learning_rate": 8.88476371558466e-07,
      "loss": 0.70059543,
      "memory(GiB)": 138.1,
      "step": 69700,
      "train_speed(iter/s)": 0.200808
    },
    {
      "acc": 0.76366444,
      "epoch": 1.6263870864258738,
      "grad_norm": 4.34375,
      "learning_rate": 8.874016709865257e-07,
      "loss": 0.83497429,
      "memory(GiB)": 138.1,
      "step": 69710,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.79546161,
      "epoch": 1.6266203939981627,
      "grad_norm": 5.59375,
      "learning_rate": 8.863275574868014e-07,
      "loss": 0.73149471,
      "memory(GiB)": 138.1,
      "step": 69720,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.77597876,
      "epoch": 1.6268537015704516,
      "grad_norm": 3.890625,
      "learning_rate": 8.852540312126256e-07,
      "loss": 0.79382687,
      "memory(GiB)": 138.1,
      "step": 69730,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.77241516,
      "epoch": 1.6270870091427405,
      "grad_norm": 6.09375,
      "learning_rate": 8.841810923172389e-07,
      "loss": 0.82993584,
      "memory(GiB)": 138.1,
      "step": 69740,
      "train_speed(iter/s)": 0.200867
    },
    {
      "acc": 0.77992134,
      "epoch": 1.6273203167150294,
      "grad_norm": 5.125,
      "learning_rate": 8.83108740953807e-07,
      "loss": 0.77837553,
      "memory(GiB)": 138.1,
      "step": 69750,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.79285231,
      "epoch": 1.6275536242873183,
      "grad_norm": 4.8125,
      "learning_rate": 8.82036977275405e-07,
      "loss": 0.74223328,
      "memory(GiB)": 138.1,
      "step": 69760,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.76947985,
      "epoch": 1.6277869318596072,
      "grad_norm": 4.75,
      "learning_rate": 8.809658014350297e-07,
      "loss": 0.82958355,
      "memory(GiB)": 138.1,
      "step": 69770,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.78920531,
      "epoch": 1.628020239431896,
      "grad_norm": 4.15625,
      "learning_rate": 8.798952135855893e-07,
      "loss": 0.76306615,
      "memory(GiB)": 138.1,
      "step": 69780,
      "train_speed(iter/s)": 0.200928
    },
    {
      "acc": 0.77562838,
      "epoch": 1.628253547004185,
      "grad_norm": 8.25,
      "learning_rate": 8.788252138799092e-07,
      "loss": 0.80673237,
      "memory(GiB)": 138.1,
      "step": 69790,
      "train_speed(iter/s)": 0.200942
    },
    {
      "acc": 0.77968626,
      "epoch": 1.6284868545764737,
      "grad_norm": 5.1875,
      "learning_rate": 8.777558024707339e-07,
      "loss": 0.78458438,
      "memory(GiB)": 138.1,
      "step": 69800,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.78913994,
      "epoch": 1.6287201621487628,
      "grad_norm": 4.375,
      "learning_rate": 8.766869795107191e-07,
      "loss": 0.75235834,
      "memory(GiB)": 138.1,
      "step": 69810,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.78043246,
      "epoch": 1.6289534697210515,
      "grad_norm": 6.1875,
      "learning_rate": 8.756187451524412e-07,
      "loss": 0.77849183,
      "memory(GiB)": 138.1,
      "step": 69820,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.77848892,
      "epoch": 1.6291867772933406,
      "grad_norm": 4.34375,
      "learning_rate": 8.745510995483892e-07,
      "loss": 0.80239801,
      "memory(GiB)": 138.1,
      "step": 69830,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.79218383,
      "epoch": 1.6294200848656293,
      "grad_norm": 5.28125,
      "learning_rate": 8.734840428509694e-07,
      "loss": 0.74597349,
      "memory(GiB)": 138.1,
      "step": 69840,
      "train_speed(iter/s)": 0.201017
    },
    {
      "acc": 0.77905416,
      "epoch": 1.6296533924379184,
      "grad_norm": 5.5,
      "learning_rate": 8.724175752125042e-07,
      "loss": 0.78995914,
      "memory(GiB)": 138.1,
      "step": 69850,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.78956504,
      "epoch": 1.629886700010207,
      "grad_norm": 5.84375,
      "learning_rate": 8.713516967852292e-07,
      "loss": 0.76276588,
      "memory(GiB)": 138.1,
      "step": 69860,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.77725039,
      "epoch": 1.6301200075824962,
      "grad_norm": 5.53125,
      "learning_rate": 8.702864077213014e-07,
      "loss": 0.80270042,
      "memory(GiB)": 138.1,
      "step": 69870,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.77587008,
      "epoch": 1.6303533151547849,
      "grad_norm": 4.65625,
      "learning_rate": 8.692217081727883e-07,
      "loss": 0.81218023,
      "memory(GiB)": 138.1,
      "step": 69880,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.76779122,
      "epoch": 1.630586622727074,
      "grad_norm": 5.625,
      "learning_rate": 8.681575982916773e-07,
      "loss": 0.85469055,
      "memory(GiB)": 138.1,
      "step": 69890,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.78068151,
      "epoch": 1.6308199302993627,
      "grad_norm": 5.65625,
      "learning_rate": 8.670940782298675e-07,
      "loss": 0.79066305,
      "memory(GiB)": 138.1,
      "step": 69900,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.76384864,
      "epoch": 1.6310532378716518,
      "grad_norm": 4.25,
      "learning_rate": 8.660311481391792e-07,
      "loss": 0.85888195,
      "memory(GiB)": 138.1,
      "step": 69910,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.79069233,
      "epoch": 1.6312865454439405,
      "grad_norm": 5.34375,
      "learning_rate": 8.649688081713431e-07,
      "loss": 0.74954491,
      "memory(GiB)": 138.1,
      "step": 69920,
      "train_speed(iter/s)": 0.201135
    },
    {
      "acc": 0.78507986,
      "epoch": 1.6315198530162296,
      "grad_norm": 5.0625,
      "learning_rate": 8.639070584780074e-07,
      "loss": 0.79639606,
      "memory(GiB)": 138.1,
      "step": 69930,
      "train_speed(iter/s)": 0.20115
    },
    {
      "acc": 0.79327669,
      "epoch": 1.6317531605885183,
      "grad_norm": 5.53125,
      "learning_rate": 8.628458992107386e-07,
      "loss": 0.74162989,
      "memory(GiB)": 138.1,
      "step": 69940,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.7880043,
      "epoch": 1.6319864681608074,
      "grad_norm": 8.3125,
      "learning_rate": 8.617853305210161e-07,
      "loss": 0.75817766,
      "memory(GiB)": 138.1,
      "step": 69950,
      "train_speed(iter/s)": 0.20118
    },
    {
      "acc": 0.77378907,
      "epoch": 1.632219775733096,
      "grad_norm": 7.6875,
      "learning_rate": 8.607253525602355e-07,
      "loss": 0.80225611,
      "memory(GiB)": 138.1,
      "step": 69960,
      "train_speed(iter/s)": 0.201195
    },
    {
      "acc": 0.77905831,
      "epoch": 1.6324530833053852,
      "grad_norm": 5.1875,
      "learning_rate": 8.596659654797068e-07,
      "loss": 0.79952292,
      "memory(GiB)": 138.1,
      "step": 69970,
      "train_speed(iter/s)": 0.201209
    },
    {
      "acc": 0.77380133,
      "epoch": 1.6326863908776739,
      "grad_norm": 4.75,
      "learning_rate": 8.586071694306602e-07,
      "loss": 0.80798283,
      "memory(GiB)": 138.1,
      "step": 69980,
      "train_speed(iter/s)": 0.201225
    },
    {
      "acc": 0.79780202,
      "epoch": 1.6329196984499628,
      "grad_norm": 4.625,
      "learning_rate": 8.575489645642371e-07,
      "loss": 0.71656194,
      "memory(GiB)": 138.1,
      "step": 69990,
      "train_speed(iter/s)": 0.201239
    },
    {
      "acc": 0.77291632,
      "epoch": 1.6331530060222517,
      "grad_norm": 11.8125,
      "learning_rate": 8.564913510314943e-07,
      "loss": 0.80805435,
      "memory(GiB)": 138.1,
      "step": 70000,
      "train_speed(iter/s)": 0.201254
    },
    {
      "epoch": 1.6331530060222517,
      "eval_acc": 0.744727211115419,
      "eval_loss": 0.8044537901878357,
      "eval_runtime": 1270.0228,
      "eval_samples_per_second": 28.339,
      "eval_steps_per_second": 14.17,
      "step": 70000
    },
    {
      "acc": 0.78201637,
      "epoch": 1.6333863135945406,
      "grad_norm": 5.75,
      "learning_rate": 8.554343289834094e-07,
      "loss": 0.80393801,
      "memory(GiB)": 138.1,
      "step": 70010,
      "train_speed(iter/s)": 0.200523
    },
    {
      "acc": 0.79293175,
      "epoch": 1.6336196211668295,
      "grad_norm": 5.0625,
      "learning_rate": 8.543778985708683e-07,
      "loss": 0.74324322,
      "memory(GiB)": 138.1,
      "step": 70020,
      "train_speed(iter/s)": 0.200537
    },
    {
      "acc": 0.78512459,
      "epoch": 1.6338529287391184,
      "grad_norm": 5.65625,
      "learning_rate": 8.533220599446789e-07,
      "loss": 0.78935528,
      "memory(GiB)": 138.1,
      "step": 70030,
      "train_speed(iter/s)": 0.200552
    },
    {
      "acc": 0.77615108,
      "epoch": 1.6340862363114073,
      "grad_norm": 6.84375,
      "learning_rate": 8.522668132555601e-07,
      "loss": 0.82629166,
      "memory(GiB)": 138.1,
      "step": 70040,
      "train_speed(iter/s)": 0.200566
    },
    {
      "acc": 0.77868662,
      "epoch": 1.6343195438836962,
      "grad_norm": 8.9375,
      "learning_rate": 8.512121586541499e-07,
      "loss": 0.79097457,
      "memory(GiB)": 138.1,
      "step": 70050,
      "train_speed(iter/s)": 0.20058
    },
    {
      "acc": 0.79038181,
      "epoch": 1.634552851455985,
      "grad_norm": 5.4375,
      "learning_rate": 8.501580962909989e-07,
      "loss": 0.73661346,
      "memory(GiB)": 138.1,
      "step": 70060,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.7649703,
      "epoch": 1.634786159028274,
      "grad_norm": 6.75,
      "learning_rate": 8.491046263165737e-07,
      "loss": 0.85525341,
      "memory(GiB)": 138.1,
      "step": 70070,
      "train_speed(iter/s)": 0.20061
    },
    {
      "acc": 0.7755065,
      "epoch": 1.6350194666005629,
      "grad_norm": 4.71875,
      "learning_rate": 8.480517488812578e-07,
      "loss": 0.79242172,
      "memory(GiB)": 138.1,
      "step": 70080,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.77949743,
      "epoch": 1.6352527741728518,
      "grad_norm": 5.40625,
      "learning_rate": 8.469994641353468e-07,
      "loss": 0.77577677,
      "memory(GiB)": 138.1,
      "step": 70090,
      "train_speed(iter/s)": 0.200639
    },
    {
      "acc": 0.77214479,
      "epoch": 1.6354860817451407,
      "grad_norm": 4.96875,
      "learning_rate": 8.459477722290577e-07,
      "loss": 0.82063198,
      "memory(GiB)": 138.1,
      "step": 70100,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.77816544,
      "epoch": 1.6357193893174296,
      "grad_norm": 14.4375,
      "learning_rate": 8.448966733125152e-07,
      "loss": 0.81221743,
      "memory(GiB)": 138.1,
      "step": 70110,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.7740953,
      "epoch": 1.6359526968897184,
      "grad_norm": 7.09375,
      "learning_rate": 8.438461675357679e-07,
      "loss": 0.81897221,
      "memory(GiB)": 138.1,
      "step": 70120,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.78639469,
      "epoch": 1.6361860044620073,
      "grad_norm": 5.15625,
      "learning_rate": 8.427962550487717e-07,
      "loss": 0.7703733,
      "memory(GiB)": 138.1,
      "step": 70130,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.77399602,
      "epoch": 1.6364193120342962,
      "grad_norm": 6.71875,
      "learning_rate": 8.417469360014019e-07,
      "loss": 0.81631184,
      "memory(GiB)": 138.1,
      "step": 70140,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.78081784,
      "epoch": 1.6366526196065851,
      "grad_norm": 5.8125,
      "learning_rate": 8.406982105434502e-07,
      "loss": 0.79938803,
      "memory(GiB)": 138.1,
      "step": 70150,
      "train_speed(iter/s)": 0.200725
    },
    {
      "acc": 0.78800888,
      "epoch": 1.636885927178874,
      "grad_norm": 5.5,
      "learning_rate": 8.396500788246192e-07,
      "loss": 0.75535965,
      "memory(GiB)": 138.1,
      "step": 70160,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.80056391,
      "epoch": 1.637119234751163,
      "grad_norm": 6.46875,
      "learning_rate": 8.38602540994532e-07,
      "loss": 0.69680567,
      "memory(GiB)": 138.1,
      "step": 70170,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.7816555,
      "epoch": 1.6373525423234518,
      "grad_norm": 4.625,
      "learning_rate": 8.375555972027233e-07,
      "loss": 0.76161251,
      "memory(GiB)": 138.1,
      "step": 70180,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.78333101,
      "epoch": 1.6375858498957405,
      "grad_norm": 4.40625,
      "learning_rate": 8.365092475986442e-07,
      "loss": 0.77230682,
      "memory(GiB)": 138.1,
      "step": 70190,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.75599384,
      "epoch": 1.6378191574680296,
      "grad_norm": 5.15625,
      "learning_rate": 8.35463492331659e-07,
      "loss": 0.91006165,
      "memory(GiB)": 138.1,
      "step": 70200,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.7760139,
      "epoch": 1.6380524650403183,
      "grad_norm": 5.5,
      "learning_rate": 8.344183315510518e-07,
      "loss": 0.807833,
      "memory(GiB)": 138.1,
      "step": 70210,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.77669868,
      "epoch": 1.6382857726126074,
      "grad_norm": 4.75,
      "learning_rate": 8.333737654060176e-07,
      "loss": 0.78656607,
      "memory(GiB)": 138.1,
      "step": 70220,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.77367878,
      "epoch": 1.6385190801848961,
      "grad_norm": 5.03125,
      "learning_rate": 8.323297940456665e-07,
      "loss": 0.82135525,
      "memory(GiB)": 138.1,
      "step": 70230,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.77797899,
      "epoch": 1.6387523877571852,
      "grad_norm": 4.78125,
      "learning_rate": 8.312864176190282e-07,
      "loss": 0.80228634,
      "memory(GiB)": 138.1,
      "step": 70240,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.7765986,
      "epoch": 1.638985695329474,
      "grad_norm": 6.8125,
      "learning_rate": 8.302436362750416e-07,
      "loss": 0.80377598,
      "memory(GiB)": 138.1,
      "step": 70250,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.79614129,
      "epoch": 1.639219002901763,
      "grad_norm": 4.375,
      "learning_rate": 8.292014501625656e-07,
      "loss": 0.70680809,
      "memory(GiB)": 138.1,
      "step": 70260,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.79076648,
      "epoch": 1.6394523104740517,
      "grad_norm": 4.375,
      "learning_rate": 8.281598594303708e-07,
      "loss": 0.74855566,
      "memory(GiB)": 138.1,
      "step": 70270,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.7758749,
      "epoch": 1.6396856180463408,
      "grad_norm": 3.984375,
      "learning_rate": 8.271188642271432e-07,
      "loss": 0.81647902,
      "memory(GiB)": 138.1,
      "step": 70280,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.79907188,
      "epoch": 1.6399189256186295,
      "grad_norm": 5.03125,
      "learning_rate": 8.260784647014864e-07,
      "loss": 0.7401226,
      "memory(GiB)": 138.1,
      "step": 70290,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.77903986,
      "epoch": 1.6401522331909186,
      "grad_norm": 5.40625,
      "learning_rate": 8.250386610019167e-07,
      "loss": 0.80523014,
      "memory(GiB)": 138.1,
      "step": 70300,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78396397,
      "epoch": 1.6403855407632073,
      "grad_norm": 5.46875,
      "learning_rate": 8.239994532768647e-07,
      "loss": 0.78131781,
      "memory(GiB)": 138.1,
      "step": 70310,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.78332367,
      "epoch": 1.6406188483354964,
      "grad_norm": 5.0,
      "learning_rate": 8.22960841674677e-07,
      "loss": 0.77818413,
      "memory(GiB)": 138.1,
      "step": 70320,
      "train_speed(iter/s)": 0.200979
    },
    {
      "acc": 0.77349386,
      "epoch": 1.640852155907785,
      "grad_norm": 4.96875,
      "learning_rate": 8.219228263436168e-07,
      "loss": 0.82582359,
      "memory(GiB)": 138.1,
      "step": 70330,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.78453584,
      "epoch": 1.6410854634800742,
      "grad_norm": 4.875,
      "learning_rate": 8.208854074318579e-07,
      "loss": 0.76499305,
      "memory(GiB)": 138.1,
      "step": 70340,
      "train_speed(iter/s)": 0.201008
    },
    {
      "acc": 0.78075652,
      "epoch": 1.641318771052363,
      "grad_norm": 9.5,
      "learning_rate": 8.198485850874943e-07,
      "loss": 0.78641806,
      "memory(GiB)": 138.1,
      "step": 70350,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.77025485,
      "epoch": 1.641552078624652,
      "grad_norm": 5.6875,
      "learning_rate": 8.188123594585312e-07,
      "loss": 0.81662521,
      "memory(GiB)": 138.1,
      "step": 70360,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.78271184,
      "epoch": 1.6417853861969407,
      "grad_norm": 4.0,
      "learning_rate": 8.177767306928875e-07,
      "loss": 0.76894827,
      "memory(GiB)": 138.1,
      "step": 70370,
      "train_speed(iter/s)": 0.201051
    },
    {
      "acc": 0.77495737,
      "epoch": 1.6420186937692296,
      "grad_norm": 6.25,
      "learning_rate": 8.16741698938402e-07,
      "loss": 0.80135536,
      "memory(GiB)": 138.1,
      "step": 70380,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.77556715,
      "epoch": 1.6422520013415185,
      "grad_norm": 3.484375,
      "learning_rate": 8.157072643428227e-07,
      "loss": 0.80013609,
      "memory(GiB)": 138.1,
      "step": 70390,
      "train_speed(iter/s)": 0.201081
    },
    {
      "acc": 0.78865757,
      "epoch": 1.6424853089138074,
      "grad_norm": 6.46875,
      "learning_rate": 8.14673427053817e-07,
      "loss": 0.7638164,
      "memory(GiB)": 138.1,
      "step": 70400,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.78071489,
      "epoch": 1.6427186164860963,
      "grad_norm": 4.875,
      "learning_rate": 8.136401872189631e-07,
      "loss": 0.7829968,
      "memory(GiB)": 138.1,
      "step": 70410,
      "train_speed(iter/s)": 0.201111
    },
    {
      "acc": 0.79857535,
      "epoch": 1.6429519240583852,
      "grad_norm": 5.96875,
      "learning_rate": 8.126075449857574e-07,
      "loss": 0.71098719,
      "memory(GiB)": 138.1,
      "step": 70420,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.76003633,
      "epoch": 1.643185231630674,
      "grad_norm": 5.09375,
      "learning_rate": 8.115755005016074e-07,
      "loss": 0.87234459,
      "memory(GiB)": 138.1,
      "step": 70430,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.76972141,
      "epoch": 1.643418539202963,
      "grad_norm": 10.25,
      "learning_rate": 8.105440539138371e-07,
      "loss": 0.82762814,
      "memory(GiB)": 138.1,
      "step": 70440,
      "train_speed(iter/s)": 0.201155
    },
    {
      "acc": 0.78295097,
      "epoch": 1.643651846775252,
      "grad_norm": 4.65625,
      "learning_rate": 8.095132053696869e-07,
      "loss": 0.79033747,
      "memory(GiB)": 138.1,
      "step": 70450,
      "train_speed(iter/s)": 0.201169
    },
    {
      "acc": 0.77744741,
      "epoch": 1.6438851543475408,
      "grad_norm": 4.90625,
      "learning_rate": 8.084829550163087e-07,
      "loss": 0.8104229,
      "memory(GiB)": 138.1,
      "step": 70460,
      "train_speed(iter/s)": 0.201184
    },
    {
      "acc": 0.79442644,
      "epoch": 1.6441184619198297,
      "grad_norm": 5.09375,
      "learning_rate": 8.074533030007714e-07,
      "loss": 0.7583622,
      "memory(GiB)": 138.1,
      "step": 70470,
      "train_speed(iter/s)": 0.2012
    },
    {
      "acc": 0.77418184,
      "epoch": 1.6443517694921186,
      "grad_norm": 4.25,
      "learning_rate": 8.064242494700581e-07,
      "loss": 0.79210043,
      "memory(GiB)": 138.1,
      "step": 70480,
      "train_speed(iter/s)": 0.201215
    },
    {
      "acc": 0.7854991,
      "epoch": 1.6445850770644075,
      "grad_norm": 9.3125,
      "learning_rate": 8.053957945710633e-07,
      "loss": 0.7795774,
      "memory(GiB)": 138.1,
      "step": 70490,
      "train_speed(iter/s)": 0.201229
    },
    {
      "acc": 0.76263819,
      "epoch": 1.6448183846366964,
      "grad_norm": 5.75,
      "learning_rate": 8.043679384506014e-07,
      "loss": 0.83604841,
      "memory(GiB)": 138.1,
      "step": 70500,
      "train_speed(iter/s)": 0.201243
    },
    {
      "epoch": 1.6448183846366964,
      "eval_acc": 0.7446782069803348,
      "eval_loss": 0.8044106364250183,
      "eval_runtime": 1269.9954,
      "eval_samples_per_second": 28.339,
      "eval_steps_per_second": 14.17,
      "step": 70500
    },
    {
      "acc": 0.77483778,
      "epoch": 1.6450516922089853,
      "grad_norm": 5.84375,
      "learning_rate": 8.033406812553962e-07,
      "loss": 0.81247177,
      "memory(GiB)": 138.1,
      "step": 70510,
      "train_speed(iter/s)": 0.200516
    },
    {
      "acc": 0.78622189,
      "epoch": 1.6452849997812742,
      "grad_norm": 7.625,
      "learning_rate": 8.023140231320919e-07,
      "loss": 0.77647934,
      "memory(GiB)": 138.1,
      "step": 70520,
      "train_speed(iter/s)": 0.200532
    },
    {
      "acc": 0.76669769,
      "epoch": 1.645518307353563,
      "grad_norm": 4.9375,
      "learning_rate": 8.012879642272392e-07,
      "loss": 0.84020367,
      "memory(GiB)": 138.1,
      "step": 70530,
      "train_speed(iter/s)": 0.200548
    },
    {
      "acc": 0.77495613,
      "epoch": 1.645751614925852,
      "grad_norm": 4.15625,
      "learning_rate": 8.002625046873114e-07,
      "loss": 0.81265011,
      "memory(GiB)": 138.1,
      "step": 70540,
      "train_speed(iter/s)": 0.200562
    },
    {
      "acc": 0.77314444,
      "epoch": 1.6459849224981409,
      "grad_norm": 4.90625,
      "learning_rate": 7.992376446586891e-07,
      "loss": 0.81458588,
      "memory(GiB)": 138.1,
      "step": 70550,
      "train_speed(iter/s)": 0.200577
    },
    {
      "acc": 0.78395252,
      "epoch": 1.6462182300704298,
      "grad_norm": 6.5625,
      "learning_rate": 7.982133842876744e-07,
      "loss": 0.7673286,
      "memory(GiB)": 138.1,
      "step": 70560,
      "train_speed(iter/s)": 0.200592
    },
    {
      "acc": 0.79952912,
      "epoch": 1.6464515376427187,
      "grad_norm": 6.5625,
      "learning_rate": 7.971897237204785e-07,
      "loss": 0.71654468,
      "memory(GiB)": 138.1,
      "step": 70570,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.77154684,
      "epoch": 1.6466848452150074,
      "grad_norm": 5.4375,
      "learning_rate": 7.961666631032273e-07,
      "loss": 0.8394556,
      "memory(GiB)": 138.1,
      "step": 70580,
      "train_speed(iter/s)": 0.200622
    },
    {
      "acc": 0.78006334,
      "epoch": 1.6469181527872965,
      "grad_norm": 5.8125,
      "learning_rate": 7.951442025819651e-07,
      "loss": 0.79687848,
      "memory(GiB)": 138.1,
      "step": 70590,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.77312818,
      "epoch": 1.6471514603595852,
      "grad_norm": 12.5625,
      "learning_rate": 7.941223423026445e-07,
      "loss": 0.82311382,
      "memory(GiB)": 138.1,
      "step": 70600,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78867445,
      "epoch": 1.6473847679318743,
      "grad_norm": 5.8125,
      "learning_rate": 7.931010824111396e-07,
      "loss": 0.76376724,
      "memory(GiB)": 138.1,
      "step": 70610,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.78184714,
      "epoch": 1.647618075504163,
      "grad_norm": 5.84375,
      "learning_rate": 7.920804230532331e-07,
      "loss": 0.77635255,
      "memory(GiB)": 138.1,
      "step": 70620,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.79582644,
      "epoch": 1.647851383076452,
      "grad_norm": 6.21875,
      "learning_rate": 7.910603643746223e-07,
      "loss": 0.72300158,
      "memory(GiB)": 138.1,
      "step": 70630,
      "train_speed(iter/s)": 0.200693
    },
    {
      "acc": 0.76628838,
      "epoch": 1.6480846906487407,
      "grad_norm": 6.71875,
      "learning_rate": 7.90040906520923e-07,
      "loss": 0.85222702,
      "memory(GiB)": 138.1,
      "step": 70640,
      "train_speed(iter/s)": 0.200709
    },
    {
      "acc": 0.7875648,
      "epoch": 1.6483179982210299,
      "grad_norm": 6.40625,
      "learning_rate": 7.890220496376616e-07,
      "loss": 0.76165543,
      "memory(GiB)": 138.1,
      "step": 70650,
      "train_speed(iter/s)": 0.200724
    },
    {
      "acc": 0.78504138,
      "epoch": 1.6485513057933185,
      "grad_norm": 4.65625,
      "learning_rate": 7.880037938702789e-07,
      "loss": 0.74155846,
      "memory(GiB)": 138.1,
      "step": 70660,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.78913331,
      "epoch": 1.6487846133656077,
      "grad_norm": 5.4375,
      "learning_rate": 7.869861393641304e-07,
      "loss": 0.75573359,
      "memory(GiB)": 138.1,
      "step": 70670,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.77992077,
      "epoch": 1.6490179209378963,
      "grad_norm": 6.5,
      "learning_rate": 7.859690862644876e-07,
      "loss": 0.78720312,
      "memory(GiB)": 138.1,
      "step": 70680,
      "train_speed(iter/s)": 0.200767
    },
    {
      "acc": 0.77604442,
      "epoch": 1.6492512285101855,
      "grad_norm": 6.6875,
      "learning_rate": 7.849526347165321e-07,
      "loss": 0.83598385,
      "memory(GiB)": 138.1,
      "step": 70690,
      "train_speed(iter/s)": 0.200782
    },
    {
      "acc": 0.78691902,
      "epoch": 1.6494845360824741,
      "grad_norm": 8.875,
      "learning_rate": 7.83936784865365e-07,
      "loss": 0.77053566,
      "memory(GiB)": 138.1,
      "step": 70700,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.79002995,
      "epoch": 1.6497178436547633,
      "grad_norm": 3.6875,
      "learning_rate": 7.829215368559967e-07,
      "loss": 0.74581184,
      "memory(GiB)": 138.1,
      "step": 70710,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.78356628,
      "epoch": 1.649951151227052,
      "grad_norm": 5.28125,
      "learning_rate": 7.819068908333532e-07,
      "loss": 0.76630058,
      "memory(GiB)": 138.1,
      "step": 70720,
      "train_speed(iter/s)": 0.200827
    },
    {
      "acc": 0.79238472,
      "epoch": 1.650184458799341,
      "grad_norm": 5.21875,
      "learning_rate": 7.808928469422766e-07,
      "loss": 0.7440876,
      "memory(GiB)": 138.1,
      "step": 70730,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.7771996,
      "epoch": 1.6504177663716297,
      "grad_norm": 5.875,
      "learning_rate": 7.798794053275193e-07,
      "loss": 0.80373068,
      "memory(GiB)": 138.1,
      "step": 70740,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.75577822,
      "epoch": 1.6506510739439189,
      "grad_norm": 5.5625,
      "learning_rate": 7.78866566133753e-07,
      "loss": 0.90629873,
      "memory(GiB)": 138.1,
      "step": 70750,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.79528246,
      "epoch": 1.6508843815162075,
      "grad_norm": 5.84375,
      "learning_rate": 7.778543295055563e-07,
      "loss": 0.73617687,
      "memory(GiB)": 138.1,
      "step": 70760,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.77932301,
      "epoch": 1.6511176890884964,
      "grad_norm": 5.53125,
      "learning_rate": 7.768426955874287e-07,
      "loss": 0.78966246,
      "memory(GiB)": 138.1,
      "step": 70770,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.78333249,
      "epoch": 1.6513509966607853,
      "grad_norm": 5.46875,
      "learning_rate": 7.758316645237791e-07,
      "loss": 0.76522617,
      "memory(GiB)": 138.1,
      "step": 70780,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.78264174,
      "epoch": 1.6515843042330742,
      "grad_norm": 4.59375,
      "learning_rate": 7.748212364589314e-07,
      "loss": 0.77815781,
      "memory(GiB)": 138.1,
      "step": 70790,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.7721468,
      "epoch": 1.6518176118053631,
      "grad_norm": 6.09375,
      "learning_rate": 7.738114115371254e-07,
      "loss": 0.82284946,
      "memory(GiB)": 138.1,
      "step": 70800,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.77733574,
      "epoch": 1.652050919377652,
      "grad_norm": 5.71875,
      "learning_rate": 7.728021899025124e-07,
      "loss": 0.80925665,
      "memory(GiB)": 138.1,
      "step": 70810,
      "train_speed(iter/s)": 0.20096
    },
    {
      "acc": 0.79648423,
      "epoch": 1.652284226949941,
      "grad_norm": 4.53125,
      "learning_rate": 7.717935716991592e-07,
      "loss": 0.72788992,
      "memory(GiB)": 138.1,
      "step": 70820,
      "train_speed(iter/s)": 0.200975
    },
    {
      "acc": 0.7860816,
      "epoch": 1.6525175345222298,
      "grad_norm": 7.125,
      "learning_rate": 7.707855570710443e-07,
      "loss": 0.75475664,
      "memory(GiB)": 138.1,
      "step": 70830,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.77979727,
      "epoch": 1.6527508420945187,
      "grad_norm": 5.03125,
      "learning_rate": 7.69778146162064e-07,
      "loss": 0.80329094,
      "memory(GiB)": 138.1,
      "step": 70840,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.77364244,
      "epoch": 1.6529841496668076,
      "grad_norm": 5.875,
      "learning_rate": 7.68771339116024e-07,
      "loss": 0.82023067,
      "memory(GiB)": 138.1,
      "step": 70850,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.77010069,
      "epoch": 1.6532174572390965,
      "grad_norm": 5.65625,
      "learning_rate": 7.677651360766453e-07,
      "loss": 0.85668049,
      "memory(GiB)": 138.1,
      "step": 70860,
      "train_speed(iter/s)": 0.201034
    },
    {
      "acc": 0.7844202,
      "epoch": 1.6534507648113854,
      "grad_norm": 7.375,
      "learning_rate": 7.667595371875663e-07,
      "loss": 0.7779212,
      "memory(GiB)": 138.1,
      "step": 70870,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.77224879,
      "epoch": 1.6536840723836743,
      "grad_norm": 5.03125,
      "learning_rate": 7.657545425923313e-07,
      "loss": 0.82133636,
      "memory(GiB)": 138.1,
      "step": 70880,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.76610975,
      "epoch": 1.6539173799559632,
      "grad_norm": 5.46875,
      "learning_rate": 7.647501524344064e-07,
      "loss": 0.8340517,
      "memory(GiB)": 138.1,
      "step": 70890,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78516321,
      "epoch": 1.6541506875282521,
      "grad_norm": 4.75,
      "learning_rate": 7.637463668571659e-07,
      "loss": 0.76711388,
      "memory(GiB)": 138.1,
      "step": 70900,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.77286749,
      "epoch": 1.654383995100541,
      "grad_norm": 5.375,
      "learning_rate": 7.627431860039019e-07,
      "loss": 0.81947851,
      "memory(GiB)": 138.1,
      "step": 70910,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.76813011,
      "epoch": 1.65461730267283,
      "grad_norm": 5.71875,
      "learning_rate": 7.617406100178171e-07,
      "loss": 0.83200006,
      "memory(GiB)": 138.1,
      "step": 70920,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.78832273,
      "epoch": 1.6548506102451188,
      "grad_norm": 5.03125,
      "learning_rate": 7.607386390420279e-07,
      "loss": 0.74536948,
      "memory(GiB)": 138.1,
      "step": 70930,
      "train_speed(iter/s)": 0.201134
    },
    {
      "acc": 0.7811326,
      "epoch": 1.6550839178174077,
      "grad_norm": 5.4375,
      "learning_rate": 7.597372732195674e-07,
      "loss": 0.78975849,
      "memory(GiB)": 138.1,
      "step": 70940,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.77318063,
      "epoch": 1.6553172253896964,
      "grad_norm": 6.75,
      "learning_rate": 7.587365126933782e-07,
      "loss": 0.8096487,
      "memory(GiB)": 138.1,
      "step": 70950,
      "train_speed(iter/s)": 0.201163
    },
    {
      "acc": 0.78359327,
      "epoch": 1.6555505329619855,
      "grad_norm": 5.28125,
      "learning_rate": 7.577363576063212e-07,
      "loss": 0.77436295,
      "memory(GiB)": 138.1,
      "step": 70960,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.78843203,
      "epoch": 1.6557838405342742,
      "grad_norm": 5.46875,
      "learning_rate": 7.567368081011656e-07,
      "loss": 0.74306149,
      "memory(GiB)": 138.1,
      "step": 70970,
      "train_speed(iter/s)": 0.201191
    },
    {
      "acc": 0.78639054,
      "epoch": 1.6560171481065633,
      "grad_norm": 7.375,
      "learning_rate": 7.557378643206003e-07,
      "loss": 0.78001995,
      "memory(GiB)": 138.1,
      "step": 70980,
      "train_speed(iter/s)": 0.201206
    },
    {
      "acc": 0.79448624,
      "epoch": 1.656250455678852,
      "grad_norm": 7.6875,
      "learning_rate": 7.547395264072193e-07,
      "loss": 0.74755363,
      "memory(GiB)": 138.1,
      "step": 70990,
      "train_speed(iter/s)": 0.20122
    },
    {
      "acc": 0.78536434,
      "epoch": 1.656483763251141,
      "grad_norm": 6.78125,
      "learning_rate": 7.537417945035391e-07,
      "loss": 0.76806111,
      "memory(GiB)": 138.1,
      "step": 71000,
      "train_speed(iter/s)": 0.201235
    },
    {
      "epoch": 1.656483763251141,
      "eval_acc": 0.7447166415960871,
      "eval_loss": 0.8044018745422363,
      "eval_runtime": 1270.2511,
      "eval_samples_per_second": 28.334,
      "eval_steps_per_second": 14.167,
      "step": 71000
    },
    {
      "acc": 0.78267994,
      "epoch": 1.6567170708234298,
      "grad_norm": 5.5625,
      "learning_rate": 7.52744668751984e-07,
      "loss": 0.78940582,
      "memory(GiB)": 138.1,
      "step": 71010,
      "train_speed(iter/s)": 0.200514
    },
    {
      "acc": 0.78548698,
      "epoch": 1.656950378395719,
      "grad_norm": 9.375,
      "learning_rate": 7.517481492948925e-07,
      "loss": 0.77665896,
      "memory(GiB)": 138.1,
      "step": 71020,
      "train_speed(iter/s)": 0.200527
    },
    {
      "acc": 0.76159725,
      "epoch": 1.6571836859680076,
      "grad_norm": 5.84375,
      "learning_rate": 7.507522362745195e-07,
      "loss": 0.8417695,
      "memory(GiB)": 138.1,
      "step": 71030,
      "train_speed(iter/s)": 0.200541
    },
    {
      "acc": 0.78062215,
      "epoch": 1.6574169935402967,
      "grad_norm": 5.59375,
      "learning_rate": 7.497569298330293e-07,
      "loss": 0.7981185,
      "memory(GiB)": 138.1,
      "step": 71040,
      "train_speed(iter/s)": 0.200557
    },
    {
      "acc": 0.76416349,
      "epoch": 1.6576503011125854,
      "grad_norm": 6.3125,
      "learning_rate": 7.487622301125041e-07,
      "loss": 0.87657032,
      "memory(GiB)": 138.1,
      "step": 71050,
      "train_speed(iter/s)": 0.200572
    },
    {
      "acc": 0.78121519,
      "epoch": 1.6578836086848745,
      "grad_norm": 4.65625,
      "learning_rate": 7.477681372549355e-07,
      "loss": 0.77680473,
      "memory(GiB)": 138.1,
      "step": 71060,
      "train_speed(iter/s)": 0.200585
    },
    {
      "acc": 0.77829385,
      "epoch": 1.6581169162571632,
      "grad_norm": 5.75,
      "learning_rate": 7.467746514022284e-07,
      "loss": 0.80236874,
      "memory(GiB)": 138.1,
      "step": 71070,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.76091347,
      "epoch": 1.6583502238294523,
      "grad_norm": 5.5625,
      "learning_rate": 7.457817726962058e-07,
      "loss": 0.86437483,
      "memory(GiB)": 138.1,
      "step": 71080,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.79476452,
      "epoch": 1.658583531401741,
      "grad_norm": 5.78125,
      "learning_rate": 7.447895012785983e-07,
      "loss": 0.73813944,
      "memory(GiB)": 138.1,
      "step": 71090,
      "train_speed(iter/s)": 0.200627
    },
    {
      "acc": 0.77920585,
      "epoch": 1.65881683897403,
      "grad_norm": 5.09375,
      "learning_rate": 7.437978372910554e-07,
      "loss": 0.80791998,
      "memory(GiB)": 138.1,
      "step": 71100,
      "train_speed(iter/s)": 0.200641
    },
    {
      "acc": 0.7747447,
      "epoch": 1.6590501465463188,
      "grad_norm": 4.9375,
      "learning_rate": 7.428067808751327e-07,
      "loss": 0.80314159,
      "memory(GiB)": 138.1,
      "step": 71110,
      "train_speed(iter/s)": 0.200656
    },
    {
      "acc": 0.78580241,
      "epoch": 1.659283454118608,
      "grad_norm": 5.71875,
      "learning_rate": 7.41816332172306e-07,
      "loss": 0.77066259,
      "memory(GiB)": 138.1,
      "step": 71120,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.78698373,
      "epoch": 1.6595167616908966,
      "grad_norm": 6.28125,
      "learning_rate": 7.408264913239598e-07,
      "loss": 0.74832349,
      "memory(GiB)": 138.1,
      "step": 71130,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.77769785,
      "epoch": 1.6597500692631855,
      "grad_norm": 5.96875,
      "learning_rate": 7.398372584713964e-07,
      "loss": 0.78736668,
      "memory(GiB)": 138.1,
      "step": 71140,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.78976698,
      "epoch": 1.6599833768354744,
      "grad_norm": 6.0625,
      "learning_rate": 7.388486337558265e-07,
      "loss": 0.75616255,
      "memory(GiB)": 138.1,
      "step": 71150,
      "train_speed(iter/s)": 0.200714
    },
    {
      "acc": 0.80499172,
      "epoch": 1.6602166844077633,
      "grad_norm": 5.5,
      "learning_rate": 7.378606173183749e-07,
      "loss": 0.70488954,
      "memory(GiB)": 138.1,
      "step": 71160,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.79924831,
      "epoch": 1.6604499919800522,
      "grad_norm": 5.5625,
      "learning_rate": 7.36873209300083e-07,
      "loss": 0.71615515,
      "memory(GiB)": 138.1,
      "step": 71170,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.77024212,
      "epoch": 1.660683299552341,
      "grad_norm": 5.1875,
      "learning_rate": 7.35886409841901e-07,
      "loss": 0.82428303,
      "memory(GiB)": 138.1,
      "step": 71180,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.78113737,
      "epoch": 1.66091660712463,
      "grad_norm": 4.3125,
      "learning_rate": 7.349002190846965e-07,
      "loss": 0.77647839,
      "memory(GiB)": 138.1,
      "step": 71190,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.78592319,
      "epoch": 1.6611499146969189,
      "grad_norm": 4.5625,
      "learning_rate": 7.339146371692468e-07,
      "loss": 0.77881441,
      "memory(GiB)": 138.1,
      "step": 71200,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.78388247,
      "epoch": 1.6613832222692078,
      "grad_norm": 7.34375,
      "learning_rate": 7.329296642362438e-07,
      "loss": 0.77189293,
      "memory(GiB)": 138.1,
      "step": 71210,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.77709417,
      "epoch": 1.6616165298414967,
      "grad_norm": 12.875,
      "learning_rate": 7.319453004262911e-07,
      "loss": 0.80911427,
      "memory(GiB)": 138.1,
      "step": 71220,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.78052092,
      "epoch": 1.6618498374137856,
      "grad_norm": 6.15625,
      "learning_rate": 7.309615458799058e-07,
      "loss": 0.80394659,
      "memory(GiB)": 138.1,
      "step": 71230,
      "train_speed(iter/s)": 0.200832
    },
    {
      "acc": 0.78070478,
      "epoch": 1.6620831449860745,
      "grad_norm": 4.84375,
      "learning_rate": 7.299784007375205e-07,
      "loss": 0.78910527,
      "memory(GiB)": 138.1,
      "step": 71240,
      "train_speed(iter/s)": 0.200846
    },
    {
      "acc": 0.79349833,
      "epoch": 1.6623164525583634,
      "grad_norm": 5.375,
      "learning_rate": 7.289958651394774e-07,
      "loss": 0.7258852,
      "memory(GiB)": 138.1,
      "step": 71250,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.76768055,
      "epoch": 1.6625497601306523,
      "grad_norm": 5.28125,
      "learning_rate": 7.280139392260344e-07,
      "loss": 0.83310871,
      "memory(GiB)": 138.1,
      "step": 71260,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.76861515,
      "epoch": 1.6627830677029412,
      "grad_norm": 7.6875,
      "learning_rate": 7.270326231373598e-07,
      "loss": 0.81961651,
      "memory(GiB)": 138.1,
      "step": 71270,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.81069078,
      "epoch": 1.66301637527523,
      "grad_norm": 6.3125,
      "learning_rate": 7.260519170135383e-07,
      "loss": 0.6584691,
      "memory(GiB)": 138.1,
      "step": 71280,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.7901927,
      "epoch": 1.663249682847519,
      "grad_norm": 6.90625,
      "learning_rate": 7.25071820994564e-07,
      "loss": 0.76735644,
      "memory(GiB)": 138.1,
      "step": 71290,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77350497,
      "epoch": 1.6634829904198078,
      "grad_norm": 6.15625,
      "learning_rate": 7.240923352203438e-07,
      "loss": 0.81890402,
      "memory(GiB)": 138.1,
      "step": 71300,
      "train_speed(iter/s)": 0.200934
    },
    {
      "acc": 0.77600427,
      "epoch": 1.6637162979920967,
      "grad_norm": 4.46875,
      "learning_rate": 7.231134598307022e-07,
      "loss": 0.80230637,
      "memory(GiB)": 138.1,
      "step": 71310,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78995695,
      "epoch": 1.6639496055643856,
      "grad_norm": 4.40625,
      "learning_rate": 7.221351949653715e-07,
      "loss": 0.73928757,
      "memory(GiB)": 138.1,
      "step": 71320,
      "train_speed(iter/s)": 0.200965
    },
    {
      "acc": 0.78438892,
      "epoch": 1.6641829131366745,
      "grad_norm": 4.8125,
      "learning_rate": 7.211575407639987e-07,
      "loss": 0.77428908,
      "memory(GiB)": 138.1,
      "step": 71330,
      "train_speed(iter/s)": 0.200979
    },
    {
      "acc": 0.77166224,
      "epoch": 1.6644162207089632,
      "grad_norm": 7.84375,
      "learning_rate": 7.20180497366143e-07,
      "loss": 0.83759232,
      "memory(GiB)": 138.1,
      "step": 71340,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.79361496,
      "epoch": 1.6646495282812523,
      "grad_norm": 4.28125,
      "learning_rate": 7.192040649112797e-07,
      "loss": 0.74620438,
      "memory(GiB)": 138.1,
      "step": 71350,
      "train_speed(iter/s)": 0.201009
    },
    {
      "acc": 0.77841673,
      "epoch": 1.664882835853541,
      "grad_norm": 5.0625,
      "learning_rate": 7.182282435387922e-07,
      "loss": 0.77743587,
      "memory(GiB)": 138.1,
      "step": 71360,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.77280235,
      "epoch": 1.6651161434258301,
      "grad_norm": 5.4375,
      "learning_rate": 7.172530333879774e-07,
      "loss": 0.83453283,
      "memory(GiB)": 138.1,
      "step": 71370,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.78664598,
      "epoch": 1.6653494509981188,
      "grad_norm": 6.0,
      "learning_rate": 7.1627843459805e-07,
      "loss": 0.78319445,
      "memory(GiB)": 138.1,
      "step": 71380,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.77737265,
      "epoch": 1.665582758570408,
      "grad_norm": 5.65625,
      "learning_rate": 7.153044473081299e-07,
      "loss": 0.79047518,
      "memory(GiB)": 138.1,
      "step": 71390,
      "train_speed(iter/s)": 0.201068
    },
    {
      "acc": 0.7710959,
      "epoch": 1.6658160661426966,
      "grad_norm": 6.0625,
      "learning_rate": 7.143310716572565e-07,
      "loss": 0.83085079,
      "memory(GiB)": 138.1,
      "step": 71400,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.797966,
      "epoch": 1.6660493737149857,
      "grad_norm": 5.75,
      "learning_rate": 7.133583077843776e-07,
      "loss": 0.70618973,
      "memory(GiB)": 138.1,
      "step": 71410,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.77575045,
      "epoch": 1.6662826812872744,
      "grad_norm": 5.34375,
      "learning_rate": 7.12386155828354e-07,
      "loss": 0.82906694,
      "memory(GiB)": 138.1,
      "step": 71420,
      "train_speed(iter/s)": 0.201114
    },
    {
      "acc": 0.75881324,
      "epoch": 1.6665159888595635,
      "grad_norm": 5.65625,
      "learning_rate": 7.114146159279622e-07,
      "loss": 0.8684248,
      "memory(GiB)": 138.1,
      "step": 71430,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.77309985,
      "epoch": 1.6667492964318522,
      "grad_norm": 6.28125,
      "learning_rate": 7.104436882218879e-07,
      "loss": 0.81412992,
      "memory(GiB)": 138.1,
      "step": 71440,
      "train_speed(iter/s)": 0.201143
    },
    {
      "acc": 0.78402615,
      "epoch": 1.6669826040041413,
      "grad_norm": 3.8125,
      "learning_rate": 7.094733728487313e-07,
      "loss": 0.78737326,
      "memory(GiB)": 138.1,
      "step": 71450,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.79735641,
      "epoch": 1.66721591157643,
      "grad_norm": 6.6875,
      "learning_rate": 7.085036699470027e-07,
      "loss": 0.71820598,
      "memory(GiB)": 138.1,
      "step": 71460,
      "train_speed(iter/s)": 0.201172
    },
    {
      "acc": 0.77270122,
      "epoch": 1.6674492191487191,
      "grad_norm": 5.03125,
      "learning_rate": 7.075345796551303e-07,
      "loss": 0.82472687,
      "memory(GiB)": 138.1,
      "step": 71470,
      "train_speed(iter/s)": 0.201185
    },
    {
      "acc": 0.79649363,
      "epoch": 1.6676825267210078,
      "grad_norm": 6.3125,
      "learning_rate": 7.065661021114478e-07,
      "loss": 0.72360072,
      "memory(GiB)": 138.1,
      "step": 71480,
      "train_speed(iter/s)": 0.201199
    },
    {
      "acc": 0.78411646,
      "epoch": 1.667915834293297,
      "grad_norm": 6.09375,
      "learning_rate": 7.055982374542086e-07,
      "loss": 0.75441995,
      "memory(GiB)": 138.1,
      "step": 71490,
      "train_speed(iter/s)": 0.201214
    },
    {
      "acc": 0.76906228,
      "epoch": 1.6681491418655856,
      "grad_norm": 5.0,
      "learning_rate": 7.046309858215733e-07,
      "loss": 0.84269314,
      "memory(GiB)": 138.1,
      "step": 71500,
      "train_speed(iter/s)": 0.201228
    },
    {
      "epoch": 1.6681491418655856,
      "eval_acc": 0.7446756446726179,
      "eval_loss": 0.8043976426124573,
      "eval_runtime": 1270.6288,
      "eval_samples_per_second": 28.325,
      "eval_steps_per_second": 14.163,
      "step": 71500
    },
    {
      "acc": 0.77692218,
      "epoch": 1.6683824494378747,
      "grad_norm": 5.90625,
      "learning_rate": 7.036643473516164e-07,
      "loss": 0.81663551,
      "memory(GiB)": 138.1,
      "step": 71510,
      "train_speed(iter/s)": 0.200512
    },
    {
      "acc": 0.77694969,
      "epoch": 1.6686157570101634,
      "grad_norm": 5.625,
      "learning_rate": 7.026983221823264e-07,
      "loss": 0.81028423,
      "memory(GiB)": 138.1,
      "step": 71520,
      "train_speed(iter/s)": 0.200526
    },
    {
      "acc": 0.77378316,
      "epoch": 1.6688490645824523,
      "grad_norm": 6.46875,
      "learning_rate": 7.017329104516013e-07,
      "loss": 0.7947732,
      "memory(GiB)": 138.1,
      "step": 71530,
      "train_speed(iter/s)": 0.200541
    },
    {
      "acc": 0.77548847,
      "epoch": 1.6690823721547412,
      "grad_norm": 4.65625,
      "learning_rate": 7.007681122972559e-07,
      "loss": 0.80890713,
      "memory(GiB)": 138.1,
      "step": 71540,
      "train_speed(iter/s)": 0.200555
    },
    {
      "acc": 0.78633842,
      "epoch": 1.66931567972703,
      "grad_norm": 5.4375,
      "learning_rate": 6.998039278570134e-07,
      "loss": 0.74881501,
      "memory(GiB)": 138.1,
      "step": 71550,
      "train_speed(iter/s)": 0.20057
    },
    {
      "acc": 0.76336951,
      "epoch": 1.669548987299319,
      "grad_norm": 5.78125,
      "learning_rate": 6.988403572685115e-07,
      "loss": 0.82800121,
      "memory(GiB)": 138.1,
      "step": 71560,
      "train_speed(iter/s)": 0.200585
    },
    {
      "acc": 0.78569932,
      "epoch": 1.669782294871608,
      "grad_norm": 4.25,
      "learning_rate": 6.978774006692984e-07,
      "loss": 0.76846313,
      "memory(GiB)": 138.1,
      "step": 71570,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.78505049,
      "epoch": 1.6700156024438968,
      "grad_norm": 4.0625,
      "learning_rate": 6.969150581968359e-07,
      "loss": 0.76446638,
      "memory(GiB)": 138.1,
      "step": 71580,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.79139848,
      "epoch": 1.6702489100161857,
      "grad_norm": 3.65625,
      "learning_rate": 6.959533299885001e-07,
      "loss": 0.75788527,
      "memory(GiB)": 138.1,
      "step": 71590,
      "train_speed(iter/s)": 0.200628
    },
    {
      "acc": 0.78600264,
      "epoch": 1.6704822175884746,
      "grad_norm": 5.28125,
      "learning_rate": 6.949922161815748e-07,
      "loss": 0.76887283,
      "memory(GiB)": 138.1,
      "step": 71600,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.77672091,
      "epoch": 1.6707155251607635,
      "grad_norm": 5.59375,
      "learning_rate": 6.94031716913261e-07,
      "loss": 0.80708447,
      "memory(GiB)": 138.1,
      "step": 71610,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.78550982,
      "epoch": 1.6709488327330524,
      "grad_norm": 6.84375,
      "learning_rate": 6.930718323206676e-07,
      "loss": 0.78040328,
      "memory(GiB)": 138.1,
      "step": 71620,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.78563089,
      "epoch": 1.6711821403053413,
      "grad_norm": 7.84375,
      "learning_rate": 6.921125625408198e-07,
      "loss": 0.75789003,
      "memory(GiB)": 138.1,
      "step": 71630,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.77634792,
      "epoch": 1.6714154478776302,
      "grad_norm": 7.25,
      "learning_rate": 6.911539077106527e-07,
      "loss": 0.81327028,
      "memory(GiB)": 138.1,
      "step": 71640,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.75646667,
      "epoch": 1.671648755449919,
      "grad_norm": 5.09375,
      "learning_rate": 6.901958679670123e-07,
      "loss": 0.90101776,
      "memory(GiB)": 138.1,
      "step": 71650,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.80443058,
      "epoch": 1.671882063022208,
      "grad_norm": 5.625,
      "learning_rate": 6.892384434466609e-07,
      "loss": 0.70652561,
      "memory(GiB)": 138.1,
      "step": 71660,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.77762895,
      "epoch": 1.6721153705944969,
      "grad_norm": 6.875,
      "learning_rate": 6.882816342862692e-07,
      "loss": 0.80890656,
      "memory(GiB)": 138.1,
      "step": 71670,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.78075476,
      "epoch": 1.6723486781667858,
      "grad_norm": 4.5,
      "learning_rate": 6.873254406224223e-07,
      "loss": 0.79064302,
      "memory(GiB)": 138.1,
      "step": 71680,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.78527708,
      "epoch": 1.6725819857390747,
      "grad_norm": 6.09375,
      "learning_rate": 6.863698625916137e-07,
      "loss": 0.74829164,
      "memory(GiB)": 138.1,
      "step": 71690,
      "train_speed(iter/s)": 0.200773
    },
    {
      "acc": 0.77983356,
      "epoch": 1.6728152933113636,
      "grad_norm": 4.25,
      "learning_rate": 6.854149003302562e-07,
      "loss": 0.77540479,
      "memory(GiB)": 138.1,
      "step": 71700,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.79154415,
      "epoch": 1.6730486008836525,
      "grad_norm": 4.90625,
      "learning_rate": 6.844605539746679e-07,
      "loss": 0.7527297,
      "memory(GiB)": 138.1,
      "step": 71710,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.81196423,
      "epoch": 1.6732819084559414,
      "grad_norm": 3.984375,
      "learning_rate": 6.835068236610809e-07,
      "loss": 0.66903067,
      "memory(GiB)": 138.1,
      "step": 71720,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.80410776,
      "epoch": 1.67351521602823,
      "grad_norm": 5.28125,
      "learning_rate": 6.825537095256418e-07,
      "loss": 0.68753047,
      "memory(GiB)": 138.1,
      "step": 71730,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.76747966,
      "epoch": 1.6737485236005192,
      "grad_norm": 4.625,
      "learning_rate": 6.816012117044052e-07,
      "loss": 0.83944092,
      "memory(GiB)": 138.1,
      "step": 71740,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.7719739,
      "epoch": 1.6739818311728079,
      "grad_norm": 4.8125,
      "learning_rate": 6.806493303333422e-07,
      "loss": 0.82619953,
      "memory(GiB)": 138.1,
      "step": 71750,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.77261276,
      "epoch": 1.674215138745097,
      "grad_norm": 8.25,
      "learning_rate": 6.796980655483315e-07,
      "loss": 0.80542164,
      "memory(GiB)": 138.1,
      "step": 71760,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.77031894,
      "epoch": 1.6744484463173857,
      "grad_norm": 4.46875,
      "learning_rate": 6.787474174851683e-07,
      "loss": 0.84091234,
      "memory(GiB)": 138.1,
      "step": 71770,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.78931608,
      "epoch": 1.6746817538896748,
      "grad_norm": 4.71875,
      "learning_rate": 6.777973862795556e-07,
      "loss": 0.76971698,
      "memory(GiB)": 138.1,
      "step": 71780,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.80558147,
      "epoch": 1.6749150614619635,
      "grad_norm": 5.0,
      "learning_rate": 6.768479720671106e-07,
      "loss": 0.69032288,
      "memory(GiB)": 138.1,
      "step": 71790,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.79586306,
      "epoch": 1.6751483690342526,
      "grad_norm": 3.1875,
      "learning_rate": 6.758991749833616e-07,
      "loss": 0.72956157,
      "memory(GiB)": 138.1,
      "step": 71800,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.77934403,
      "epoch": 1.6753816766065412,
      "grad_norm": 4.65625,
      "learning_rate": 6.749509951637484e-07,
      "loss": 0.78227091,
      "memory(GiB)": 138.1,
      "step": 71810,
      "train_speed(iter/s)": 0.200947
    },
    {
      "acc": 0.7744832,
      "epoch": 1.6756149841788304,
      "grad_norm": 7.0625,
      "learning_rate": 6.740034327436251e-07,
      "loss": 0.81382933,
      "memory(GiB)": 138.1,
      "step": 71820,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.77508135,
      "epoch": 1.675848291751119,
      "grad_norm": 5.84375,
      "learning_rate": 6.730564878582535e-07,
      "loss": 0.81892805,
      "memory(GiB)": 138.1,
      "step": 71830,
      "train_speed(iter/s)": 0.200975
    },
    {
      "acc": 0.80630512,
      "epoch": 1.6760815993234082,
      "grad_norm": 6.15625,
      "learning_rate": 6.721101606428132e-07,
      "loss": 0.70892029,
      "memory(GiB)": 138.1,
      "step": 71840,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.7737956,
      "epoch": 1.6763149068956968,
      "grad_norm": 5.8125,
      "learning_rate": 6.711644512323895e-07,
      "loss": 0.83846264,
      "memory(GiB)": 138.1,
      "step": 71850,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.76679764,
      "epoch": 1.676548214467986,
      "grad_norm": 4.3125,
      "learning_rate": 6.702193597619821e-07,
      "loss": 0.84609222,
      "memory(GiB)": 138.1,
      "step": 71860,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.75910959,
      "epoch": 1.6767815220402746,
      "grad_norm": 5.40625,
      "learning_rate": 6.692748863665044e-07,
      "loss": 0.85825319,
      "memory(GiB)": 138.1,
      "step": 71870,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.78103266,
      "epoch": 1.6770148296125638,
      "grad_norm": 5.34375,
      "learning_rate": 6.683310311807772e-07,
      "loss": 0.78897181,
      "memory(GiB)": 138.1,
      "step": 71880,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.78252711,
      "epoch": 1.6772481371848524,
      "grad_norm": 5.71875,
      "learning_rate": 6.673877943395385e-07,
      "loss": 0.7663425,
      "memory(GiB)": 138.1,
      "step": 71890,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.79238405,
      "epoch": 1.6774814447571416,
      "grad_norm": 6.40625,
      "learning_rate": 6.664451759774332e-07,
      "loss": 0.7801724,
      "memory(GiB)": 138.1,
      "step": 71900,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.78016853,
      "epoch": 1.6777147523294302,
      "grad_norm": 5.625,
      "learning_rate": 6.655031762290203e-07,
      "loss": 0.77946939,
      "memory(GiB)": 138.1,
      "step": 71910,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.75844526,
      "epoch": 1.6779480599017191,
      "grad_norm": 4.71875,
      "learning_rate": 6.645617952287686e-07,
      "loss": 0.85980473,
      "memory(GiB)": 138.1,
      "step": 71920,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.77880216,
      "epoch": 1.678181367474008,
      "grad_norm": 6.15625,
      "learning_rate": 6.636210331110621e-07,
      "loss": 0.78121481,
      "memory(GiB)": 138.1,
      "step": 71930,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.76943226,
      "epoch": 1.678414675046297,
      "grad_norm": 4.6875,
      "learning_rate": 6.626808900101939e-07,
      "loss": 0.82415466,
      "memory(GiB)": 138.1,
      "step": 71940,
      "train_speed(iter/s)": 0.201134
    },
    {
      "acc": 0.76650505,
      "epoch": 1.6786479826185858,
      "grad_norm": 5.59375,
      "learning_rate": 6.617413660603672e-07,
      "loss": 0.86484928,
      "memory(GiB)": 138.1,
      "step": 71950,
      "train_speed(iter/s)": 0.201149
    },
    {
      "acc": 0.78949447,
      "epoch": 1.6788812901908747,
      "grad_norm": 5.0,
      "learning_rate": 6.608024613957015e-07,
      "loss": 0.77047791,
      "memory(GiB)": 138.1,
      "step": 71960,
      "train_speed(iter/s)": 0.201164
    },
    {
      "acc": 0.77228599,
      "epoch": 1.6791145977631636,
      "grad_norm": 6.4375,
      "learning_rate": 6.598641761502222e-07,
      "loss": 0.82509651,
      "memory(GiB)": 138.1,
      "step": 71970,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.75860147,
      "epoch": 1.6793479053354525,
      "grad_norm": 5.4375,
      "learning_rate": 6.58926510457873e-07,
      "loss": 0.88990307,
      "memory(GiB)": 138.1,
      "step": 71980,
      "train_speed(iter/s)": 0.201192
    },
    {
      "acc": 0.77799234,
      "epoch": 1.6795812129077414,
      "grad_norm": 4.5625,
      "learning_rate": 6.579894644525026e-07,
      "loss": 0.83457985,
      "memory(GiB)": 138.1,
      "step": 71990,
      "train_speed(iter/s)": 0.201207
    },
    {
      "acc": 0.80525131,
      "epoch": 1.6798145204800303,
      "grad_norm": 3.59375,
      "learning_rate": 6.570530382678741e-07,
      "loss": 0.69075212,
      "memory(GiB)": 138.1,
      "step": 72000,
      "train_speed(iter/s)": 0.201221
    },
    {
      "epoch": 1.6798145204800303,
      "eval_acc": 0.7446879757785052,
      "eval_loss": 0.8043897747993469,
      "eval_runtime": 1271.86,
      "eval_samples_per_second": 28.298,
      "eval_steps_per_second": 14.149,
      "step": 72000
    },
    {
      "acc": 0.77684412,
      "epoch": 1.6800478280523192,
      "grad_norm": 7.34375,
      "learning_rate": 6.561172320376647e-07,
      "loss": 0.79601202,
      "memory(GiB)": 138.1,
      "step": 72010,
      "train_speed(iter/s)": 0.200509
    },
    {
      "acc": 0.76988397,
      "epoch": 1.6802811356246081,
      "grad_norm": 5.65625,
      "learning_rate": 6.551820458954561e-07,
      "loss": 0.80573511,
      "memory(GiB)": 138.1,
      "step": 72020,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.77471781,
      "epoch": 1.680514443196897,
      "grad_norm": 4.53125,
      "learning_rate": 6.5424747997475e-07,
      "loss": 0.79388332,
      "memory(GiB)": 138.1,
      "step": 72030,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.76678681,
      "epoch": 1.680747750769186,
      "grad_norm": 4.5,
      "learning_rate": 6.53313534408952e-07,
      "loss": 0.85442133,
      "memory(GiB)": 138.1,
      "step": 72040,
      "train_speed(iter/s)": 0.200554
    },
    {
      "acc": 0.78160801,
      "epoch": 1.6809810583414748,
      "grad_norm": 7.375,
      "learning_rate": 6.523802093313857e-07,
      "loss": 0.75970039,
      "memory(GiB)": 138.1,
      "step": 72050,
      "train_speed(iter/s)": 0.20057
    },
    {
      "acc": 0.75993977,
      "epoch": 1.6812143659137637,
      "grad_norm": 6.5625,
      "learning_rate": 6.514475048752805e-07,
      "loss": 0.88980999,
      "memory(GiB)": 138.1,
      "step": 72060,
      "train_speed(iter/s)": 0.200585
    },
    {
      "acc": 0.78098273,
      "epoch": 1.6814476734860526,
      "grad_norm": 4.4375,
      "learning_rate": 6.505154211737813e-07,
      "loss": 0.7791935,
      "memory(GiB)": 138.1,
      "step": 72070,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.78494759,
      "epoch": 1.6816809810583415,
      "grad_norm": 6.375,
      "learning_rate": 6.495839583599428e-07,
      "loss": 0.765938,
      "memory(GiB)": 138.1,
      "step": 72080,
      "train_speed(iter/s)": 0.200614
    },
    {
      "acc": 0.78756342,
      "epoch": 1.6819142886306304,
      "grad_norm": 4.90625,
      "learning_rate": 6.486531165667292e-07,
      "loss": 0.74883633,
      "memory(GiB)": 138.1,
      "step": 72090,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.78392973,
      "epoch": 1.6821475962029193,
      "grad_norm": 5.03125,
      "learning_rate": 6.477228959270199e-07,
      "loss": 0.78099174,
      "memory(GiB)": 138.1,
      "step": 72100,
      "train_speed(iter/s)": 0.200643
    },
    {
      "acc": 0.78292713,
      "epoch": 1.6823809037752082,
      "grad_norm": 5.46875,
      "learning_rate": 6.467932965736024e-07,
      "loss": 0.75630102,
      "memory(GiB)": 138.1,
      "step": 72110,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.78047161,
      "epoch": 1.682614211347497,
      "grad_norm": 5.125,
      "learning_rate": 6.458643186391789e-07,
      "loss": 0.76697688,
      "memory(GiB)": 138.1,
      "step": 72120,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.78350782,
      "epoch": 1.682847518919786,
      "grad_norm": 7.0,
      "learning_rate": 6.449359622563567e-07,
      "loss": 0.76841931,
      "memory(GiB)": 138.1,
      "step": 72130,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.79610729,
      "epoch": 1.6830808264920747,
      "grad_norm": 4.375,
      "learning_rate": 6.44008227557662e-07,
      "loss": 0.74552526,
      "memory(GiB)": 138.1,
      "step": 72140,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.75990648,
      "epoch": 1.6833141340643638,
      "grad_norm": 5.34375,
      "learning_rate": 6.430811146755272e-07,
      "loss": 0.85680218,
      "memory(GiB)": 138.1,
      "step": 72150,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.78126965,
      "epoch": 1.6835474416366525,
      "grad_norm": 6.0625,
      "learning_rate": 6.421546237422971e-07,
      "loss": 0.78917265,
      "memory(GiB)": 138.1,
      "step": 72160,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.77440681,
      "epoch": 1.6837807492089416,
      "grad_norm": 4.53125,
      "learning_rate": 6.412287548902291e-07,
      "loss": 0.82259197,
      "memory(GiB)": 138.1,
      "step": 72170,
      "train_speed(iter/s)": 0.200744
    },
    {
      "acc": 0.79978375,
      "epoch": 1.6840140567812303,
      "grad_norm": 9.0625,
      "learning_rate": 6.403035082514891e-07,
      "loss": 0.71059504,
      "memory(GiB)": 138.1,
      "step": 72180,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.78088136,
      "epoch": 1.6842473643535194,
      "grad_norm": 6.4375,
      "learning_rate": 6.393788839581578e-07,
      "loss": 0.78694282,
      "memory(GiB)": 138.1,
      "step": 72190,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.76632071,
      "epoch": 1.684480671925808,
      "grad_norm": 7.75,
      "learning_rate": 6.384548821422243e-07,
      "loss": 0.8362587,
      "memory(GiB)": 138.1,
      "step": 72200,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.75542583,
      "epoch": 1.6847139794980972,
      "grad_norm": 5.25,
      "learning_rate": 6.375315029355883e-07,
      "loss": 0.88243284,
      "memory(GiB)": 138.1,
      "step": 72210,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.79456859,
      "epoch": 1.6849472870703859,
      "grad_norm": 6.1875,
      "learning_rate": 6.366087464700637e-07,
      "loss": 0.7337564,
      "memory(GiB)": 138.1,
      "step": 72220,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.79592638,
      "epoch": 1.685180594642675,
      "grad_norm": 5.0,
      "learning_rate": 6.35686612877372e-07,
      "loss": 0.7540689,
      "memory(GiB)": 138.1,
      "step": 72230,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.76712937,
      "epoch": 1.6854139022149637,
      "grad_norm": 4.84375,
      "learning_rate": 6.34765102289151e-07,
      "loss": 0.85317764,
      "memory(GiB)": 138.1,
      "step": 72240,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.77703066,
      "epoch": 1.6856472097872528,
      "grad_norm": 4.875,
      "learning_rate": 6.338442148369406e-07,
      "loss": 0.77372303,
      "memory(GiB)": 138.1,
      "step": 72250,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.78281322,
      "epoch": 1.6858805173595415,
      "grad_norm": 4.5,
      "learning_rate": 6.329239506522017e-07,
      "loss": 0.77638769,
      "memory(GiB)": 138.1,
      "step": 72260,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.78430648,
      "epoch": 1.6861138249318306,
      "grad_norm": 7.125,
      "learning_rate": 6.320043098662992e-07,
      "loss": 0.7674891,
      "memory(GiB)": 138.1,
      "step": 72270,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.77010136,
      "epoch": 1.6863471325041193,
      "grad_norm": 4.6875,
      "learning_rate": 6.310852926105138e-07,
      "loss": 0.82719336,
      "memory(GiB)": 138.1,
      "step": 72280,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.78210573,
      "epoch": 1.6865804400764084,
      "grad_norm": 5.4375,
      "learning_rate": 6.301668990160331e-07,
      "loss": 0.79944468,
      "memory(GiB)": 138.1,
      "step": 72290,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.80032415,
      "epoch": 1.686813747648697,
      "grad_norm": 4.9375,
      "learning_rate": 6.292491292139574e-07,
      "loss": 0.72096748,
      "memory(GiB)": 138.1,
      "step": 72300,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.7697011,
      "epoch": 1.687047055220986,
      "grad_norm": 7.03125,
      "learning_rate": 6.283319833353002e-07,
      "loss": 0.82724371,
      "memory(GiB)": 138.1,
      "step": 72310,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.79757509,
      "epoch": 1.6872803627932749,
      "grad_norm": 6.6875,
      "learning_rate": 6.274154615109812e-07,
      "loss": 0.70658522,
      "memory(GiB)": 138.1,
      "step": 72320,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.79727383,
      "epoch": 1.6875136703655638,
      "grad_norm": 6.875,
      "learning_rate": 6.264995638718352e-07,
      "loss": 0.7160409,
      "memory(GiB)": 138.1,
      "step": 72330,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.76207228,
      "epoch": 1.6877469779378527,
      "grad_norm": 9.125,
      "learning_rate": 6.255842905486065e-07,
      "loss": 0.85418339,
      "memory(GiB)": 138.1,
      "step": 72340,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.77744265,
      "epoch": 1.6879802855101416,
      "grad_norm": 5.125,
      "learning_rate": 6.246696416719495e-07,
      "loss": 0.80833817,
      "memory(GiB)": 138.1,
      "step": 72350,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.80837536,
      "epoch": 1.6882135930824305,
      "grad_norm": 5.4375,
      "learning_rate": 6.237556173724291e-07,
      "loss": 0.70104604,
      "memory(GiB)": 138.1,
      "step": 72360,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.79856534,
      "epoch": 1.6884469006547194,
      "grad_norm": 4.8125,
      "learning_rate": 6.228422177805244e-07,
      "loss": 0.70726795,
      "memory(GiB)": 138.1,
      "step": 72370,
      "train_speed(iter/s)": 0.20103
    },
    {
      "acc": 0.76452069,
      "epoch": 1.6886802082270083,
      "grad_norm": 8.4375,
      "learning_rate": 6.21929443026621e-07,
      "loss": 0.84759407,
      "memory(GiB)": 138.1,
      "step": 72380,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.78385258,
      "epoch": 1.6889135157992972,
      "grad_norm": 4.90625,
      "learning_rate": 6.210172932410169e-07,
      "loss": 0.76335421,
      "memory(GiB)": 138.1,
      "step": 72390,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.77586222,
      "epoch": 1.689146823371586,
      "grad_norm": 5.9375,
      "learning_rate": 6.20105768553923e-07,
      "loss": 0.79805136,
      "memory(GiB)": 138.1,
      "step": 72400,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.78417902,
      "epoch": 1.689380130943875,
      "grad_norm": 5.375,
      "learning_rate": 6.191948690954575e-07,
      "loss": 0.75983038,
      "memory(GiB)": 138.1,
      "step": 72410,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.78113298,
      "epoch": 1.6896134385161639,
      "grad_norm": 4.6875,
      "learning_rate": 6.182845949956523e-07,
      "loss": 0.80795841,
      "memory(GiB)": 138.1,
      "step": 72420,
      "train_speed(iter/s)": 0.201104
    },
    {
      "acc": 0.78343716,
      "epoch": 1.6898467460884528,
      "grad_norm": 5.1875,
      "learning_rate": 6.173749463844486e-07,
      "loss": 0.76608415,
      "memory(GiB)": 138.1,
      "step": 72430,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.79046288,
      "epoch": 1.6900800536607417,
      "grad_norm": 5.1875,
      "learning_rate": 6.164659233916976e-07,
      "loss": 0.75967202,
      "memory(GiB)": 138.1,
      "step": 72440,
      "train_speed(iter/s)": 0.201131
    },
    {
      "acc": 0.76052351,
      "epoch": 1.6903133612330306,
      "grad_norm": 6.0625,
      "learning_rate": 6.15557526147163e-07,
      "loss": 0.86396046,
      "memory(GiB)": 138.1,
      "step": 72450,
      "train_speed(iter/s)": 0.201145
    },
    {
      "acc": 0.7680316,
      "epoch": 1.6905466688053195,
      "grad_norm": 4.875,
      "learning_rate": 6.146497547805169e-07,
      "loss": 0.83324471,
      "memory(GiB)": 138.1,
      "step": 72460,
      "train_speed(iter/s)": 0.201159
    },
    {
      "acc": 0.78809872,
      "epoch": 1.6907799763776084,
      "grad_norm": 3.703125,
      "learning_rate": 6.137426094213466e-07,
      "loss": 0.74023752,
      "memory(GiB)": 138.1,
      "step": 72470,
      "train_speed(iter/s)": 0.201174
    },
    {
      "acc": 0.77969913,
      "epoch": 1.6910132839498973,
      "grad_norm": 4.71875,
      "learning_rate": 6.128360901991426e-07,
      "loss": 0.79022303,
      "memory(GiB)": 138.1,
      "step": 72480,
      "train_speed(iter/s)": 0.201189
    },
    {
      "acc": 0.80677795,
      "epoch": 1.691246591522186,
      "grad_norm": 10.1875,
      "learning_rate": 6.119301972433128e-07,
      "loss": 0.69705276,
      "memory(GiB)": 138.1,
      "step": 72490,
      "train_speed(iter/s)": 0.201203
    },
    {
      "acc": 0.76214151,
      "epoch": 1.691479899094475,
      "grad_norm": 7.3125,
      "learning_rate": 6.110249306831733e-07,
      "loss": 0.86750393,
      "memory(GiB)": 138.1,
      "step": 72500,
      "train_speed(iter/s)": 0.201218
    },
    {
      "epoch": 1.691479899094475,
      "eval_acc": 0.7446806091438193,
      "eval_loss": 0.8044036626815796,
      "eval_runtime": 1270.6076,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "step": 72500
    },
    {
      "acc": 0.80648994,
      "epoch": 1.6917132066667637,
      "grad_norm": 6.1875,
      "learning_rate": 6.101202906479487e-07,
      "loss": 0.67632608,
      "memory(GiB)": 138.1,
      "step": 72510,
      "train_speed(iter/s)": 0.200512
    },
    {
      "acc": 0.77256756,
      "epoch": 1.6919465142390528,
      "grad_norm": 3.859375,
      "learning_rate": 6.092162772667781e-07,
      "loss": 0.80494852,
      "memory(GiB)": 138.1,
      "step": 72520,
      "train_speed(iter/s)": 0.200527
    },
    {
      "acc": 0.77806387,
      "epoch": 1.6921798218113415,
      "grad_norm": 6.625,
      "learning_rate": 6.083128906687074e-07,
      "loss": 0.78679771,
      "memory(GiB)": 138.1,
      "step": 72530,
      "train_speed(iter/s)": 0.200542
    },
    {
      "acc": 0.77233944,
      "epoch": 1.6924131293836306,
      "grad_norm": 5.09375,
      "learning_rate": 6.074101309826968e-07,
      "loss": 0.80660133,
      "memory(GiB)": 138.1,
      "step": 72540,
      "train_speed(iter/s)": 0.200557
    },
    {
      "acc": 0.79014969,
      "epoch": 1.6926464369559193,
      "grad_norm": 6.875,
      "learning_rate": 6.065079983376132e-07,
      "loss": 0.76289606,
      "memory(GiB)": 138.1,
      "step": 72550,
      "train_speed(iter/s)": 0.200571
    },
    {
      "acc": 0.78846631,
      "epoch": 1.6928797445282084,
      "grad_norm": 5.75,
      "learning_rate": 6.056064928622374e-07,
      "loss": 0.77690454,
      "memory(GiB)": 138.1,
      "step": 72560,
      "train_speed(iter/s)": 0.200586
    },
    {
      "acc": 0.79145346,
      "epoch": 1.6931130521004971,
      "grad_norm": 6.15625,
      "learning_rate": 6.047056146852575e-07,
      "loss": 0.74644814,
      "memory(GiB)": 138.1,
      "step": 72570,
      "train_speed(iter/s)": 0.2006
    },
    {
      "acc": 0.78505545,
      "epoch": 1.6933463596727862,
      "grad_norm": 4.125,
      "learning_rate": 6.038053639352754e-07,
      "loss": 0.76110735,
      "memory(GiB)": 138.1,
      "step": 72580,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.78503914,
      "epoch": 1.693579667245075,
      "grad_norm": 4.21875,
      "learning_rate": 6.029057407407995e-07,
      "loss": 0.7781271,
      "memory(GiB)": 138.1,
      "step": 72590,
      "train_speed(iter/s)": 0.200628
    },
    {
      "acc": 0.82557945,
      "epoch": 1.693812974817364,
      "grad_norm": 4.21875,
      "learning_rate": 6.020067452302514e-07,
      "loss": 0.6177803,
      "memory(GiB)": 138.1,
      "step": 72600,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.79121065,
      "epoch": 1.6940462823896527,
      "grad_norm": 5.78125,
      "learning_rate": 6.011083775319637e-07,
      "loss": 0.73865213,
      "memory(GiB)": 138.1,
      "step": 72610,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.77857733,
      "epoch": 1.6942795899619418,
      "grad_norm": 4.15625,
      "learning_rate": 6.002106377741762e-07,
      "loss": 0.78328385,
      "memory(GiB)": 138.1,
      "step": 72620,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.7810358,
      "epoch": 1.6945128975342305,
      "grad_norm": 6.53125,
      "learning_rate": 5.99313526085043e-07,
      "loss": 0.7782196,
      "memory(GiB)": 138.1,
      "step": 72630,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.79591274,
      "epoch": 1.6947462051065196,
      "grad_norm": 16.125,
      "learning_rate": 5.984170425926256e-07,
      "loss": 0.72331409,
      "memory(GiB)": 138.1,
      "step": 72640,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.77169046,
      "epoch": 1.6949795126788083,
      "grad_norm": 5.09375,
      "learning_rate": 5.975211874248954e-07,
      "loss": 0.83417196,
      "memory(GiB)": 138.1,
      "step": 72650,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.80202818,
      "epoch": 1.6952128202510974,
      "grad_norm": 6.21875,
      "learning_rate": 5.96625960709738e-07,
      "loss": 0.72186661,
      "memory(GiB)": 138.1,
      "step": 72660,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78016157,
      "epoch": 1.695446127823386,
      "grad_norm": 5.53125,
      "learning_rate": 5.957313625749445e-07,
      "loss": 0.80601139,
      "memory(GiB)": 138.1,
      "step": 72670,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.78957376,
      "epoch": 1.695679435395675,
      "grad_norm": 4.75,
      "learning_rate": 5.948373931482204e-07,
      "loss": 0.75643826,
      "memory(GiB)": 138.1,
      "step": 72680,
      "train_speed(iter/s)": 0.200758
    },
    {
      "acc": 0.75675497,
      "epoch": 1.695912742967964,
      "grad_norm": 6.875,
      "learning_rate": 5.939440525571788e-07,
      "loss": 0.88091316,
      "memory(GiB)": 138.1,
      "step": 72690,
      "train_speed(iter/s)": 0.200771
    },
    {
      "acc": 0.77509961,
      "epoch": 1.6961460505402528,
      "grad_norm": 6.21875,
      "learning_rate": 5.930513409293437e-07,
      "loss": 0.79928541,
      "memory(GiB)": 138.1,
      "step": 72700,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.77624722,
      "epoch": 1.6963793581125417,
      "grad_norm": 6.1875,
      "learning_rate": 5.921592583921488e-07,
      "loss": 0.7726511,
      "memory(GiB)": 138.1,
      "step": 72710,
      "train_speed(iter/s)": 0.2008
    },
    {
      "acc": 0.77374887,
      "epoch": 1.6966126656848306,
      "grad_norm": 5.9375,
      "learning_rate": 5.912678050729398e-07,
      "loss": 0.7831955,
      "memory(GiB)": 138.1,
      "step": 72720,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.7693531,
      "epoch": 1.6968459732571195,
      "grad_norm": 4.34375,
      "learning_rate": 5.903769810989713e-07,
      "loss": 0.82154694,
      "memory(GiB)": 138.1,
      "step": 72730,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.77684445,
      "epoch": 1.6970792808294084,
      "grad_norm": 4.4375,
      "learning_rate": 5.894867865974064e-07,
      "loss": 0.80565586,
      "memory(GiB)": 138.1,
      "step": 72740,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.78173313,
      "epoch": 1.6973125884016973,
      "grad_norm": 4.25,
      "learning_rate": 5.885972216953223e-07,
      "loss": 0.78659649,
      "memory(GiB)": 138.1,
      "step": 72750,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.78354235,
      "epoch": 1.6975458959739862,
      "grad_norm": 5.09375,
      "learning_rate": 5.877082865197026e-07,
      "loss": 0.80655441,
      "memory(GiB)": 138.1,
      "step": 72760,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.79456587,
      "epoch": 1.697779203546275,
      "grad_norm": 6.65625,
      "learning_rate": 5.86819981197444e-07,
      "loss": 0.73504858,
      "memory(GiB)": 138.1,
      "step": 72770,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.79108696,
      "epoch": 1.698012511118564,
      "grad_norm": 8.5,
      "learning_rate": 5.859323058553512e-07,
      "loss": 0.75745306,
      "memory(GiB)": 138.1,
      "step": 72780,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.7710947,
      "epoch": 1.698245818690853,
      "grad_norm": 4.25,
      "learning_rate": 5.850452606201384e-07,
      "loss": 0.81637497,
      "memory(GiB)": 138.1,
      "step": 72790,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.8023735,
      "epoch": 1.6984791262631418,
      "grad_norm": 5.125,
      "learning_rate": 5.841588456184333e-07,
      "loss": 0.69645176,
      "memory(GiB)": 138.1,
      "step": 72800,
      "train_speed(iter/s)": 0.20093
    },
    {
      "acc": 0.78629084,
      "epoch": 1.6987124338354307,
      "grad_norm": 7.1875,
      "learning_rate": 5.8327306097677e-07,
      "loss": 0.75006866,
      "memory(GiB)": 138.1,
      "step": 72810,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.77999191,
      "epoch": 1.6989457414077196,
      "grad_norm": 10.5625,
      "learning_rate": 5.823879068215943e-07,
      "loss": 0.80349598,
      "memory(GiB)": 138.1,
      "step": 72820,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.77250009,
      "epoch": 1.6991790489800085,
      "grad_norm": 6.09375,
      "learning_rate": 5.815033832792605e-07,
      "loss": 0.79956465,
      "memory(GiB)": 138.1,
      "step": 72830,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.78692751,
      "epoch": 1.6994123565522974,
      "grad_norm": 4.90625,
      "learning_rate": 5.806194904760365e-07,
      "loss": 0.75956717,
      "memory(GiB)": 138.1,
      "step": 72840,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.7845644,
      "epoch": 1.6996456641245863,
      "grad_norm": 5.75,
      "learning_rate": 5.797362285380948e-07,
      "loss": 0.77232275,
      "memory(GiB)": 138.1,
      "step": 72850,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.78012753,
      "epoch": 1.6998789716968752,
      "grad_norm": 6.8125,
      "learning_rate": 5.788535975915239e-07,
      "loss": 0.77585268,
      "memory(GiB)": 138.1,
      "step": 72860,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.77515068,
      "epoch": 1.700112279269164,
      "grad_norm": 5.34375,
      "learning_rate": 5.779715977623168e-07,
      "loss": 0.81616535,
      "memory(GiB)": 138.1,
      "step": 72870,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.79815941,
      "epoch": 1.7003455868414528,
      "grad_norm": 5.625,
      "learning_rate": 5.770902291763791e-07,
      "loss": 0.71203232,
      "memory(GiB)": 138.1,
      "step": 72880,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.80398407,
      "epoch": 1.7005788944137419,
      "grad_norm": 5.8125,
      "learning_rate": 5.762094919595274e-07,
      "loss": 0.71322527,
      "memory(GiB)": 138.1,
      "step": 72890,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.79169493,
      "epoch": 1.7008122019860306,
      "grad_norm": 5.125,
      "learning_rate": 5.753293862374842e-07,
      "loss": 0.73782015,
      "memory(GiB)": 138.1,
      "step": 72900,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.78701067,
      "epoch": 1.7010455095583197,
      "grad_norm": 7.125,
      "learning_rate": 5.744499121358871e-07,
      "loss": 0.77875328,
      "memory(GiB)": 138.1,
      "step": 72910,
      "train_speed(iter/s)": 0.201081
    },
    {
      "acc": 0.79096022,
      "epoch": 1.7012788171306084,
      "grad_norm": 4.4375,
      "learning_rate": 5.735710697802793e-07,
      "loss": 0.73712015,
      "memory(GiB)": 138.1,
      "step": 72920,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.77540636,
      "epoch": 1.7015121247028975,
      "grad_norm": 4.84375,
      "learning_rate": 5.726928592961156e-07,
      "loss": 0.80175991,
      "memory(GiB)": 138.1,
      "step": 72930,
      "train_speed(iter/s)": 0.20111
    },
    {
      "acc": 0.77421494,
      "epoch": 1.7017454322751862,
      "grad_norm": 7.53125,
      "learning_rate": 5.718152808087601e-07,
      "loss": 0.81656933,
      "memory(GiB)": 138.1,
      "step": 72940,
      "train_speed(iter/s)": 0.201125
    },
    {
      "acc": 0.76136036,
      "epoch": 1.7019787398474753,
      "grad_norm": 4.5,
      "learning_rate": 5.709383344434854e-07,
      "loss": 0.8624939,
      "memory(GiB)": 138.1,
      "step": 72950,
      "train_speed(iter/s)": 0.201139
    },
    {
      "acc": 0.790942,
      "epoch": 1.702212047419764,
      "grad_norm": 7.21875,
      "learning_rate": 5.700620203254781e-07,
      "loss": 0.74097085,
      "memory(GiB)": 138.1,
      "step": 72960,
      "train_speed(iter/s)": 0.201154
    },
    {
      "acc": 0.77590914,
      "epoch": 1.702445354992053,
      "grad_norm": 5.875,
      "learning_rate": 5.691863385798296e-07,
      "loss": 0.81017885,
      "memory(GiB)": 138.1,
      "step": 72970,
      "train_speed(iter/s)": 0.201168
    },
    {
      "acc": 0.7668539,
      "epoch": 1.7026786625643417,
      "grad_norm": 5.21875,
      "learning_rate": 5.683112893315451e-07,
      "loss": 0.86975269,
      "memory(GiB)": 138.1,
      "step": 72980,
      "train_speed(iter/s)": 0.201183
    },
    {
      "acc": 0.78860574,
      "epoch": 1.7029119701366309,
      "grad_norm": 5.53125,
      "learning_rate": 5.674368727055351e-07,
      "loss": 0.74698706,
      "memory(GiB)": 138.1,
      "step": 72990,
      "train_speed(iter/s)": 0.201197
    },
    {
      "acc": 0.77102966,
      "epoch": 1.7031452777089195,
      "grad_norm": 5.3125,
      "learning_rate": 5.665630888266254e-07,
      "loss": 0.82879314,
      "memory(GiB)": 138.1,
      "step": 73000,
      "train_speed(iter/s)": 0.201211
    },
    {
      "epoch": 1.7031452777089195,
      "eval_acc": 0.7447366596251248,
      "eval_loss": 0.8043876886367798,
      "eval_runtime": 1271.875,
      "eval_samples_per_second": 28.298,
      "eval_steps_per_second": 14.149,
      "step": 73000
    },
    {
      "acc": 0.78933325,
      "epoch": 1.7033785852812087,
      "grad_norm": 5.0625,
      "learning_rate": 5.656899378195468e-07,
      "loss": 0.75085564,
      "memory(GiB)": 138.1,
      "step": 73010,
      "train_speed(iter/s)": 0.200509
    },
    {
      "acc": 0.79252825,
      "epoch": 1.7036118928534973,
      "grad_norm": 4.34375,
      "learning_rate": 5.648174198089407e-07,
      "loss": 0.74576683,
      "memory(GiB)": 138.1,
      "step": 73020,
      "train_speed(iter/s)": 0.200523
    },
    {
      "acc": 0.79472671,
      "epoch": 1.7038452004257865,
      "grad_norm": 5.8125,
      "learning_rate": 5.639455349193602e-07,
      "loss": 0.7537653,
      "memory(GiB)": 138.1,
      "step": 73030,
      "train_speed(iter/s)": 0.200537
    },
    {
      "acc": 0.78937383,
      "epoch": 1.7040785079980751,
      "grad_norm": 5.28125,
      "learning_rate": 5.630742832752655e-07,
      "loss": 0.74292388,
      "memory(GiB)": 138.1,
      "step": 73040,
      "train_speed(iter/s)": 0.200551
    },
    {
      "acc": 0.78601589,
      "epoch": 1.7043118155703643,
      "grad_norm": 3.828125,
      "learning_rate": 5.622036650010281e-07,
      "loss": 0.77539072,
      "memory(GiB)": 138.1,
      "step": 73050,
      "train_speed(iter/s)": 0.200565
    },
    {
      "acc": 0.80177393,
      "epoch": 1.704545123142653,
      "grad_norm": 4.65625,
      "learning_rate": 5.613336802209274e-07,
      "loss": 0.7048872,
      "memory(GiB)": 138.1,
      "step": 73060,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.76926432,
      "epoch": 1.7047784307149418,
      "grad_norm": 5.78125,
      "learning_rate": 5.604643290591555e-07,
      "loss": 0.82517433,
      "memory(GiB)": 138.1,
      "step": 73070,
      "train_speed(iter/s)": 0.200594
    },
    {
      "acc": 0.78017607,
      "epoch": 1.7050117382872307,
      "grad_norm": 5.5625,
      "learning_rate": 5.595956116398111e-07,
      "loss": 0.75298491,
      "memory(GiB)": 138.1,
      "step": 73080,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.77815399,
      "epoch": 1.7052450458595196,
      "grad_norm": 5.1875,
      "learning_rate": 5.58727528086902e-07,
      "loss": 0.78736992,
      "memory(GiB)": 138.1,
      "step": 73090,
      "train_speed(iter/s)": 0.200623
    },
    {
      "acc": 0.77575245,
      "epoch": 1.7054783534318085,
      "grad_norm": 5.375,
      "learning_rate": 5.578600785243493e-07,
      "loss": 0.81773357,
      "memory(GiB)": 138.1,
      "step": 73100,
      "train_speed(iter/s)": 0.200638
    },
    {
      "acc": 0.79463716,
      "epoch": 1.7057116610040974,
      "grad_norm": 4.3125,
      "learning_rate": 5.569932630759789e-07,
      "loss": 0.74758506,
      "memory(GiB)": 138.1,
      "step": 73110,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.78639936,
      "epoch": 1.7059449685763863,
      "grad_norm": 4.90625,
      "learning_rate": 5.561270818655301e-07,
      "loss": 0.76089525,
      "memory(GiB)": 138.1,
      "step": 73120,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.7882925,
      "epoch": 1.7061782761486752,
      "grad_norm": 10.375,
      "learning_rate": 5.552615350166496e-07,
      "loss": 0.76413431,
      "memory(GiB)": 138.1,
      "step": 73130,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.78352389,
      "epoch": 1.7064115837209641,
      "grad_norm": 5.375,
      "learning_rate": 5.54396622652893e-07,
      "loss": 0.77671185,
      "memory(GiB)": 138.1,
      "step": 73140,
      "train_speed(iter/s)": 0.200694
    },
    {
      "acc": 0.78264346,
      "epoch": 1.706644891293253,
      "grad_norm": 8.0625,
      "learning_rate": 5.535323448977275e-07,
      "loss": 0.76960349,
      "memory(GiB)": 138.1,
      "step": 73150,
      "train_speed(iter/s)": 0.200709
    },
    {
      "acc": 0.8065589,
      "epoch": 1.706878198865542,
      "grad_norm": 6.84375,
      "learning_rate": 5.526687018745286e-07,
      "loss": 0.69353619,
      "memory(GiB)": 138.1,
      "step": 73160,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.78679342,
      "epoch": 1.7071115064378308,
      "grad_norm": 5.0625,
      "learning_rate": 5.518056937065802e-07,
      "loss": 0.77192287,
      "memory(GiB)": 138.1,
      "step": 73170,
      "train_speed(iter/s)": 0.200737
    },
    {
      "acc": 0.7717082,
      "epoch": 1.7073448140101197,
      "grad_norm": 4.15625,
      "learning_rate": 5.509433205170761e-07,
      "loss": 0.82198477,
      "memory(GiB)": 138.1,
      "step": 73180,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.78816519,
      "epoch": 1.7075781215824086,
      "grad_norm": 6.03125,
      "learning_rate": 5.500815824291216e-07,
      "loss": 0.78270626,
      "memory(GiB)": 138.1,
      "step": 73190,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.76701832,
      "epoch": 1.7078114291546975,
      "grad_norm": 6.4375,
      "learning_rate": 5.492204795657274e-07,
      "loss": 0.82847338,
      "memory(GiB)": 138.1,
      "step": 73200,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.7891695,
      "epoch": 1.7080447367269864,
      "grad_norm": 4.59375,
      "learning_rate": 5.483600120498178e-07,
      "loss": 0.74464188,
      "memory(GiB)": 138.1,
      "step": 73210,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.76037254,
      "epoch": 1.7082780442992753,
      "grad_norm": 5.0,
      "learning_rate": 5.475001800042228e-07,
      "loss": 0.87401991,
      "memory(GiB)": 138.1,
      "step": 73220,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.78426857,
      "epoch": 1.7085113518715642,
      "grad_norm": 5.15625,
      "learning_rate": 5.466409835516834e-07,
      "loss": 0.7634037,
      "memory(GiB)": 138.1,
      "step": 73230,
      "train_speed(iter/s)": 0.20082
    },
    {
      "acc": 0.80567493,
      "epoch": 1.7087446594438531,
      "grad_norm": 4.78125,
      "learning_rate": 5.457824228148506e-07,
      "loss": 0.68272343,
      "memory(GiB)": 138.1,
      "step": 73240,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.77797771,
      "epoch": 1.708977967016142,
      "grad_norm": 4.78125,
      "learning_rate": 5.449244979162816e-07,
      "loss": 0.80399952,
      "memory(GiB)": 138.1,
      "step": 73250,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.78294101,
      "epoch": 1.709211274588431,
      "grad_norm": 6.6875,
      "learning_rate": 5.440672089784476e-07,
      "loss": 0.78328276,
      "memory(GiB)": 138.1,
      "step": 73260,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.79575491,
      "epoch": 1.7094445821607196,
      "grad_norm": 5.5625,
      "learning_rate": 5.43210556123725e-07,
      "loss": 0.73883858,
      "memory(GiB)": 138.1,
      "step": 73270,
      "train_speed(iter/s)": 0.200876
    },
    {
      "acc": 0.78026762,
      "epoch": 1.7096778897330087,
      "grad_norm": 5.625,
      "learning_rate": 5.423545394744012e-07,
      "loss": 0.79718757,
      "memory(GiB)": 138.1,
      "step": 73280,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.79239721,
      "epoch": 1.7099111973052974,
      "grad_norm": 5.5,
      "learning_rate": 5.414991591526714e-07,
      "loss": 0.74911137,
      "memory(GiB)": 138.1,
      "step": 73290,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.78526192,
      "epoch": 1.7101445048775865,
      "grad_norm": 9.625,
      "learning_rate": 5.406444152806406e-07,
      "loss": 0.75527325,
      "memory(GiB)": 138.1,
      "step": 73300,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77781644,
      "epoch": 1.7103778124498752,
      "grad_norm": 5.28125,
      "learning_rate": 5.397903079803251e-07,
      "loss": 0.81301641,
      "memory(GiB)": 138.1,
      "step": 73310,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.78796282,
      "epoch": 1.7106111200221643,
      "grad_norm": 6.09375,
      "learning_rate": 5.389368373736464e-07,
      "loss": 0.75768142,
      "memory(GiB)": 138.1,
      "step": 73320,
      "train_speed(iter/s)": 0.200947
    },
    {
      "acc": 0.80338306,
      "epoch": 1.710844427594453,
      "grad_norm": 7.4375,
      "learning_rate": 5.380840035824397e-07,
      "loss": 0.6969276,
      "memory(GiB)": 138.1,
      "step": 73330,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.78724689,
      "epoch": 1.711077735166742,
      "grad_norm": 5.84375,
      "learning_rate": 5.372318067284438e-07,
      "loss": 0.76626501,
      "memory(GiB)": 138.1,
      "step": 73340,
      "train_speed(iter/s)": 0.200976
    },
    {
      "acc": 0.78899498,
      "epoch": 1.7113110427390308,
      "grad_norm": 5.71875,
      "learning_rate": 5.363802469333118e-07,
      "loss": 0.75074024,
      "memory(GiB)": 138.1,
      "step": 73350,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.77315474,
      "epoch": 1.71154435031132,
      "grad_norm": 6.8125,
      "learning_rate": 5.355293243186033e-07,
      "loss": 0.82559099,
      "memory(GiB)": 138.1,
      "step": 73360,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.77842388,
      "epoch": 1.7117776578836086,
      "grad_norm": 6.96875,
      "learning_rate": 5.34679039005786e-07,
      "loss": 0.8188942,
      "memory(GiB)": 138.1,
      "step": 73370,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.76217799,
      "epoch": 1.7120109654558977,
      "grad_norm": 5.8125,
      "learning_rate": 5.338293911162401e-07,
      "loss": 0.85048714,
      "memory(GiB)": 138.1,
      "step": 73380,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.76450157,
      "epoch": 1.7122442730281864,
      "grad_norm": 7.03125,
      "learning_rate": 5.329803807712497e-07,
      "loss": 0.85439548,
      "memory(GiB)": 138.1,
      "step": 73390,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.78162451,
      "epoch": 1.7124775806004755,
      "grad_norm": 7.53125,
      "learning_rate": 5.321320080920128e-07,
      "loss": 0.78223362,
      "memory(GiB)": 138.1,
      "step": 73400,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.77304645,
      "epoch": 1.7127108881727642,
      "grad_norm": 5.15625,
      "learning_rate": 5.312842731996332e-07,
      "loss": 0.82387562,
      "memory(GiB)": 138.1,
      "step": 73410,
      "train_speed(iter/s)": 0.201076
    },
    {
      "acc": 0.77667975,
      "epoch": 1.7129441957450533,
      "grad_norm": 4.03125,
      "learning_rate": 5.304371762151261e-07,
      "loss": 0.80046349,
      "memory(GiB)": 138.1,
      "step": 73420,
      "train_speed(iter/s)": 0.20109
    },
    {
      "acc": 0.79003797,
      "epoch": 1.713177503317342,
      "grad_norm": 10.625,
      "learning_rate": 5.295907172594139e-07,
      "loss": 0.75672364,
      "memory(GiB)": 138.1,
      "step": 73430,
      "train_speed(iter/s)": 0.201105
    },
    {
      "acc": 0.79200077,
      "epoch": 1.713410810889631,
      "grad_norm": 5.71875,
      "learning_rate": 5.287448964533276e-07,
      "loss": 0.7386096,
      "memory(GiB)": 138.1,
      "step": 73440,
      "train_speed(iter/s)": 0.201119
    },
    {
      "acc": 0.77990046,
      "epoch": 1.7136441184619198,
      "grad_norm": 6.75,
      "learning_rate": 5.278997139176084e-07,
      "loss": 0.79957833,
      "memory(GiB)": 138.1,
      "step": 73450,
      "train_speed(iter/s)": 0.201132
    },
    {
      "acc": 0.74390507,
      "epoch": 1.7138774260342087,
      "grad_norm": 4.1875,
      "learning_rate": 5.270551697729059e-07,
      "loss": 0.94712648,
      "memory(GiB)": 138.1,
      "step": 73460,
      "train_speed(iter/s)": 0.201147
    },
    {
      "acc": 0.77517686,
      "epoch": 1.7141107336064976,
      "grad_norm": 4.875,
      "learning_rate": 5.262112641397788e-07,
      "loss": 0.80620651,
      "memory(GiB)": 138.1,
      "step": 73470,
      "train_speed(iter/s)": 0.201161
    },
    {
      "acc": 0.76355195,
      "epoch": 1.7143440411787865,
      "grad_norm": 7.25,
      "learning_rate": 5.253679971386932e-07,
      "loss": 0.84303713,
      "memory(GiB)": 138.1,
      "step": 73480,
      "train_speed(iter/s)": 0.201173
    },
    {
      "acc": 0.78939095,
      "epoch": 1.7145773487510754,
      "grad_norm": 5.09375,
      "learning_rate": 5.245253688900287e-07,
      "loss": 0.75973749,
      "memory(GiB)": 138.1,
      "step": 73490,
      "train_speed(iter/s)": 0.201187
    },
    {
      "acc": 0.78993163,
      "epoch": 1.7148106563233643,
      "grad_norm": 5.375,
      "learning_rate": 5.236833795140651e-07,
      "loss": 0.73632402,
      "memory(GiB)": 138.1,
      "step": 73500,
      "train_speed(iter/s)": 0.201202
    },
    {
      "epoch": 1.7148106563233643,
      "eval_acc": 0.7447105561152596,
      "eval_loss": 0.8043909072875977,
      "eval_runtime": 1271.1227,
      "eval_samples_per_second": 28.314,
      "eval_steps_per_second": 14.158,
      "step": 73500
    },
    {
      "acc": 0.79015398,
      "epoch": 1.7150439638956532,
      "grad_norm": 5.1875,
      "learning_rate": 5.228420291309999e-07,
      "loss": 0.74385405,
      "memory(GiB)": 138.1,
      "step": 73510,
      "train_speed(iter/s)": 0.200504
    },
    {
      "acc": 0.7631135,
      "epoch": 1.715277271467942,
      "grad_norm": 5.5625,
      "learning_rate": 5.22001317860934e-07,
      "loss": 0.84219952,
      "memory(GiB)": 138.1,
      "step": 73520,
      "train_speed(iter/s)": 0.200519
    },
    {
      "acc": 0.76704111,
      "epoch": 1.715510579040231,
      "grad_norm": 3.984375,
      "learning_rate": 5.211612458238785e-07,
      "loss": 0.87439299,
      "memory(GiB)": 138.1,
      "step": 73530,
      "train_speed(iter/s)": 0.200533
    },
    {
      "acc": 0.79748411,
      "epoch": 1.7157438866125199,
      "grad_norm": 6.0625,
      "learning_rate": 5.203218131397553e-07,
      "loss": 0.73548927,
      "memory(GiB)": 138.1,
      "step": 73540,
      "train_speed(iter/s)": 0.200547
    },
    {
      "acc": 0.77375269,
      "epoch": 1.7159771941848088,
      "grad_norm": 4.875,
      "learning_rate": 5.194830199283907e-07,
      "loss": 0.80748701,
      "memory(GiB)": 138.1,
      "step": 73550,
      "train_speed(iter/s)": 0.200561
    },
    {
      "acc": 0.77218609,
      "epoch": 1.7162105017570977,
      "grad_norm": 6.5,
      "learning_rate": 5.18644866309524e-07,
      "loss": 0.81236382,
      "memory(GiB)": 138.1,
      "step": 73560,
      "train_speed(iter/s)": 0.200575
    },
    {
      "acc": 0.79794626,
      "epoch": 1.7164438093293866,
      "grad_norm": 3.90625,
      "learning_rate": 5.178073524028016e-07,
      "loss": 0.71432037,
      "memory(GiB)": 138.1,
      "step": 73570,
      "train_speed(iter/s)": 0.200589
    },
    {
      "acc": 0.75880527,
      "epoch": 1.7166771169016755,
      "grad_norm": 6.28125,
      "learning_rate": 5.169704783277756e-07,
      "loss": 0.84431038,
      "memory(GiB)": 138.1,
      "step": 73580,
      "train_speed(iter/s)": 0.200603
    },
    {
      "acc": 0.78814268,
      "epoch": 1.7169104244739644,
      "grad_norm": 5.375,
      "learning_rate": 5.16134244203913e-07,
      "loss": 0.75521536,
      "memory(GiB)": 138.1,
      "step": 73590,
      "train_speed(iter/s)": 0.200618
    },
    {
      "acc": 0.78718247,
      "epoch": 1.7171437320462533,
      "grad_norm": 6.625,
      "learning_rate": 5.152986501505835e-07,
      "loss": 0.76596746,
      "memory(GiB)": 138.1,
      "step": 73600,
      "train_speed(iter/s)": 0.200633
    },
    {
      "acc": 0.77716827,
      "epoch": 1.7173770396185422,
      "grad_norm": 3.65625,
      "learning_rate": 5.144636962870708e-07,
      "loss": 0.80403767,
      "memory(GiB)": 138.1,
      "step": 73610,
      "train_speed(iter/s)": 0.200646
    },
    {
      "acc": 0.75140824,
      "epoch": 1.717610347190831,
      "grad_norm": 5.5625,
      "learning_rate": 5.136293827325606e-07,
      "loss": 0.89990616,
      "memory(GiB)": 138.1,
      "step": 73620,
      "train_speed(iter/s)": 0.20066
    },
    {
      "acc": 0.77322063,
      "epoch": 1.71784365476312,
      "grad_norm": 6.125,
      "learning_rate": 5.127957096061537e-07,
      "loss": 0.83046427,
      "memory(GiB)": 138.1,
      "step": 73630,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.76129708,
      "epoch": 1.7180769623354089,
      "grad_norm": 5.90625,
      "learning_rate": 5.119626770268543e-07,
      "loss": 0.86177006,
      "memory(GiB)": 138.1,
      "step": 73640,
      "train_speed(iter/s)": 0.200688
    },
    {
      "acc": 0.78243065,
      "epoch": 1.7183102699076978,
      "grad_norm": 3.5625,
      "learning_rate": 5.111302851135802e-07,
      "loss": 0.79324946,
      "memory(GiB)": 138.1,
      "step": 73650,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.79360733,
      "epoch": 1.7185435774799864,
      "grad_norm": 4.84375,
      "learning_rate": 5.10298533985154e-07,
      "loss": 0.74446688,
      "memory(GiB)": 138.1,
      "step": 73660,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.78374872,
      "epoch": 1.7187768850522755,
      "grad_norm": 5.25,
      "learning_rate": 5.09467423760307e-07,
      "loss": 0.74643164,
      "memory(GiB)": 138.1,
      "step": 73670,
      "train_speed(iter/s)": 0.200731
    },
    {
      "acc": 0.78896942,
      "epoch": 1.7190101926245642,
      "grad_norm": 4.84375,
      "learning_rate": 5.086369545576814e-07,
      "loss": 0.75824614,
      "memory(GiB)": 138.1,
      "step": 73680,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.78774047,
      "epoch": 1.7192435001968533,
      "grad_norm": 5.375,
      "learning_rate": 5.078071264958245e-07,
      "loss": 0.74300013,
      "memory(GiB)": 138.1,
      "step": 73690,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.79266844,
      "epoch": 1.719476807769142,
      "grad_norm": 4.125,
      "learning_rate": 5.069779396931967e-07,
      "loss": 0.74105115,
      "memory(GiB)": 138.1,
      "step": 73700,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.76223621,
      "epoch": 1.7197101153414311,
      "grad_norm": 13.625,
      "learning_rate": 5.061493942681628e-07,
      "loss": 0.88398075,
      "memory(GiB)": 138.1,
      "step": 73710,
      "train_speed(iter/s)": 0.200788
    },
    {
      "acc": 0.77189965,
      "epoch": 1.7199434229137198,
      "grad_norm": 5.71875,
      "learning_rate": 5.053214903389975e-07,
      "loss": 0.81558266,
      "memory(GiB)": 138.1,
      "step": 73720,
      "train_speed(iter/s)": 0.200802
    },
    {
      "acc": 0.78657942,
      "epoch": 1.720176730486009,
      "grad_norm": 5.71875,
      "learning_rate": 5.044942280238835e-07,
      "loss": 0.7773448,
      "memory(GiB)": 138.1,
      "step": 73730,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.76862726,
      "epoch": 1.7204100380582976,
      "grad_norm": 4.40625,
      "learning_rate": 5.036676074409114e-07,
      "loss": 0.85371304,
      "memory(GiB)": 138.1,
      "step": 73740,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.77896681,
      "epoch": 1.7206433456305867,
      "grad_norm": 7.53125,
      "learning_rate": 5.028416287080834e-07,
      "loss": 0.7777874,
      "memory(GiB)": 138.1,
      "step": 73750,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.77873478,
      "epoch": 1.7208766532028754,
      "grad_norm": 5.28125,
      "learning_rate": 5.020162919433053e-07,
      "loss": 0.79268093,
      "memory(GiB)": 138.1,
      "step": 73760,
      "train_speed(iter/s)": 0.200858
    },
    {
      "acc": 0.78746619,
      "epoch": 1.7211099607751645,
      "grad_norm": 7.21875,
      "learning_rate": 5.01191597264396e-07,
      "loss": 0.75564404,
      "memory(GiB)": 138.1,
      "step": 73770,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.78620825,
      "epoch": 1.7213432683474532,
      "grad_norm": 6.28125,
      "learning_rate": 5.003675447890782e-07,
      "loss": 0.75644531,
      "memory(GiB)": 138.1,
      "step": 73780,
      "train_speed(iter/s)": 0.200885
    },
    {
      "acc": 0.76999698,
      "epoch": 1.7215765759197423,
      "grad_norm": 4.8125,
      "learning_rate": 4.995441346349872e-07,
      "loss": 0.83167076,
      "memory(GiB)": 138.1,
      "step": 73790,
      "train_speed(iter/s)": 0.200899
    },
    {
      "acc": 0.80347652,
      "epoch": 1.721809883492031,
      "grad_norm": 4.28125,
      "learning_rate": 4.987213669196639e-07,
      "loss": 0.68127127,
      "memory(GiB)": 138.1,
      "step": 73800,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.79354105,
      "epoch": 1.7220431910643201,
      "grad_norm": 4.5625,
      "learning_rate": 4.978992417605566e-07,
      "loss": 0.73909149,
      "memory(GiB)": 138.1,
      "step": 73810,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.8010478,
      "epoch": 1.7222764986366088,
      "grad_norm": 5.46875,
      "learning_rate": 4.970777592750253e-07,
      "loss": 0.69659333,
      "memory(GiB)": 138.1,
      "step": 73820,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.76971216,
      "epoch": 1.722509806208898,
      "grad_norm": 5.09375,
      "learning_rate": 4.962569195803352e-07,
      "loss": 0.8232933,
      "memory(GiB)": 138.1,
      "step": 73830,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.77865553,
      "epoch": 1.7227431137811866,
      "grad_norm": 4.15625,
      "learning_rate": 4.954367227936635e-07,
      "loss": 0.7898417,
      "memory(GiB)": 138.1,
      "step": 73840,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.77204628,
      "epoch": 1.7229764213534755,
      "grad_norm": 6.71875,
      "learning_rate": 4.946171690320889e-07,
      "loss": 0.82885933,
      "memory(GiB)": 138.1,
      "step": 73850,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.79157839,
      "epoch": 1.7232097289257644,
      "grad_norm": 5.5,
      "learning_rate": 4.937982584126055e-07,
      "loss": 0.73304939,
      "memory(GiB)": 138.1,
      "step": 73860,
      "train_speed(iter/s)": 0.200995
    },
    {
      "acc": 0.7956171,
      "epoch": 1.7234430364980533,
      "grad_norm": 5.5625,
      "learning_rate": 4.929799910521116e-07,
      "loss": 0.73145752,
      "memory(GiB)": 138.1,
      "step": 73870,
      "train_speed(iter/s)": 0.201009
    },
    {
      "acc": 0.76525025,
      "epoch": 1.7236763440703422,
      "grad_norm": 4.96875,
      "learning_rate": 4.921623670674142e-07,
      "loss": 0.84171267,
      "memory(GiB)": 138.1,
      "step": 73880,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.80335445,
      "epoch": 1.723909651642631,
      "grad_norm": 4.9375,
      "learning_rate": 4.913453865752299e-07,
      "loss": 0.69952669,
      "memory(GiB)": 138.1,
      "step": 73890,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.76818728,
      "epoch": 1.72414295921492,
      "grad_norm": 11.875,
      "learning_rate": 4.905290496921811e-07,
      "loss": 0.84597569,
      "memory(GiB)": 138.1,
      "step": 73900,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.79364624,
      "epoch": 1.724376266787209,
      "grad_norm": 5.875,
      "learning_rate": 4.897133565348012e-07,
      "loss": 0.73016977,
      "memory(GiB)": 138.1,
      "step": 73910,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.79641161,
      "epoch": 1.7246095743594978,
      "grad_norm": 3.640625,
      "learning_rate": 4.888983072195296e-07,
      "loss": 0.72878189,
      "memory(GiB)": 138.1,
      "step": 73920,
      "train_speed(iter/s)": 0.201082
    },
    {
      "acc": 0.7892786,
      "epoch": 1.7248428819317867,
      "grad_norm": 5.34375,
      "learning_rate": 4.880839018627132e-07,
      "loss": 0.74224253,
      "memory(GiB)": 138.1,
      "step": 73930,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.78093824,
      "epoch": 1.7250761895040756,
      "grad_norm": 5.65625,
      "learning_rate": 4.872701405806101e-07,
      "loss": 0.77963114,
      "memory(GiB)": 138.1,
      "step": 73940,
      "train_speed(iter/s)": 0.20111
    },
    {
      "acc": 0.77672095,
      "epoch": 1.7253094970763645,
      "grad_norm": 7.0625,
      "learning_rate": 4.864570234893834e-07,
      "loss": 0.78376875,
      "memory(GiB)": 138.1,
      "step": 73950,
      "train_speed(iter/s)": 0.201123
    },
    {
      "acc": 0.78407612,
      "epoch": 1.7255428046486534,
      "grad_norm": 6.125,
      "learning_rate": 4.856445507051049e-07,
      "loss": 0.77355375,
      "memory(GiB)": 138.1,
      "step": 73960,
      "train_speed(iter/s)": 0.201137
    },
    {
      "acc": 0.7757288,
      "epoch": 1.7257761122209423,
      "grad_norm": 4.5625,
      "learning_rate": 4.848327223437554e-07,
      "loss": 0.80582609,
      "memory(GiB)": 138.1,
      "step": 73970,
      "train_speed(iter/s)": 0.201151
    },
    {
      "acc": 0.76491833,
      "epoch": 1.7260094197932312,
      "grad_norm": 6.34375,
      "learning_rate": 4.840215385212232e-07,
      "loss": 0.86212664,
      "memory(GiB)": 138.1,
      "step": 73980,
      "train_speed(iter/s)": 0.201165
    },
    {
      "acc": 0.77588701,
      "epoch": 1.72624272736552,
      "grad_norm": 7.25,
      "learning_rate": 4.832109993533041e-07,
      "loss": 0.80320539,
      "memory(GiB)": 138.1,
      "step": 73990,
      "train_speed(iter/s)": 0.201178
    },
    {
      "acc": 0.7706954,
      "epoch": 1.726476034937809,
      "grad_norm": 5.3125,
      "learning_rate": 4.824011049557037e-07,
      "loss": 0.83672714,
      "memory(GiB)": 138.1,
      "step": 74000,
      "train_speed(iter/s)": 0.201192
    },
    {
      "epoch": 1.726476034937809,
      "eval_acc": 0.7447166415960871,
      "eval_loss": 0.8044154047966003,
      "eval_runtime": 1272.3059,
      "eval_samples_per_second": 28.288,
      "eval_steps_per_second": 14.144,
      "step": 74000
    },
    {
      "acc": 0.77697601,
      "epoch": 1.7267093425100979,
      "grad_norm": 7.1875,
      "learning_rate": 4.815918554440324e-07,
      "loss": 0.802526,
      "memory(GiB)": 138.1,
      "step": 74010,
      "train_speed(iter/s)": 0.200499
    },
    {
      "acc": 0.78431935,
      "epoch": 1.7269426500823868,
      "grad_norm": 4.71875,
      "learning_rate": 4.807832509338112e-07,
      "loss": 0.771733,
      "memory(GiB)": 138.1,
      "step": 74020,
      "train_speed(iter/s)": 0.200512
    },
    {
      "acc": 0.78355474,
      "epoch": 1.7271759576546755,
      "grad_norm": 4.40625,
      "learning_rate": 4.799752915404682e-07,
      "loss": 0.77948675,
      "memory(GiB)": 138.1,
      "step": 74030,
      "train_speed(iter/s)": 0.200525
    },
    {
      "acc": 0.77410603,
      "epoch": 1.7274092652269646,
      "grad_norm": 4.9375,
      "learning_rate": 4.791679773793389e-07,
      "loss": 0.82328606,
      "memory(GiB)": 138.1,
      "step": 74040,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.77805958,
      "epoch": 1.7276425727992533,
      "grad_norm": 5.96875,
      "learning_rate": 4.78361308565668e-07,
      "loss": 0.80983696,
      "memory(GiB)": 138.1,
      "step": 74050,
      "train_speed(iter/s)": 0.200553
    },
    {
      "acc": 0.76219211,
      "epoch": 1.7278758803715424,
      "grad_norm": 8.875,
      "learning_rate": 4.77555285214607e-07,
      "loss": 0.85069485,
      "memory(GiB)": 138.1,
      "step": 74060,
      "train_speed(iter/s)": 0.200567
    },
    {
      "acc": 0.77103195,
      "epoch": 1.728109187943831,
      "grad_norm": 3.90625,
      "learning_rate": 4.767499074412152e-07,
      "loss": 0.82191906,
      "memory(GiB)": 138.1,
      "step": 74070,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.80303154,
      "epoch": 1.7283424955161202,
      "grad_norm": 7.125,
      "learning_rate": 4.7594517536045936e-07,
      "loss": 0.72213902,
      "memory(GiB)": 138.1,
      "step": 74080,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.76572065,
      "epoch": 1.7285758030884089,
      "grad_norm": 6.1875,
      "learning_rate": 4.751410890872166e-07,
      "loss": 0.85232534,
      "memory(GiB)": 138.1,
      "step": 74090,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.77745919,
      "epoch": 1.728809110660698,
      "grad_norm": 5.53125,
      "learning_rate": 4.743376487362683e-07,
      "loss": 0.81853027,
      "memory(GiB)": 138.1,
      "step": 74100,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.79782553,
      "epoch": 1.7290424182329867,
      "grad_norm": 5.03125,
      "learning_rate": 4.7353485442230507e-07,
      "loss": 0.70489302,
      "memory(GiB)": 138.1,
      "step": 74110,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.77856512,
      "epoch": 1.7292757258052758,
      "grad_norm": 11.5,
      "learning_rate": 4.7273270625992794e-07,
      "loss": 0.80788822,
      "memory(GiB)": 138.1,
      "step": 74120,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.76892738,
      "epoch": 1.7295090333775645,
      "grad_norm": 6.59375,
      "learning_rate": 4.719312043636404e-07,
      "loss": 0.83906326,
      "memory(GiB)": 138.1,
      "step": 74130,
      "train_speed(iter/s)": 0.200663
    },
    {
      "acc": 0.77776084,
      "epoch": 1.7297423409498536,
      "grad_norm": 6.0,
      "learning_rate": 4.711303488478591e-07,
      "loss": 0.78263907,
      "memory(GiB)": 138.1,
      "step": 74140,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.7878387,
      "epoch": 1.7299756485221423,
      "grad_norm": 5.28125,
      "learning_rate": 4.7033013982690446e-07,
      "loss": 0.76044273,
      "memory(GiB)": 138.1,
      "step": 74150,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.77629614,
      "epoch": 1.7302089560944314,
      "grad_norm": 7.53125,
      "learning_rate": 4.695305774150061e-07,
      "loss": 0.79856586,
      "memory(GiB)": 138.1,
      "step": 74160,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.7986949,
      "epoch": 1.73044226366672,
      "grad_norm": 6.40625,
      "learning_rate": 4.687316617263027e-07,
      "loss": 0.73556232,
      "memory(GiB)": 138.1,
      "step": 74170,
      "train_speed(iter/s)": 0.20072
    },
    {
      "acc": 0.79174337,
      "epoch": 1.7306755712390092,
      "grad_norm": 6.375,
      "learning_rate": 4.679333928748375e-07,
      "loss": 0.73341341,
      "memory(GiB)": 138.1,
      "step": 74180,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.77004156,
      "epoch": 1.7309088788112978,
      "grad_norm": 7.09375,
      "learning_rate": 4.671357709745644e-07,
      "loss": 0.83492775,
      "memory(GiB)": 138.1,
      "step": 74190,
      "train_speed(iter/s)": 0.200748
    },
    {
      "acc": 0.79907241,
      "epoch": 1.731142186383587,
      "grad_norm": 5.375,
      "learning_rate": 4.6633879613934227e-07,
      "loss": 0.6985404,
      "memory(GiB)": 138.1,
      "step": 74200,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.75945568,
      "epoch": 1.7313754939558756,
      "grad_norm": 4.78125,
      "learning_rate": 4.6554246848294127e-07,
      "loss": 0.86438541,
      "memory(GiB)": 138.1,
      "step": 74210,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.78431191,
      "epoch": 1.7316088015281648,
      "grad_norm": 5.1875,
      "learning_rate": 4.6474678811903483e-07,
      "loss": 0.77434368,
      "memory(GiB)": 138.1,
      "step": 74220,
      "train_speed(iter/s)": 0.200791
    },
    {
      "acc": 0.7858922,
      "epoch": 1.7318421091004534,
      "grad_norm": 6.09375,
      "learning_rate": 4.639517551612066e-07,
      "loss": 0.74613075,
      "memory(GiB)": 138.1,
      "step": 74230,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.78975163,
      "epoch": 1.7320754166727423,
      "grad_norm": 5.65625,
      "learning_rate": 4.63157369722948e-07,
      "loss": 0.74640374,
      "memory(GiB)": 138.1,
      "step": 74240,
      "train_speed(iter/s)": 0.200817
    },
    {
      "acc": 0.78697009,
      "epoch": 1.7323087242450312,
      "grad_norm": 4.96875,
      "learning_rate": 4.6236363191765666e-07,
      "loss": 0.76076813,
      "memory(GiB)": 138.1,
      "step": 74250,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.774368,
      "epoch": 1.7325420318173201,
      "grad_norm": 6.96875,
      "learning_rate": 4.615705418586391e-07,
      "loss": 0.84603786,
      "memory(GiB)": 138.1,
      "step": 74260,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.78506193,
      "epoch": 1.732775339389609,
      "grad_norm": 6.28125,
      "learning_rate": 4.6077809965910716e-07,
      "loss": 0.76752243,
      "memory(GiB)": 138.1,
      "step": 74270,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.77089567,
      "epoch": 1.733008646961898,
      "grad_norm": 4.875,
      "learning_rate": 4.59986305432184e-07,
      "loss": 0.81474075,
      "memory(GiB)": 138.1,
      "step": 74280,
      "train_speed(iter/s)": 0.200873
    },
    {
      "acc": 0.7874424,
      "epoch": 1.7332419545341868,
      "grad_norm": 6.3125,
      "learning_rate": 4.591951592908972e-07,
      "loss": 0.76511889,
      "memory(GiB)": 138.1,
      "step": 74290,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.77571363,
      "epoch": 1.7334752621064757,
      "grad_norm": 4.1875,
      "learning_rate": 4.5840466134818184e-07,
      "loss": 0.8134244,
      "memory(GiB)": 138.1,
      "step": 74300,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.77631946,
      "epoch": 1.7337085696787646,
      "grad_norm": 5.90625,
      "learning_rate": 4.576148117168816e-07,
      "loss": 0.78171606,
      "memory(GiB)": 138.1,
      "step": 74310,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.79241109,
      "epoch": 1.7339418772510535,
      "grad_norm": 8.0,
      "learning_rate": 4.568256105097468e-07,
      "loss": 0.72888813,
      "memory(GiB)": 138.1,
      "step": 74320,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.78169894,
      "epoch": 1.7341751848233424,
      "grad_norm": 4.9375,
      "learning_rate": 4.560370578394374e-07,
      "loss": 0.76154728,
      "memory(GiB)": 138.1,
      "step": 74330,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.79366865,
      "epoch": 1.7344084923956313,
      "grad_norm": 5.875,
      "learning_rate": 4.5524915381851663e-07,
      "loss": 0.74610434,
      "memory(GiB)": 138.1,
      "step": 74340,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.78711934,
      "epoch": 1.7346417999679202,
      "grad_norm": 7.71875,
      "learning_rate": 4.544618985594601e-07,
      "loss": 0.75808601,
      "memory(GiB)": 138.1,
      "step": 74350,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.7786828,
      "epoch": 1.7348751075402091,
      "grad_norm": 5.40625,
      "learning_rate": 4.5367529217464733e-07,
      "loss": 0.78535357,
      "memory(GiB)": 138.1,
      "step": 74360,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.78281074,
      "epoch": 1.735108415112498,
      "grad_norm": 5.78125,
      "learning_rate": 4.5288933477636466e-07,
      "loss": 0.77032428,
      "memory(GiB)": 138.1,
      "step": 74370,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.77319117,
      "epoch": 1.735341722684787,
      "grad_norm": 5.96875,
      "learning_rate": 4.5210402647680895e-07,
      "loss": 0.81330729,
      "memory(GiB)": 138.1,
      "step": 74380,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.78810053,
      "epoch": 1.7355750302570758,
      "grad_norm": 4.8125,
      "learning_rate": 4.5131936738808225e-07,
      "loss": 0.76743989,
      "memory(GiB)": 138.1,
      "step": 74390,
      "train_speed(iter/s)": 0.201028
    },
    {
      "acc": 0.79243145,
      "epoch": 1.7358083378293647,
      "grad_norm": 6.375,
      "learning_rate": 4.5053535762219494e-07,
      "loss": 0.74746494,
      "memory(GiB)": 138.1,
      "step": 74400,
      "train_speed(iter/s)": 0.201042
    },
    {
      "acc": 0.7838623,
      "epoch": 1.7360416454016536,
      "grad_norm": 5.4375,
      "learning_rate": 4.4975199729106355e-07,
      "loss": 0.78794641,
      "memory(GiB)": 138.1,
      "step": 74410,
      "train_speed(iter/s)": 0.201056
    },
    {
      "acc": 0.8007185,
      "epoch": 1.7362749529739423,
      "grad_norm": 6.4375,
      "learning_rate": 4.489692865065126e-07,
      "loss": 0.72447138,
      "memory(GiB)": 138.1,
      "step": 74420,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.78932304,
      "epoch": 1.7365082605462314,
      "grad_norm": 4.125,
      "learning_rate": 4.4818722538027326e-07,
      "loss": 0.75672827,
      "memory(GiB)": 138.1,
      "step": 74430,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.78215332,
      "epoch": 1.73674156811852,
      "grad_norm": 4.75,
      "learning_rate": 4.474058140239862e-07,
      "loss": 0.77884374,
      "memory(GiB)": 138.1,
      "step": 74440,
      "train_speed(iter/s)": 0.2011
    },
    {
      "acc": 0.78150392,
      "epoch": 1.7369748756908092,
      "grad_norm": 6.09375,
      "learning_rate": 4.4662505254919665e-07,
      "loss": 0.75879669,
      "memory(GiB)": 138.1,
      "step": 74450,
      "train_speed(iter/s)": 0.201114
    },
    {
      "acc": 0.79613762,
      "epoch": 1.737208183263098,
      "grad_norm": 5.8125,
      "learning_rate": 4.4584494106735707e-07,
      "loss": 0.73280921,
      "memory(GiB)": 138.1,
      "step": 74460,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.78308897,
      "epoch": 1.737441490835387,
      "grad_norm": 5.84375,
      "learning_rate": 4.4506547968983016e-07,
      "loss": 0.78369331,
      "memory(GiB)": 138.1,
      "step": 74470,
      "train_speed(iter/s)": 0.201143
    },
    {
      "acc": 0.77581263,
      "epoch": 1.7376747984076757,
      "grad_norm": 8.125,
      "learning_rate": 4.4428666852788126e-07,
      "loss": 0.80782318,
      "memory(GiB)": 138.1,
      "step": 74480,
      "train_speed(iter/s)": 0.201157
    },
    {
      "acc": 0.76641207,
      "epoch": 1.7379081059799648,
      "grad_norm": 5.375,
      "learning_rate": 4.4350850769268874e-07,
      "loss": 0.84937601,
      "memory(GiB)": 138.1,
      "step": 74490,
      "train_speed(iter/s)": 0.20117
    },
    {
      "acc": 0.78670392,
      "epoch": 1.7381414135522535,
      "grad_norm": 4.65625,
      "learning_rate": 4.4273099729533255e-07,
      "loss": 0.76552391,
      "memory(GiB)": 138.1,
      "step": 74500,
      "train_speed(iter/s)": 0.201183
    },
    {
      "epoch": 1.7381414135522535,
      "eval_acc": 0.7446762852495471,
      "eval_loss": 0.804404079914093,
      "eval_runtime": 1271.2914,
      "eval_samples_per_second": 28.311,
      "eval_steps_per_second": 14.156,
      "step": 74500
    },
    {
      "acc": 0.7867198,
      "epoch": 1.7383747211245426,
      "grad_norm": 4.5,
      "learning_rate": 4.419541374468023e-07,
      "loss": 0.76824493,
      "memory(GiB)": 138.1,
      "step": 74510,
      "train_speed(iter/s)": 0.200496
    },
    {
      "acc": 0.77137146,
      "epoch": 1.7386080286968313,
      "grad_norm": 5.75,
      "learning_rate": 4.411779282579959e-07,
      "loss": 0.82607307,
      "memory(GiB)": 138.1,
      "step": 74520,
      "train_speed(iter/s)": 0.200509
    },
    {
      "acc": 0.78419704,
      "epoch": 1.7388413362691204,
      "grad_norm": 4.25,
      "learning_rate": 4.4040236983971476e-07,
      "loss": 0.76362429,
      "memory(GiB)": 138.1,
      "step": 74530,
      "train_speed(iter/s)": 0.200523
    },
    {
      "acc": 0.78158565,
      "epoch": 1.739074643841409,
      "grad_norm": 4.8125,
      "learning_rate": 4.3962746230267084e-07,
      "loss": 0.77455659,
      "memory(GiB)": 138.1,
      "step": 74540,
      "train_speed(iter/s)": 0.200537
    },
    {
      "acc": 0.7929049,
      "epoch": 1.7393079514136982,
      "grad_norm": 6.9375,
      "learning_rate": 4.388532057574818e-07,
      "loss": 0.73814707,
      "memory(GiB)": 138.1,
      "step": 74550,
      "train_speed(iter/s)": 0.200551
    },
    {
      "acc": 0.79585075,
      "epoch": 1.7395412589859869,
      "grad_norm": 4.21875,
      "learning_rate": 4.380796003146731e-07,
      "loss": 0.71769018,
      "memory(GiB)": 138.1,
      "step": 74560,
      "train_speed(iter/s)": 0.200565
    },
    {
      "acc": 0.76547136,
      "epoch": 1.739774566558276,
      "grad_norm": 4.71875,
      "learning_rate": 4.3730664608467534e-07,
      "loss": 0.81900177,
      "memory(GiB)": 138.1,
      "step": 74570,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.77250733,
      "epoch": 1.7400078741305647,
      "grad_norm": 10.375,
      "learning_rate": 4.3653434317782905e-07,
      "loss": 0.79657941,
      "memory(GiB)": 138.1,
      "step": 74580,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.77615132,
      "epoch": 1.7402411817028538,
      "grad_norm": 5.3125,
      "learning_rate": 4.3576269170438e-07,
      "loss": 0.81013956,
      "memory(GiB)": 138.1,
      "step": 74590,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.80237694,
      "epoch": 1.7404744892751425,
      "grad_norm": 5.53125,
      "learning_rate": 4.349916917744801e-07,
      "loss": 0.69212132,
      "memory(GiB)": 138.1,
      "step": 74600,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.77684078,
      "epoch": 1.7407077968474314,
      "grad_norm": 5.1875,
      "learning_rate": 4.3422134349819014e-07,
      "loss": 0.82745504,
      "memory(GiB)": 138.1,
      "step": 74610,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.7814332,
      "epoch": 1.7409411044197203,
      "grad_norm": 6.59375,
      "learning_rate": 4.334516469854766e-07,
      "loss": 0.75724936,
      "memory(GiB)": 138.1,
      "step": 74620,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.76991339,
      "epoch": 1.7411744119920092,
      "grad_norm": 7.21875,
      "learning_rate": 4.3268260234621497e-07,
      "loss": 0.82924404,
      "memory(GiB)": 138.1,
      "step": 74630,
      "train_speed(iter/s)": 0.200663
    },
    {
      "acc": 0.78513436,
      "epoch": 1.741407719564298,
      "grad_norm": 6.34375,
      "learning_rate": 4.319142096901846e-07,
      "loss": 0.76789331,
      "memory(GiB)": 138.1,
      "step": 74640,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.77668796,
      "epoch": 1.741641027136587,
      "grad_norm": 5.15625,
      "learning_rate": 4.3114646912707394e-07,
      "loss": 0.8108284,
      "memory(GiB)": 138.1,
      "step": 74650,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.7892621,
      "epoch": 1.7418743347088759,
      "grad_norm": 5.0625,
      "learning_rate": 4.30379380766478e-07,
      "loss": 0.76283331,
      "memory(GiB)": 138.1,
      "step": 74660,
      "train_speed(iter/s)": 0.200705
    },
    {
      "acc": 0.80962782,
      "epoch": 1.7421076422811648,
      "grad_norm": 4.96875,
      "learning_rate": 4.296129447178965e-07,
      "loss": 0.68484173,
      "memory(GiB)": 138.1,
      "step": 74670,
      "train_speed(iter/s)": 0.200718
    },
    {
      "acc": 0.78058996,
      "epoch": 1.7423409498534537,
      "grad_norm": 5.53125,
      "learning_rate": 4.288471610907402e-07,
      "loss": 0.80650549,
      "memory(GiB)": 138.1,
      "step": 74680,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.77258387,
      "epoch": 1.7425742574257426,
      "grad_norm": 7.625,
      "learning_rate": 4.2808202999432335e-07,
      "loss": 0.80087252,
      "memory(GiB)": 138.1,
      "step": 74690,
      "train_speed(iter/s)": 0.200746
    },
    {
      "acc": 0.77971539,
      "epoch": 1.7428075649980315,
      "grad_norm": 4.90625,
      "learning_rate": 4.2731755153786915e-07,
      "loss": 0.79446883,
      "memory(GiB)": 138.1,
      "step": 74700,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.78136635,
      "epoch": 1.7430408725703204,
      "grad_norm": 4.4375,
      "learning_rate": 4.2655372583050525e-07,
      "loss": 0.78863945,
      "memory(GiB)": 138.1,
      "step": 74710,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.78437352,
      "epoch": 1.7432741801426093,
      "grad_norm": 4.15625,
      "learning_rate": 4.2579055298126945e-07,
      "loss": 0.7832253,
      "memory(GiB)": 138.1,
      "step": 74720,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.77852464,
      "epoch": 1.7435074877148982,
      "grad_norm": 5.1875,
      "learning_rate": 4.2502803309910354e-07,
      "loss": 0.79541359,
      "memory(GiB)": 138.1,
      "step": 74730,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.77891903,
      "epoch": 1.743740795287187,
      "grad_norm": 5.28125,
      "learning_rate": 4.2426616629285544e-07,
      "loss": 0.79765043,
      "memory(GiB)": 138.1,
      "step": 74740,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.76886244,
      "epoch": 1.743974102859476,
      "grad_norm": 29.25,
      "learning_rate": 4.235049526712848e-07,
      "loss": 0.8290019,
      "memory(GiB)": 138.1,
      "step": 74750,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.78655448,
      "epoch": 1.7442074104317649,
      "grad_norm": 14.1875,
      "learning_rate": 4.227443923430513e-07,
      "loss": 0.75835438,
      "memory(GiB)": 138.1,
      "step": 74760,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.78014526,
      "epoch": 1.7444407180040538,
      "grad_norm": 6.5,
      "learning_rate": 4.2198448541672654e-07,
      "loss": 0.7994524,
      "memory(GiB)": 138.1,
      "step": 74770,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.77601213,
      "epoch": 1.7446740255763427,
      "grad_norm": 5.1875,
      "learning_rate": 4.212252320007859e-07,
      "loss": 0.83280792,
      "memory(GiB)": 138.1,
      "step": 74780,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.79946394,
      "epoch": 1.7449073331486316,
      "grad_norm": 4.96875,
      "learning_rate": 4.204666322036138e-07,
      "loss": 0.70037394,
      "memory(GiB)": 138.1,
      "step": 74790,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.78642559,
      "epoch": 1.7451406407209205,
      "grad_norm": 6.40625,
      "learning_rate": 4.197086861334998e-07,
      "loss": 0.78597708,
      "memory(GiB)": 138.1,
      "step": 74800,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.76866412,
      "epoch": 1.7453739482932091,
      "grad_norm": 5.5625,
      "learning_rate": 4.189513938986395e-07,
      "loss": 0.82813854,
      "memory(GiB)": 138.1,
      "step": 74810,
      "train_speed(iter/s)": 0.200917
    },
    {
      "acc": 0.77864561,
      "epoch": 1.7456072558654983,
      "grad_norm": 7.96875,
      "learning_rate": 4.181947556071381e-07,
      "loss": 0.79182148,
      "memory(GiB)": 138.1,
      "step": 74820,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.77730217,
      "epoch": 1.745840563437787,
      "grad_norm": 6.6875,
      "learning_rate": 4.1743877136700307e-07,
      "loss": 0.80669651,
      "memory(GiB)": 138.1,
      "step": 74830,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.78707137,
      "epoch": 1.746073871010076,
      "grad_norm": 5.8125,
      "learning_rate": 4.1668344128615254e-07,
      "loss": 0.77385092,
      "memory(GiB)": 138.1,
      "step": 74840,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.79038467,
      "epoch": 1.7463071785823647,
      "grad_norm": 6.09375,
      "learning_rate": 4.1592876547241035e-07,
      "loss": 0.74476175,
      "memory(GiB)": 138.1,
      "step": 74850,
      "train_speed(iter/s)": 0.200971
    },
    {
      "acc": 0.78900266,
      "epoch": 1.7465404861546538,
      "grad_norm": 5.34375,
      "learning_rate": 4.151747440335047e-07,
      "loss": 0.74805627,
      "memory(GiB)": 138.1,
      "step": 74860,
      "train_speed(iter/s)": 0.200984
    },
    {
      "acc": 0.77410288,
      "epoch": 1.7467737937269425,
      "grad_norm": 4.40625,
      "learning_rate": 4.144213770770711e-07,
      "loss": 0.82338734,
      "memory(GiB)": 138.1,
      "step": 74870,
      "train_speed(iter/s)": 0.200996
    },
    {
      "acc": 0.7824645,
      "epoch": 1.7470071012992316,
      "grad_norm": 4.5,
      "learning_rate": 4.1366866471065524e-07,
      "loss": 0.77526579,
      "memory(GiB)": 138.1,
      "step": 74880,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.76652699,
      "epoch": 1.7472404088715203,
      "grad_norm": 4.90625,
      "learning_rate": 4.129166070417051e-07,
      "loss": 0.85960999,
      "memory(GiB)": 138.1,
      "step": 74890,
      "train_speed(iter/s)": 0.201024
    },
    {
      "acc": 0.78324442,
      "epoch": 1.7474737164438094,
      "grad_norm": 6.5625,
      "learning_rate": 4.121652041775759e-07,
      "loss": 0.75736356,
      "memory(GiB)": 138.1,
      "step": 74900,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.78140106,
      "epoch": 1.7477070240160981,
      "grad_norm": 5.78125,
      "learning_rate": 4.1141445622553175e-07,
      "loss": 0.80148916,
      "memory(GiB)": 138.1,
      "step": 74910,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.79169626,
      "epoch": 1.7479403315883872,
      "grad_norm": 5.59375,
      "learning_rate": 4.106643632927404e-07,
      "loss": 0.7412683,
      "memory(GiB)": 138.1,
      "step": 74920,
      "train_speed(iter/s)": 0.201066
    },
    {
      "acc": 0.77393618,
      "epoch": 1.748173639160676,
      "grad_norm": 6.71875,
      "learning_rate": 4.099149254862783e-07,
      "loss": 0.78559313,
      "memory(GiB)": 138.1,
      "step": 74930,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.78671503,
      "epoch": 1.748406946732965,
      "grad_norm": 7.40625,
      "learning_rate": 4.091661429131277e-07,
      "loss": 0.77732067,
      "memory(GiB)": 138.1,
      "step": 74940,
      "train_speed(iter/s)": 0.201095
    },
    {
      "acc": 0.76118011,
      "epoch": 1.7486402543052537,
      "grad_norm": 5.65625,
      "learning_rate": 4.0841801568017534e-07,
      "loss": 0.86246977,
      "memory(GiB)": 138.1,
      "step": 74950,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.76429033,
      "epoch": 1.7488735618775428,
      "grad_norm": 5.375,
      "learning_rate": 4.0767054389421857e-07,
      "loss": 0.85855999,
      "memory(GiB)": 138.1,
      "step": 74960,
      "train_speed(iter/s)": 0.201121
    },
    {
      "acc": 0.78033648,
      "epoch": 1.7491068694498315,
      "grad_norm": 4.15625,
      "learning_rate": 4.069237276619564e-07,
      "loss": 0.76560135,
      "memory(GiB)": 138.1,
      "step": 74970,
      "train_speed(iter/s)": 0.201134
    },
    {
      "acc": 0.76945028,
      "epoch": 1.7493401770221206,
      "grad_norm": 4.71875,
      "learning_rate": 4.0617756708999976e-07,
      "loss": 0.82791023,
      "memory(GiB)": 138.1,
      "step": 74980,
      "train_speed(iter/s)": 0.201148
    },
    {
      "acc": 0.76738582,
      "epoch": 1.7495734845944093,
      "grad_norm": 5.46875,
      "learning_rate": 4.05432062284859e-07,
      "loss": 0.834793,
      "memory(GiB)": 138.1,
      "step": 74990,
      "train_speed(iter/s)": 0.201161
    },
    {
      "acc": 0.78293915,
      "epoch": 1.7498067921666982,
      "grad_norm": 6.5,
      "learning_rate": 4.0468721335295726e-07,
      "loss": 0.78276381,
      "memory(GiB)": 138.1,
      "step": 75000,
      "train_speed(iter/s)": 0.201175
    },
    {
      "epoch": 1.7498067921666982,
      "eval_acc": 0.7446785272687994,
      "eval_loss": 0.8044347763061523,
      "eval_runtime": 1272.0014,
      "eval_samples_per_second": 28.295,
      "eval_steps_per_second": 14.148,
      "step": 75000
    },
    {
      "acc": 0.77675529,
      "epoch": 1.750040099738987,
      "grad_norm": 5.9375,
      "learning_rate": 4.0394302040062117e-07,
      "loss": 0.78634806,
      "memory(GiB)": 138.1,
      "step": 75010,
      "train_speed(iter/s)": 0.200492
    },
    {
      "acc": 0.79533587,
      "epoch": 1.750273407311276,
      "grad_norm": 4.46875,
      "learning_rate": 4.03199483534083e-07,
      "loss": 0.73570852,
      "memory(GiB)": 138.1,
      "step": 75020,
      "train_speed(iter/s)": 0.200505
    },
    {
      "acc": 0.78597369,
      "epoch": 1.750506714883565,
      "grad_norm": 6.53125,
      "learning_rate": 4.0245660285948394e-07,
      "loss": 0.75488353,
      "memory(GiB)": 138.1,
      "step": 75030,
      "train_speed(iter/s)": 0.200519
    },
    {
      "acc": 0.8074131,
      "epoch": 1.7507400224558538,
      "grad_norm": 11.6875,
      "learning_rate": 4.0171437848286867e-07,
      "loss": 0.698316,
      "memory(GiB)": 138.1,
      "step": 75040,
      "train_speed(iter/s)": 0.200532
    },
    {
      "acc": 0.77187395,
      "epoch": 1.7509733300281427,
      "grad_norm": 6.53125,
      "learning_rate": 4.009728105101901e-07,
      "loss": 0.81652546,
      "memory(GiB)": 138.1,
      "step": 75050,
      "train_speed(iter/s)": 0.200547
    },
    {
      "acc": 0.77719297,
      "epoch": 1.7512066376004316,
      "grad_norm": 6.0,
      "learning_rate": 4.00231899047307e-07,
      "loss": 0.8079525,
      "memory(GiB)": 138.1,
      "step": 75060,
      "train_speed(iter/s)": 0.20056
    },
    {
      "acc": 0.77472291,
      "epoch": 1.7514399451727205,
      "grad_norm": 6.15625,
      "learning_rate": 3.994916441999841e-07,
      "loss": 0.80752106,
      "memory(GiB)": 138.1,
      "step": 75070,
      "train_speed(iter/s)": 0.200573
    },
    {
      "acc": 0.79288235,
      "epoch": 1.7516732527450094,
      "grad_norm": 6.0625,
      "learning_rate": 3.9875204607389304e-07,
      "loss": 0.76388402,
      "memory(GiB)": 138.1,
      "step": 75080,
      "train_speed(iter/s)": 0.200587
    },
    {
      "acc": 0.77766056,
      "epoch": 1.7519065603172983,
      "grad_norm": 4.25,
      "learning_rate": 3.980131047746105e-07,
      "loss": 0.80273018,
      "memory(GiB)": 138.1,
      "step": 75090,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.78243771,
      "epoch": 1.7521398678895872,
      "grad_norm": 4.875,
      "learning_rate": 3.9727482040762044e-07,
      "loss": 0.77541404,
      "memory(GiB)": 138.1,
      "step": 75100,
      "train_speed(iter/s)": 0.200615
    },
    {
      "acc": 0.78391609,
      "epoch": 1.752373175461876,
      "grad_norm": 5.59375,
      "learning_rate": 3.965371930783113e-07,
      "loss": 0.78062081,
      "memory(GiB)": 138.1,
      "step": 75110,
      "train_speed(iter/s)": 0.200628
    },
    {
      "acc": 0.79337425,
      "epoch": 1.752606483034165,
      "grad_norm": 5.0,
      "learning_rate": 3.958002228919822e-07,
      "loss": 0.73935184,
      "memory(GiB)": 138.1,
      "step": 75120,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.80228825,
      "epoch": 1.752839790606454,
      "grad_norm": 3.96875,
      "learning_rate": 3.9506390995383225e-07,
      "loss": 0.70532112,
      "memory(GiB)": 138.1,
      "step": 75130,
      "train_speed(iter/s)": 0.200656
    },
    {
      "acc": 0.79691057,
      "epoch": 1.7530730981787428,
      "grad_norm": 5.5,
      "learning_rate": 3.943282543689725e-07,
      "loss": 0.73715506,
      "memory(GiB)": 138.1,
      "step": 75140,
      "train_speed(iter/s)": 0.20067
    },
    {
      "acc": 0.78184276,
      "epoch": 1.7533064057510317,
      "grad_norm": 5.0625,
      "learning_rate": 3.935932562424166e-07,
      "loss": 0.78103657,
      "memory(GiB)": 138.1,
      "step": 75150,
      "train_speed(iter/s)": 0.200681
    },
    {
      "acc": 0.78166604,
      "epoch": 1.7535397133233206,
      "grad_norm": 4.6875,
      "learning_rate": 3.9285891567908465e-07,
      "loss": 0.78287544,
      "memory(GiB)": 138.1,
      "step": 75160,
      "train_speed(iter/s)": 0.200695
    },
    {
      "acc": 0.78169422,
      "epoch": 1.7537730208956095,
      "grad_norm": 6.28125,
      "learning_rate": 3.9212523278380434e-07,
      "loss": 0.75878248,
      "memory(GiB)": 138.1,
      "step": 75170,
      "train_speed(iter/s)": 0.200707
    },
    {
      "acc": 0.78662972,
      "epoch": 1.7540063284678984,
      "grad_norm": 7.4375,
      "learning_rate": 3.9139220766130803e-07,
      "loss": 0.75259981,
      "memory(GiB)": 138.1,
      "step": 75180,
      "train_speed(iter/s)": 0.20072
    },
    {
      "acc": 0.77813616,
      "epoch": 1.7542396360401873,
      "grad_norm": 9.8125,
      "learning_rate": 3.9065984041623594e-07,
      "loss": 0.7942369,
      "memory(GiB)": 138.1,
      "step": 75190,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.77014294,
      "epoch": 1.754472943612476,
      "grad_norm": 4.09375,
      "learning_rate": 3.8992813115313164e-07,
      "loss": 0.83160582,
      "memory(GiB)": 138.1,
      "step": 75200,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.77763653,
      "epoch": 1.754706251184765,
      "grad_norm": 5.28125,
      "learning_rate": 3.8919707997644884e-07,
      "loss": 0.8163887,
      "memory(GiB)": 138.1,
      "step": 75210,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.77771187,
      "epoch": 1.7549395587570538,
      "grad_norm": 4.6875,
      "learning_rate": 3.8846668699054233e-07,
      "loss": 0.80659122,
      "memory(GiB)": 138.1,
      "step": 75220,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.78318548,
      "epoch": 1.7551728663293429,
      "grad_norm": 5.90625,
      "learning_rate": 3.877369522996771e-07,
      "loss": 0.80368767,
      "memory(GiB)": 138.1,
      "step": 75230,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.7785296,
      "epoch": 1.7554061739016316,
      "grad_norm": 4.625,
      "learning_rate": 3.8700787600802203e-07,
      "loss": 0.78689766,
      "memory(GiB)": 138.1,
      "step": 75240,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.77669411,
      "epoch": 1.7556394814739207,
      "grad_norm": 6.90625,
      "learning_rate": 3.862794582196522e-07,
      "loss": 0.8024128,
      "memory(GiB)": 138.1,
      "step": 75250,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.77892799,
      "epoch": 1.7558727890462094,
      "grad_norm": 5.96875,
      "learning_rate": 3.8555169903854993e-07,
      "loss": 0.7945653,
      "memory(GiB)": 138.1,
      "step": 75260,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.79608464,
      "epoch": 1.7561060966184985,
      "grad_norm": 6.03125,
      "learning_rate": 3.848245985686011e-07,
      "loss": 0.72459307,
      "memory(GiB)": 138.1,
      "step": 75270,
      "train_speed(iter/s)": 0.200843
    },
    {
      "acc": 0.77168674,
      "epoch": 1.7563394041907872,
      "grad_norm": 5.96875,
      "learning_rate": 3.84098156913601e-07,
      "loss": 0.82376051,
      "memory(GiB)": 138.1,
      "step": 75280,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.79666181,
      "epoch": 1.7565727117630763,
      "grad_norm": 6.6875,
      "learning_rate": 3.8337237417724827e-07,
      "loss": 0.72730808,
      "memory(GiB)": 138.1,
      "step": 75290,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.8078371,
      "epoch": 1.756806019335365,
      "grad_norm": 5.8125,
      "learning_rate": 3.826472504631473e-07,
      "loss": 0.68620291,
      "memory(GiB)": 138.1,
      "step": 75300,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.78211536,
      "epoch": 1.757039326907654,
      "grad_norm": 6.28125,
      "learning_rate": 3.8192278587481035e-07,
      "loss": 0.77496066,
      "memory(GiB)": 138.1,
      "step": 75310,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.77927465,
      "epoch": 1.7572726344799428,
      "grad_norm": 5.21875,
      "learning_rate": 3.811989805156546e-07,
      "loss": 0.80956287,
      "memory(GiB)": 138.1,
      "step": 75320,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.79260778,
      "epoch": 1.7575059420522319,
      "grad_norm": 7.46875,
      "learning_rate": 3.804758344890025e-07,
      "loss": 0.74879017,
      "memory(GiB)": 138.1,
      "step": 75330,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.78099527,
      "epoch": 1.7577392496245205,
      "grad_norm": 5.75,
      "learning_rate": 3.7975334789808194e-07,
      "loss": 0.78341789,
      "memory(GiB)": 138.1,
      "step": 75340,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.77331486,
      "epoch": 1.7579725571968097,
      "grad_norm": 4.84375,
      "learning_rate": 3.790315208460299e-07,
      "loss": 0.82067747,
      "memory(GiB)": 138.1,
      "step": 75350,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.77612276,
      "epoch": 1.7582058647690983,
      "grad_norm": 5.25,
      "learning_rate": 3.783103534358845e-07,
      "loss": 0.79735045,
      "memory(GiB)": 138.1,
      "step": 75360,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.77384906,
      "epoch": 1.7584391723413875,
      "grad_norm": 6.6875,
      "learning_rate": 3.775898457705951e-07,
      "loss": 0.81172314,
      "memory(GiB)": 138.1,
      "step": 75370,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.77728834,
      "epoch": 1.7586724799136761,
      "grad_norm": 5.8125,
      "learning_rate": 3.768699979530122e-07,
      "loss": 0.77499533,
      "memory(GiB)": 138.1,
      "step": 75380,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.78484507,
      "epoch": 1.758905787485965,
      "grad_norm": 4.65625,
      "learning_rate": 3.761508100858929e-07,
      "loss": 0.7881773,
      "memory(GiB)": 138.1,
      "step": 75390,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.76613727,
      "epoch": 1.759139095058254,
      "grad_norm": 7.375,
      "learning_rate": 3.7543228227190307e-07,
      "loss": 0.84683084,
      "memory(GiB)": 138.1,
      "step": 75400,
      "train_speed(iter/s)": 0.201021
    },
    {
      "acc": 0.80847015,
      "epoch": 1.7593724026305428,
      "grad_norm": 4.125,
      "learning_rate": 3.747144146136111e-07,
      "loss": 0.67406015,
      "memory(GiB)": 138.1,
      "step": 75410,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.79223614,
      "epoch": 1.7596057102028317,
      "grad_norm": 3.734375,
      "learning_rate": 3.739972072134934e-07,
      "loss": 0.75303974,
      "memory(GiB)": 138.1,
      "step": 75420,
      "train_speed(iter/s)": 0.201048
    },
    {
      "acc": 0.7955615,
      "epoch": 1.7598390177751206,
      "grad_norm": 5.40625,
      "learning_rate": 3.7328066017393025e-07,
      "loss": 0.72555103,
      "memory(GiB)": 138.1,
      "step": 75430,
      "train_speed(iter/s)": 0.201061
    },
    {
      "acc": 0.7834197,
      "epoch": 1.7600723253474095,
      "grad_norm": 6.5625,
      "learning_rate": 3.725647735972093e-07,
      "loss": 0.77190681,
      "memory(GiB)": 138.1,
      "step": 75440,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.78522491,
      "epoch": 1.7603056329196984,
      "grad_norm": 3.6875,
      "learning_rate": 3.7184954758552215e-07,
      "loss": 0.75268602,
      "memory(GiB)": 138.1,
      "step": 75450,
      "train_speed(iter/s)": 0.201089
    },
    {
      "acc": 0.77963495,
      "epoch": 1.7605389404919873,
      "grad_norm": 5.65625,
      "learning_rate": 3.711349822409671e-07,
      "loss": 0.78821554,
      "memory(GiB)": 138.1,
      "step": 75460,
      "train_speed(iter/s)": 0.201103
    },
    {
      "acc": 0.77956514,
      "epoch": 1.7607722480642762,
      "grad_norm": 5.875,
      "learning_rate": 3.7042107766554925e-07,
      "loss": 0.78396311,
      "memory(GiB)": 138.1,
      "step": 75470,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.79036522,
      "epoch": 1.7610055556365651,
      "grad_norm": 5.40625,
      "learning_rate": 3.6970783396117706e-07,
      "loss": 0.77081327,
      "memory(GiB)": 138.1,
      "step": 75480,
      "train_speed(iter/s)": 0.201128
    },
    {
      "acc": 0.77387805,
      "epoch": 1.761238863208854,
      "grad_norm": 6.625,
      "learning_rate": 3.689952512296674e-07,
      "loss": 0.83468933,
      "memory(GiB)": 138.1,
      "step": 75490,
      "train_speed(iter/s)": 0.201142
    },
    {
      "acc": 0.78259726,
      "epoch": 1.761472170781143,
      "grad_norm": 5.53125,
      "learning_rate": 3.682833295727389e-07,
      "loss": 0.76884212,
      "memory(GiB)": 138.1,
      "step": 75500,
      "train_speed(iter/s)": 0.201156
    },
    {
      "epoch": 1.761472170781143,
      "eval_acc": 0.7446898975092928,
      "eval_loss": 0.8043956756591797,
      "eval_runtime": 1270.8807,
      "eval_samples_per_second": 28.32,
      "eval_steps_per_second": 14.16,
      "step": 75500
    },
    {
      "acc": 0.775494,
      "epoch": 1.7617054783534318,
      "grad_norm": 3.890625,
      "learning_rate": 3.675720690920209e-07,
      "loss": 0.7966898,
      "memory(GiB)": 138.1,
      "step": 75510,
      "train_speed(iter/s)": 0.200477
    },
    {
      "acc": 0.77051344,
      "epoch": 1.7619387859257207,
      "grad_norm": 4.96875,
      "learning_rate": 3.668614698890444e-07,
      "loss": 0.8196578,
      "memory(GiB)": 138.1,
      "step": 75520,
      "train_speed(iter/s)": 0.200491
    },
    {
      "acc": 0.77491951,
      "epoch": 1.7621720934980096,
      "grad_norm": 4.5,
      "learning_rate": 3.661515320652459e-07,
      "loss": 0.82127848,
      "memory(GiB)": 138.1,
      "step": 75530,
      "train_speed(iter/s)": 0.200505
    },
    {
      "acc": 0.80416489,
      "epoch": 1.7624054010702985,
      "grad_norm": 5.28125,
      "learning_rate": 3.654422557219711e-07,
      "loss": 0.6906394,
      "memory(GiB)": 138.1,
      "step": 75540,
      "train_speed(iter/s)": 0.200519
    },
    {
      "acc": 0.76411433,
      "epoch": 1.7626387086425874,
      "grad_norm": 5.46875,
      "learning_rate": 3.6473364096046795e-07,
      "loss": 0.85491076,
      "memory(GiB)": 138.1,
      "step": 75550,
      "train_speed(iter/s)": 0.200534
    },
    {
      "acc": 0.7577879,
      "epoch": 1.7628720162148763,
      "grad_norm": 5.0625,
      "learning_rate": 3.64025687881891e-07,
      "loss": 0.87597389,
      "memory(GiB)": 138.1,
      "step": 75560,
      "train_speed(iter/s)": 0.200548
    },
    {
      "acc": 0.78338909,
      "epoch": 1.763105323787165,
      "grad_norm": 4.5,
      "learning_rate": 3.633183965872994e-07,
      "loss": 0.78178053,
      "memory(GiB)": 138.1,
      "step": 75570,
      "train_speed(iter/s)": 0.200561
    },
    {
      "acc": 0.80338173,
      "epoch": 1.7633386313594541,
      "grad_norm": 5.21875,
      "learning_rate": 3.6261176717766076e-07,
      "loss": 0.71166687,
      "memory(GiB)": 138.1,
      "step": 75580,
      "train_speed(iter/s)": 0.200575
    },
    {
      "acc": 0.77883034,
      "epoch": 1.7635719389317428,
      "grad_norm": 6.46875,
      "learning_rate": 3.6190579975384497e-07,
      "loss": 0.81209888,
      "memory(GiB)": 138.1,
      "step": 75590,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.78879728,
      "epoch": 1.763805246504032,
      "grad_norm": 4.4375,
      "learning_rate": 3.6120049441662805e-07,
      "loss": 0.76921787,
      "memory(GiB)": 138.1,
      "step": 75600,
      "train_speed(iter/s)": 0.200602
    },
    {
      "acc": 0.78142576,
      "epoch": 1.7640385540763206,
      "grad_norm": 5.46875,
      "learning_rate": 3.6049585126669395e-07,
      "loss": 0.79211183,
      "memory(GiB)": 138.1,
      "step": 75610,
      "train_speed(iter/s)": 0.200615
    },
    {
      "acc": 0.78412828,
      "epoch": 1.7642718616486097,
      "grad_norm": 6.3125,
      "learning_rate": 3.5979187040462883e-07,
      "loss": 0.77902346,
      "memory(GiB)": 138.1,
      "step": 75620,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.75922632,
      "epoch": 1.7645051692208984,
      "grad_norm": 6.3125,
      "learning_rate": 3.5908855193092617e-07,
      "loss": 0.87735043,
      "memory(GiB)": 138.1,
      "step": 75630,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.79961157,
      "epoch": 1.7647384767931875,
      "grad_norm": 5.46875,
      "learning_rate": 3.5838589594598514e-07,
      "loss": 0.7121191,
      "memory(GiB)": 138.1,
      "step": 75640,
      "train_speed(iter/s)": 0.200658
    },
    {
      "acc": 0.7941186,
      "epoch": 1.7649717843654762,
      "grad_norm": 5.125,
      "learning_rate": 3.5768390255010833e-07,
      "loss": 0.71949234,
      "memory(GiB)": 138.1,
      "step": 75650,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.77251234,
      "epoch": 1.7652050919377653,
      "grad_norm": 5.5625,
      "learning_rate": 3.569825718435066e-07,
      "loss": 0.82141552,
      "memory(GiB)": 138.1,
      "step": 75660,
      "train_speed(iter/s)": 0.200686
    },
    {
      "acc": 0.7619668,
      "epoch": 1.765438399510054,
      "grad_norm": 4.65625,
      "learning_rate": 3.562819039262938e-07,
      "loss": 0.87379723,
      "memory(GiB)": 138.1,
      "step": 75670,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.80607891,
      "epoch": 1.7656717070823431,
      "grad_norm": 7.5,
      "learning_rate": 3.5558189889848995e-07,
      "loss": 0.67555485,
      "memory(GiB)": 138.1,
      "step": 75680,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.78483949,
      "epoch": 1.7659050146546318,
      "grad_norm": 4.0625,
      "learning_rate": 3.5488255686002005e-07,
      "loss": 0.80067978,
      "memory(GiB)": 138.1,
      "step": 75690,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.77885909,
      "epoch": 1.766138322226921,
      "grad_norm": 5.28125,
      "learning_rate": 3.5418387791071706e-07,
      "loss": 0.81857595,
      "memory(GiB)": 138.1,
      "step": 75700,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.77617831,
      "epoch": 1.7663716297992096,
      "grad_norm": 5.125,
      "learning_rate": 3.5348586215031457e-07,
      "loss": 0.79372025,
      "memory(GiB)": 138.1,
      "step": 75710,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.78535709,
      "epoch": 1.7666049373714987,
      "grad_norm": 6.03125,
      "learning_rate": 3.527885096784567e-07,
      "loss": 0.76806793,
      "memory(GiB)": 138.1,
      "step": 75720,
      "train_speed(iter/s)": 0.200769
    },
    {
      "acc": 0.76860857,
      "epoch": 1.7668382449437874,
      "grad_norm": 5.59375,
      "learning_rate": 3.520918205946883e-07,
      "loss": 0.81936932,
      "memory(GiB)": 138.1,
      "step": 75730,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.77166557,
      "epoch": 1.7670715525160765,
      "grad_norm": 5.375,
      "learning_rate": 3.51395794998462e-07,
      "loss": 0.82154722,
      "memory(GiB)": 138.1,
      "step": 75740,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.78490963,
      "epoch": 1.7673048600883652,
      "grad_norm": 6.53125,
      "learning_rate": 3.507004329891367e-07,
      "loss": 0.76154852,
      "memory(GiB)": 138.1,
      "step": 75750,
      "train_speed(iter/s)": 0.200812
    },
    {
      "acc": 0.78819742,
      "epoch": 1.7675381676606543,
      "grad_norm": 5.96875,
      "learning_rate": 3.5000573466597243e-07,
      "loss": 0.74255095,
      "memory(GiB)": 138.1,
      "step": 75760,
      "train_speed(iter/s)": 0.200825
    },
    {
      "acc": 0.79479284,
      "epoch": 1.767771475232943,
      "grad_norm": 6.65625,
      "learning_rate": 3.4931170012813985e-07,
      "loss": 0.73127079,
      "memory(GiB)": 138.1,
      "step": 75770,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.79238229,
      "epoch": 1.7680047828052319,
      "grad_norm": 6.1875,
      "learning_rate": 3.4861832947471086e-07,
      "loss": 0.73390646,
      "memory(GiB)": 138.1,
      "step": 75780,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.78550787,
      "epoch": 1.7682380903775208,
      "grad_norm": 5.0625,
      "learning_rate": 3.479256228046646e-07,
      "loss": 0.75722055,
      "memory(GiB)": 138.1,
      "step": 75790,
      "train_speed(iter/s)": 0.200866
    },
    {
      "acc": 0.777526,
      "epoch": 1.7684713979498097,
      "grad_norm": 9.1875,
      "learning_rate": 3.4723358021688303e-07,
      "loss": 0.78414106,
      "memory(GiB)": 138.1,
      "step": 75800,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.77671871,
      "epoch": 1.7687047055220986,
      "grad_norm": 4.90625,
      "learning_rate": 3.465422018101572e-07,
      "loss": 0.80290337,
      "memory(GiB)": 138.1,
      "step": 75810,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.79054794,
      "epoch": 1.7689380130943875,
      "grad_norm": 5.28125,
      "learning_rate": 3.4585148768317975e-07,
      "loss": 0.74591765,
      "memory(GiB)": 138.1,
      "step": 75820,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.77464447,
      "epoch": 1.7691713206666764,
      "grad_norm": 6.28125,
      "learning_rate": 3.4516143793455027e-07,
      "loss": 0.83038616,
      "memory(GiB)": 138.1,
      "step": 75830,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.78378277,
      "epoch": 1.7694046282389653,
      "grad_norm": 4.34375,
      "learning_rate": 3.4447205266277373e-07,
      "loss": 0.75231061,
      "memory(GiB)": 138.1,
      "step": 75840,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.78303156,
      "epoch": 1.7696379358112542,
      "grad_norm": 4.53125,
      "learning_rate": 3.437833319662587e-07,
      "loss": 0.78000083,
      "memory(GiB)": 138.1,
      "step": 75850,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.79331565,
      "epoch": 1.769871243383543,
      "grad_norm": 5.65625,
      "learning_rate": 3.430952759433209e-07,
      "loss": 0.74238071,
      "memory(GiB)": 138.1,
      "step": 75860,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.78225794,
      "epoch": 1.770104550955832,
      "grad_norm": 5.5625,
      "learning_rate": 3.4240788469217966e-07,
      "loss": 0.78348684,
      "memory(GiB)": 138.1,
      "step": 75870,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.79212933,
      "epoch": 1.7703378585281209,
      "grad_norm": 6.96875,
      "learning_rate": 3.417211583109592e-07,
      "loss": 0.73944921,
      "memory(GiB)": 138.1,
      "step": 75880,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.77557068,
      "epoch": 1.7705711661004098,
      "grad_norm": 5.03125,
      "learning_rate": 3.4103509689769165e-07,
      "loss": 0.78513107,
      "memory(GiB)": 138.1,
      "step": 75890,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.77322273,
      "epoch": 1.7708044736726987,
      "grad_norm": 6.78125,
      "learning_rate": 3.4034970055030923e-07,
      "loss": 0.81071148,
      "memory(GiB)": 138.1,
      "step": 75900,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.79248972,
      "epoch": 1.7710377812449876,
      "grad_norm": 4.28125,
      "learning_rate": 3.396649693666537e-07,
      "loss": 0.74013586,
      "memory(GiB)": 138.1,
      "step": 75910,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.79005871,
      "epoch": 1.7712710888172765,
      "grad_norm": 6.03125,
      "learning_rate": 3.3898090344446966e-07,
      "loss": 0.73053083,
      "memory(GiB)": 138.1,
      "step": 75920,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.76683903,
      "epoch": 1.7715043963895654,
      "grad_norm": 7.09375,
      "learning_rate": 3.382975028814078e-07,
      "loss": 0.8392231,
      "memory(GiB)": 138.1,
      "step": 75930,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.78439379,
      "epoch": 1.7717377039618543,
      "grad_norm": 6.59375,
      "learning_rate": 3.3761476777502355e-07,
      "loss": 0.77608676,
      "memory(GiB)": 138.1,
      "step": 75940,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.78523664,
      "epoch": 1.7719710115341432,
      "grad_norm": 4.9375,
      "learning_rate": 3.369326982227761e-07,
      "loss": 0.76700554,
      "memory(GiB)": 138.1,
      "step": 75950,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.78907003,
      "epoch": 1.7722043191064318,
      "grad_norm": 4.4375,
      "learning_rate": 3.3625129432203197e-07,
      "loss": 0.75762873,
      "memory(GiB)": 138.1,
      "step": 75960,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.77812548,
      "epoch": 1.772437626678721,
      "grad_norm": 4.28125,
      "learning_rate": 3.3557055617006006e-07,
      "loss": 0.76968746,
      "memory(GiB)": 138.1,
      "step": 75970,
      "train_speed(iter/s)": 0.201113
    },
    {
      "acc": 0.77837658,
      "epoch": 1.7726709342510096,
      "grad_norm": 4.34375,
      "learning_rate": 3.348904838640371e-07,
      "loss": 0.77840204,
      "memory(GiB)": 138.1,
      "step": 75980,
      "train_speed(iter/s)": 0.201127
    },
    {
      "acc": 0.77046366,
      "epoch": 1.7729042418232988,
      "grad_norm": 5.40625,
      "learning_rate": 3.3421107750104155e-07,
      "loss": 0.83179445,
      "memory(GiB)": 138.1,
      "step": 75990,
      "train_speed(iter/s)": 0.201141
    },
    {
      "acc": 0.77653723,
      "epoch": 1.7731375493955874,
      "grad_norm": 5.96875,
      "learning_rate": 3.3353233717805967e-07,
      "loss": 0.80116844,
      "memory(GiB)": 138.1,
      "step": 76000,
      "train_speed(iter/s)": 0.201154
    },
    {
      "epoch": 1.7731375493955874,
      "eval_acc": 0.744712317701815,
      "eval_loss": 0.804376482963562,
      "eval_runtime": 1270.1032,
      "eval_samples_per_second": 28.337,
      "eval_steps_per_second": 14.169,
      "step": 76000
    },
    {
      "acc": 0.77820454,
      "epoch": 1.7733708569678766,
      "grad_norm": 4.59375,
      "learning_rate": 3.3285426299198175e-07,
      "loss": 0.81404266,
      "memory(GiB)": 138.1,
      "step": 76010,
      "train_speed(iter/s)": 0.200481
    },
    {
      "acc": 0.77452345,
      "epoch": 1.7736041645401652,
      "grad_norm": 4.4375,
      "learning_rate": 3.321768550396015e-07,
      "loss": 0.81841717,
      "memory(GiB)": 138.1,
      "step": 76020,
      "train_speed(iter/s)": 0.200495
    },
    {
      "acc": 0.78945813,
      "epoch": 1.7738374721124543,
      "grad_norm": 4.96875,
      "learning_rate": 3.3150011341761933e-07,
      "loss": 0.75137343,
      "memory(GiB)": 138.1,
      "step": 76030,
      "train_speed(iter/s)": 0.200509
    },
    {
      "acc": 0.7791081,
      "epoch": 1.774070779684743,
      "grad_norm": 5.375,
      "learning_rate": 3.308240382226385e-07,
      "loss": 0.7906178,
      "memory(GiB)": 138.1,
      "step": 76040,
      "train_speed(iter/s)": 0.200523
    },
    {
      "acc": 0.78823118,
      "epoch": 1.7743040872570321,
      "grad_norm": 4.59375,
      "learning_rate": 3.301486295511713e-07,
      "loss": 0.77563086,
      "memory(GiB)": 138.1,
      "step": 76050,
      "train_speed(iter/s)": 0.200537
    },
    {
      "acc": 0.78934536,
      "epoch": 1.7745373948293208,
      "grad_norm": 7.5,
      "learning_rate": 3.294738874996295e-07,
      "loss": 0.7632966,
      "memory(GiB)": 138.1,
      "step": 76060,
      "train_speed(iter/s)": 0.200552
    },
    {
      "acc": 0.77486506,
      "epoch": 1.77477070240161,
      "grad_norm": 7.0,
      "learning_rate": 3.2879981216433433e-07,
      "loss": 0.85285072,
      "memory(GiB)": 138.1,
      "step": 76070,
      "train_speed(iter/s)": 0.200566
    },
    {
      "acc": 0.7874649,
      "epoch": 1.7750040099738986,
      "grad_norm": 4.25,
      "learning_rate": 3.281264036415088e-07,
      "loss": 0.75169353,
      "memory(GiB)": 138.1,
      "step": 76080,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.77508116,
      "epoch": 1.7752373175461877,
      "grad_norm": 6.125,
      "learning_rate": 3.274536620272811e-07,
      "loss": 0.80775452,
      "memory(GiB)": 138.1,
      "step": 76090,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.77886529,
      "epoch": 1.7754706251184764,
      "grad_norm": 4.8125,
      "learning_rate": 3.267815874176866e-07,
      "loss": 0.80793839,
      "memory(GiB)": 138.1,
      "step": 76100,
      "train_speed(iter/s)": 0.200607
    },
    {
      "acc": 0.77209039,
      "epoch": 1.7757039326907655,
      "grad_norm": 7.125,
      "learning_rate": 3.2611017990866244e-07,
      "loss": 0.82538052,
      "memory(GiB)": 138.1,
      "step": 76110,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.77398095,
      "epoch": 1.7759372402630542,
      "grad_norm": 4.96875,
      "learning_rate": 3.254394395960536e-07,
      "loss": 0.81047382,
      "memory(GiB)": 138.1,
      "step": 76120,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.76875162,
      "epoch": 1.7761705478353433,
      "grad_norm": 5.0,
      "learning_rate": 3.247693665756052e-07,
      "loss": 0.84657841,
      "memory(GiB)": 138.1,
      "step": 76130,
      "train_speed(iter/s)": 0.200648
    },
    {
      "acc": 0.7959794,
      "epoch": 1.776403855407632,
      "grad_norm": 4.8125,
      "learning_rate": 3.2409996094297294e-07,
      "loss": 0.74632068,
      "memory(GiB)": 138.1,
      "step": 76140,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.77729673,
      "epoch": 1.776637162979921,
      "grad_norm": 6.25,
      "learning_rate": 3.234312227937114e-07,
      "loss": 0.82577114,
      "memory(GiB)": 138.1,
      "step": 76150,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.80631809,
      "epoch": 1.7768704705522098,
      "grad_norm": 6.125,
      "learning_rate": 3.2276315222328547e-07,
      "loss": 0.71360807,
      "memory(GiB)": 138.1,
      "step": 76160,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.80361118,
      "epoch": 1.7771037781244987,
      "grad_norm": 4.1875,
      "learning_rate": 3.2209574932706043e-07,
      "loss": 0.69395614,
      "memory(GiB)": 138.1,
      "step": 76170,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77744942,
      "epoch": 1.7773370856967876,
      "grad_norm": 5.0625,
      "learning_rate": 3.214290142003079e-07,
      "loss": 0.80556965,
      "memory(GiB)": 138.1,
      "step": 76180,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.78394184,
      "epoch": 1.7775703932690765,
      "grad_norm": 5.875,
      "learning_rate": 3.207629469382051e-07,
      "loss": 0.78173552,
      "memory(GiB)": 138.1,
      "step": 76190,
      "train_speed(iter/s)": 0.20073
    },
    {
      "acc": 0.7806911,
      "epoch": 1.7778037008413654,
      "grad_norm": 4.875,
      "learning_rate": 3.2009754763583143e-07,
      "loss": 0.79640622,
      "memory(GiB)": 138.1,
      "step": 76200,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.78159871,
      "epoch": 1.7780370084136543,
      "grad_norm": 5.34375,
      "learning_rate": 3.194328163881738e-07,
      "loss": 0.77030783,
      "memory(GiB)": 138.1,
      "step": 76210,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.78937597,
      "epoch": 1.7782703159859432,
      "grad_norm": 4.5,
      "learning_rate": 3.1876875329012235e-07,
      "loss": 0.74239864,
      "memory(GiB)": 138.1,
      "step": 76220,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.76783009,
      "epoch": 1.778503623558232,
      "grad_norm": 5.03125,
      "learning_rate": 3.181053584364707e-07,
      "loss": 0.84220142,
      "memory(GiB)": 138.1,
      "step": 76230,
      "train_speed(iter/s)": 0.200783
    },
    {
      "acc": 0.77021403,
      "epoch": 1.778736931130521,
      "grad_norm": 5.5,
      "learning_rate": 3.174426319219204e-07,
      "loss": 0.83477764,
      "memory(GiB)": 138.1,
      "step": 76240,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.78976417,
      "epoch": 1.77897023870281,
      "grad_norm": 5.03125,
      "learning_rate": 3.167805738410723e-07,
      "loss": 0.75705347,
      "memory(GiB)": 138.1,
      "step": 76250,
      "train_speed(iter/s)": 0.200812
    },
    {
      "acc": 0.78457289,
      "epoch": 1.7792035462750988,
      "grad_norm": 6.6875,
      "learning_rate": 3.161191842884381e-07,
      "loss": 0.76162119,
      "memory(GiB)": 138.1,
      "step": 76260,
      "train_speed(iter/s)": 0.200827
    },
    {
      "acc": 0.78860617,
      "epoch": 1.7794368538473877,
      "grad_norm": 4.5625,
      "learning_rate": 3.1545846335842843e-07,
      "loss": 0.76644807,
      "memory(GiB)": 138.1,
      "step": 76270,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.78241735,
      "epoch": 1.7796701614196766,
      "grad_norm": 5.59375,
      "learning_rate": 3.1479841114536334e-07,
      "loss": 0.7895534,
      "memory(GiB)": 138.1,
      "step": 76280,
      "train_speed(iter/s)": 0.200854
    },
    {
      "acc": 0.76458092,
      "epoch": 1.7799034689919655,
      "grad_norm": 5.59375,
      "learning_rate": 3.1413902774346305e-07,
      "loss": 0.85796795,
      "memory(GiB)": 138.1,
      "step": 76290,
      "train_speed(iter/s)": 0.200868
    },
    {
      "acc": 0.7956975,
      "epoch": 1.7801367765642544,
      "grad_norm": 6.59375,
      "learning_rate": 3.134803132468561e-07,
      "loss": 0.73320594,
      "memory(GiB)": 138.1,
      "step": 76300,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.80051413,
      "epoch": 1.7803700841365433,
      "grad_norm": 5.4375,
      "learning_rate": 3.128222677495729e-07,
      "loss": 0.71121464,
      "memory(GiB)": 138.1,
      "step": 76310,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.78982344,
      "epoch": 1.7806033917088322,
      "grad_norm": 5.9375,
      "learning_rate": 3.1216489134554886e-07,
      "loss": 0.76041822,
      "memory(GiB)": 138.1,
      "step": 76320,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.78644514,
      "epoch": 1.780836699281121,
      "grad_norm": 5.5625,
      "learning_rate": 3.115081841286255e-07,
      "loss": 0.76248388,
      "memory(GiB)": 138.1,
      "step": 76330,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.78216338,
      "epoch": 1.78107000685341,
      "grad_norm": 5.21875,
      "learning_rate": 3.108521461925457e-07,
      "loss": 0.80187321,
      "memory(GiB)": 138.1,
      "step": 76340,
      "train_speed(iter/s)": 0.200937
    },
    {
      "acc": 0.75998535,
      "epoch": 1.7813033144256987,
      "grad_norm": 5.125,
      "learning_rate": 3.101967776309617e-07,
      "loss": 0.8771349,
      "memory(GiB)": 138.1,
      "step": 76350,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.80899925,
      "epoch": 1.7815366219979878,
      "grad_norm": 5.0625,
      "learning_rate": 3.095420785374237e-07,
      "loss": 0.66126566,
      "memory(GiB)": 138.1,
      "step": 76360,
      "train_speed(iter/s)": 0.200965
    },
    {
      "acc": 0.76196551,
      "epoch": 1.7817699295702765,
      "grad_norm": 6.78125,
      "learning_rate": 3.088880490053925e-07,
      "loss": 0.87400131,
      "memory(GiB)": 138.1,
      "step": 76370,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.76612358,
      "epoch": 1.7820032371425656,
      "grad_norm": 6.21875,
      "learning_rate": 3.0823468912822895e-07,
      "loss": 0.84581919,
      "memory(GiB)": 138.1,
      "step": 76380,
      "train_speed(iter/s)": 0.200992
    },
    {
      "acc": 0.78834362,
      "epoch": 1.7822365447148543,
      "grad_norm": 11.0,
      "learning_rate": 3.0758199899920014e-07,
      "loss": 0.74992781,
      "memory(GiB)": 138.1,
      "step": 76390,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.77249632,
      "epoch": 1.7824698522871434,
      "grad_norm": 4.71875,
      "learning_rate": 3.069299787114782e-07,
      "loss": 0.83343058,
      "memory(GiB)": 138.1,
      "step": 76400,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.78593674,
      "epoch": 1.782703159859432,
      "grad_norm": 4.46875,
      "learning_rate": 3.0627862835813814e-07,
      "loss": 0.75706806,
      "memory(GiB)": 138.1,
      "step": 76410,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.77988739,
      "epoch": 1.7829364674317212,
      "grad_norm": 4.8125,
      "learning_rate": 3.0562794803216114e-07,
      "loss": 0.77060046,
      "memory(GiB)": 138.1,
      "step": 76420,
      "train_speed(iter/s)": 0.201045
    },
    {
      "acc": 0.80068836,
      "epoch": 1.7831697750040099,
      "grad_norm": 6.96875,
      "learning_rate": 3.0497793782642946e-07,
      "loss": 0.69606977,
      "memory(GiB)": 138.1,
      "step": 76430,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.80669575,
      "epoch": 1.783403082576299,
      "grad_norm": 4.6875,
      "learning_rate": 3.043285978337346e-07,
      "loss": 0.69010477,
      "memory(GiB)": 138.1,
      "step": 76440,
      "train_speed(iter/s)": 0.201073
    },
    {
      "acc": 0.78510504,
      "epoch": 1.7836363901485877,
      "grad_norm": 6.0,
      "learning_rate": 3.036799281467678e-07,
      "loss": 0.75290413,
      "memory(GiB)": 138.1,
      "step": 76450,
      "train_speed(iter/s)": 0.201086
    },
    {
      "acc": 0.78098297,
      "epoch": 1.7838696977208768,
      "grad_norm": 4.96875,
      "learning_rate": 3.0303192885812737e-07,
      "loss": 0.77756319,
      "memory(GiB)": 138.1,
      "step": 76460,
      "train_speed(iter/s)": 0.2011
    },
    {
      "acc": 0.79312925,
      "epoch": 1.7841030052931655,
      "grad_norm": 4.8125,
      "learning_rate": 3.023846000603148e-07,
      "loss": 0.733179,
      "memory(GiB)": 138.1,
      "step": 76470,
      "train_speed(iter/s)": 0.201115
    },
    {
      "acc": 0.79176178,
      "epoch": 1.7843363128654546,
      "grad_norm": 5.375,
      "learning_rate": 3.0173794184573444e-07,
      "loss": 0.75639806,
      "memory(GiB)": 138.1,
      "step": 76480,
      "train_speed(iter/s)": 0.201129
    },
    {
      "acc": 0.78836441,
      "epoch": 1.7845696204377433,
      "grad_norm": 5.28125,
      "learning_rate": 3.0109195430669925e-07,
      "loss": 0.76128173,
      "memory(GiB)": 138.1,
      "step": 76490,
      "train_speed(iter/s)": 0.201142
    },
    {
      "acc": 0.78055067,
      "epoch": 1.7848029280100324,
      "grad_norm": 5.78125,
      "learning_rate": 3.00446637535422e-07,
      "loss": 0.79550552,
      "memory(GiB)": 138.1,
      "step": 76500,
      "train_speed(iter/s)": 0.201156
    },
    {
      "epoch": 1.7848029280100324,
      "eval_acc": 0.7446987054420694,
      "eval_loss": 0.8044025897979736,
      "eval_runtime": 1269.883,
      "eval_samples_per_second": 28.342,
      "eval_steps_per_second": 14.171,
      "step": 76500
    },
    {
      "acc": 0.78949766,
      "epoch": 1.785036235582321,
      "grad_norm": 4.96875,
      "learning_rate": 2.9980199162402245e-07,
      "loss": 0.77280111,
      "memory(GiB)": 138.1,
      "step": 76510,
      "train_speed(iter/s)": 0.200486
    },
    {
      "acc": 0.7705359,
      "epoch": 1.7852695431546102,
      "grad_norm": 5.9375,
      "learning_rate": 2.9915801666452307e-07,
      "loss": 0.82575054,
      "memory(GiB)": 138.1,
      "step": 76520,
      "train_speed(iter/s)": 0.200499
    },
    {
      "acc": 0.79377584,
      "epoch": 1.7855028507268988,
      "grad_norm": 5.96875,
      "learning_rate": 2.985147127488508e-07,
      "loss": 0.73605633,
      "memory(GiB)": 138.1,
      "step": 76530,
      "train_speed(iter/s)": 0.200512
    },
    {
      "acc": 0.77772698,
      "epoch": 1.7857361582991877,
      "grad_norm": 6.75,
      "learning_rate": 2.978720799688378e-07,
      "loss": 0.77994833,
      "memory(GiB)": 138.1,
      "step": 76540,
      "train_speed(iter/s)": 0.200526
    },
    {
      "acc": 0.78651562,
      "epoch": 1.7859694658714766,
      "grad_norm": 7.78125,
      "learning_rate": 2.9723011841621905e-07,
      "loss": 0.74728775,
      "memory(GiB)": 138.1,
      "step": 76550,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.76202259,
      "epoch": 1.7862027734437655,
      "grad_norm": 8.3125,
      "learning_rate": 2.965888281826357e-07,
      "loss": 0.84415607,
      "memory(GiB)": 138.1,
      "step": 76560,
      "train_speed(iter/s)": 0.200552
    },
    {
      "acc": 0.77949867,
      "epoch": 1.7864360810160544,
      "grad_norm": 4.3125,
      "learning_rate": 2.9594820935963e-07,
      "loss": 0.76835766,
      "memory(GiB)": 138.1,
      "step": 76570,
      "train_speed(iter/s)": 0.200565
    },
    {
      "acc": 0.78318701,
      "epoch": 1.7866693885883433,
      "grad_norm": 4.03125,
      "learning_rate": 2.953082620386516e-07,
      "loss": 0.77278767,
      "memory(GiB)": 138.1,
      "step": 76580,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.78765154,
      "epoch": 1.7869026961606322,
      "grad_norm": 7.375,
      "learning_rate": 2.946689863110508e-07,
      "loss": 0.74817243,
      "memory(GiB)": 138.1,
      "step": 76590,
      "train_speed(iter/s)": 0.200593
    },
    {
      "acc": 0.79166756,
      "epoch": 1.7871360037329211,
      "grad_norm": 4.875,
      "learning_rate": 2.9403038226808625e-07,
      "loss": 0.74761534,
      "memory(GiB)": 138.1,
      "step": 76600,
      "train_speed(iter/s)": 0.200606
    },
    {
      "acc": 0.77928109,
      "epoch": 1.78736931130521,
      "grad_norm": 5.9375,
      "learning_rate": 2.9339245000091776e-07,
      "loss": 0.80569696,
      "memory(GiB)": 138.1,
      "step": 76610,
      "train_speed(iter/s)": 0.20062
    },
    {
      "acc": 0.78152971,
      "epoch": 1.787602618877499,
      "grad_norm": 5.90625,
      "learning_rate": 2.9275518960060867e-07,
      "loss": 0.78764749,
      "memory(GiB)": 138.1,
      "step": 76620,
      "train_speed(iter/s)": 0.200633
    },
    {
      "acc": 0.78837299,
      "epoch": 1.7878359264497878,
      "grad_norm": 6.40625,
      "learning_rate": 2.9211860115813005e-07,
      "loss": 0.7382781,
      "memory(GiB)": 138.1,
      "step": 76630,
      "train_speed(iter/s)": 0.200647
    },
    {
      "acc": 0.77095346,
      "epoch": 1.7880692340220767,
      "grad_norm": 5.0,
      "learning_rate": 2.9148268476435206e-07,
      "loss": 0.82100334,
      "memory(GiB)": 138.1,
      "step": 76640,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.77619829,
      "epoch": 1.7883025415943656,
      "grad_norm": 7.4375,
      "learning_rate": 2.908474405100542e-07,
      "loss": 0.82527828,
      "memory(GiB)": 138.1,
      "step": 76650,
      "train_speed(iter/s)": 0.200675
    },
    {
      "acc": 0.77772455,
      "epoch": 1.7885358491666545,
      "grad_norm": 3.921875,
      "learning_rate": 2.9021286848591626e-07,
      "loss": 0.79985743,
      "memory(GiB)": 138.1,
      "step": 76660,
      "train_speed(iter/s)": 0.200688
    },
    {
      "acc": 0.78497052,
      "epoch": 1.7887691567389434,
      "grad_norm": 6.3125,
      "learning_rate": 2.895789687825218e-07,
      "loss": 0.77265472,
      "memory(GiB)": 138.1,
      "step": 76670,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.78477559,
      "epoch": 1.7890024643112323,
      "grad_norm": 5.6875,
      "learning_rate": 2.8894574149036237e-07,
      "loss": 0.77249074,
      "memory(GiB)": 138.1,
      "step": 76680,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.78217597,
      "epoch": 1.7892357718835212,
      "grad_norm": 5.3125,
      "learning_rate": 2.8831318669982956e-07,
      "loss": 0.78129129,
      "memory(GiB)": 138.1,
      "step": 76690,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.78776398,
      "epoch": 1.7894690794558101,
      "grad_norm": 4.5,
      "learning_rate": 2.876813045012211e-07,
      "loss": 0.76364908,
      "memory(GiB)": 138.1,
      "step": 76700,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.78187065,
      "epoch": 1.789702387028099,
      "grad_norm": 6.8125,
      "learning_rate": 2.8705009498473604e-07,
      "loss": 0.77821112,
      "memory(GiB)": 138.1,
      "step": 76710,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.78065195,
      "epoch": 1.789935694600388,
      "grad_norm": 12.5625,
      "learning_rate": 2.8641955824048216e-07,
      "loss": 0.78781352,
      "memory(GiB)": 138.1,
      "step": 76720,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.75964537,
      "epoch": 1.7901690021726768,
      "grad_norm": 5.03125,
      "learning_rate": 2.857896943584665e-07,
      "loss": 0.86047745,
      "memory(GiB)": 138.1,
      "step": 76730,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.76801662,
      "epoch": 1.7904023097449655,
      "grad_norm": 7.28125,
      "learning_rate": 2.85160503428602e-07,
      "loss": 0.84608536,
      "memory(GiB)": 138.1,
      "step": 76740,
      "train_speed(iter/s)": 0.200797
    },
    {
      "acc": 0.78553681,
      "epoch": 1.7906356173172546,
      "grad_norm": 5.1875,
      "learning_rate": 2.8453198554070694e-07,
      "loss": 0.78472672,
      "memory(GiB)": 138.1,
      "step": 76750,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.7960628,
      "epoch": 1.7908689248895433,
      "grad_norm": 5.34375,
      "learning_rate": 2.8390414078450003e-07,
      "loss": 0.71955853,
      "memory(GiB)": 138.1,
      "step": 76760,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.78666868,
      "epoch": 1.7911022324618324,
      "grad_norm": 13.3125,
      "learning_rate": 2.8327696924960737e-07,
      "loss": 0.76741972,
      "memory(GiB)": 138.1,
      "step": 76770,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.7942173,
      "epoch": 1.791335540034121,
      "grad_norm": 5.59375,
      "learning_rate": 2.8265047102555733e-07,
      "loss": 0.75203381,
      "memory(GiB)": 138.1,
      "step": 76780,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.79992652,
      "epoch": 1.7915688476064102,
      "grad_norm": 6.03125,
      "learning_rate": 2.8202464620178225e-07,
      "loss": 0.70470304,
      "memory(GiB)": 138.1,
      "step": 76790,
      "train_speed(iter/s)": 0.200864
    },
    {
      "acc": 0.80589857,
      "epoch": 1.791802155178699,
      "grad_norm": 4.15625,
      "learning_rate": 2.8139949486761853e-07,
      "loss": 0.67734222,
      "memory(GiB)": 138.1,
      "step": 76800,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.78081474,
      "epoch": 1.792035462750988,
      "grad_norm": 5.9375,
      "learning_rate": 2.807750171123058e-07,
      "loss": 0.78283153,
      "memory(GiB)": 138.1,
      "step": 76810,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.78747606,
      "epoch": 1.7922687703232767,
      "grad_norm": 6.5625,
      "learning_rate": 2.8015121302498894e-07,
      "loss": 0.75321217,
      "memory(GiB)": 138.1,
      "step": 76820,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.78687334,
      "epoch": 1.7925020778955658,
      "grad_norm": 3.875,
      "learning_rate": 2.7952808269471445e-07,
      "loss": 0.75753322,
      "memory(GiB)": 138.1,
      "step": 76830,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.77859888,
      "epoch": 1.7927353854678545,
      "grad_norm": 6.28125,
      "learning_rate": 2.7890562621043503e-07,
      "loss": 0.7866539,
      "memory(GiB)": 138.1,
      "step": 76840,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.78868618,
      "epoch": 1.7929686930401436,
      "grad_norm": 4.1875,
      "learning_rate": 2.782838436610058e-07,
      "loss": 0.75083804,
      "memory(GiB)": 138.1,
      "step": 76850,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.76023216,
      "epoch": 1.7932020006124323,
      "grad_norm": 6.53125,
      "learning_rate": 2.776627351351868e-07,
      "loss": 0.87442131,
      "memory(GiB)": 138.1,
      "step": 76860,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.7890811,
      "epoch": 1.7934353081847214,
      "grad_norm": 4.96875,
      "learning_rate": 2.770423007216411e-07,
      "loss": 0.75472221,
      "memory(GiB)": 138.1,
      "step": 76870,
      "train_speed(iter/s)": 0.200973
    },
    {
      "acc": 0.77964153,
      "epoch": 1.79366861575701,
      "grad_norm": 5.875,
      "learning_rate": 2.764225405089332e-07,
      "loss": 0.80402193,
      "memory(GiB)": 138.1,
      "step": 76880,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.76858959,
      "epoch": 1.7939019233292992,
      "grad_norm": 5.3125,
      "learning_rate": 2.7580345458553705e-07,
      "loss": 0.84677277,
      "memory(GiB)": 138.1,
      "step": 76890,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.78171978,
      "epoch": 1.7941352309015879,
      "grad_norm": 6.0625,
      "learning_rate": 2.75185043039824e-07,
      "loss": 0.78139849,
      "memory(GiB)": 138.1,
      "step": 76900,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.76607041,
      "epoch": 1.794368538473877,
      "grad_norm": 4.9375,
      "learning_rate": 2.7456730596007454e-07,
      "loss": 0.83766766,
      "memory(GiB)": 138.1,
      "step": 76910,
      "train_speed(iter/s)": 0.201027
    },
    {
      "acc": 0.79166379,
      "epoch": 1.7946018460461657,
      "grad_norm": 4.75,
      "learning_rate": 2.739502434344693e-07,
      "loss": 0.73072734,
      "memory(GiB)": 138.1,
      "step": 76920,
      "train_speed(iter/s)": 0.201041
    },
    {
      "acc": 0.7861433,
      "epoch": 1.7948351536184546,
      "grad_norm": 5.5,
      "learning_rate": 2.733338555510939e-07,
      "loss": 0.77780027,
      "memory(GiB)": 138.1,
      "step": 76930,
      "train_speed(iter/s)": 0.201054
    },
    {
      "acc": 0.78073063,
      "epoch": 1.7950684611907435,
      "grad_norm": 4.625,
      "learning_rate": 2.7271814239793693e-07,
      "loss": 0.77235231,
      "memory(GiB)": 138.1,
      "step": 76940,
      "train_speed(iter/s)": 0.201067
    },
    {
      "acc": 0.79449263,
      "epoch": 1.7953017687630324,
      "grad_norm": 6.34375,
      "learning_rate": 2.721031040628924e-07,
      "loss": 0.74123163,
      "memory(GiB)": 138.1,
      "step": 76950,
      "train_speed(iter/s)": 0.20108
    },
    {
      "acc": 0.77358427,
      "epoch": 1.7955350763353213,
      "grad_norm": 6.84375,
      "learning_rate": 2.714887406337563e-07,
      "loss": 0.8185647,
      "memory(GiB)": 138.1,
      "step": 76960,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.77752705,
      "epoch": 1.7957683839076102,
      "grad_norm": 5.125,
      "learning_rate": 2.708750521982284e-07,
      "loss": 0.80426521,
      "memory(GiB)": 138.1,
      "step": 76970,
      "train_speed(iter/s)": 0.201107
    },
    {
      "acc": 0.7755044,
      "epoch": 1.796001691479899,
      "grad_norm": 4.78125,
      "learning_rate": 2.7026203884391313e-07,
      "loss": 0.81011715,
      "memory(GiB)": 138.1,
      "step": 76980,
      "train_speed(iter/s)": 0.201121
    },
    {
      "acc": 0.78045125,
      "epoch": 1.796234999052188,
      "grad_norm": 5.21875,
      "learning_rate": 2.696497006583176e-07,
      "loss": 0.79744205,
      "memory(GiB)": 138.1,
      "step": 76990,
      "train_speed(iter/s)": 0.201135
    },
    {
      "acc": 0.77946234,
      "epoch": 1.7964683066244769,
      "grad_norm": 5.09375,
      "learning_rate": 2.6903803772885375e-07,
      "loss": 0.79059172,
      "memory(GiB)": 138.1,
      "step": 77000,
      "train_speed(iter/s)": 0.201149
    },
    {
      "epoch": 1.7964683066244769,
      "eval_acc": 0.7447200046249655,
      "eval_loss": 0.8044254779815674,
      "eval_runtime": 1271.9514,
      "eval_samples_per_second": 28.296,
      "eval_steps_per_second": 14.148,
      "step": 77000
    },
    {
      "acc": 0.78546667,
      "epoch": 1.7967016141967658,
      "grad_norm": 6.96875,
      "learning_rate": 2.6842705014283545e-07,
      "loss": 0.76823864,
      "memory(GiB)": 138.1,
      "step": 77010,
      "train_speed(iter/s)": 0.200482
    },
    {
      "acc": 0.77572174,
      "epoch": 1.7969349217690547,
      "grad_norm": 4.21875,
      "learning_rate": 2.6781673798748074e-07,
      "loss": 0.8007781,
      "memory(GiB)": 138.1,
      "step": 77020,
      "train_speed(iter/s)": 0.200497
    },
    {
      "acc": 0.77363772,
      "epoch": 1.7971682293413436,
      "grad_norm": 6.09375,
      "learning_rate": 2.672071013499122e-07,
      "loss": 0.82122746,
      "memory(GiB)": 138.1,
      "step": 77030,
      "train_speed(iter/s)": 0.200509
    },
    {
      "acc": 0.76963663,
      "epoch": 1.7974015369136325,
      "grad_norm": 5.40625,
      "learning_rate": 2.665981403171558e-07,
      "loss": 0.84492531,
      "memory(GiB)": 138.1,
      "step": 77040,
      "train_speed(iter/s)": 0.200522
    },
    {
      "acc": 0.78084383,
      "epoch": 1.7976348444859214,
      "grad_norm": 23.875,
      "learning_rate": 2.6598985497613915e-07,
      "loss": 0.7569222,
      "memory(GiB)": 138.1,
      "step": 77050,
      "train_speed(iter/s)": 0.200536
    },
    {
      "acc": 0.77769785,
      "epoch": 1.7978681520582103,
      "grad_norm": 5.46875,
      "learning_rate": 2.653822454136951e-07,
      "loss": 0.78896704,
      "memory(GiB)": 138.1,
      "step": 77060,
      "train_speed(iter/s)": 0.20055
    },
    {
      "acc": 0.80396709,
      "epoch": 1.7981014596304992,
      "grad_norm": 4.40625,
      "learning_rate": 2.647753117165608e-07,
      "loss": 0.6958436,
      "memory(GiB)": 138.1,
      "step": 77070,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.79132442,
      "epoch": 1.798334767202788,
      "grad_norm": 3.875,
      "learning_rate": 2.641690539713743e-07,
      "loss": 0.75024261,
      "memory(GiB)": 138.1,
      "step": 77080,
      "train_speed(iter/s)": 0.200577
    },
    {
      "acc": 0.78543453,
      "epoch": 1.798568074775077,
      "grad_norm": 5.8125,
      "learning_rate": 2.635634722646807e-07,
      "loss": 0.77864475,
      "memory(GiB)": 138.1,
      "step": 77090,
      "train_speed(iter/s)": 0.200591
    },
    {
      "acc": 0.77233229,
      "epoch": 1.7988013823473659,
      "grad_norm": 4.65625,
      "learning_rate": 2.6295856668292487e-07,
      "loss": 0.81739025,
      "memory(GiB)": 138.1,
      "step": 77100,
      "train_speed(iter/s)": 0.200605
    },
    {
      "acc": 0.76904373,
      "epoch": 1.7990346899196545,
      "grad_norm": 5.65625,
      "learning_rate": 2.623543373124571e-07,
      "loss": 0.81224604,
      "memory(GiB)": 138.1,
      "step": 77110,
      "train_speed(iter/s)": 0.200618
    },
    {
      "acc": 0.78083372,
      "epoch": 1.7992679974919437,
      "grad_norm": 5.90625,
      "learning_rate": 2.6175078423953225e-07,
      "loss": 0.79091768,
      "memory(GiB)": 138.1,
      "step": 77120,
      "train_speed(iter/s)": 0.200631
    },
    {
      "acc": 0.80353069,
      "epoch": 1.7995013050642323,
      "grad_norm": 5.625,
      "learning_rate": 2.6114790755030593e-07,
      "loss": 0.72558646,
      "memory(GiB)": 138.1,
      "step": 77130,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.77595344,
      "epoch": 1.7997346126365215,
      "grad_norm": 6.28125,
      "learning_rate": 2.605457073308393e-07,
      "loss": 0.79515977,
      "memory(GiB)": 138.1,
      "step": 77140,
      "train_speed(iter/s)": 0.200658
    },
    {
      "acc": 0.7906909,
      "epoch": 1.7999679202088101,
      "grad_norm": 6.53125,
      "learning_rate": 2.5994418366709686e-07,
      "loss": 0.75885439,
      "memory(GiB)": 138.1,
      "step": 77150,
      "train_speed(iter/s)": 0.200672
    },
    {
      "acc": 0.76935129,
      "epoch": 1.8002012277810993,
      "grad_norm": 6.25,
      "learning_rate": 2.5934333664494436e-07,
      "loss": 0.84422007,
      "memory(GiB)": 138.1,
      "step": 77160,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.77682958,
      "epoch": 1.800434535353388,
      "grad_norm": 5.28125,
      "learning_rate": 2.5874316635015383e-07,
      "loss": 0.82697449,
      "memory(GiB)": 138.1,
      "step": 77170,
      "train_speed(iter/s)": 0.200699
    },
    {
      "acc": 0.78309669,
      "epoch": 1.800667842925677,
      "grad_norm": 6.25,
      "learning_rate": 2.581436728683984e-07,
      "loss": 0.76987929,
      "memory(GiB)": 138.1,
      "step": 77180,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.79188871,
      "epoch": 1.8009011504979657,
      "grad_norm": 6.625,
      "learning_rate": 2.575448562852567e-07,
      "loss": 0.76328211,
      "memory(GiB)": 138.1,
      "step": 77190,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.7809392,
      "epoch": 1.8011344580702549,
      "grad_norm": 6.65625,
      "learning_rate": 2.569467166862083e-07,
      "loss": 0.79415689,
      "memory(GiB)": 138.1,
      "step": 77200,
      "train_speed(iter/s)": 0.200738
    },
    {
      "acc": 0.77853994,
      "epoch": 1.8013677656425435,
      "grad_norm": 4.875,
      "learning_rate": 2.563492541566387e-07,
      "loss": 0.80777569,
      "memory(GiB)": 138.1,
      "step": 77210,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.78512239,
      "epoch": 1.8016010732148326,
      "grad_norm": 5.09375,
      "learning_rate": 2.557524687818347e-07,
      "loss": 0.76266022,
      "memory(GiB)": 138.1,
      "step": 77220,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.78263688,
      "epoch": 1.8018343807871213,
      "grad_norm": 6.0625,
      "learning_rate": 2.551563606469881e-07,
      "loss": 0.7762639,
      "memory(GiB)": 138.1,
      "step": 77230,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.79786873,
      "epoch": 1.8020676883594104,
      "grad_norm": 5.09375,
      "learning_rate": 2.545609298371926e-07,
      "loss": 0.72444544,
      "memory(GiB)": 138.1,
      "step": 77240,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.78850036,
      "epoch": 1.8023009959316991,
      "grad_norm": 4.3125,
      "learning_rate": 2.539661764374457e-07,
      "loss": 0.75326676,
      "memory(GiB)": 138.1,
      "step": 77250,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.77271185,
      "epoch": 1.8025343035039882,
      "grad_norm": 4.65625,
      "learning_rate": 2.5337210053264893e-07,
      "loss": 0.80799408,
      "memory(GiB)": 138.1,
      "step": 77260,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.76806946,
      "epoch": 1.802767611076277,
      "grad_norm": 5.59375,
      "learning_rate": 2.5277870220760504e-07,
      "loss": 0.85381813,
      "memory(GiB)": 138.1,
      "step": 77270,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.77597227,
      "epoch": 1.803000918648566,
      "grad_norm": 7.71875,
      "learning_rate": 2.521859815470229e-07,
      "loss": 0.81069126,
      "memory(GiB)": 138.1,
      "step": 77280,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.785149,
      "epoch": 1.8032342262208547,
      "grad_norm": 7.34375,
      "learning_rate": 2.515939386355121e-07,
      "loss": 0.76063423,
      "memory(GiB)": 138.1,
      "step": 77290,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.77492638,
      "epoch": 1.8034675337931438,
      "grad_norm": 4.59375,
      "learning_rate": 2.5100257355758715e-07,
      "loss": 0.78454142,
      "memory(GiB)": 138.1,
      "step": 77300,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.77450662,
      "epoch": 1.8037008413654325,
      "grad_norm": 4.96875,
      "learning_rate": 2.5041188639766624e-07,
      "loss": 0.78545609,
      "memory(GiB)": 138.1,
      "step": 77310,
      "train_speed(iter/s)": 0.200885
    },
    {
      "acc": 0.77065268,
      "epoch": 1.8039341489377214,
      "grad_norm": 4.5625,
      "learning_rate": 2.498218772400673e-07,
      "loss": 0.83770046,
      "memory(GiB)": 138.1,
      "step": 77320,
      "train_speed(iter/s)": 0.200898
    },
    {
      "acc": 0.77967792,
      "epoch": 1.8041674565100103,
      "grad_norm": 10.25,
      "learning_rate": 2.4923254616901646e-07,
      "loss": 0.80178738,
      "memory(GiB)": 138.1,
      "step": 77330,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.74775844,
      "epoch": 1.8044007640822992,
      "grad_norm": 5.40625,
      "learning_rate": 2.486438932686386e-07,
      "loss": 0.90301533,
      "memory(GiB)": 138.1,
      "step": 77340,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.7848135,
      "epoch": 1.8046340716545881,
      "grad_norm": 10.9375,
      "learning_rate": 2.4805591862296587e-07,
      "loss": 0.75431414,
      "memory(GiB)": 138.1,
      "step": 77350,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.78480129,
      "epoch": 1.804867379226877,
      "grad_norm": 5.21875,
      "learning_rate": 2.4746862231593006e-07,
      "loss": 0.76673164,
      "memory(GiB)": 138.1,
      "step": 77360,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.77168055,
      "epoch": 1.805100686799166,
      "grad_norm": 6.15625,
      "learning_rate": 2.468820044313669e-07,
      "loss": 0.83345318,
      "memory(GiB)": 138.1,
      "step": 77370,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.77629547,
      "epoch": 1.8053339943714548,
      "grad_norm": 4.65625,
      "learning_rate": 2.4629606505301775e-07,
      "loss": 0.81766853,
      "memory(GiB)": 138.1,
      "step": 77380,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.80229378,
      "epoch": 1.8055673019437437,
      "grad_norm": 4.25,
      "learning_rate": 2.457108042645245e-07,
      "loss": 0.71003008,
      "memory(GiB)": 138.1,
      "step": 77390,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.76214561,
      "epoch": 1.8058006095160326,
      "grad_norm": 5.4375,
      "learning_rate": 2.4512622214943274e-07,
      "loss": 0.87743969,
      "memory(GiB)": 138.1,
      "step": 77400,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.78265681,
      "epoch": 1.8060339170883215,
      "grad_norm": 4.21875,
      "learning_rate": 2.445423187911905e-07,
      "loss": 0.78319178,
      "memory(GiB)": 138.1,
      "step": 77410,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.77975445,
      "epoch": 1.8062672246606104,
      "grad_norm": 6.09375,
      "learning_rate": 2.439590942731518e-07,
      "loss": 0.79706392,
      "memory(GiB)": 138.1,
      "step": 77420,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.80120268,
      "epoch": 1.8065005322328993,
      "grad_norm": 5.5,
      "learning_rate": 2.433765486785694e-07,
      "loss": 0.70945711,
      "memory(GiB)": 138.1,
      "step": 77430,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.76658869,
      "epoch": 1.8067338398051882,
      "grad_norm": 4.1875,
      "learning_rate": 2.427946820906041e-07,
      "loss": 0.84224901,
      "memory(GiB)": 138.1,
      "step": 77440,
      "train_speed(iter/s)": 0.201059
    },
    {
      "acc": 0.78487082,
      "epoch": 1.806967147377477,
      "grad_norm": 7.28125,
      "learning_rate": 2.42213494592316e-07,
      "loss": 0.76289835,
      "memory(GiB)": 138.1,
      "step": 77450,
      "train_speed(iter/s)": 0.201072
    },
    {
      "acc": 0.78675003,
      "epoch": 1.807200454949766,
      "grad_norm": 5.84375,
      "learning_rate": 2.4163298626666885e-07,
      "loss": 0.76830788,
      "memory(GiB)": 138.1,
      "step": 77460,
      "train_speed(iter/s)": 0.201085
    },
    {
      "acc": 0.79696965,
      "epoch": 1.807433762522055,
      "grad_norm": 5.125,
      "learning_rate": 2.410531571965308e-07,
      "loss": 0.73548222,
      "memory(GiB)": 138.1,
      "step": 77470,
      "train_speed(iter/s)": 0.201099
    },
    {
      "acc": 0.76947789,
      "epoch": 1.8076670700943438,
      "grad_norm": 5.34375,
      "learning_rate": 2.4047400746467165e-07,
      "loss": 0.84600887,
      "memory(GiB)": 138.1,
      "step": 77480,
      "train_speed(iter/s)": 0.201112
    },
    {
      "acc": 0.78702273,
      "epoch": 1.8079003776666327,
      "grad_norm": 4.0625,
      "learning_rate": 2.398955371537665e-07,
      "loss": 0.76369286,
      "memory(GiB)": 138.1,
      "step": 77490,
      "train_speed(iter/s)": 0.201126
    },
    {
      "acc": 0.7907228,
      "epoch": 1.8081336852389214,
      "grad_norm": 7.03125,
      "learning_rate": 2.393177463463897e-07,
      "loss": 0.76782665,
      "memory(GiB)": 138.1,
      "step": 77500,
      "train_speed(iter/s)": 0.201137
    },
    {
      "epoch": 1.8081336852389214,
      "eval_acc": 0.7447478697213858,
      "eval_loss": 0.8044267296791077,
      "eval_runtime": 1270.6205,
      "eval_samples_per_second": 28.326,
      "eval_steps_per_second": 14.163,
      "step": 77500
    },
    {
      "acc": 0.79640326,
      "epoch": 1.8083669928112105,
      "grad_norm": 12.0625,
      "learning_rate": 2.387406351250221e-07,
      "loss": 0.71954288,
      "memory(GiB)": 138.1,
      "step": 77510,
      "train_speed(iter/s)": 0.200476
    },
    {
      "acc": 0.78035679,
      "epoch": 1.8086003003834992,
      "grad_norm": 9.3125,
      "learning_rate": 2.3816420357204495e-07,
      "loss": 0.79402161,
      "memory(GiB)": 138.1,
      "step": 77520,
      "train_speed(iter/s)": 0.200489
    },
    {
      "acc": 0.80683222,
      "epoch": 1.8088336079557883,
      "grad_norm": 4.125,
      "learning_rate": 2.3758845176974465e-07,
      "loss": 0.70322952,
      "memory(GiB)": 138.1,
      "step": 77530,
      "train_speed(iter/s)": 0.200502
    },
    {
      "acc": 0.78853745,
      "epoch": 1.809066915528077,
      "grad_norm": 6.34375,
      "learning_rate": 2.3701337980030993e-07,
      "loss": 0.76086707,
      "memory(GiB)": 138.1,
      "step": 77540,
      "train_speed(iter/s)": 0.200515
    },
    {
      "acc": 0.77532892,
      "epoch": 1.809300223100366,
      "grad_norm": 5.6875,
      "learning_rate": 2.3643898774583064e-07,
      "loss": 0.8250165,
      "memory(GiB)": 138.1,
      "step": 77550,
      "train_speed(iter/s)": 0.200528
    },
    {
      "acc": 0.79044323,
      "epoch": 1.8095335306726548,
      "grad_norm": 4.625,
      "learning_rate": 2.3586527568830286e-07,
      "loss": 0.76854286,
      "memory(GiB)": 138.1,
      "step": 77560,
      "train_speed(iter/s)": 0.200543
    },
    {
      "acc": 0.76759367,
      "epoch": 1.8097668382449439,
      "grad_norm": 6.0625,
      "learning_rate": 2.3529224370962223e-07,
      "loss": 0.82133694,
      "memory(GiB)": 138.1,
      "step": 77570,
      "train_speed(iter/s)": 0.200557
    },
    {
      "acc": 0.79010715,
      "epoch": 1.8100001458172326,
      "grad_norm": 5.75,
      "learning_rate": 2.347198918915905e-07,
      "loss": 0.74337749,
      "memory(GiB)": 138.1,
      "step": 77580,
      "train_speed(iter/s)": 0.200571
    },
    {
      "acc": 0.78925729,
      "epoch": 1.8102334533895217,
      "grad_norm": 4.21875,
      "learning_rate": 2.3414822031590956e-07,
      "loss": 0.76391687,
      "memory(GiB)": 138.1,
      "step": 77590,
      "train_speed(iter/s)": 0.200584
    },
    {
      "acc": 0.78881116,
      "epoch": 1.8104667609618104,
      "grad_norm": 5.3125,
      "learning_rate": 2.3357722906418523e-07,
      "loss": 0.74917927,
      "memory(GiB)": 138.1,
      "step": 77600,
      "train_speed(iter/s)": 0.200597
    },
    {
      "acc": 0.77421007,
      "epoch": 1.8107000685340995,
      "grad_norm": 8.5,
      "learning_rate": 2.3300691821792788e-07,
      "loss": 0.81043053,
      "memory(GiB)": 138.1,
      "step": 77610,
      "train_speed(iter/s)": 0.20061
    },
    {
      "acc": 0.79148536,
      "epoch": 1.8109333761063882,
      "grad_norm": 6.21875,
      "learning_rate": 2.3243728785854737e-07,
      "loss": 0.7413094,
      "memory(GiB)": 138.1,
      "step": 77620,
      "train_speed(iter/s)": 0.200622
    },
    {
      "acc": 0.75671234,
      "epoch": 1.8111666836786773,
      "grad_norm": 5.15625,
      "learning_rate": 2.318683380673592e-07,
      "loss": 0.88378201,
      "memory(GiB)": 138.1,
      "step": 77630,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.77831712,
      "epoch": 1.811399991250966,
      "grad_norm": 6.625,
      "learning_rate": 2.313000689255801e-07,
      "loss": 0.779283,
      "memory(GiB)": 138.1,
      "step": 77640,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.77758551,
      "epoch": 1.811633298823255,
      "grad_norm": 5.40625,
      "learning_rate": 2.3073248051433127e-07,
      "loss": 0.79231982,
      "memory(GiB)": 138.1,
      "step": 77650,
      "train_speed(iter/s)": 0.200664
    },
    {
      "acc": 0.79954634,
      "epoch": 1.8118666063955438,
      "grad_norm": 8.3125,
      "learning_rate": 2.301655729146357e-07,
      "loss": 0.70971565,
      "memory(GiB)": 138.1,
      "step": 77660,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.79472017,
      "epoch": 1.8120999139678329,
      "grad_norm": 3.671875,
      "learning_rate": 2.295993462074181e-07,
      "loss": 0.72714438,
      "memory(GiB)": 138.1,
      "step": 77670,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.78361807,
      "epoch": 1.8123332215401216,
      "grad_norm": 9.3125,
      "learning_rate": 2.2903380047350876e-07,
      "loss": 0.76192293,
      "memory(GiB)": 138.1,
      "step": 77680,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.75392261,
      "epoch": 1.8125665291124105,
      "grad_norm": 5.75,
      "learning_rate": 2.284689357936376e-07,
      "loss": 0.89919205,
      "memory(GiB)": 138.1,
      "step": 77690,
      "train_speed(iter/s)": 0.200718
    },
    {
      "acc": 0.77323561,
      "epoch": 1.8127998366846994,
      "grad_norm": 7.1875,
      "learning_rate": 2.2790475224844067e-07,
      "loss": 0.82149096,
      "memory(GiB)": 138.1,
      "step": 77700,
      "train_speed(iter/s)": 0.200732
    },
    {
      "acc": 0.7900878,
      "epoch": 1.8130331442569882,
      "grad_norm": 5.84375,
      "learning_rate": 2.2734124991845352e-07,
      "loss": 0.74978971,
      "memory(GiB)": 138.1,
      "step": 77710,
      "train_speed(iter/s)": 0.200745
    },
    {
      "acc": 0.77167468,
      "epoch": 1.8132664518292771,
      "grad_norm": 6.09375,
      "learning_rate": 2.2677842888411738e-07,
      "loss": 0.83236217,
      "memory(GiB)": 138.1,
      "step": 77720,
      "train_speed(iter/s)": 0.200759
    },
    {
      "acc": 0.75530715,
      "epoch": 1.813499759401566,
      "grad_norm": 4.59375,
      "learning_rate": 2.26216289225773e-07,
      "loss": 0.8862648,
      "memory(GiB)": 138.1,
      "step": 77730,
      "train_speed(iter/s)": 0.200772
    },
    {
      "acc": 0.79051895,
      "epoch": 1.813733066973855,
      "grad_norm": 5.15625,
      "learning_rate": 2.256548310236667e-07,
      "loss": 0.74030609,
      "memory(GiB)": 138.1,
      "step": 77740,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.79147606,
      "epoch": 1.8139663745461438,
      "grad_norm": 4.6875,
      "learning_rate": 2.2509405435794662e-07,
      "loss": 0.72963748,
      "memory(GiB)": 138.1,
      "step": 77750,
      "train_speed(iter/s)": 0.200797
    },
    {
      "acc": 0.77560072,
      "epoch": 1.8141996821184327,
      "grad_norm": 6.15625,
      "learning_rate": 2.2453395930866262e-07,
      "loss": 0.8010334,
      "memory(GiB)": 138.1,
      "step": 77760,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.7697216,
      "epoch": 1.8144329896907216,
      "grad_norm": 5.25,
      "learning_rate": 2.2397454595576906e-07,
      "loss": 0.82412586,
      "memory(GiB)": 138.1,
      "step": 77770,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.79263906,
      "epoch": 1.8146662972630105,
      "grad_norm": 4.71875,
      "learning_rate": 2.2341581437912097e-07,
      "loss": 0.73385839,
      "memory(GiB)": 138.1,
      "step": 77780,
      "train_speed(iter/s)": 0.200837
    },
    {
      "acc": 0.77369003,
      "epoch": 1.8148996048352994,
      "grad_norm": 9.0625,
      "learning_rate": 2.2285776465847842e-07,
      "loss": 0.78555937,
      "memory(GiB)": 138.1,
      "step": 77790,
      "train_speed(iter/s)": 0.20085
    },
    {
      "acc": 0.81455336,
      "epoch": 1.8151329124075883,
      "grad_norm": 5.625,
      "learning_rate": 2.2230039687350212e-07,
      "loss": 0.65586133,
      "memory(GiB)": 138.1,
      "step": 77800,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.79386964,
      "epoch": 1.8153662199798772,
      "grad_norm": 5.46875,
      "learning_rate": 2.217437111037557e-07,
      "loss": 0.74161119,
      "memory(GiB)": 138.1,
      "step": 77810,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.784272,
      "epoch": 1.8155995275521661,
      "grad_norm": 5.90625,
      "learning_rate": 2.211877074287072e-07,
      "loss": 0.76136417,
      "memory(GiB)": 138.1,
      "step": 77820,
      "train_speed(iter/s)": 0.200892
    },
    {
      "acc": 0.78744574,
      "epoch": 1.815832835124455,
      "grad_norm": 7.90625,
      "learning_rate": 2.206323859277254e-07,
      "loss": 0.77874732,
      "memory(GiB)": 138.1,
      "step": 77830,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.78655195,
      "epoch": 1.816066142696744,
      "grad_norm": 4.03125,
      "learning_rate": 2.200777466800813e-07,
      "loss": 0.76195555,
      "memory(GiB)": 138.1,
      "step": 77840,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.77641582,
      "epoch": 1.8162994502690328,
      "grad_norm": 7.09375,
      "learning_rate": 2.1952378976495048e-07,
      "loss": 0.82479191,
      "memory(GiB)": 138.1,
      "step": 77850,
      "train_speed(iter/s)": 0.200932
    },
    {
      "acc": 0.78586106,
      "epoch": 1.8165327578413217,
      "grad_norm": 5.34375,
      "learning_rate": 2.1897051526141022e-07,
      "loss": 0.76681871,
      "memory(GiB)": 138.1,
      "step": 77860,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.78540492,
      "epoch": 1.8167660654136106,
      "grad_norm": 7.5625,
      "learning_rate": 2.1841792324843958e-07,
      "loss": 0.78315992,
      "memory(GiB)": 138.1,
      "step": 77870,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.78267908,
      "epoch": 1.8169993729858995,
      "grad_norm": 6.71875,
      "learning_rate": 2.1786601380492156e-07,
      "loss": 0.79527521,
      "memory(GiB)": 138.1,
      "step": 77880,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.78243675,
      "epoch": 1.8172326805581882,
      "grad_norm": 6.21875,
      "learning_rate": 2.1731478700964093e-07,
      "loss": 0.78953333,
      "memory(GiB)": 138.1,
      "step": 77890,
      "train_speed(iter/s)": 0.200986
    },
    {
      "acc": 0.77621989,
      "epoch": 1.8174659881304773,
      "grad_norm": 7.5625,
      "learning_rate": 2.1676424294128474e-07,
      "loss": 0.78211274,
      "memory(GiB)": 138.1,
      "step": 77900,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.79703307,
      "epoch": 1.817699295702766,
      "grad_norm": 6.5,
      "learning_rate": 2.16214381678444e-07,
      "loss": 0.70052767,
      "memory(GiB)": 138.1,
      "step": 77910,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.78100514,
      "epoch": 1.8179326032750551,
      "grad_norm": 6.1875,
      "learning_rate": 2.1566520329960928e-07,
      "loss": 0.76739492,
      "memory(GiB)": 138.1,
      "step": 77920,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.78988457,
      "epoch": 1.8181659108473438,
      "grad_norm": 6.75,
      "learning_rate": 2.1511670788317839e-07,
      "loss": 0.75334415,
      "memory(GiB)": 138.1,
      "step": 77930,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.79587126,
      "epoch": 1.818399218419633,
      "grad_norm": 6.25,
      "learning_rate": 2.1456889550744707e-07,
      "loss": 0.7218152,
      "memory(GiB)": 138.1,
      "step": 77940,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.76245584,
      "epoch": 1.8186325259919216,
      "grad_norm": 6.53125,
      "learning_rate": 2.1402176625061554e-07,
      "loss": 0.88262806,
      "memory(GiB)": 138.1,
      "step": 77950,
      "train_speed(iter/s)": 0.201066
    },
    {
      "acc": 0.77558794,
      "epoch": 1.8188658335642107,
      "grad_norm": 6.09375,
      "learning_rate": 2.1347532019078686e-07,
      "loss": 0.80598965,
      "memory(GiB)": 138.1,
      "step": 77960,
      "train_speed(iter/s)": 0.201079
    },
    {
      "acc": 0.77870779,
      "epoch": 1.8190991411364994,
      "grad_norm": 6.96875,
      "learning_rate": 2.1292955740596478e-07,
      "loss": 0.79289889,
      "memory(GiB)": 138.1,
      "step": 77970,
      "train_speed(iter/s)": 0.201093
    },
    {
      "acc": 0.77847414,
      "epoch": 1.8193324487087885,
      "grad_norm": 5.71875,
      "learning_rate": 2.123844779740586e-07,
      "loss": 0.80017395,
      "memory(GiB)": 138.1,
      "step": 77980,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.77644243,
      "epoch": 1.8195657562810772,
      "grad_norm": 8.125,
      "learning_rate": 2.118400819728772e-07,
      "loss": 0.80827723,
      "memory(GiB)": 138.1,
      "step": 77990,
      "train_speed(iter/s)": 0.20112
    },
    {
      "acc": 0.79161921,
      "epoch": 1.8197990638533663,
      "grad_norm": 8.25,
      "learning_rate": 2.1129636948013287e-07,
      "loss": 0.75301976,
      "memory(GiB)": 138.1,
      "step": 78000,
      "train_speed(iter/s)": 0.201134
    },
    {
      "epoch": 1.8197990638533663,
      "eval_acc": 0.7446887764996667,
      "eval_loss": 0.8044082522392273,
      "eval_runtime": 1271.6729,
      "eval_samples_per_second": 28.302,
      "eval_steps_per_second": 14.151,
      "step": 78000
    },
    {
      "acc": 0.78124561,
      "epoch": 1.820032371425655,
      "grad_norm": 5.125,
      "learning_rate": 2.1075334057344077e-07,
      "loss": 0.79524374,
      "memory(GiB)": 138.1,
      "step": 78010,
      "train_speed(iter/s)": 0.200477
    },
    {
      "acc": 0.80066538,
      "epoch": 1.8202656789979441,
      "grad_norm": 4.03125,
      "learning_rate": 2.102109953303183e-07,
      "loss": 0.70290136,
      "memory(GiB)": 138.1,
      "step": 78020,
      "train_speed(iter/s)": 0.20049
    },
    {
      "acc": 0.76704898,
      "epoch": 1.8204989865702328,
      "grad_norm": 7.71875,
      "learning_rate": 2.0966933382818465e-07,
      "loss": 0.86864662,
      "memory(GiB)": 138.1,
      "step": 78030,
      "train_speed(iter/s)": 0.200504
    },
    {
      "acc": 0.78443837,
      "epoch": 1.820732294142522,
      "grad_norm": 6.0,
      "learning_rate": 2.0912835614436132e-07,
      "loss": 0.78189793,
      "memory(GiB)": 138.1,
      "step": 78040,
      "train_speed(iter/s)": 0.200517
    },
    {
      "acc": 0.76247149,
      "epoch": 1.8209656017148106,
      "grad_norm": 5.9375,
      "learning_rate": 2.085880623560743e-07,
      "loss": 0.85706472,
      "memory(GiB)": 138.1,
      "step": 78050,
      "train_speed(iter/s)": 0.200531
    },
    {
      "acc": 0.79499025,
      "epoch": 1.8211989092870997,
      "grad_norm": 5.46875,
      "learning_rate": 2.080484525404497e-07,
      "loss": 0.73801851,
      "memory(GiB)": 138.1,
      "step": 78060,
      "train_speed(iter/s)": 0.200544
    },
    {
      "acc": 0.77147074,
      "epoch": 1.8214322168593884,
      "grad_norm": 4.8125,
      "learning_rate": 2.0750952677451585e-07,
      "loss": 0.82764015,
      "memory(GiB)": 138.1,
      "step": 78070,
      "train_speed(iter/s)": 0.200558
    },
    {
      "acc": 0.78612347,
      "epoch": 1.8216655244316773,
      "grad_norm": 8.4375,
      "learning_rate": 2.069712851352046e-07,
      "loss": 0.74784603,
      "memory(GiB)": 138.1,
      "step": 78080,
      "train_speed(iter/s)": 0.200572
    },
    {
      "acc": 0.79965754,
      "epoch": 1.8218988320039662,
      "grad_norm": 7.21875,
      "learning_rate": 2.0643372769935055e-07,
      "loss": 0.71098585,
      "memory(GiB)": 138.1,
      "step": 78090,
      "train_speed(iter/s)": 0.200585
    },
    {
      "acc": 0.79127426,
      "epoch": 1.822132139576255,
      "grad_norm": 8.9375,
      "learning_rate": 2.0589685454368957e-07,
      "loss": 0.75095186,
      "memory(GiB)": 138.1,
      "step": 78100,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.79673557,
      "epoch": 1.822365447148544,
      "grad_norm": 6.96875,
      "learning_rate": 2.0536066574485868e-07,
      "loss": 0.72113428,
      "memory(GiB)": 138.1,
      "step": 78110,
      "train_speed(iter/s)": 0.200612
    },
    {
      "acc": 0.79205379,
      "epoch": 1.8225987547208329,
      "grad_norm": 4.1875,
      "learning_rate": 2.0482516137940113e-07,
      "loss": 0.73350883,
      "memory(GiB)": 138.1,
      "step": 78120,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.76067524,
      "epoch": 1.8228320622931218,
      "grad_norm": 5.125,
      "learning_rate": 2.0429034152375793e-07,
      "loss": 0.87135782,
      "memory(GiB)": 138.1,
      "step": 78130,
      "train_speed(iter/s)": 0.200639
    },
    {
      "acc": 0.77361522,
      "epoch": 1.8230653698654107,
      "grad_norm": 4.78125,
      "learning_rate": 2.0375620625427584e-07,
      "loss": 0.79716196,
      "memory(GiB)": 138.1,
      "step": 78140,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.76916037,
      "epoch": 1.8232986774376996,
      "grad_norm": 4.46875,
      "learning_rate": 2.0322275564720163e-07,
      "loss": 0.81197529,
      "memory(GiB)": 138.1,
      "step": 78150,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.78346782,
      "epoch": 1.8235319850099885,
      "grad_norm": 6.6875,
      "learning_rate": 2.0268998977868603e-07,
      "loss": 0.79479795,
      "memory(GiB)": 138.1,
      "step": 78160,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.78680763,
      "epoch": 1.8237652925822774,
      "grad_norm": 6.25,
      "learning_rate": 2.0215790872478048e-07,
      "loss": 0.75535355,
      "memory(GiB)": 138.1,
      "step": 78170,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.78493023,
      "epoch": 1.8239986001545663,
      "grad_norm": 7.21875,
      "learning_rate": 2.016265125614403e-07,
      "loss": 0.79210768,
      "memory(GiB)": 138.1,
      "step": 78180,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77242117,
      "epoch": 1.8242319077268552,
      "grad_norm": 5.4375,
      "learning_rate": 2.010958013645209e-07,
      "loss": 0.8096137,
      "memory(GiB)": 138.1,
      "step": 78190,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.79208431,
      "epoch": 1.824465215299144,
      "grad_norm": 4.40625,
      "learning_rate": 2.0056577520978171e-07,
      "loss": 0.7458868,
      "memory(GiB)": 138.1,
      "step": 78200,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.76901507,
      "epoch": 1.824698522871433,
      "grad_norm": 4.90625,
      "learning_rate": 2.0003643417288386e-07,
      "loss": 0.82966356,
      "memory(GiB)": 138.1,
      "step": 78210,
      "train_speed(iter/s)": 0.200743
    },
    {
      "acc": 0.7897049,
      "epoch": 1.8249318304437219,
      "grad_norm": 7.875,
      "learning_rate": 1.9950777832939029e-07,
      "loss": 0.74858961,
      "memory(GiB)": 138.1,
      "step": 78220,
      "train_speed(iter/s)": 0.200756
    },
    {
      "acc": 0.76055508,
      "epoch": 1.8251651380160108,
      "grad_norm": 5.96875,
      "learning_rate": 1.9897980775476778e-07,
      "loss": 0.87515984,
      "memory(GiB)": 138.1,
      "step": 78230,
      "train_speed(iter/s)": 0.20077
    },
    {
      "acc": 0.79980593,
      "epoch": 1.8253984455882997,
      "grad_norm": 6.59375,
      "learning_rate": 1.9845252252438274e-07,
      "loss": 0.70951848,
      "memory(GiB)": 138.1,
      "step": 78240,
      "train_speed(iter/s)": 0.200783
    },
    {
      "acc": 0.79019027,
      "epoch": 1.8256317531605886,
      "grad_norm": 5.4375,
      "learning_rate": 1.9792592271350443e-07,
      "loss": 0.73835411,
      "memory(GiB)": 138.1,
      "step": 78250,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.76858778,
      "epoch": 1.8258650607328775,
      "grad_norm": 5.375,
      "learning_rate": 1.9740000839730656e-07,
      "loss": 0.82827644,
      "memory(GiB)": 138.1,
      "step": 78260,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.77280111,
      "epoch": 1.8260983683051664,
      "grad_norm": 5.46875,
      "learning_rate": 1.9687477965086132e-07,
      "loss": 0.81801071,
      "memory(GiB)": 138.1,
      "step": 78270,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.7763041,
      "epoch": 1.826331675877455,
      "grad_norm": 6.71875,
      "learning_rate": 1.963502365491471e-07,
      "loss": 0.82029505,
      "memory(GiB)": 138.1,
      "step": 78280,
      "train_speed(iter/s)": 0.200838
    },
    {
      "acc": 0.77627983,
      "epoch": 1.8265649834497442,
      "grad_norm": 4.65625,
      "learning_rate": 1.958263791670406e-07,
      "loss": 0.80829754,
      "memory(GiB)": 138.1,
      "step": 78290,
      "train_speed(iter/s)": 0.200851
    },
    {
      "acc": 0.80785828,
      "epoch": 1.8267982910220328,
      "grad_norm": 5.59375,
      "learning_rate": 1.953032075793232e-07,
      "loss": 0.68023977,
      "memory(GiB)": 138.1,
      "step": 78300,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.75986423,
      "epoch": 1.827031598594322,
      "grad_norm": 10.4375,
      "learning_rate": 1.9478072186067676e-07,
      "loss": 0.88966808,
      "memory(GiB)": 138.1,
      "step": 78310,
      "train_speed(iter/s)": 0.200879
    },
    {
      "acc": 0.77503624,
      "epoch": 1.8272649061666106,
      "grad_norm": 4.15625,
      "learning_rate": 1.9425892208568665e-07,
      "loss": 0.80476112,
      "memory(GiB)": 138.1,
      "step": 78320,
      "train_speed(iter/s)": 0.200892
    },
    {
      "acc": 0.77604232,
      "epoch": 1.8274982137388998,
      "grad_norm": 6.875,
      "learning_rate": 1.9373780832883937e-07,
      "loss": 0.82170696,
      "memory(GiB)": 138.1,
      "step": 78330,
      "train_speed(iter/s)": 0.200905
    },
    {
      "acc": 0.77696238,
      "epoch": 1.8277315213111884,
      "grad_norm": 4.34375,
      "learning_rate": 1.9321738066452266e-07,
      "loss": 0.80580139,
      "memory(GiB)": 138.1,
      "step": 78340,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.77858658,
      "epoch": 1.8279648288834776,
      "grad_norm": 9.375,
      "learning_rate": 1.9269763916702988e-07,
      "loss": 0.80092487,
      "memory(GiB)": 138.1,
      "step": 78350,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.78533554,
      "epoch": 1.8281981364557662,
      "grad_norm": 7.9375,
      "learning_rate": 1.9217858391055167e-07,
      "loss": 0.75122461,
      "memory(GiB)": 138.1,
      "step": 78360,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.7888196,
      "epoch": 1.8284314440280554,
      "grad_norm": 5.0625,
      "learning_rate": 1.916602149691843e-07,
      "loss": 0.74484324,
      "memory(GiB)": 138.1,
      "step": 78370,
      "train_speed(iter/s)": 0.200958
    },
    {
      "acc": 0.76858897,
      "epoch": 1.828664751600344,
      "grad_norm": 4.90625,
      "learning_rate": 1.911425324169247e-07,
      "loss": 0.82479372,
      "memory(GiB)": 138.1,
      "step": 78380,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.77135515,
      "epoch": 1.8288980591726332,
      "grad_norm": 7.0,
      "learning_rate": 1.90625536327671e-07,
      "loss": 0.83313236,
      "memory(GiB)": 138.1,
      "step": 78390,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.79445343,
      "epoch": 1.8291313667449218,
      "grad_norm": 5.5625,
      "learning_rate": 1.9010922677522525e-07,
      "loss": 0.74220562,
      "memory(GiB)": 138.1,
      "step": 78400,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.78015499,
      "epoch": 1.829364674317211,
      "grad_norm": 5.59375,
      "learning_rate": 1.8959360383329017e-07,
      "loss": 0.79001665,
      "memory(GiB)": 138.1,
      "step": 78410,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.76368427,
      "epoch": 1.8295979818894996,
      "grad_norm": 7.40625,
      "learning_rate": 1.8907866757547077e-07,
      "loss": 0.86557426,
      "memory(GiB)": 138.1,
      "step": 78420,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.79951019,
      "epoch": 1.8298312894617887,
      "grad_norm": 6.25,
      "learning_rate": 1.8856441807527325e-07,
      "loss": 0.71126404,
      "memory(GiB)": 138.1,
      "step": 78430,
      "train_speed(iter/s)": 0.201039
    },
    {
      "acc": 0.77727337,
      "epoch": 1.8300645970340774,
      "grad_norm": 5.78125,
      "learning_rate": 1.8805085540610836e-07,
      "loss": 0.78865066,
      "memory(GiB)": 138.1,
      "step": 78440,
      "train_speed(iter/s)": 0.201053
    },
    {
      "acc": 0.8113802,
      "epoch": 1.8302979046063665,
      "grad_norm": 14.6875,
      "learning_rate": 1.8753797964128573e-07,
      "loss": 0.68480577,
      "memory(GiB)": 138.1,
      "step": 78450,
      "train_speed(iter/s)": 0.201065
    },
    {
      "acc": 0.78258839,
      "epoch": 1.8305312121786552,
      "grad_norm": 5.0625,
      "learning_rate": 1.870257908540174e-07,
      "loss": 0.79557185,
      "memory(GiB)": 138.1,
      "step": 78460,
      "train_speed(iter/s)": 0.201078
    },
    {
      "acc": 0.78330555,
      "epoch": 1.8307645197509441,
      "grad_norm": 7.375,
      "learning_rate": 1.8651428911742043e-07,
      "loss": 0.77810469,
      "memory(GiB)": 138.1,
      "step": 78470,
      "train_speed(iter/s)": 0.201092
    },
    {
      "acc": 0.78139677,
      "epoch": 1.830997827323233,
      "grad_norm": 4.90625,
      "learning_rate": 1.8600347450450974e-07,
      "loss": 0.77297707,
      "memory(GiB)": 138.1,
      "step": 78480,
      "train_speed(iter/s)": 0.201106
    },
    {
      "acc": 0.79237366,
      "epoch": 1.831231134895522,
      "grad_norm": 7.4375,
      "learning_rate": 1.8549334708820476e-07,
      "loss": 0.73578663,
      "memory(GiB)": 138.1,
      "step": 78490,
      "train_speed(iter/s)": 0.201118
    },
    {
      "acc": 0.78118787,
      "epoch": 1.8314644424678108,
      "grad_norm": 3.578125,
      "learning_rate": 1.8498390694132562e-07,
      "loss": 0.77322454,
      "memory(GiB)": 138.1,
      "step": 78500,
      "train_speed(iter/s)": 0.201132
    },
    {
      "epoch": 1.8314644424678108,
      "eval_acc": 0.7446793279899608,
      "eval_loss": 0.8043855428695679,
      "eval_runtime": 1271.7476,
      "eval_samples_per_second": 28.3,
      "eval_steps_per_second": 14.151,
      "step": 78500
    },
    {
      "acc": 0.79847651,
      "epoch": 1.8316977500400997,
      "grad_norm": 6.1875,
      "learning_rate": 1.8447515413659578e-07,
      "loss": 0.71289806,
      "memory(GiB)": 138.1,
      "step": 78510,
      "train_speed(iter/s)": 0.200478
    },
    {
      "acc": 0.76541691,
      "epoch": 1.8319310576123886,
      "grad_norm": 5.40625,
      "learning_rate": 1.8396708874663826e-07,
      "loss": 0.84451513,
      "memory(GiB)": 138.1,
      "step": 78520,
      "train_speed(iter/s)": 0.200492
    },
    {
      "acc": 0.75975571,
      "epoch": 1.8321643651846775,
      "grad_norm": 5.875,
      "learning_rate": 1.8345971084398007e-07,
      "loss": 0.87615738,
      "memory(GiB)": 138.1,
      "step": 78530,
      "train_speed(iter/s)": 0.200504
    },
    {
      "acc": 0.76963463,
      "epoch": 1.8323976727569664,
      "grad_norm": 6.0625,
      "learning_rate": 1.829530205010488e-07,
      "loss": 0.83870296,
      "memory(GiB)": 138.1,
      "step": 78540,
      "train_speed(iter/s)": 0.200518
    },
    {
      "acc": 0.7904026,
      "epoch": 1.8326309803292553,
      "grad_norm": 5.1875,
      "learning_rate": 1.8244701779017438e-07,
      "loss": 0.74433708,
      "memory(GiB)": 138.1,
      "step": 78550,
      "train_speed(iter/s)": 0.200531
    },
    {
      "acc": 0.78499613,
      "epoch": 1.8328642879015442,
      "grad_norm": 6.28125,
      "learning_rate": 1.8194170278358847e-07,
      "loss": 0.77916923,
      "memory(GiB)": 138.1,
      "step": 78560,
      "train_speed(iter/s)": 0.200544
    },
    {
      "acc": 0.76727276,
      "epoch": 1.833097595473833,
      "grad_norm": 4.5,
      "learning_rate": 1.8143707555342504e-07,
      "loss": 0.84287758,
      "memory(GiB)": 138.1,
      "step": 78570,
      "train_speed(iter/s)": 0.200557
    },
    {
      "acc": 0.78795671,
      "epoch": 1.833330903046122,
      "grad_norm": 3.9375,
      "learning_rate": 1.8093313617171927e-07,
      "loss": 0.77323265,
      "memory(GiB)": 138.1,
      "step": 78580,
      "train_speed(iter/s)": 0.20057
    },
    {
      "acc": 0.76881981,
      "epoch": 1.833564210618411,
      "grad_norm": 6.40625,
      "learning_rate": 1.8042988471040856e-07,
      "loss": 0.84690332,
      "memory(GiB)": 138.1,
      "step": 78590,
      "train_speed(iter/s)": 0.200583
    },
    {
      "acc": 0.78443508,
      "epoch": 1.8337975181906998,
      "grad_norm": 4.84375,
      "learning_rate": 1.7992732124133106e-07,
      "loss": 0.78077679,
      "memory(GiB)": 138.1,
      "step": 78600,
      "train_speed(iter/s)": 0.200596
    },
    {
      "acc": 0.79543123,
      "epoch": 1.8340308257629887,
      "grad_norm": 6.59375,
      "learning_rate": 1.7942544583622878e-07,
      "loss": 0.73268194,
      "memory(GiB)": 138.1,
      "step": 78610,
      "train_speed(iter/s)": 0.200609
    },
    {
      "acc": 0.77256894,
      "epoch": 1.8342641333352776,
      "grad_norm": 4.75,
      "learning_rate": 1.7892425856674334e-07,
      "loss": 0.83725958,
      "memory(GiB)": 138.1,
      "step": 78620,
      "train_speed(iter/s)": 0.200622
    },
    {
      "acc": 0.78704338,
      "epoch": 1.8344974409075665,
      "grad_norm": 6.09375,
      "learning_rate": 1.7842375950442025e-07,
      "loss": 0.75788522,
      "memory(GiB)": 138.1,
      "step": 78630,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.76816063,
      "epoch": 1.8347307484798554,
      "grad_norm": 5.96875,
      "learning_rate": 1.7792394872070407e-07,
      "loss": 0.82397146,
      "memory(GiB)": 138.1,
      "step": 78640,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.79935145,
      "epoch": 1.834964056052144,
      "grad_norm": 6.21875,
      "learning_rate": 1.7742482628694379e-07,
      "loss": 0.72347355,
      "memory(GiB)": 138.1,
      "step": 78650,
      "train_speed(iter/s)": 0.200663
    },
    {
      "acc": 0.78173332,
      "epoch": 1.8351973636244332,
      "grad_norm": 6.75,
      "learning_rate": 1.76926392274388e-07,
      "loss": 0.78438444,
      "memory(GiB)": 138.1,
      "step": 78660,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.79420872,
      "epoch": 1.8354306711967219,
      "grad_norm": 5.6875,
      "learning_rate": 1.7642864675418925e-07,
      "loss": 0.74262061,
      "memory(GiB)": 138.1,
      "step": 78670,
      "train_speed(iter/s)": 0.200688
    },
    {
      "acc": 0.78799567,
      "epoch": 1.835663978769011,
      "grad_norm": 4.625,
      "learning_rate": 1.7593158979739955e-07,
      "loss": 0.75903831,
      "memory(GiB)": 138.1,
      "step": 78680,
      "train_speed(iter/s)": 0.200701
    },
    {
      "acc": 0.77767072,
      "epoch": 1.8358972863412997,
      "grad_norm": 5.71875,
      "learning_rate": 1.7543522147497382e-07,
      "loss": 0.79137626,
      "memory(GiB)": 138.1,
      "step": 78690,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.79292173,
      "epoch": 1.8361305939135888,
      "grad_norm": 4.84375,
      "learning_rate": 1.7493954185776928e-07,
      "loss": 0.74533262,
      "memory(GiB)": 138.1,
      "step": 78700,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.79466348,
      "epoch": 1.8363639014858775,
      "grad_norm": 4.1875,
      "learning_rate": 1.7444455101654267e-07,
      "loss": 0.73135486,
      "memory(GiB)": 138.1,
      "step": 78710,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.78591375,
      "epoch": 1.8365972090581666,
      "grad_norm": 3.9375,
      "learning_rate": 1.7395024902195522e-07,
      "loss": 0.77585912,
      "memory(GiB)": 138.1,
      "step": 78720,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.77571945,
      "epoch": 1.8368305166304553,
      "grad_norm": 5.375,
      "learning_rate": 1.7345663594456775e-07,
      "loss": 0.79873123,
      "memory(GiB)": 138.1,
      "step": 78730,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.80318031,
      "epoch": 1.8370638242027444,
      "grad_norm": 6.53125,
      "learning_rate": 1.7296371185484328e-07,
      "loss": 0.6877492,
      "memory(GiB)": 138.1,
      "step": 78740,
      "train_speed(iter/s)": 0.20078
    },
    {
      "acc": 0.77101789,
      "epoch": 1.837297131775033,
      "grad_norm": 9.375,
      "learning_rate": 1.7247147682314724e-07,
      "loss": 0.8241971,
      "memory(GiB)": 138.1,
      "step": 78750,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.78852234,
      "epoch": 1.8375304393473222,
      "grad_norm": 6.8125,
      "learning_rate": 1.7197993091974452e-07,
      "loss": 0.75911512,
      "memory(GiB)": 138.1,
      "step": 78760,
      "train_speed(iter/s)": 0.200806
    },
    {
      "acc": 0.77903395,
      "epoch": 1.8377637469196109,
      "grad_norm": 5.3125,
      "learning_rate": 1.7148907421480455e-07,
      "loss": 0.7937254,
      "memory(GiB)": 138.1,
      "step": 78770,
      "train_speed(iter/s)": 0.20082
    },
    {
      "acc": 0.78879142,
      "epoch": 1.8379970544919,
      "grad_norm": 4.03125,
      "learning_rate": 1.7099890677839626e-07,
      "loss": 0.77068634,
      "memory(GiB)": 138.1,
      "step": 78780,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.78774767,
      "epoch": 1.8382303620641887,
      "grad_norm": 7.6875,
      "learning_rate": 1.7050942868049147e-07,
      "loss": 0.75463233,
      "memory(GiB)": 138.1,
      "step": 78790,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.78863096,
      "epoch": 1.8384636696364778,
      "grad_norm": 4.09375,
      "learning_rate": 1.7002063999096208e-07,
      "loss": 0.74385252,
      "memory(GiB)": 138.1,
      "step": 78800,
      "train_speed(iter/s)": 0.20086
    },
    {
      "acc": 0.77340384,
      "epoch": 1.8386969772087665,
      "grad_norm": 4.53125,
      "learning_rate": 1.695325407795839e-07,
      "loss": 0.82360716,
      "memory(GiB)": 138.1,
      "step": 78810,
      "train_speed(iter/s)": 0.200874
    },
    {
      "acc": 0.78802562,
      "epoch": 1.8389302847810556,
      "grad_norm": 5.9375,
      "learning_rate": 1.6904513111603238e-07,
      "loss": 0.76730204,
      "memory(GiB)": 138.1,
      "step": 78820,
      "train_speed(iter/s)": 0.200886
    },
    {
      "acc": 0.78368859,
      "epoch": 1.8391635923533443,
      "grad_norm": 4.46875,
      "learning_rate": 1.685584110698846e-07,
      "loss": 0.77572279,
      "memory(GiB)": 138.1,
      "step": 78830,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.79617405,
      "epoch": 1.8393968999256334,
      "grad_norm": 5.34375,
      "learning_rate": 1.680723807106205e-07,
      "loss": 0.70224414,
      "memory(GiB)": 138.1,
      "step": 78840,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.77466588,
      "epoch": 1.839630207497922,
      "grad_norm": 5.46875,
      "learning_rate": 1.675870401076196e-07,
      "loss": 0.81798153,
      "memory(GiB)": 138.1,
      "step": 78850,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.78734894,
      "epoch": 1.839863515070211,
      "grad_norm": 4.5625,
      "learning_rate": 1.6710238933016597e-07,
      "loss": 0.7503355,
      "memory(GiB)": 138.1,
      "step": 78860,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.79105768,
      "epoch": 1.8400968226424999,
      "grad_norm": 5.1875,
      "learning_rate": 1.6661842844744148e-07,
      "loss": 0.74174986,
      "memory(GiB)": 138.1,
      "step": 78870,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.76048555,
      "epoch": 1.8403301302147888,
      "grad_norm": 5.09375,
      "learning_rate": 1.6613515752853303e-07,
      "loss": 0.8763361,
      "memory(GiB)": 138.1,
      "step": 78880,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.77112312,
      "epoch": 1.8405634377870776,
      "grad_norm": 6.9375,
      "learning_rate": 1.6565257664242606e-07,
      "loss": 0.79881568,
      "memory(GiB)": 138.1,
      "step": 78890,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.76842785,
      "epoch": 1.8407967453593665,
      "grad_norm": 5.03125,
      "learning_rate": 1.6517068585800932e-07,
      "loss": 0.82623138,
      "memory(GiB)": 138.1,
      "step": 78900,
      "train_speed(iter/s)": 0.200991
    },
    {
      "acc": 0.78424797,
      "epoch": 1.8410300529316554,
      "grad_norm": 5.28125,
      "learning_rate": 1.646894852440728e-07,
      "loss": 0.76516953,
      "memory(GiB)": 138.1,
      "step": 78910,
      "train_speed(iter/s)": 0.201004
    },
    {
      "acc": 0.79357691,
      "epoch": 1.8412633605039443,
      "grad_norm": 5.59375,
      "learning_rate": 1.642089748693071e-07,
      "loss": 0.73923221,
      "memory(GiB)": 138.1,
      "step": 78920,
      "train_speed(iter/s)": 0.201018
    },
    {
      "acc": 0.77890062,
      "epoch": 1.8414966680762332,
      "grad_norm": 4.4375,
      "learning_rate": 1.6372915480230622e-07,
      "loss": 0.81259813,
      "memory(GiB)": 138.1,
      "step": 78930,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.80381775,
      "epoch": 1.8417299756485221,
      "grad_norm": 6.09375,
      "learning_rate": 1.6325002511156262e-07,
      "loss": 0.7070673,
      "memory(GiB)": 138.1,
      "step": 78940,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.77546864,
      "epoch": 1.841963283220811,
      "grad_norm": 4.71875,
      "learning_rate": 1.6277158586547325e-07,
      "loss": 0.81202116,
      "memory(GiB)": 138.1,
      "step": 78950,
      "train_speed(iter/s)": 0.201056
    },
    {
      "acc": 0.78609886,
      "epoch": 1.8421965907931,
      "grad_norm": 5.53125,
      "learning_rate": 1.6229383713233516e-07,
      "loss": 0.76256151,
      "memory(GiB)": 138.1,
      "step": 78960,
      "train_speed(iter/s)": 0.20107
    },
    {
      "acc": 0.77121072,
      "epoch": 1.8424298983653888,
      "grad_norm": 5.15625,
      "learning_rate": 1.6181677898034597e-07,
      "loss": 0.82044077,
      "memory(GiB)": 138.1,
      "step": 78970,
      "train_speed(iter/s)": 0.201083
    },
    {
      "acc": 0.7971004,
      "epoch": 1.8426632059376777,
      "grad_norm": 7.03125,
      "learning_rate": 1.6134041147760738e-07,
      "loss": 0.70855894,
      "memory(GiB)": 138.1,
      "step": 78980,
      "train_speed(iter/s)": 0.201096
    },
    {
      "acc": 0.76304283,
      "epoch": 1.8428965135099666,
      "grad_norm": 6.03125,
      "learning_rate": 1.608647346921177e-07,
      "loss": 0.84548931,
      "memory(GiB)": 138.1,
      "step": 78990,
      "train_speed(iter/s)": 0.201108
    },
    {
      "acc": 0.76878996,
      "epoch": 1.8431298210822555,
      "grad_norm": 6.9375,
      "learning_rate": 1.6038974869178214e-07,
      "loss": 0.8415431,
      "memory(GiB)": 138.1,
      "step": 79000,
      "train_speed(iter/s)": 0.201122
    },
    {
      "epoch": 1.8431298210822555,
      "eval_acc": 0.744701748182483,
      "eval_loss": 0.8044163584709167,
      "eval_runtime": 1270.82,
      "eval_samples_per_second": 28.321,
      "eval_steps_per_second": 14.161,
      "step": 79000
    },
    {
      "acc": 0.7843946,
      "epoch": 1.8433631286545444,
      "grad_norm": 4.59375,
      "learning_rate": 1.5991545354440363e-07,
      "loss": 0.77073593,
      "memory(GiB)": 138.1,
      "step": 79010,
      "train_speed(iter/s)": 0.200473
    },
    {
      "acc": 0.78733177,
      "epoch": 1.8435964362268333,
      "grad_norm": 5.3125,
      "learning_rate": 1.594418493176886e-07,
      "loss": 0.76721573,
      "memory(GiB)": 138.1,
      "step": 79020,
      "train_speed(iter/s)": 0.200486
    },
    {
      "acc": 0.772614,
      "epoch": 1.8438297437991222,
      "grad_norm": 4.53125,
      "learning_rate": 1.5896893607924346e-07,
      "loss": 0.82392635,
      "memory(GiB)": 138.1,
      "step": 79030,
      "train_speed(iter/s)": 0.200499
    },
    {
      "acc": 0.78404236,
      "epoch": 1.844063051371411,
      "grad_norm": 5.25,
      "learning_rate": 1.5849671389657594e-07,
      "loss": 0.7704998,
      "memory(GiB)": 138.1,
      "step": 79040,
      "train_speed(iter/s)": 0.200512
    },
    {
      "acc": 0.77602654,
      "epoch": 1.8442963589437,
      "grad_norm": 6.9375,
      "learning_rate": 1.58025182837096e-07,
      "loss": 0.80820026,
      "memory(GiB)": 138.1,
      "step": 79050,
      "train_speed(iter/s)": 0.200526
    },
    {
      "acc": 0.79040246,
      "epoch": 1.8445296665159887,
      "grad_norm": 4.125,
      "learning_rate": 1.5755434296811478e-07,
      "loss": 0.75322657,
      "memory(GiB)": 138.1,
      "step": 79060,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.76732149,
      "epoch": 1.8447629740882778,
      "grad_norm": 4.59375,
      "learning_rate": 1.5708419435684463e-07,
      "loss": 0.85027685,
      "memory(GiB)": 138.1,
      "step": 79070,
      "train_speed(iter/s)": 0.200552
    },
    {
      "acc": 0.77900348,
      "epoch": 1.8449962816605665,
      "grad_norm": 7.15625,
      "learning_rate": 1.5661473707039852e-07,
      "loss": 0.7863019,
      "memory(GiB)": 138.1,
      "step": 79080,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.79656229,
      "epoch": 1.8452295892328556,
      "grad_norm": 4.15625,
      "learning_rate": 1.561459711757918e-07,
      "loss": 0.73082271,
      "memory(GiB)": 138.1,
      "step": 79090,
      "train_speed(iter/s)": 0.200578
    },
    {
      "acc": 0.78649101,
      "epoch": 1.8454628968051443,
      "grad_norm": 5.34375,
      "learning_rate": 1.5567789673994026e-07,
      "loss": 0.76644926,
      "memory(GiB)": 138.1,
      "step": 79100,
      "train_speed(iter/s)": 0.20059
    },
    {
      "acc": 0.76249952,
      "epoch": 1.8456962043774334,
      "grad_norm": 4.8125,
      "learning_rate": 1.5521051382966224e-07,
      "loss": 0.85472183,
      "memory(GiB)": 138.1,
      "step": 79110,
      "train_speed(iter/s)": 0.200604
    },
    {
      "acc": 0.79876027,
      "epoch": 1.845929511949722,
      "grad_norm": 5.15625,
      "learning_rate": 1.5474382251167597e-07,
      "loss": 0.70353813,
      "memory(GiB)": 138.1,
      "step": 79120,
      "train_speed(iter/s)": 0.200616
    },
    {
      "acc": 0.77245951,
      "epoch": 1.8461628195220112,
      "grad_norm": 5.0,
      "learning_rate": 1.542778228526004e-07,
      "loss": 0.79053502,
      "memory(GiB)": 138.1,
      "step": 79130,
      "train_speed(iter/s)": 0.20063
    },
    {
      "acc": 0.76596804,
      "epoch": 1.8463961270943,
      "grad_norm": 5.21875,
      "learning_rate": 1.53812514918959e-07,
      "loss": 0.83255501,
      "memory(GiB)": 138.1,
      "step": 79140,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.76959438,
      "epoch": 1.846629434666589,
      "grad_norm": 4.5625,
      "learning_rate": 1.5334789877717248e-07,
      "loss": 0.82380829,
      "memory(GiB)": 138.1,
      "step": 79150,
      "train_speed(iter/s)": 0.200657
    },
    {
      "acc": 0.76817465,
      "epoch": 1.8468627422388777,
      "grad_norm": 5.03125,
      "learning_rate": 1.5288397449356617e-07,
      "loss": 0.84938889,
      "memory(GiB)": 138.1,
      "step": 79160,
      "train_speed(iter/s)": 0.20067
    },
    {
      "acc": 0.77281103,
      "epoch": 1.8470960498111668,
      "grad_norm": 4.96875,
      "learning_rate": 1.524207421343643e-07,
      "loss": 0.81317339,
      "memory(GiB)": 138.1,
      "step": 79170,
      "train_speed(iter/s)": 0.200683
    },
    {
      "acc": 0.78833442,
      "epoch": 1.8473293573834555,
      "grad_norm": 4.78125,
      "learning_rate": 1.5195820176569288e-07,
      "loss": 0.73116922,
      "memory(GiB)": 138.1,
      "step": 79180,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.77847328,
      "epoch": 1.8475626649557446,
      "grad_norm": 5.8125,
      "learning_rate": 1.5149635345358017e-07,
      "loss": 0.80921946,
      "memory(GiB)": 138.1,
      "step": 79190,
      "train_speed(iter/s)": 0.200708
    },
    {
      "acc": 0.77257528,
      "epoch": 1.8477959725280333,
      "grad_norm": 5.40625,
      "learning_rate": 1.51035197263954e-07,
      "loss": 0.83006363,
      "memory(GiB)": 138.1,
      "step": 79200,
      "train_speed(iter/s)": 0.200722
    },
    {
      "acc": 0.78382149,
      "epoch": 1.8480292801003224,
      "grad_norm": 4.9375,
      "learning_rate": 1.5057473326264614e-07,
      "loss": 0.76725683,
      "memory(GiB)": 138.1,
      "step": 79210,
      "train_speed(iter/s)": 0.200735
    },
    {
      "acc": 0.79515305,
      "epoch": 1.848262587672611,
      "grad_norm": 4.6875,
      "learning_rate": 1.5011496151538462e-07,
      "loss": 0.73181725,
      "memory(GiB)": 138.1,
      "step": 79220,
      "train_speed(iter/s)": 0.200748
    },
    {
      "acc": 0.77747626,
      "epoch": 1.8484958952449,
      "grad_norm": 5.5,
      "learning_rate": 1.4965588208780468e-07,
      "loss": 0.77987266,
      "memory(GiB)": 138.1,
      "step": 79230,
      "train_speed(iter/s)": 0.200761
    },
    {
      "acc": 0.77652779,
      "epoch": 1.8487292028171889,
      "grad_norm": 5.75,
      "learning_rate": 1.491974950454378e-07,
      "loss": 0.80380249,
      "memory(GiB)": 138.1,
      "step": 79240,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.78755636,
      "epoch": 1.8489625103894778,
      "grad_norm": 4.4375,
      "learning_rate": 1.4873980045371938e-07,
      "loss": 0.75492868,
      "memory(GiB)": 138.1,
      "step": 79250,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.80663643,
      "epoch": 1.8491958179617667,
      "grad_norm": 4.90625,
      "learning_rate": 1.4828279837798553e-07,
      "loss": 0.68906574,
      "memory(GiB)": 138.1,
      "step": 79260,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.78660917,
      "epoch": 1.8494291255340556,
      "grad_norm": 4.96875,
      "learning_rate": 1.478264888834724e-07,
      "loss": 0.76002073,
      "memory(GiB)": 138.1,
      "step": 79270,
      "train_speed(iter/s)": 0.200814
    },
    {
      "acc": 0.77109876,
      "epoch": 1.8496624331063445,
      "grad_norm": 6.40625,
      "learning_rate": 1.4737087203531896e-07,
      "loss": 0.81629725,
      "memory(GiB)": 138.1,
      "step": 79280,
      "train_speed(iter/s)": 0.200827
    },
    {
      "acc": 0.78523669,
      "epoch": 1.8498957406786334,
      "grad_norm": 7.78125,
      "learning_rate": 1.4691594789856268e-07,
      "loss": 0.76274586,
      "memory(GiB)": 138.1,
      "step": 79290,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.78136597,
      "epoch": 1.8501290482509223,
      "grad_norm": 5.65625,
      "learning_rate": 1.46461716538146e-07,
      "loss": 0.79236145,
      "memory(GiB)": 138.1,
      "step": 79300,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.77388811,
      "epoch": 1.8503623558232112,
      "grad_norm": 6.90625,
      "learning_rate": 1.4600817801890933e-07,
      "loss": 0.82036371,
      "memory(GiB)": 138.1,
      "step": 79310,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.74084215,
      "epoch": 1.8505956633955,
      "grad_norm": 5.65625,
      "learning_rate": 1.4555533240559526e-07,
      "loss": 0.93160048,
      "memory(GiB)": 138.1,
      "step": 79320,
      "train_speed(iter/s)": 0.200878
    },
    {
      "acc": 0.77508183,
      "epoch": 1.850828970967789,
      "grad_norm": 6.1875,
      "learning_rate": 1.4510317976284715e-07,
      "loss": 0.81556673,
      "memory(GiB)": 138.1,
      "step": 79330,
      "train_speed(iter/s)": 0.200891
    },
    {
      "acc": 0.7719595,
      "epoch": 1.8510622785400779,
      "grad_norm": 9.0,
      "learning_rate": 1.4465172015520945e-07,
      "loss": 0.82613068,
      "memory(GiB)": 138.1,
      "step": 79340,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.79376431,
      "epoch": 1.8512955861123668,
      "grad_norm": 4.75,
      "learning_rate": 1.4420095364712838e-07,
      "loss": 0.72482729,
      "memory(GiB)": 138.1,
      "step": 79350,
      "train_speed(iter/s)": 0.200915
    },
    {
      "acc": 0.76974692,
      "epoch": 1.8515288936846557,
      "grad_norm": 6.25,
      "learning_rate": 1.4375088030295027e-07,
      "loss": 0.83236866,
      "memory(GiB)": 138.1,
      "step": 79360,
      "train_speed(iter/s)": 0.200928
    },
    {
      "acc": 0.79396596,
      "epoch": 1.8517622012569446,
      "grad_norm": 4.0,
      "learning_rate": 1.433015001869237e-07,
      "loss": 0.74162588,
      "memory(GiB)": 138.1,
      "step": 79370,
      "train_speed(iter/s)": 0.20094
    },
    {
      "acc": 0.76310349,
      "epoch": 1.8519955088292335,
      "grad_norm": 5.6875,
      "learning_rate": 1.428528133631968e-07,
      "loss": 0.85866594,
      "memory(GiB)": 138.1,
      "step": 79380,
      "train_speed(iter/s)": 0.200953
    },
    {
      "acc": 0.75772862,
      "epoch": 1.8522288164015224,
      "grad_norm": 6.75,
      "learning_rate": 1.4240481989581944e-07,
      "loss": 0.87741089,
      "memory(GiB)": 138.1,
      "step": 79390,
      "train_speed(iter/s)": 0.200967
    },
    {
      "acc": 0.77879143,
      "epoch": 1.8524621239738113,
      "grad_norm": 5.125,
      "learning_rate": 1.4195751984874383e-07,
      "loss": 0.79758248,
      "memory(GiB)": 138.1,
      "step": 79400,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.78456173,
      "epoch": 1.8526954315461002,
      "grad_norm": 5.5,
      "learning_rate": 1.4151091328582e-07,
      "loss": 0.75210719,
      "memory(GiB)": 138.1,
      "step": 79410,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.77924018,
      "epoch": 1.852928739118389,
      "grad_norm": 4.03125,
      "learning_rate": 1.410650002708025e-07,
      "loss": 0.80060263,
      "memory(GiB)": 138.1,
      "step": 79420,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.78018885,
      "epoch": 1.8531620466906777,
      "grad_norm": 5.03125,
      "learning_rate": 1.4061978086734484e-07,
      "loss": 0.77695847,
      "memory(GiB)": 138.1,
      "step": 79430,
      "train_speed(iter/s)": 0.20102
    },
    {
      "acc": 0.76331925,
      "epoch": 1.8533953542629669,
      "grad_norm": 6.84375,
      "learning_rate": 1.4017525513900175e-07,
      "loss": 0.85669298,
      "memory(GiB)": 138.1,
      "step": 79440,
      "train_speed(iter/s)": 0.201033
    },
    {
      "acc": 0.79373131,
      "epoch": 1.8536286618352555,
      "grad_norm": 10.5,
      "learning_rate": 1.3973142314922862e-07,
      "loss": 0.73819771,
      "memory(GiB)": 138.1,
      "step": 79450,
      "train_speed(iter/s)": 0.201046
    },
    {
      "acc": 0.77359915,
      "epoch": 1.8538619694075447,
      "grad_norm": 6.625,
      "learning_rate": 1.3928828496138358e-07,
      "loss": 0.83192682,
      "memory(GiB)": 138.1,
      "step": 79460,
      "train_speed(iter/s)": 0.201058
    },
    {
      "acc": 0.79424295,
      "epoch": 1.8540952769798333,
      "grad_norm": 7.75,
      "learning_rate": 1.3884584063872386e-07,
      "loss": 0.72979498,
      "memory(GiB)": 138.1,
      "step": 79470,
      "train_speed(iter/s)": 0.201071
    },
    {
      "acc": 0.79742351,
      "epoch": 1.8543285845521225,
      "grad_norm": 4.625,
      "learning_rate": 1.3840409024440726e-07,
      "loss": 0.72036037,
      "memory(GiB)": 138.1,
      "step": 79480,
      "train_speed(iter/s)": 0.201084
    },
    {
      "acc": 0.76095319,
      "epoch": 1.8545618921244111,
      "grad_norm": 6.125,
      "learning_rate": 1.3796303384149557e-07,
      "loss": 0.855832,
      "memory(GiB)": 138.1,
      "step": 79490,
      "train_speed(iter/s)": 0.201098
    },
    {
      "acc": 0.768857,
      "epoch": 1.8547951996967003,
      "grad_norm": 4.84375,
      "learning_rate": 1.375226714929473e-07,
      "loss": 0.83169584,
      "memory(GiB)": 138.1,
      "step": 79500,
      "train_speed(iter/s)": 0.201111
    },
    {
      "epoch": 1.8547951996967003,
      "eval_acc": 0.744718723471107,
      "eval_loss": 0.8043954968452454,
      "eval_runtime": 1271.2392,
      "eval_samples_per_second": 28.312,
      "eval_steps_per_second": 14.156,
      "step": 79500
    },
    {
      "acc": 0.80413227,
      "epoch": 1.855028507268989,
      "grad_norm": 4.5625,
      "learning_rate": 1.3708300326162605e-07,
      "loss": 0.69383755,
      "memory(GiB)": 138.1,
      "step": 79510,
      "train_speed(iter/s)": 0.200467
    },
    {
      "acc": 0.77757564,
      "epoch": 1.855261814841278,
      "grad_norm": 5.1875,
      "learning_rate": 1.3664402921029328e-07,
      "loss": 0.78825073,
      "memory(GiB)": 138.1,
      "step": 79520,
      "train_speed(iter/s)": 0.20048
    },
    {
      "acc": 0.75611916,
      "epoch": 1.8554951224135667,
      "grad_norm": 4.21875,
      "learning_rate": 1.3620574940161168e-07,
      "loss": 0.88193607,
      "memory(GiB)": 138.1,
      "step": 79530,
      "train_speed(iter/s)": 0.200492
    },
    {
      "acc": 0.78447027,
      "epoch": 1.8557284299858559,
      "grad_norm": 5.59375,
      "learning_rate": 1.357681638981473e-07,
      "loss": 0.78287034,
      "memory(GiB)": 138.1,
      "step": 79540,
      "train_speed(iter/s)": 0.200504
    },
    {
      "acc": 0.78698115,
      "epoch": 1.8559617375581445,
      "grad_norm": 6.875,
      "learning_rate": 1.3533127276236458e-07,
      "loss": 0.78185997,
      "memory(GiB)": 138.1,
      "step": 79550,
      "train_speed(iter/s)": 0.200518
    },
    {
      "acc": 0.77826419,
      "epoch": 1.8561950451304337,
      "grad_norm": 4.4375,
      "learning_rate": 1.348950760566292e-07,
      "loss": 0.79370699,
      "memory(GiB)": 138.1,
      "step": 79560,
      "train_speed(iter/s)": 0.200531
    },
    {
      "acc": 0.78184977,
      "epoch": 1.8564283527027223,
      "grad_norm": 5.78125,
      "learning_rate": 1.3445957384320808e-07,
      "loss": 0.77437334,
      "memory(GiB)": 138.1,
      "step": 79570,
      "train_speed(iter/s)": 0.200545
    },
    {
      "acc": 0.76415167,
      "epoch": 1.8566616602750114,
      "grad_norm": 6.125,
      "learning_rate": 1.340247661842692e-07,
      "loss": 0.85826664,
      "memory(GiB)": 138.1,
      "step": 79580,
      "train_speed(iter/s)": 0.200557
    },
    {
      "acc": 0.76513901,
      "epoch": 1.8568949678473001,
      "grad_norm": 5.625,
      "learning_rate": 1.335906531418818e-07,
      "loss": 0.84353552,
      "memory(GiB)": 138.1,
      "step": 79590,
      "train_speed(iter/s)": 0.20057
    },
    {
      "acc": 0.79837427,
      "epoch": 1.8571282754195892,
      "grad_norm": 4.8125,
      "learning_rate": 1.3315723477801467e-07,
      "loss": 0.71099424,
      "memory(GiB)": 138.1,
      "step": 79600,
      "train_speed(iter/s)": 0.200583
    },
    {
      "acc": 0.77831469,
      "epoch": 1.857361582991878,
      "grad_norm": 4.90625,
      "learning_rate": 1.3272451115453888e-07,
      "loss": 0.80567703,
      "memory(GiB)": 138.1,
      "step": 79610,
      "train_speed(iter/s)": 0.200597
    },
    {
      "acc": 0.77662849,
      "epoch": 1.8575948905641668,
      "grad_norm": 5.875,
      "learning_rate": 1.32292482333225e-07,
      "loss": 0.80015821,
      "memory(GiB)": 138.1,
      "step": 79620,
      "train_speed(iter/s)": 0.20061
    },
    {
      "acc": 0.78590217,
      "epoch": 1.8578281981364557,
      "grad_norm": 5.0625,
      "learning_rate": 1.3186114837574538e-07,
      "loss": 0.75417018,
      "memory(GiB)": 138.1,
      "step": 79630,
      "train_speed(iter/s)": 0.200624
    },
    {
      "acc": 0.77556219,
      "epoch": 1.8580615057087446,
      "grad_norm": 4.78125,
      "learning_rate": 1.3143050934367187e-07,
      "loss": 0.81039896,
      "memory(GiB)": 138.1,
      "step": 79640,
      "train_speed(iter/s)": 0.200637
    },
    {
      "acc": 0.78844309,
      "epoch": 1.8582948132810335,
      "grad_norm": 6.4375,
      "learning_rate": 1.310005652984797e-07,
      "loss": 0.75361748,
      "memory(GiB)": 138.1,
      "step": 79650,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78295665,
      "epoch": 1.8585281208533224,
      "grad_norm": 11.5,
      "learning_rate": 1.3057131630154208e-07,
      "loss": 0.7822053,
      "memory(GiB)": 138.1,
      "step": 79660,
      "train_speed(iter/s)": 0.200663
    },
    {
      "acc": 0.76080747,
      "epoch": 1.8587614284256113,
      "grad_norm": 5.53125,
      "learning_rate": 1.3014276241413438e-07,
      "loss": 0.84956684,
      "memory(GiB)": 138.1,
      "step": 79670,
      "train_speed(iter/s)": 0.200676
    },
    {
      "acc": 0.78363895,
      "epoch": 1.8589947359979002,
      "grad_norm": 4.53125,
      "learning_rate": 1.2971490369743323e-07,
      "loss": 0.79574165,
      "memory(GiB)": 138.1,
      "step": 79680,
      "train_speed(iter/s)": 0.200689
    },
    {
      "acc": 0.77177901,
      "epoch": 1.8592280435701891,
      "grad_norm": 6.0,
      "learning_rate": 1.2928774021251368e-07,
      "loss": 0.8176857,
      "memory(GiB)": 138.1,
      "step": 79690,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.75965395,
      "epoch": 1.859461351142478,
      "grad_norm": 5.53125,
      "learning_rate": 1.288612720203547e-07,
      "loss": 0.86375732,
      "memory(GiB)": 138.1,
      "step": 79700,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.76926231,
      "epoch": 1.859694658714767,
      "grad_norm": 7.09375,
      "learning_rate": 1.284354991818343e-07,
      "loss": 0.83586464,
      "memory(GiB)": 138.1,
      "step": 79710,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78583107,
      "epoch": 1.8599279662870558,
      "grad_norm": 5.125,
      "learning_rate": 1.2801042175773104e-07,
      "loss": 0.76661096,
      "memory(GiB)": 138.1,
      "step": 79720,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.7856154,
      "epoch": 1.8601612738593447,
      "grad_norm": 4.96875,
      "learning_rate": 1.2758603980872419e-07,
      "loss": 0.76946712,
      "memory(GiB)": 138.1,
      "step": 79730,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.78945646,
      "epoch": 1.8603945814316336,
      "grad_norm": 5.3125,
      "learning_rate": 1.2716235339539585e-07,
      "loss": 0.76702728,
      "memory(GiB)": 138.1,
      "step": 79740,
      "train_speed(iter/s)": 0.200768
    },
    {
      "acc": 0.78451262,
      "epoch": 1.8606278890039225,
      "grad_norm": 5.5625,
      "learning_rate": 1.267393625782254e-07,
      "loss": 0.75838819,
      "memory(GiB)": 138.1,
      "step": 79750,
      "train_speed(iter/s)": 0.20078
    },
    {
      "acc": 0.77074156,
      "epoch": 1.8608611965762114,
      "grad_norm": 4.71875,
      "learning_rate": 1.263170674175951e-07,
      "loss": 0.83515434,
      "memory(GiB)": 138.1,
      "step": 79760,
      "train_speed(iter/s)": 0.200793
    },
    {
      "acc": 0.77783523,
      "epoch": 1.8610945041485003,
      "grad_norm": 5.5625,
      "learning_rate": 1.2589546797378783e-07,
      "loss": 0.7946856,
      "memory(GiB)": 138.1,
      "step": 79770,
      "train_speed(iter/s)": 0.200807
    },
    {
      "acc": 0.76884985,
      "epoch": 1.8613278117207892,
      "grad_norm": 6.03125,
      "learning_rate": 1.2547456430698656e-07,
      "loss": 0.82223921,
      "memory(GiB)": 138.1,
      "step": 79780,
      "train_speed(iter/s)": 0.20082
    },
    {
      "acc": 0.7526144,
      "epoch": 1.861561119293078,
      "grad_norm": 5.15625,
      "learning_rate": 1.2505435647727548e-07,
      "loss": 0.89265423,
      "memory(GiB)": 138.1,
      "step": 79790,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.77072687,
      "epoch": 1.861794426865367,
      "grad_norm": 5.75,
      "learning_rate": 1.2463484454463826e-07,
      "loss": 0.82332354,
      "memory(GiB)": 138.1,
      "step": 79800,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.7775485,
      "epoch": 1.862027734437656,
      "grad_norm": 5.34375,
      "learning_rate": 1.2421602856896087e-07,
      "loss": 0.7913826,
      "memory(GiB)": 138.1,
      "step": 79810,
      "train_speed(iter/s)": 0.200858
    },
    {
      "acc": 0.78050547,
      "epoch": 1.8622610420099446,
      "grad_norm": 4.65625,
      "learning_rate": 1.237979086100294e-07,
      "loss": 0.7908287,
      "memory(GiB)": 138.1,
      "step": 79820,
      "train_speed(iter/s)": 0.200871
    },
    {
      "acc": 0.78652105,
      "epoch": 1.8624943495822337,
      "grad_norm": 4.96875,
      "learning_rate": 1.233804847275294e-07,
      "loss": 0.76471519,
      "memory(GiB)": 138.1,
      "step": 79830,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.78220301,
      "epoch": 1.8627276571545224,
      "grad_norm": 5.34375,
      "learning_rate": 1.2296375698104878e-07,
      "loss": 0.77280107,
      "memory(GiB)": 138.1,
      "step": 79840,
      "train_speed(iter/s)": 0.200897
    },
    {
      "acc": 0.78212929,
      "epoch": 1.8629609647268115,
      "grad_norm": 6.0625,
      "learning_rate": 1.2254772543007442e-07,
      "loss": 0.80081415,
      "memory(GiB)": 138.1,
      "step": 79850,
      "train_speed(iter/s)": 0.20091
    },
    {
      "acc": 0.78351145,
      "epoch": 1.8631942722991002,
      "grad_norm": 6.125,
      "learning_rate": 1.2213239013399602e-07,
      "loss": 0.77549877,
      "memory(GiB)": 138.1,
      "step": 79860,
      "train_speed(iter/s)": 0.200923
    },
    {
      "acc": 0.77944012,
      "epoch": 1.8634275798713893,
      "grad_norm": 5.3125,
      "learning_rate": 1.217177511521017e-07,
      "loss": 0.78213992,
      "memory(GiB)": 138.1,
      "step": 79870,
      "train_speed(iter/s)": 0.200936
    },
    {
      "acc": 0.78833656,
      "epoch": 1.863660887443678,
      "grad_norm": 4.78125,
      "learning_rate": 1.2130380854358136e-07,
      "loss": 0.74980211,
      "memory(GiB)": 138.1,
      "step": 79880,
      "train_speed(iter/s)": 0.200949
    },
    {
      "acc": 0.78198175,
      "epoch": 1.863894195015967,
      "grad_norm": 4.9375,
      "learning_rate": 1.208905623675255e-07,
      "loss": 0.80464964,
      "memory(GiB)": 138.1,
      "step": 79890,
      "train_speed(iter/s)": 0.200961
    },
    {
      "acc": 0.77077436,
      "epoch": 1.8641275025882558,
      "grad_norm": 6.78125,
      "learning_rate": 1.2047801268292414e-07,
      "loss": 0.82589788,
      "memory(GiB)": 138.1,
      "step": 79900,
      "train_speed(iter/s)": 0.200974
    },
    {
      "acc": 0.77802629,
      "epoch": 1.864360810160545,
      "grad_norm": 5.9375,
      "learning_rate": 1.2006615954866906e-07,
      "loss": 0.77997022,
      "memory(GiB)": 138.1,
      "step": 79910,
      "train_speed(iter/s)": 0.200987
    },
    {
      "acc": 0.77522287,
      "epoch": 1.8645941177328336,
      "grad_norm": 5.71875,
      "learning_rate": 1.196550030235516e-07,
      "loss": 0.80558329,
      "memory(GiB)": 138.1,
      "step": 79920,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.76346202,
      "epoch": 1.8648274253051227,
      "grad_norm": 6.40625,
      "learning_rate": 1.1924454316626478e-07,
      "loss": 0.8442524,
      "memory(GiB)": 138.1,
      "step": 79930,
      "train_speed(iter/s)": 0.201012
    },
    {
      "acc": 0.77351999,
      "epoch": 1.8650607328774114,
      "grad_norm": 8.8125,
      "learning_rate": 1.1883478003540172e-07,
      "loss": 0.82075424,
      "memory(GiB)": 138.1,
      "step": 79940,
      "train_speed(iter/s)": 0.201026
    },
    {
      "acc": 0.7649683,
      "epoch": 1.8652940404497005,
      "grad_norm": 3.84375,
      "learning_rate": 1.1842571368945566e-07,
      "loss": 0.85019636,
      "memory(GiB)": 138.1,
      "step": 79950,
      "train_speed(iter/s)": 0.201038
    },
    {
      "acc": 0.78048048,
      "epoch": 1.8655273480219892,
      "grad_norm": 5.09375,
      "learning_rate": 1.1801734418682154e-07,
      "loss": 0.79374285,
      "memory(GiB)": 138.1,
      "step": 79960,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.78092804,
      "epoch": 1.8657606555942783,
      "grad_norm": 6.125,
      "learning_rate": 1.1760967158579217e-07,
      "loss": 0.80186243,
      "memory(GiB)": 138.1,
      "step": 79970,
      "train_speed(iter/s)": 0.201065
    },
    {
      "acc": 0.79199791,
      "epoch": 1.865993963166567,
      "grad_norm": 5.71875,
      "learning_rate": 1.172026959445649e-07,
      "loss": 0.74338923,
      "memory(GiB)": 138.1,
      "step": 79980,
      "train_speed(iter/s)": 0.201077
    },
    {
      "acc": 0.79580288,
      "epoch": 1.866227270738856,
      "grad_norm": 5.21875,
      "learning_rate": 1.1679641732123382e-07,
      "loss": 0.71093869,
      "memory(GiB)": 138.1,
      "step": 79990,
      "train_speed(iter/s)": 0.201091
    },
    {
      "acc": 0.77390499,
      "epoch": 1.8664605783111448,
      "grad_norm": 6.25,
      "learning_rate": 1.163908357737964e-07,
      "loss": 0.81473513,
      "memory(GiB)": 138.1,
      "step": 80000,
      "train_speed(iter/s)": 0.201105
    },
    {
      "epoch": 1.8664605783111448,
      "eval_acc": 0.7447302538558327,
      "eval_loss": 0.8044291138648987,
      "eval_runtime": 1271.074,
      "eval_samples_per_second": 28.315,
      "eval_steps_per_second": 14.158,
      "step": 80000
    },
    {
      "acc": 0.79017296,
      "epoch": 1.8666938858834337,
      "grad_norm": 5.59375,
      "learning_rate": 1.1598595136014745e-07,
      "loss": 0.75285716,
      "memory(GiB)": 138.1,
      "step": 80010,
      "train_speed(iter/s)": 0.200465
    },
    {
      "acc": 0.76734552,
      "epoch": 1.8669271934557226,
      "grad_norm": 5.46875,
      "learning_rate": 1.1558176413808519e-07,
      "loss": 0.864007,
      "memory(GiB)": 138.1,
      "step": 80020,
      "train_speed(iter/s)": 0.200477
    },
    {
      "acc": 0.76343198,
      "epoch": 1.8671605010280115,
      "grad_norm": 5.34375,
      "learning_rate": 1.1517827416530736e-07,
      "loss": 0.8714529,
      "memory(GiB)": 138.1,
      "step": 80030,
      "train_speed(iter/s)": 0.200489
    },
    {
      "acc": 0.7741498,
      "epoch": 1.8673938086003004,
      "grad_norm": 4.0625,
      "learning_rate": 1.1477548149941176e-07,
      "loss": 0.79983683,
      "memory(GiB)": 138.1,
      "step": 80040,
      "train_speed(iter/s)": 0.200503
    },
    {
      "acc": 0.79464846,
      "epoch": 1.8676271161725893,
      "grad_norm": 5.71875,
      "learning_rate": 1.143733861978974e-07,
      "loss": 0.72727637,
      "memory(GiB)": 138.1,
      "step": 80050,
      "train_speed(iter/s)": 0.200516
    },
    {
      "acc": 0.7844614,
      "epoch": 1.8678604237448782,
      "grad_norm": 4.5,
      "learning_rate": 1.1397198831816226e-07,
      "loss": 0.73518605,
      "memory(GiB)": 138.1,
      "step": 80060,
      "train_speed(iter/s)": 0.200529
    },
    {
      "acc": 0.78245525,
      "epoch": 1.868093731317167,
      "grad_norm": 5.3125,
      "learning_rate": 1.1357128791750716e-07,
      "loss": 0.78568835,
      "memory(GiB)": 138.1,
      "step": 80070,
      "train_speed(iter/s)": 0.200543
    },
    {
      "acc": 0.79324665,
      "epoch": 1.868327038889456,
      "grad_norm": 5.28125,
      "learning_rate": 1.1317128505313024e-07,
      "loss": 0.75692196,
      "memory(GiB)": 138.1,
      "step": 80080,
      "train_speed(iter/s)": 0.200556
    },
    {
      "acc": 0.79333744,
      "epoch": 1.8685603464617448,
      "grad_norm": 5.6875,
      "learning_rate": 1.1277197978213362e-07,
      "loss": 0.71830125,
      "memory(GiB)": 138.1,
      "step": 80090,
      "train_speed(iter/s)": 0.200569
    },
    {
      "acc": 0.75221357,
      "epoch": 1.8687936540340337,
      "grad_norm": 5.5625,
      "learning_rate": 1.1237337216151723e-07,
      "loss": 0.89948711,
      "memory(GiB)": 138.1,
      "step": 80100,
      "train_speed(iter/s)": 0.200582
    },
    {
      "acc": 0.77064877,
      "epoch": 1.8690269616063226,
      "grad_norm": 6.15625,
      "learning_rate": 1.1197546224818112e-07,
      "loss": 0.82701855,
      "memory(GiB)": 138.1,
      "step": 80110,
      "train_speed(iter/s)": 0.200595
    },
    {
      "acc": 0.79913635,
      "epoch": 1.8692602691786115,
      "grad_norm": 5.125,
      "learning_rate": 1.1157825009892931e-07,
      "loss": 0.71697931,
      "memory(GiB)": 138.1,
      "step": 80120,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.76889296,
      "epoch": 1.8694935767509004,
      "grad_norm": 6.0,
      "learning_rate": 1.1118173577046088e-07,
      "loss": 0.82190304,
      "memory(GiB)": 138.1,
      "step": 80130,
      "train_speed(iter/s)": 0.200621
    },
    {
      "acc": 0.78341179,
      "epoch": 1.8697268843231893,
      "grad_norm": 5.15625,
      "learning_rate": 1.1078591931937999e-07,
      "loss": 0.77279377,
      "memory(GiB)": 138.1,
      "step": 80140,
      "train_speed(iter/s)": 0.200633
    },
    {
      "acc": 0.78050628,
      "epoch": 1.8699601918954782,
      "grad_norm": 5.25,
      "learning_rate": 1.1039080080218811e-07,
      "loss": 0.80319948,
      "memory(GiB)": 138.1,
      "step": 80150,
      "train_speed(iter/s)": 0.200646
    },
    {
      "acc": 0.78934498,
      "epoch": 1.8701934994677671,
      "grad_norm": 7.625,
      "learning_rate": 1.0999638027528959e-07,
      "loss": 0.74584684,
      "memory(GiB)": 138.1,
      "step": 80160,
      "train_speed(iter/s)": 0.200659
    },
    {
      "acc": 0.78847446,
      "epoch": 1.870426807040056,
      "grad_norm": 4.8125,
      "learning_rate": 1.0960265779498769e-07,
      "loss": 0.73895617,
      "memory(GiB)": 138.1,
      "step": 80170,
      "train_speed(iter/s)": 0.200671
    },
    {
      "acc": 0.77396088,
      "epoch": 1.870660114612345,
      "grad_norm": 5.4375,
      "learning_rate": 1.092096334174847e-07,
      "loss": 0.81397905,
      "memory(GiB)": 138.1,
      "step": 80180,
      "train_speed(iter/s)": 0.200684
    },
    {
      "acc": 0.78955622,
      "epoch": 1.8708934221846336,
      "grad_norm": 5.21875,
      "learning_rate": 1.0881730719888628e-07,
      "loss": 0.76100597,
      "memory(GiB)": 138.1,
      "step": 80190,
      "train_speed(iter/s)": 0.200696
    },
    {
      "acc": 0.80156097,
      "epoch": 1.8711267297569227,
      "grad_norm": 3.828125,
      "learning_rate": 1.0842567919519597e-07,
      "loss": 0.69538174,
      "memory(GiB)": 138.1,
      "step": 80200,
      "train_speed(iter/s)": 0.200709
    },
    {
      "acc": 0.78851342,
      "epoch": 1.8713600373292114,
      "grad_norm": 6.34375,
      "learning_rate": 1.0803474946231963e-07,
      "loss": 0.76109829,
      "memory(GiB)": 138.1,
      "step": 80210,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.78874664,
      "epoch": 1.8715933449015005,
      "grad_norm": 5.71875,
      "learning_rate": 1.0764451805606091e-07,
      "loss": 0.73757567,
      "memory(GiB)": 138.1,
      "step": 80220,
      "train_speed(iter/s)": 0.200735
    },
    {
      "acc": 0.77023153,
      "epoch": 1.8718266524737892,
      "grad_norm": 3.984375,
      "learning_rate": 1.0725498503212694e-07,
      "loss": 0.81957588,
      "memory(GiB)": 138.1,
      "step": 80230,
      "train_speed(iter/s)": 0.200748
    },
    {
      "acc": 0.77765932,
      "epoch": 1.8720599600460783,
      "grad_norm": 6.6875,
      "learning_rate": 1.0686615044612159e-07,
      "loss": 0.77763338,
      "memory(GiB)": 138.1,
      "step": 80240,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.7891181,
      "epoch": 1.872293267618367,
      "grad_norm": 4.84375,
      "learning_rate": 1.0647801435355264e-07,
      "loss": 0.75162325,
      "memory(GiB)": 138.1,
      "step": 80250,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.76905499,
      "epoch": 1.8725265751906561,
      "grad_norm": 4.5625,
      "learning_rate": 1.0609057680982527e-07,
      "loss": 0.82668858,
      "memory(GiB)": 138.1,
      "step": 80260,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.78598719,
      "epoch": 1.8727598827629448,
      "grad_norm": 4.9375,
      "learning_rate": 1.0570383787024574e-07,
      "loss": 0.76735053,
      "memory(GiB)": 138.1,
      "step": 80270,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.78083096,
      "epoch": 1.872993190335234,
      "grad_norm": 5.46875,
      "learning_rate": 1.0531779759002214e-07,
      "loss": 0.79012175,
      "memory(GiB)": 138.1,
      "step": 80280,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.7970469,
      "epoch": 1.8732264979075226,
      "grad_norm": 6.25,
      "learning_rate": 1.0493245602426095e-07,
      "loss": 0.71222925,
      "memory(GiB)": 138.1,
      "step": 80290,
      "train_speed(iter/s)": 0.200823
    },
    {
      "acc": 0.78838396,
      "epoch": 1.8734598054798117,
      "grad_norm": 6.0625,
      "learning_rate": 1.0454781322796981e-07,
      "loss": 0.74144182,
      "memory(GiB)": 138.1,
      "step": 80300,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.80160275,
      "epoch": 1.8736931130521004,
      "grad_norm": 4.78125,
      "learning_rate": 1.0416386925605592e-07,
      "loss": 0.71015835,
      "memory(GiB)": 138.1,
      "step": 80310,
      "train_speed(iter/s)": 0.200849
    },
    {
      "acc": 0.80436602,
      "epoch": 1.8739264206243895,
      "grad_norm": 3.828125,
      "learning_rate": 1.0378062416332712e-07,
      "loss": 0.70317974,
      "memory(GiB)": 138.1,
      "step": 80320,
      "train_speed(iter/s)": 0.200862
    },
    {
      "acc": 0.7857511,
      "epoch": 1.8741597281966782,
      "grad_norm": 4.625,
      "learning_rate": 1.0339807800449241e-07,
      "loss": 0.76099644,
      "memory(GiB)": 138.1,
      "step": 80330,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.77934189,
      "epoch": 1.8743930357689673,
      "grad_norm": 5.90625,
      "learning_rate": 1.0301623083415924e-07,
      "loss": 0.79090543,
      "memory(GiB)": 138.1,
      "step": 80340,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.79418821,
      "epoch": 1.874626343341256,
      "grad_norm": 5.65625,
      "learning_rate": 1.0263508270683731e-07,
      "loss": 0.73172894,
      "memory(GiB)": 138.1,
      "step": 80350,
      "train_speed(iter/s)": 0.200899
    },
    {
      "acc": 0.76580453,
      "epoch": 1.8748596509135451,
      "grad_norm": 5.71875,
      "learning_rate": 1.0225463367693367e-07,
      "loss": 0.84161091,
      "memory(GiB)": 138.1,
      "step": 80360,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.80100813,
      "epoch": 1.8750929584858338,
      "grad_norm": 5.4375,
      "learning_rate": 1.0187488379875876e-07,
      "loss": 0.71097794,
      "memory(GiB)": 138.1,
      "step": 80370,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.79370942,
      "epoch": 1.875326266058123,
      "grad_norm": 6.65625,
      "learning_rate": 1.0149583312652089e-07,
      "loss": 0.73821492,
      "memory(GiB)": 138.1,
      "step": 80380,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.79008064,
      "epoch": 1.8755595736304116,
      "grad_norm": 6.1875,
      "learning_rate": 1.0111748171433067e-07,
      "loss": 0.76331358,
      "memory(GiB)": 138.1,
      "step": 80390,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.76334047,
      "epoch": 1.8757928812027005,
      "grad_norm": 6.1875,
      "learning_rate": 1.00739829616196e-07,
      "loss": 0.85111656,
      "memory(GiB)": 138.1,
      "step": 80400,
      "train_speed(iter/s)": 0.200962
    },
    {
      "acc": 0.78283181,
      "epoch": 1.8760261887749894,
      "grad_norm": 4.78125,
      "learning_rate": 1.0036287688602764e-07,
      "loss": 0.76812816,
      "memory(GiB)": 138.1,
      "step": 80410,
      "train_speed(iter/s)": 0.200976
    },
    {
      "acc": 0.76459651,
      "epoch": 1.8762594963472783,
      "grad_norm": 5.71875,
      "learning_rate": 9.998662357763534e-08,
      "loss": 0.85813875,
      "memory(GiB)": 138.1,
      "step": 80420,
      "train_speed(iter/s)": 0.200988
    },
    {
      "acc": 0.77645798,
      "epoch": 1.8764928039195672,
      "grad_norm": 4.84375,
      "learning_rate": 9.961106974472834e-08,
      "loss": 0.80161037,
      "memory(GiB)": 138.1,
      "step": 80430,
      "train_speed(iter/s)": 0.201
    },
    {
      "acc": 0.77092047,
      "epoch": 1.876726111491856,
      "grad_norm": 5.625,
      "learning_rate": 9.923621544091877e-08,
      "loss": 0.8250824,
      "memory(GiB)": 138.1,
      "step": 80440,
      "train_speed(iter/s)": 0.201013
    },
    {
      "acc": 0.79298916,
      "epoch": 1.876959419064145,
      "grad_norm": 6.1875,
      "learning_rate": 9.886206071971493e-08,
      "loss": 0.75579348,
      "memory(GiB)": 138.1,
      "step": 80450,
      "train_speed(iter/s)": 0.201025
    },
    {
      "acc": 0.77953978,
      "epoch": 1.8771927266364339,
      "grad_norm": 7.0625,
      "learning_rate": 9.848860563452855e-08,
      "loss": 0.78975258,
      "memory(GiB)": 138.1,
      "step": 80460,
      "train_speed(iter/s)": 0.201037
    },
    {
      "acc": 0.78567953,
      "epoch": 1.8774260342087228,
      "grad_norm": 3.96875,
      "learning_rate": 9.811585023866976e-08,
      "loss": 0.76402936,
      "memory(GiB)": 138.1,
      "step": 80470,
      "train_speed(iter/s)": 0.20105
    },
    {
      "acc": 0.78225698,
      "epoch": 1.8776593417810117,
      "grad_norm": 4.875,
      "learning_rate": 9.774379458534933e-08,
      "loss": 0.78416901,
      "memory(GiB)": 138.1,
      "step": 80480,
      "train_speed(iter/s)": 0.201063
    },
    {
      "acc": 0.77153597,
      "epoch": 1.8778926493533006,
      "grad_norm": 5.28125,
      "learning_rate": 9.73724387276781e-08,
      "loss": 0.84029179,
      "memory(GiB)": 138.1,
      "step": 80490,
      "train_speed(iter/s)": 0.201075
    },
    {
      "acc": 0.77242804,
      "epoch": 1.8781259569255895,
      "grad_norm": 5.1875,
      "learning_rate": 9.700178271866645e-08,
      "loss": 0.81851034,
      "memory(GiB)": 138.1,
      "step": 80500,
      "train_speed(iter/s)": 0.201087
    },
    {
      "epoch": 1.8781259569255895,
      "eval_acc": 0.7446892569323635,
      "eval_loss": 0.804384708404541,
      "eval_runtime": 1272.9616,
      "eval_samples_per_second": 28.273,
      "eval_steps_per_second": 14.137,
      "step": 80500
    },
    {
      "acc": 0.78566909,
      "epoch": 1.8783592644978784,
      "grad_norm": 5.0625,
      "learning_rate": 9.66318266112265e-08,
      "loss": 0.78130226,
      "memory(GiB)": 138.1,
      "step": 80510,
      "train_speed(iter/s)": 0.20045
    },
    {
      "acc": 0.78306484,
      "epoch": 1.8785925720701673,
      "grad_norm": 7.53125,
      "learning_rate": 9.626257045816879e-08,
      "loss": 0.77127929,
      "memory(GiB)": 138.1,
      "step": 80520,
      "train_speed(iter/s)": 0.200462
    },
    {
      "acc": 0.80443401,
      "epoch": 1.8788258796424562,
      "grad_norm": 4.59375,
      "learning_rate": 9.589401431220502e-08,
      "loss": 0.70631161,
      "memory(GiB)": 138.1,
      "step": 80530,
      "train_speed(iter/s)": 0.200474
    },
    {
      "acc": 0.78915844,
      "epoch": 1.879059187214745,
      "grad_norm": 5.4375,
      "learning_rate": 9.552615822594536e-08,
      "loss": 0.75501652,
      "memory(GiB)": 138.1,
      "step": 80540,
      "train_speed(iter/s)": 0.200486
    },
    {
      "acc": 0.77199645,
      "epoch": 1.879292494787034,
      "grad_norm": 7.125,
      "learning_rate": 9.515900225190222e-08,
      "loss": 0.8397831,
      "memory(GiB)": 138.1,
      "step": 80550,
      "train_speed(iter/s)": 0.200499
    },
    {
      "acc": 0.77649212,
      "epoch": 1.8795258023593229,
      "grad_norm": 8.0,
      "learning_rate": 9.479254644248648e-08,
      "loss": 0.80428057,
      "memory(GiB)": 138.1,
      "step": 80560,
      "train_speed(iter/s)": 0.200511
    },
    {
      "acc": 0.79153538,
      "epoch": 1.8797591099316118,
      "grad_norm": 5.5625,
      "learning_rate": 9.442679085000961e-08,
      "loss": 0.74753222,
      "memory(GiB)": 138.1,
      "step": 80570,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.77919335,
      "epoch": 1.8799924175039004,
      "grad_norm": 4.9375,
      "learning_rate": 9.40617355266843e-08,
      "loss": 0.80626259,
      "memory(GiB)": 138.1,
      "step": 80580,
      "train_speed(iter/s)": 0.200538
    },
    {
      "acc": 0.76336107,
      "epoch": 1.8802257250761896,
      "grad_norm": 5.1875,
      "learning_rate": 9.369738052461996e-08,
      "loss": 0.86293964,
      "memory(GiB)": 138.1,
      "step": 80590,
      "train_speed(iter/s)": 0.20055
    },
    {
      "acc": 0.79171991,
      "epoch": 1.8804590326484782,
      "grad_norm": 23.5,
      "learning_rate": 9.333372589583e-08,
      "loss": 0.75359583,
      "memory(GiB)": 138.1,
      "step": 80600,
      "train_speed(iter/s)": 0.200563
    },
    {
      "acc": 0.79271598,
      "epoch": 1.8806923402207674,
      "grad_norm": 8.5,
      "learning_rate": 9.297077169222513e-08,
      "loss": 0.74482245,
      "memory(GiB)": 138.1,
      "step": 80610,
      "train_speed(iter/s)": 0.200576
    },
    {
      "acc": 0.77063508,
      "epoch": 1.880925647793056,
      "grad_norm": 4.40625,
      "learning_rate": 9.260851796561609e-08,
      "loss": 0.84915276,
      "memory(GiB)": 138.1,
      "step": 80620,
      "train_speed(iter/s)": 0.20059
    },
    {
      "acc": 0.79158163,
      "epoch": 1.8811589553653452,
      "grad_norm": 3.96875,
      "learning_rate": 9.224696476771655e-08,
      "loss": 0.78680096,
      "memory(GiB)": 138.1,
      "step": 80630,
      "train_speed(iter/s)": 0.200602
    },
    {
      "acc": 0.77896152,
      "epoch": 1.8813922629376338,
      "grad_norm": 5.6875,
      "learning_rate": 9.188611215013631e-08,
      "loss": 0.79821224,
      "memory(GiB)": 138.1,
      "step": 80640,
      "train_speed(iter/s)": 0.200614
    },
    {
      "acc": 0.77518044,
      "epoch": 1.881625570509923,
      "grad_norm": 6.5,
      "learning_rate": 9.152596016438864e-08,
      "loss": 0.80904598,
      "memory(GiB)": 138.1,
      "step": 80650,
      "train_speed(iter/s)": 0.200627
    },
    {
      "acc": 0.7727581,
      "epoch": 1.8818588780822116,
      "grad_norm": 6.8125,
      "learning_rate": 9.11665088618835e-08,
      "loss": 0.83627472,
      "memory(GiB)": 138.1,
      "step": 80660,
      "train_speed(iter/s)": 0.20064
    },
    {
      "acc": 0.79154997,
      "epoch": 1.8820921856545008,
      "grad_norm": 4.25,
      "learning_rate": 9.080775829393373e-08,
      "loss": 0.75480595,
      "memory(GiB)": 138.1,
      "step": 80670,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.78853526,
      "epoch": 1.8823254932267894,
      "grad_norm": 6.28125,
      "learning_rate": 9.044970851175006e-08,
      "loss": 0.75794249,
      "memory(GiB)": 138.1,
      "step": 80680,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.77822475,
      "epoch": 1.8825588007990786,
      "grad_norm": 4.6875,
      "learning_rate": 9.009235956644491e-08,
      "loss": 0.79094992,
      "memory(GiB)": 138.1,
      "step": 80690,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.7686842,
      "epoch": 1.8827921083713672,
      "grad_norm": 4.75,
      "learning_rate": 8.97357115090286e-08,
      "loss": 0.84376907,
      "memory(GiB)": 138.1,
      "step": 80700,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.7691422,
      "epoch": 1.8830254159436564,
      "grad_norm": 5.625,
      "learning_rate": 8.937976439041263e-08,
      "loss": 0.83207283,
      "memory(GiB)": 138.1,
      "step": 80710,
      "train_speed(iter/s)": 0.200702
    },
    {
      "acc": 0.78920188,
      "epoch": 1.883258723515945,
      "grad_norm": 4.90625,
      "learning_rate": 8.902451826140911e-08,
      "loss": 0.74556618,
      "memory(GiB)": 138.1,
      "step": 80720,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.78106089,
      "epoch": 1.8834920310882342,
      "grad_norm": 6.59375,
      "learning_rate": 8.866997317272863e-08,
      "loss": 0.78032007,
      "memory(GiB)": 138.1,
      "step": 80730,
      "train_speed(iter/s)": 0.200726
    },
    {
      "acc": 0.80264349,
      "epoch": 1.8837253386605228,
      "grad_norm": 7.53125,
      "learning_rate": 8.831612917498288e-08,
      "loss": 0.69003105,
      "memory(GiB)": 138.1,
      "step": 80740,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.78006659,
      "epoch": 1.883958646232812,
      "grad_norm": 4.78125,
      "learning_rate": 8.796298631868317e-08,
      "loss": 0.79409275,
      "memory(GiB)": 138.1,
      "step": 80750,
      "train_speed(iter/s)": 0.200752
    },
    {
      "acc": 0.77235012,
      "epoch": 1.8841919538051006,
      "grad_norm": 6.09375,
      "learning_rate": 8.761054465423969e-08,
      "loss": 0.83254261,
      "memory(GiB)": 138.1,
      "step": 80760,
      "train_speed(iter/s)": 0.200764
    },
    {
      "acc": 0.79439788,
      "epoch": 1.8844252613773895,
      "grad_norm": 4.875,
      "learning_rate": 8.725880423196442e-08,
      "loss": 0.75280304,
      "memory(GiB)": 138.1,
      "step": 80770,
      "train_speed(iter/s)": 0.200777
    },
    {
      "acc": 0.7835988,
      "epoch": 1.8846585689496784,
      "grad_norm": 4.375,
      "learning_rate": 8.690776510206723e-08,
      "loss": 0.77280979,
      "memory(GiB)": 138.1,
      "step": 80780,
      "train_speed(iter/s)": 0.20079
    },
    {
      "acc": 0.79621954,
      "epoch": 1.8848918765219673,
      "grad_norm": 4.9375,
      "learning_rate": 8.655742731465966e-08,
      "loss": 0.74539442,
      "memory(GiB)": 138.1,
      "step": 80790,
      "train_speed(iter/s)": 0.200803
    },
    {
      "acc": 0.77132044,
      "epoch": 1.8851251840942562,
      "grad_norm": 6.1875,
      "learning_rate": 8.62077909197523e-08,
      "loss": 0.80378342,
      "memory(GiB)": 138.1,
      "step": 80800,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.79714594,
      "epoch": 1.8853584916665451,
      "grad_norm": 5.78125,
      "learning_rate": 8.585885596725518e-08,
      "loss": 0.71226463,
      "memory(GiB)": 138.1,
      "step": 80810,
      "train_speed(iter/s)": 0.200829
    },
    {
      "acc": 0.76155009,
      "epoch": 1.885591799238834,
      "grad_norm": 9.1875,
      "learning_rate": 8.551062250697795e-08,
      "loss": 0.89277039,
      "memory(GiB)": 138.1,
      "step": 80820,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.79220524,
      "epoch": 1.885825106811123,
      "grad_norm": 3.609375,
      "learning_rate": 8.516309058863249e-08,
      "loss": 0.72857275,
      "memory(GiB)": 138.1,
      "step": 80830,
      "train_speed(iter/s)": 0.200855
    },
    {
      "acc": 0.77188721,
      "epoch": 1.8860584143834118,
      "grad_norm": 4.71875,
      "learning_rate": 8.481626026182798e-08,
      "loss": 0.84043217,
      "memory(GiB)": 138.1,
      "step": 80840,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.79511795,
      "epoch": 1.8862917219557007,
      "grad_norm": 5.8125,
      "learning_rate": 8.447013157607431e-08,
      "loss": 0.7310348,
      "memory(GiB)": 138.1,
      "step": 80850,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.79568253,
      "epoch": 1.8865250295279896,
      "grad_norm": 6.03125,
      "learning_rate": 8.412470458078137e-08,
      "loss": 0.72156382,
      "memory(GiB)": 138.1,
      "step": 80860,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.75479288,
      "epoch": 1.8867583371002785,
      "grad_norm": 8.25,
      "learning_rate": 8.37799793252586e-08,
      "loss": 0.89089012,
      "memory(GiB)": 138.1,
      "step": 80870,
      "train_speed(iter/s)": 0.200906
    },
    {
      "acc": 0.79052868,
      "epoch": 1.8869916446725674,
      "grad_norm": 5.53125,
      "learning_rate": 8.343595585871611e-08,
      "loss": 0.74636126,
      "memory(GiB)": 138.1,
      "step": 80880,
      "train_speed(iter/s)": 0.200919
    },
    {
      "acc": 0.77063046,
      "epoch": 1.8872249522448563,
      "grad_norm": 4.78125,
      "learning_rate": 8.309263423026237e-08,
      "loss": 0.81843338,
      "memory(GiB)": 138.1,
      "step": 80890,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.77630882,
      "epoch": 1.8874582598171452,
      "grad_norm": 5.8125,
      "learning_rate": 8.275001448890651e-08,
      "loss": 0.81249247,
      "memory(GiB)": 138.1,
      "step": 80900,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.75715265,
      "epoch": 1.887691567389434,
      "grad_norm": 5.65625,
      "learning_rate": 8.240809668355776e-08,
      "loss": 0.88787174,
      "memory(GiB)": 138.1,
      "step": 80910,
      "train_speed(iter/s)": 0.200959
    },
    {
      "acc": 0.7806591,
      "epoch": 1.887924874961723,
      "grad_norm": 6.15625,
      "learning_rate": 8.206688086302483e-08,
      "loss": 0.80691376,
      "memory(GiB)": 138.1,
      "step": 80920,
      "train_speed(iter/s)": 0.200972
    },
    {
      "acc": 0.78252301,
      "epoch": 1.888158182534012,
      "grad_norm": 5.25,
      "learning_rate": 8.1726367076016e-08,
      "loss": 0.78188124,
      "memory(GiB)": 138.1,
      "step": 80930,
      "train_speed(iter/s)": 0.200985
    },
    {
      "acc": 0.77645779,
      "epoch": 1.8883914901063008,
      "grad_norm": 5.90625,
      "learning_rate": 8.138655537113904e-08,
      "loss": 0.82004871,
      "memory(GiB)": 138.1,
      "step": 80940,
      "train_speed(iter/s)": 0.200998
    },
    {
      "acc": 0.79138808,
      "epoch": 1.8886247976785897,
      "grad_norm": 4.375,
      "learning_rate": 8.104744579690294e-08,
      "loss": 0.75213003,
      "memory(GiB)": 138.1,
      "step": 80950,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.77159157,
      "epoch": 1.8888581052508786,
      "grad_norm": 5.15625,
      "learning_rate": 8.070903840171451e-08,
      "loss": 0.84294567,
      "memory(GiB)": 138.1,
      "step": 80960,
      "train_speed(iter/s)": 0.201022
    },
    {
      "acc": 0.77069464,
      "epoch": 1.8890914128231673,
      "grad_norm": 7.9375,
      "learning_rate": 8.037133323388236e-08,
      "loss": 0.82081795,
      "memory(GiB)": 138.1,
      "step": 80970,
      "train_speed(iter/s)": 0.201035
    },
    {
      "acc": 0.79293261,
      "epoch": 1.8893247203954564,
      "grad_norm": 5.75,
      "learning_rate": 8.00343303416129e-08,
      "loss": 0.74969621,
      "memory(GiB)": 138.1,
      "step": 80980,
      "train_speed(iter/s)": 0.201049
    },
    {
      "acc": 0.78427496,
      "epoch": 1.889558027967745,
      "grad_norm": 6.34375,
      "learning_rate": 7.969802977301322e-08,
      "loss": 0.77531166,
      "memory(GiB)": 138.1,
      "step": 80990,
      "train_speed(iter/s)": 0.201062
    },
    {
      "acc": 0.77599058,
      "epoch": 1.8897913355400342,
      "grad_norm": 5.4375,
      "learning_rate": 7.936243157609103e-08,
      "loss": 0.79406672,
      "memory(GiB)": 138.1,
      "step": 81000,
      "train_speed(iter/s)": 0.201076
    },
    {
      "epoch": 1.8897913355400342,
      "eval_acc": 0.7446983851536048,
      "eval_loss": 0.804445207118988,
      "eval_runtime": 1271.5518,
      "eval_samples_per_second": 28.305,
      "eval_steps_per_second": 14.153,
      "step": 81000
    },
    {
      "acc": 0.7871388,
      "epoch": 1.8900246431123229,
      "grad_norm": 5.40625,
      "learning_rate": 7.902753579875189e-08,
      "loss": 0.78215551,
      "memory(GiB)": 138.1,
      "step": 81010,
      "train_speed(iter/s)": 0.200443
    },
    {
      "acc": 0.76717043,
      "epoch": 1.890257950684612,
      "grad_norm": 11.125,
      "learning_rate": 7.869334248880256e-08,
      "loss": 0.82725344,
      "memory(GiB)": 138.1,
      "step": 81020,
      "train_speed(iter/s)": 0.200456
    },
    {
      "acc": 0.77815256,
      "epoch": 1.8904912582569007,
      "grad_norm": 5.96875,
      "learning_rate": 7.835985169394933e-08,
      "loss": 0.7883893,
      "memory(GiB)": 138.1,
      "step": 81030,
      "train_speed(iter/s)": 0.200469
    },
    {
      "acc": 0.76834126,
      "epoch": 1.8907245658291898,
      "grad_norm": 4.8125,
      "learning_rate": 7.802706346179744e-08,
      "loss": 0.84197254,
      "memory(GiB)": 138.1,
      "step": 81040,
      "train_speed(iter/s)": 0.200482
    },
    {
      "acc": 0.7763195,
      "epoch": 1.8909578734014785,
      "grad_norm": 5.53125,
      "learning_rate": 7.769497783985224e-08,
      "loss": 0.81596165,
      "memory(GiB)": 138.1,
      "step": 81050,
      "train_speed(iter/s)": 0.200494
    },
    {
      "acc": 0.77682486,
      "epoch": 1.8911911809737676,
      "grad_norm": 4.71875,
      "learning_rate": 7.736359487551859e-08,
      "loss": 0.80225954,
      "memory(GiB)": 138.1,
      "step": 81060,
      "train_speed(iter/s)": 0.200507
    },
    {
      "acc": 0.77334614,
      "epoch": 1.8914244885460563,
      "grad_norm": 5.71875,
      "learning_rate": 7.703291461610252e-08,
      "loss": 0.81839981,
      "memory(GiB)": 138.1,
      "step": 81070,
      "train_speed(iter/s)": 0.20052
    },
    {
      "acc": 0.76000309,
      "epoch": 1.8916577961183454,
      "grad_norm": 7.65625,
      "learning_rate": 7.670293710880683e-08,
      "loss": 0.86824694,
      "memory(GiB)": 138.1,
      "step": 81080,
      "train_speed(iter/s)": 0.200533
    },
    {
      "acc": 0.75734501,
      "epoch": 1.891891103690634,
      "grad_norm": 4.4375,
      "learning_rate": 7.637366240073717e-08,
      "loss": 0.89891548,
      "memory(GiB)": 138.1,
      "step": 81090,
      "train_speed(iter/s)": 0.200546
    },
    {
      "acc": 0.78540058,
      "epoch": 1.8921244112629232,
      "grad_norm": 4.4375,
      "learning_rate": 7.604509053889708e-08,
      "loss": 0.7534976,
      "memory(GiB)": 138.1,
      "step": 81100,
      "train_speed(iter/s)": 0.200559
    },
    {
      "acc": 0.76539426,
      "epoch": 1.8923577188352119,
      "grad_norm": 6.15625,
      "learning_rate": 7.571722157018957e-08,
      "loss": 0.84462833,
      "memory(GiB)": 138.1,
      "step": 81110,
      "train_speed(iter/s)": 0.200571
    },
    {
      "acc": 0.79330382,
      "epoch": 1.892591026407501,
      "grad_norm": 4.5625,
      "learning_rate": 7.539005554141831e-08,
      "loss": 0.72500844,
      "memory(GiB)": 138.1,
      "step": 81120,
      "train_speed(iter/s)": 0.200584
    },
    {
      "acc": 0.80007896,
      "epoch": 1.8928243339797897,
      "grad_norm": 3.875,
      "learning_rate": 7.506359249928542e-08,
      "loss": 0.71526747,
      "memory(GiB)": 138.1,
      "step": 81130,
      "train_speed(iter/s)": 0.200596
    },
    {
      "acc": 0.77556849,
      "epoch": 1.8930576415520788,
      "grad_norm": 5.8125,
      "learning_rate": 7.473783249039468e-08,
      "loss": 0.80747452,
      "memory(GiB)": 138.1,
      "step": 81140,
      "train_speed(iter/s)": 0.200609
    },
    {
      "acc": 0.78302164,
      "epoch": 1.8932909491243675,
      "grad_norm": 5.8125,
      "learning_rate": 7.441277556124781e-08,
      "loss": 0.77237792,
      "memory(GiB)": 138.1,
      "step": 81150,
      "train_speed(iter/s)": 0.200622
    },
    {
      "acc": 0.78645997,
      "epoch": 1.8935242566966564,
      "grad_norm": 5.84375,
      "learning_rate": 7.408842175824604e-08,
      "loss": 0.76683159,
      "memory(GiB)": 138.1,
      "step": 81160,
      "train_speed(iter/s)": 0.200635
    },
    {
      "acc": 0.78576627,
      "epoch": 1.8937575642689453,
      "grad_norm": 5.28125,
      "learning_rate": 7.376477112769064e-08,
      "loss": 0.77226114,
      "memory(GiB)": 138.1,
      "step": 81170,
      "train_speed(iter/s)": 0.200648
    },
    {
      "acc": 0.77925673,
      "epoch": 1.8939908718412342,
      "grad_norm": 5.0,
      "learning_rate": 7.344182371578356e-08,
      "loss": 0.79336381,
      "memory(GiB)": 138.1,
      "step": 81180,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.79144859,
      "epoch": 1.894224179413523,
      "grad_norm": 4.46875,
      "learning_rate": 7.311957956862459e-08,
      "loss": 0.73583159,
      "memory(GiB)": 138.1,
      "step": 81190,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.7666995,
      "epoch": 1.894457486985812,
      "grad_norm": 5.5625,
      "learning_rate": 7.279803873221469e-08,
      "loss": 0.84476967,
      "memory(GiB)": 138.1,
      "step": 81200,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.80386267,
      "epoch": 1.8946907945581009,
      "grad_norm": 5.375,
      "learning_rate": 7.247720125245328e-08,
      "loss": 0.68902006,
      "memory(GiB)": 138.1,
      "step": 81210,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.79350948,
      "epoch": 1.8949241021303898,
      "grad_norm": 5.8125,
      "learning_rate": 7.215706717513982e-08,
      "loss": 0.72404013,
      "memory(GiB)": 138.1,
      "step": 81220,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.76372366,
      "epoch": 1.8951574097026787,
      "grad_norm": 5.375,
      "learning_rate": 7.183763654597387e-08,
      "loss": 0.8393321,
      "memory(GiB)": 138.1,
      "step": 81230,
      "train_speed(iter/s)": 0.200725
    },
    {
      "acc": 0.79516706,
      "epoch": 1.8953907172749676,
      "grad_norm": 6.53125,
      "learning_rate": 7.15189094105534e-08,
      "loss": 0.72492518,
      "memory(GiB)": 138.1,
      "step": 81240,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.76616926,
      "epoch": 1.8956240248472565,
      "grad_norm": 4.40625,
      "learning_rate": 7.120088581437645e-08,
      "loss": 0.85155525,
      "memory(GiB)": 138.1,
      "step": 81250,
      "train_speed(iter/s)": 0.200751
    },
    {
      "acc": 0.79208398,
      "epoch": 1.8958573324195453,
      "grad_norm": 4.1875,
      "learning_rate": 7.088356580284228e-08,
      "loss": 0.73564825,
      "memory(GiB)": 138.1,
      "step": 81260,
      "train_speed(iter/s)": 0.200763
    },
    {
      "acc": 0.77262831,
      "epoch": 1.8960906399918342,
      "grad_norm": 5.5625,
      "learning_rate": 7.05669494212463e-08,
      "loss": 0.818857,
      "memory(GiB)": 138.1,
      "step": 81270,
      "train_speed(iter/s)": 0.200776
    },
    {
      "acc": 0.77157354,
      "epoch": 1.8963239475641231,
      "grad_norm": 6.28125,
      "learning_rate": 7.025103671478684e-08,
      "loss": 0.81951456,
      "memory(GiB)": 138.1,
      "step": 81280,
      "train_speed(iter/s)": 0.200789
    },
    {
      "acc": 0.78546553,
      "epoch": 1.896557255136412,
      "grad_norm": 4.53125,
      "learning_rate": 6.993582772855889e-08,
      "loss": 0.77391062,
      "memory(GiB)": 138.1,
      "step": 81290,
      "train_speed(iter/s)": 0.200801
    },
    {
      "acc": 0.77268543,
      "epoch": 1.896790562708701,
      "grad_norm": 4.59375,
      "learning_rate": 6.962132250756037e-08,
      "loss": 0.82842484,
      "memory(GiB)": 138.1,
      "step": 81300,
      "train_speed(iter/s)": 0.200815
    },
    {
      "acc": 0.7767509,
      "epoch": 1.8970238702809898,
      "grad_norm": 5.53125,
      "learning_rate": 6.930752109668481e-08,
      "loss": 0.7966526,
      "memory(GiB)": 138.1,
      "step": 81310,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.7757802,
      "epoch": 1.8972571778532787,
      "grad_norm": 4.3125,
      "learning_rate": 6.899442354072916e-08,
      "loss": 0.79777894,
      "memory(GiB)": 138.1,
      "step": 81320,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.77078934,
      "epoch": 1.8974904854255676,
      "grad_norm": 5.84375,
      "learning_rate": 6.868202988438655e-08,
      "loss": 0.82341976,
      "memory(GiB)": 138.1,
      "step": 81330,
      "train_speed(iter/s)": 0.200852
    },
    {
      "acc": 0.78582072,
      "epoch": 1.8977237929978565,
      "grad_norm": 5.65625,
      "learning_rate": 6.837034017225186e-08,
      "loss": 0.78013291,
      "memory(GiB)": 138.1,
      "step": 81340,
      "train_speed(iter/s)": 0.200864
    },
    {
      "acc": 0.77980404,
      "epoch": 1.8979571005701454,
      "grad_norm": 4.96875,
      "learning_rate": 6.80593544488184e-08,
      "loss": 0.78744049,
      "memory(GiB)": 138.1,
      "step": 81350,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.77276716,
      "epoch": 1.8981904081424341,
      "grad_norm": 5.625,
      "learning_rate": 6.774907275847898e-08,
      "loss": 0.81919241,
      "memory(GiB)": 138.1,
      "step": 81360,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.76729136,
      "epoch": 1.8984237157147232,
      "grad_norm": 5.21875,
      "learning_rate": 6.743949514552706e-08,
      "loss": 0.83895359,
      "memory(GiB)": 138.1,
      "step": 81370,
      "train_speed(iter/s)": 0.200903
    },
    {
      "acc": 0.78235312,
      "epoch": 1.898657023287012,
      "grad_norm": 4.375,
      "learning_rate": 6.713062165415451e-08,
      "loss": 0.77152939,
      "memory(GiB)": 138.1,
      "step": 81380,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.79524784,
      "epoch": 1.898890330859301,
      "grad_norm": 7.625,
      "learning_rate": 6.682245232845219e-08,
      "loss": 0.74803662,
      "memory(GiB)": 138.1,
      "step": 81390,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.78636365,
      "epoch": 1.8991236384315897,
      "grad_norm": 6.09375,
      "learning_rate": 6.651498721241212e-08,
      "loss": 0.78164091,
      "memory(GiB)": 138.1,
      "step": 81400,
      "train_speed(iter/s)": 0.200941
    },
    {
      "acc": 0.7818541,
      "epoch": 1.8993569460038788,
      "grad_norm": 4.1875,
      "learning_rate": 6.620822634992419e-08,
      "loss": 0.79073544,
      "memory(GiB)": 138.1,
      "step": 81410,
      "train_speed(iter/s)": 0.200954
    },
    {
      "acc": 0.7828022,
      "epoch": 1.8995902535761675,
      "grad_norm": 5.71875,
      "learning_rate": 6.590216978477836e-08,
      "loss": 0.77449131,
      "memory(GiB)": 138.1,
      "step": 81420,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.78085957,
      "epoch": 1.8998235611484566,
      "grad_norm": 6.0625,
      "learning_rate": 6.559681756066471e-08,
      "loss": 0.7848876,
      "memory(GiB)": 138.1,
      "step": 81430,
      "train_speed(iter/s)": 0.200978
    },
    {
      "acc": 0.78292851,
      "epoch": 1.9000568687207453,
      "grad_norm": 5.15625,
      "learning_rate": 6.529216972117225e-08,
      "loss": 0.77457685,
      "memory(GiB)": 138.1,
      "step": 81440,
      "train_speed(iter/s)": 0.20099
    },
    {
      "acc": 0.76646495,
      "epoch": 1.9002901762930344,
      "grad_norm": 5.03125,
      "learning_rate": 6.49882263097884e-08,
      "loss": 0.8381773,
      "memory(GiB)": 138.1,
      "step": 81450,
      "train_speed(iter/s)": 0.201003
    },
    {
      "acc": 0.78961086,
      "epoch": 1.900523483865323,
      "grad_norm": 4.78125,
      "learning_rate": 6.468498736990181e-08,
      "loss": 0.73249388,
      "memory(GiB)": 138.1,
      "step": 81460,
      "train_speed(iter/s)": 0.201015
    },
    {
      "acc": 0.77733517,
      "epoch": 1.9007567914376122,
      "grad_norm": 5.0,
      "learning_rate": 6.438245294480006e-08,
      "loss": 0.78677988,
      "memory(GiB)": 138.1,
      "step": 81470,
      "train_speed(iter/s)": 0.201028
    },
    {
      "acc": 0.79435854,
      "epoch": 1.900990099009901,
      "grad_norm": 4.96875,
      "learning_rate": 6.40806230776686e-08,
      "loss": 0.71958585,
      "memory(GiB)": 138.1,
      "step": 81480,
      "train_speed(iter/s)": 0.20104
    },
    {
      "acc": 0.7977644,
      "epoch": 1.90122340658219,
      "grad_norm": 6.25,
      "learning_rate": 6.37794978115952e-08,
      "loss": 0.73467293,
      "memory(GiB)": 138.1,
      "step": 81490,
      "train_speed(iter/s)": 0.201052
    },
    {
      "acc": 0.78441677,
      "epoch": 1.9014567141544787,
      "grad_norm": 7.0625,
      "learning_rate": 6.347907718956381e-08,
      "loss": 0.77021914,
      "memory(GiB)": 138.1,
      "step": 81500,
      "train_speed(iter/s)": 0.201064
    },
    {
      "epoch": 1.9014567141544787,
      "eval_acc": 0.7447350581828017,
      "eval_loss": 0.8044015765190125,
      "eval_runtime": 1270.7097,
      "eval_samples_per_second": 28.324,
      "eval_steps_per_second": 14.162,
      "step": 81500
    },
    {
      "acc": 0.78291373,
      "epoch": 1.9016900217267678,
      "grad_norm": 5.34375,
      "learning_rate": 6.317936125446012e-08,
      "loss": 0.78251152,
      "memory(GiB)": 138.1,
      "step": 81510,
      "train_speed(iter/s)": 0.200436
    },
    {
      "acc": 0.76142359,
      "epoch": 1.9019233292990565,
      "grad_norm": 6.21875,
      "learning_rate": 6.288035004906878e-08,
      "loss": 0.87129774,
      "memory(GiB)": 138.1,
      "step": 81520,
      "train_speed(iter/s)": 0.200449
    },
    {
      "acc": 0.80396862,
      "epoch": 1.9021566368713456,
      "grad_norm": 6.1875,
      "learning_rate": 6.258204361607289e-08,
      "loss": 0.69574623,
      "memory(GiB)": 138.1,
      "step": 81530,
      "train_speed(iter/s)": 0.200461
    },
    {
      "acc": 0.78162894,
      "epoch": 1.9023899444436343,
      "grad_norm": 4.46875,
      "learning_rate": 6.228444199805617e-08,
      "loss": 0.77852755,
      "memory(GiB)": 138.1,
      "step": 81540,
      "train_speed(iter/s)": 0.200474
    },
    {
      "acc": 0.77957335,
      "epoch": 1.9026232520159232,
      "grad_norm": 4.875,
      "learning_rate": 6.198754523750072e-08,
      "loss": 0.80971165,
      "memory(GiB)": 138.1,
      "step": 81550,
      "train_speed(iter/s)": 0.200487
    },
    {
      "acc": 0.7833252,
      "epoch": 1.902856559588212,
      "grad_norm": 5.84375,
      "learning_rate": 6.169135337678878e-08,
      "loss": 0.77970705,
      "memory(GiB)": 138.1,
      "step": 81560,
      "train_speed(iter/s)": 0.2005
    },
    {
      "acc": 0.77797418,
      "epoch": 1.903089867160501,
      "grad_norm": 7.125,
      "learning_rate": 6.13958664582015e-08,
      "loss": 0.79811144,
      "memory(GiB)": 138.1,
      "step": 81570,
      "train_speed(iter/s)": 0.200513
    },
    {
      "acc": 0.77659426,
      "epoch": 1.90332317473279,
      "grad_norm": 5.9375,
      "learning_rate": 6.11010845239196e-08,
      "loss": 0.80996256,
      "memory(GiB)": 138.1,
      "step": 81580,
      "train_speed(iter/s)": 0.200525
    },
    {
      "acc": 0.79570951,
      "epoch": 1.9035564823050788,
      "grad_norm": 4.21875,
      "learning_rate": 6.080700761602331e-08,
      "loss": 0.72712016,
      "memory(GiB)": 138.1,
      "step": 81590,
      "train_speed(iter/s)": 0.200538
    },
    {
      "acc": 0.77094784,
      "epoch": 1.9037897898773677,
      "grad_norm": 5.1875,
      "learning_rate": 6.051363577649238e-08,
      "loss": 0.83303509,
      "memory(GiB)": 138.1,
      "step": 81600,
      "train_speed(iter/s)": 0.200551
    },
    {
      "acc": 0.77590933,
      "epoch": 1.9040230974496566,
      "grad_norm": 5.25,
      "learning_rate": 6.022096904720388e-08,
      "loss": 0.79160347,
      "memory(GiB)": 138.1,
      "step": 81610,
      "train_speed(iter/s)": 0.200562
    },
    {
      "acc": 0.77262087,
      "epoch": 1.9042564050219455,
      "grad_norm": 4.75,
      "learning_rate": 5.992900746993768e-08,
      "loss": 0.82788048,
      "memory(GiB)": 138.1,
      "step": 81620,
      "train_speed(iter/s)": 0.200575
    },
    {
      "acc": 0.77261982,
      "epoch": 1.9044897125942344,
      "grad_norm": 8.3125,
      "learning_rate": 5.963775108637048e-08,
      "loss": 0.83426991,
      "memory(GiB)": 138.1,
      "step": 81630,
      "train_speed(iter/s)": 0.200588
    },
    {
      "acc": 0.76702747,
      "epoch": 1.9047230201665233,
      "grad_norm": 5.75,
      "learning_rate": 5.9347199938079026e-08,
      "loss": 0.8470952,
      "memory(GiB)": 138.1,
      "step": 81640,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.77795296,
      "epoch": 1.9049563277388122,
      "grad_norm": 4.84375,
      "learning_rate": 5.9057354066539564e-08,
      "loss": 0.79301691,
      "memory(GiB)": 138.1,
      "step": 81650,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.77648511,
      "epoch": 1.905189635311101,
      "grad_norm": 5.40625,
      "learning_rate": 5.876821351312734e-08,
      "loss": 0.81472111,
      "memory(GiB)": 138.1,
      "step": 81660,
      "train_speed(iter/s)": 0.200626
    },
    {
      "acc": 0.78697462,
      "epoch": 1.90542294288339,
      "grad_norm": 5.625,
      "learning_rate": 5.8479778319117665e-08,
      "loss": 0.76727009,
      "memory(GiB)": 138.1,
      "step": 81670,
      "train_speed(iter/s)": 0.200639
    },
    {
      "acc": 0.76950393,
      "epoch": 1.9056562504556789,
      "grad_norm": 4.59375,
      "learning_rate": 5.819204852568372e-08,
      "loss": 0.82860451,
      "memory(GiB)": 138.1,
      "step": 81680,
      "train_speed(iter/s)": 0.200652
    },
    {
      "acc": 0.76736298,
      "epoch": 1.9058895580279678,
      "grad_norm": 6.875,
      "learning_rate": 5.790502417389876e-08,
      "loss": 0.82692204,
      "memory(GiB)": 138.1,
      "step": 81690,
      "train_speed(iter/s)": 0.200665
    },
    {
      "acc": 0.77684221,
      "epoch": 1.9061228656002567,
      "grad_norm": 5.15625,
      "learning_rate": 5.7618705304736676e-08,
      "loss": 0.79659696,
      "memory(GiB)": 138.1,
      "step": 81700,
      "train_speed(iter/s)": 0.200677
    },
    {
      "acc": 0.76629496,
      "epoch": 1.9063561731725456,
      "grad_norm": 4.25,
      "learning_rate": 5.733309195906811e-08,
      "loss": 0.8493578,
      "memory(GiB)": 138.1,
      "step": 81710,
      "train_speed(iter/s)": 0.20069
    },
    {
      "acc": 0.76272459,
      "epoch": 1.9065894807448345,
      "grad_norm": 5.40625,
      "learning_rate": 5.7048184177666e-08,
      "loss": 0.84202557,
      "memory(GiB)": 138.1,
      "step": 81720,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.77872486,
      "epoch": 1.9068227883171232,
      "grad_norm": 7.9375,
      "learning_rate": 5.676398200119837e-08,
      "loss": 0.78867388,
      "memory(GiB)": 138.1,
      "step": 81730,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.79240322,
      "epoch": 1.9070560958894123,
      "grad_norm": 4.65625,
      "learning_rate": 5.648048547023666e-08,
      "loss": 0.71631851,
      "memory(GiB)": 138.1,
      "step": 81740,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.77871118,
      "epoch": 1.907289403461701,
      "grad_norm": 5.1875,
      "learning_rate": 5.61976946252496e-08,
      "loss": 0.79847703,
      "memory(GiB)": 138.1,
      "step": 81750,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.77508669,
      "epoch": 1.90752271103399,
      "grad_norm": 4.71875,
      "learning_rate": 5.591560950660546e-08,
      "loss": 0.8124464,
      "memory(GiB)": 138.1,
      "step": 81760,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.79896374,
      "epoch": 1.9077560186062787,
      "grad_norm": 5.15625,
      "learning_rate": 5.563423015457203e-08,
      "loss": 0.72898488,
      "memory(GiB)": 138.1,
      "step": 81770,
      "train_speed(iter/s)": 0.200767
    },
    {
      "acc": 0.80807238,
      "epoch": 1.9079893261785679,
      "grad_norm": 4.5,
      "learning_rate": 5.535355660931552e-08,
      "loss": 0.68237019,
      "memory(GiB)": 138.1,
      "step": 81780,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.77174668,
      "epoch": 1.9082226337508565,
      "grad_norm": 4.6875,
      "learning_rate": 5.50735889109022e-08,
      "loss": 0.82769318,
      "memory(GiB)": 138.1,
      "step": 81790,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.75991964,
      "epoch": 1.9084559413231457,
      "grad_norm": 4.3125,
      "learning_rate": 5.4794327099297887e-08,
      "loss": 0.86439896,
      "memory(GiB)": 138.1,
      "step": 81800,
      "train_speed(iter/s)": 0.200805
    },
    {
      "acc": 0.76893587,
      "epoch": 1.9086892488954343,
      "grad_norm": 5.15625,
      "learning_rate": 5.45157712143668e-08,
      "loss": 0.82247448,
      "memory(GiB)": 138.1,
      "step": 81810,
      "train_speed(iter/s)": 0.200817
    },
    {
      "acc": 0.79243269,
      "epoch": 1.9089225564677235,
      "grad_norm": 5.03125,
      "learning_rate": 5.423792129587269e-08,
      "loss": 0.71618028,
      "memory(GiB)": 138.1,
      "step": 81820,
      "train_speed(iter/s)": 0.20083
    },
    {
      "acc": 0.79535136,
      "epoch": 1.9091558640400121,
      "grad_norm": 4.21875,
      "learning_rate": 5.396077738347882e-08,
      "loss": 0.74078579,
      "memory(GiB)": 138.1,
      "step": 81830,
      "train_speed(iter/s)": 0.200842
    },
    {
      "acc": 0.79126263,
      "epoch": 1.9093891716123013,
      "grad_norm": 5.90625,
      "learning_rate": 5.36843395167469e-08,
      "loss": 0.76003933,
      "memory(GiB)": 138.1,
      "step": 81840,
      "train_speed(iter/s)": 0.200855
    },
    {
      "acc": 0.78736057,
      "epoch": 1.90962247918459,
      "grad_norm": 5.1875,
      "learning_rate": 5.340860773513812e-08,
      "loss": 0.76994362,
      "memory(GiB)": 138.1,
      "step": 81850,
      "train_speed(iter/s)": 0.200867
    },
    {
      "acc": 0.77512932,
      "epoch": 1.909855786756879,
      "grad_norm": 6.625,
      "learning_rate": 5.313358207801433e-08,
      "loss": 0.81734104,
      "memory(GiB)": 138.1,
      "step": 81860,
      "train_speed(iter/s)": 0.20088
    },
    {
      "acc": 0.77062392,
      "epoch": 1.9100890943291677,
      "grad_norm": 5.1875,
      "learning_rate": 5.285926258463414e-08,
      "loss": 0.80548515,
      "memory(GiB)": 138.1,
      "step": 81870,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.78182507,
      "epoch": 1.9103224019014569,
      "grad_norm": 5.4375,
      "learning_rate": 5.2585649294157326e-08,
      "loss": 0.79319897,
      "memory(GiB)": 138.1,
      "step": 81880,
      "train_speed(iter/s)": 0.200904
    },
    {
      "acc": 0.78957319,
      "epoch": 1.9105557094737455,
      "grad_norm": 7.28125,
      "learning_rate": 5.231274224564154e-08,
      "loss": 0.7621007,
      "memory(GiB)": 138.1,
      "step": 81890,
      "train_speed(iter/s)": 0.200916
    },
    {
      "acc": 0.77407713,
      "epoch": 1.9107890170460347,
      "grad_norm": 5.0625,
      "learning_rate": 5.2040541478044496e-08,
      "loss": 0.82251482,
      "memory(GiB)": 138.1,
      "step": 81900,
      "train_speed(iter/s)": 0.200929
    },
    {
      "acc": 0.75633006,
      "epoch": 1.9110223246183233,
      "grad_norm": 6.5,
      "learning_rate": 5.176904703022345e-08,
      "loss": 0.87226925,
      "memory(GiB)": 138.1,
      "step": 81910,
      "train_speed(iter/s)": 0.200942
    },
    {
      "acc": 0.77395215,
      "epoch": 1.9112556321906125,
      "grad_norm": 6.0625,
      "learning_rate": 5.149825894093241e-08,
      "loss": 0.80833683,
      "memory(GiB)": 138.1,
      "step": 81920,
      "train_speed(iter/s)": 0.200955
    },
    {
      "acc": 0.76668892,
      "epoch": 1.9114889397629011,
      "grad_norm": 7.75,
      "learning_rate": 5.1228177248828224e-08,
      "loss": 0.82612123,
      "memory(GiB)": 138.1,
      "step": 81930,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.7853972,
      "epoch": 1.91172224733519,
      "grad_norm": 6.59375,
      "learning_rate": 5.0958801992463944e-08,
      "loss": 0.76109056,
      "memory(GiB)": 138.1,
      "step": 81940,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.78190913,
      "epoch": 1.911955554907479,
      "grad_norm": 8.0,
      "learning_rate": 5.069013321029326e-08,
      "loss": 0.7810389,
      "memory(GiB)": 138.1,
      "step": 81950,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.76182289,
      "epoch": 1.9121888624797678,
      "grad_norm": 5.125,
      "learning_rate": 5.0422170940667147e-08,
      "loss": 0.85948324,
      "memory(GiB)": 138.1,
      "step": 81960,
      "train_speed(iter/s)": 0.201006
    },
    {
      "acc": 0.78014874,
      "epoch": 1.9124221700520567,
      "grad_norm": 7.4375,
      "learning_rate": 5.015491522183946e-08,
      "loss": 0.78254461,
      "memory(GiB)": 138.1,
      "step": 81970,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.77584715,
      "epoch": 1.9126554776243456,
      "grad_norm": 6.25,
      "learning_rate": 4.9888366091959125e-08,
      "loss": 0.8201292,
      "memory(GiB)": 138.1,
      "step": 81980,
      "train_speed(iter/s)": 0.201031
    },
    {
      "acc": 0.76424618,
      "epoch": 1.9128887851966345,
      "grad_norm": 5.1875,
      "learning_rate": 4.962252358907627e-08,
      "loss": 0.85319223,
      "memory(GiB)": 138.1,
      "step": 81990,
      "train_speed(iter/s)": 0.201044
    },
    {
      "acc": 0.76576996,
      "epoch": 1.9131220927689234,
      "grad_norm": 5.125,
      "learning_rate": 4.935738775114052e-08,
      "loss": 0.84802933,
      "memory(GiB)": 138.1,
      "step": 82000,
      "train_speed(iter/s)": 0.201056
    },
    {
      "epoch": 1.9131220927689234,
      "eval_acc": 0.7447390617886093,
      "eval_loss": 0.8044153451919556,
      "eval_runtime": 1273.2096,
      "eval_samples_per_second": 28.268,
      "eval_steps_per_second": 14.134,
      "step": 82000
    },
    {
      "acc": 0.78968439,
      "epoch": 1.9133554003412123,
      "grad_norm": 5.59375,
      "learning_rate": 4.9092958615999385e-08,
      "loss": 0.77019253,
      "memory(GiB)": 138.1,
      "step": 82010,
      "train_speed(iter/s)": 0.20043
    },
    {
      "acc": 0.75527754,
      "epoch": 1.9135887079135012,
      "grad_norm": 14.875,
      "learning_rate": 4.882923622140046e-08,
      "loss": 0.87390118,
      "memory(GiB)": 138.1,
      "step": 82020,
      "train_speed(iter/s)": 0.200442
    },
    {
      "acc": 0.77405939,
      "epoch": 1.9138220154857901,
      "grad_norm": 5.5625,
      "learning_rate": 4.856622060498972e-08,
      "loss": 0.80371208,
      "memory(GiB)": 138.1,
      "step": 82030,
      "train_speed(iter/s)": 0.200455
    },
    {
      "acc": 0.78950992,
      "epoch": 1.914055323058079,
      "grad_norm": 3.859375,
      "learning_rate": 4.83039118043116e-08,
      "loss": 0.744206,
      "memory(GiB)": 138.1,
      "step": 82040,
      "train_speed(iter/s)": 0.200467
    },
    {
      "acc": 0.78136091,
      "epoch": 1.914288630630368,
      "grad_norm": 9.9375,
      "learning_rate": 4.80423098568128e-08,
      "loss": 0.781563,
      "memory(GiB)": 138.1,
      "step": 82050,
      "train_speed(iter/s)": 0.20048
    },
    {
      "acc": 0.76963954,
      "epoch": 1.9145219382026568,
      "grad_norm": 4.9375,
      "learning_rate": 4.7781414799835116e-08,
      "loss": 0.82037382,
      "memory(GiB)": 138.1,
      "step": 82060,
      "train_speed(iter/s)": 0.200492
    },
    {
      "acc": 0.80206232,
      "epoch": 1.9147552457749457,
      "grad_norm": 5.6875,
      "learning_rate": 4.752122667062209e-08,
      "loss": 0.70895433,
      "memory(GiB)": 138.1,
      "step": 82070,
      "train_speed(iter/s)": 0.200505
    },
    {
      "acc": 0.79812136,
      "epoch": 1.9149885533472346,
      "grad_norm": 4.40625,
      "learning_rate": 4.726174550631513e-08,
      "loss": 0.68837681,
      "memory(GiB)": 138.1,
      "step": 82080,
      "train_speed(iter/s)": 0.200516
    },
    {
      "acc": 0.78133001,
      "epoch": 1.9152218609195235,
      "grad_norm": 4.71875,
      "learning_rate": 4.7002971343955153e-08,
      "loss": 0.74567728,
      "memory(GiB)": 138.1,
      "step": 82090,
      "train_speed(iter/s)": 0.200529
    },
    {
      "acc": 0.80090256,
      "epoch": 1.9154551684918124,
      "grad_norm": 4.21875,
      "learning_rate": 4.6744904220482056e-08,
      "loss": 0.71008487,
      "memory(GiB)": 138.1,
      "step": 82100,
      "train_speed(iter/s)": 0.200542
    },
    {
      "acc": 0.7789093,
      "epoch": 1.9156884760641013,
      "grad_norm": 4.59375,
      "learning_rate": 4.648754417273526e-08,
      "loss": 0.78899159,
      "memory(GiB)": 138.1,
      "step": 82110,
      "train_speed(iter/s)": 0.200554
    },
    {
      "acc": 0.77255368,
      "epoch": 1.91592178363639,
      "grad_norm": 7.25,
      "learning_rate": 4.623089123745261e-08,
      "loss": 0.82642889,
      "memory(GiB)": 138.1,
      "step": 82120,
      "train_speed(iter/s)": 0.200566
    },
    {
      "acc": 0.79890847,
      "epoch": 1.916155091208679,
      "grad_norm": 5.15625,
      "learning_rate": 4.597494545127035e-08,
      "loss": 0.71255264,
      "memory(GiB)": 138.1,
      "step": 82130,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.77682538,
      "epoch": 1.9163883987809678,
      "grad_norm": 8.0625,
      "learning_rate": 4.571970685072646e-08,
      "loss": 0.7757277,
      "memory(GiB)": 138.1,
      "step": 82140,
      "train_speed(iter/s)": 0.200591
    },
    {
      "acc": 0.78598347,
      "epoch": 1.916621706353257,
      "grad_norm": 3.734375,
      "learning_rate": 4.5465175472254594e-08,
      "loss": 0.75754504,
      "memory(GiB)": 138.1,
      "step": 82150,
      "train_speed(iter/s)": 0.200604
    },
    {
      "acc": 0.78972554,
      "epoch": 1.9168550139255456,
      "grad_norm": 8.1875,
      "learning_rate": 4.521135135218957e-08,
      "loss": 0.7571701,
      "memory(GiB)": 138.1,
      "step": 82160,
      "train_speed(iter/s)": 0.200615
    },
    {
      "acc": 0.77707663,
      "epoch": 1.9170883214978347,
      "grad_norm": 4.84375,
      "learning_rate": 4.495823452676518e-08,
      "loss": 0.82133408,
      "memory(GiB)": 138.1,
      "step": 82170,
      "train_speed(iter/s)": 0.200628
    },
    {
      "acc": 0.76835198,
      "epoch": 1.9173216290701234,
      "grad_norm": 5.5,
      "learning_rate": 4.4705825032113624e-08,
      "loss": 0.83846817,
      "memory(GiB)": 138.1,
      "step": 82180,
      "train_speed(iter/s)": 0.20064
    },
    {
      "acc": 0.77485347,
      "epoch": 1.9175549366424125,
      "grad_norm": 6.78125,
      "learning_rate": 4.445412290426554e-08,
      "loss": 0.81650391,
      "memory(GiB)": 138.1,
      "step": 82190,
      "train_speed(iter/s)": 0.200653
    },
    {
      "acc": 0.77240715,
      "epoch": 1.9177882442147012,
      "grad_norm": 7.0625,
      "learning_rate": 4.420312817915162e-08,
      "loss": 0.81943703,
      "memory(GiB)": 138.1,
      "step": 82200,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.7817626,
      "epoch": 1.9180215517869903,
      "grad_norm": 6.25,
      "learning_rate": 4.39528408926021e-08,
      "loss": 0.78492823,
      "memory(GiB)": 138.1,
      "step": 82210,
      "train_speed(iter/s)": 0.200678
    },
    {
      "acc": 0.77328238,
      "epoch": 1.918254859359279,
      "grad_norm": 6.84375,
      "learning_rate": 4.370326108034395e-08,
      "loss": 0.81196861,
      "memory(GiB)": 138.1,
      "step": 82220,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.76239452,
      "epoch": 1.918488166931568,
      "grad_norm": 7.53125,
      "learning_rate": 4.3454388778005894e-08,
      "loss": 0.88454895,
      "memory(GiB)": 138.1,
      "step": 82230,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.76898079,
      "epoch": 1.9187214745038568,
      "grad_norm": 5.25,
      "learning_rate": 4.3206224021113966e-08,
      "loss": 0.84615307,
      "memory(GiB)": 138.1,
      "step": 82240,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.765411,
      "epoch": 1.918954782076146,
      "grad_norm": 6.9375,
      "learning_rate": 4.2958766845093704e-08,
      "loss": 0.85209064,
      "memory(GiB)": 138.1,
      "step": 82250,
      "train_speed(iter/s)": 0.200728
    },
    {
      "acc": 0.77735281,
      "epoch": 1.9191880896484346,
      "grad_norm": 6.15625,
      "learning_rate": 4.271201728526963e-08,
      "loss": 0.79499168,
      "memory(GiB)": 138.1,
      "step": 82260,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.78337259,
      "epoch": 1.9194213972207237,
      "grad_norm": 4.71875,
      "learning_rate": 4.24659753768647e-08,
      "loss": 0.80247488,
      "memory(GiB)": 138.1,
      "step": 82270,
      "train_speed(iter/s)": 0.200755
    },
    {
      "acc": 0.78709459,
      "epoch": 1.9196547047930124,
      "grad_norm": 5.21875,
      "learning_rate": 4.222064115500191e-08,
      "loss": 0.76674123,
      "memory(GiB)": 138.1,
      "step": 82280,
      "train_speed(iter/s)": 0.200767
    },
    {
      "acc": 0.77503223,
      "epoch": 1.9198880123653015,
      "grad_norm": 11.125,
      "learning_rate": 4.1976014654702135e-08,
      "loss": 0.8402914,
      "memory(GiB)": 138.1,
      "step": 82290,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.77781281,
      "epoch": 1.9201213199375902,
      "grad_norm": 7.5,
      "learning_rate": 4.1732095910885785e-08,
      "loss": 0.78994112,
      "memory(GiB)": 138.1,
      "step": 82300,
      "train_speed(iter/s)": 0.200792
    },
    {
      "acc": 0.79021931,
      "epoch": 1.920354627509879,
      "grad_norm": 4.34375,
      "learning_rate": 4.148888495837222e-08,
      "loss": 0.7459506,
      "memory(GiB)": 138.1,
      "step": 82310,
      "train_speed(iter/s)": 0.200805
    },
    {
      "acc": 0.79193068,
      "epoch": 1.920587935082168,
      "grad_norm": 4.40625,
      "learning_rate": 4.1246381831880345e-08,
      "loss": 0.73066068,
      "memory(GiB)": 138.1,
      "step": 82320,
      "train_speed(iter/s)": 0.200818
    },
    {
      "acc": 0.7667829,
      "epoch": 1.9208212426544569,
      "grad_norm": 4.65625,
      "learning_rate": 4.1004586566026904e-08,
      "loss": 0.80857086,
      "memory(GiB)": 138.1,
      "step": 82330,
      "train_speed(iter/s)": 0.200831
    },
    {
      "acc": 0.78284898,
      "epoch": 1.9210545502267458,
      "grad_norm": 6.53125,
      "learning_rate": 4.076349919532763e-08,
      "loss": 0.79354701,
      "memory(GiB)": 138.1,
      "step": 82340,
      "train_speed(iter/s)": 0.200844
    },
    {
      "acc": 0.79704752,
      "epoch": 1.9212878577990347,
      "grad_norm": 6.46875,
      "learning_rate": 4.052311975419887e-08,
      "loss": 0.75222983,
      "memory(GiB)": 138.1,
      "step": 82350,
      "train_speed(iter/s)": 0.200856
    },
    {
      "acc": 0.78382096,
      "epoch": 1.9215211653713236,
      "grad_norm": 4.25,
      "learning_rate": 4.0283448276953186e-08,
      "loss": 0.79394064,
      "memory(GiB)": 138.1,
      "step": 82360,
      "train_speed(iter/s)": 0.200869
    },
    {
      "acc": 0.78321552,
      "epoch": 1.9217544729436125,
      "grad_norm": 3.859375,
      "learning_rate": 4.004448479780543e-08,
      "loss": 0.77877493,
      "memory(GiB)": 138.1,
      "step": 82370,
      "train_speed(iter/s)": 0.200881
    },
    {
      "acc": 0.78489828,
      "epoch": 1.9219877805159014,
      "grad_norm": 3.953125,
      "learning_rate": 3.9806229350865534e-08,
      "loss": 0.76953378,
      "memory(GiB)": 138.1,
      "step": 82380,
      "train_speed(iter/s)": 0.200893
    },
    {
      "acc": 0.77053556,
      "epoch": 1.9222210880881903,
      "grad_norm": 4.78125,
      "learning_rate": 3.95686819701463e-08,
      "loss": 0.85348272,
      "memory(GiB)": 138.1,
      "step": 82390,
      "train_speed(iter/s)": 0.200906
    },
    {
      "acc": 0.77399349,
      "epoch": 1.9224543956604792,
      "grad_norm": 7.875,
      "learning_rate": 3.93318426895567e-08,
      "loss": 0.81090708,
      "memory(GiB)": 138.1,
      "step": 82400,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.79830041,
      "epoch": 1.922687703232768,
      "grad_norm": 5.28125,
      "learning_rate": 3.9095711542905257e-08,
      "loss": 0.71653395,
      "memory(GiB)": 138.1,
      "step": 82410,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.81262779,
      "epoch": 1.922921010805057,
      "grad_norm": 5.4375,
      "learning_rate": 3.8860288563900006e-08,
      "loss": 0.65959501,
      "memory(GiB)": 138.1,
      "step": 82420,
      "train_speed(iter/s)": 0.200944
    },
    {
      "acc": 0.77822075,
      "epoch": 1.9231543183773459,
      "grad_norm": 5.53125,
      "learning_rate": 3.862557378614684e-08,
      "loss": 0.79374828,
      "memory(GiB)": 138.1,
      "step": 82430,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.7975666,
      "epoch": 1.9233876259496347,
      "grad_norm": 7.4375,
      "learning_rate": 3.839156724315174e-08,
      "loss": 0.73447914,
      "memory(GiB)": 138.1,
      "step": 82440,
      "train_speed(iter/s)": 0.200969
    },
    {
      "acc": 0.78388729,
      "epoch": 1.9236209335219236,
      "grad_norm": 3.90625,
      "learning_rate": 3.815826896831909e-08,
      "loss": 0.79074955,
      "memory(GiB)": 138.1,
      "step": 82450,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.78147302,
      "epoch": 1.9238542410942125,
      "grad_norm": 5.875,
      "learning_rate": 3.792567899495281e-08,
      "loss": 0.78904285,
      "memory(GiB)": 138.1,
      "step": 82460,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.79638376,
      "epoch": 1.9240875486665014,
      "grad_norm": 6.28125,
      "learning_rate": 3.7693797356254115e-08,
      "loss": 0.72602043,
      "memory(GiB)": 138.1,
      "step": 82470,
      "train_speed(iter/s)": 0.201007
    },
    {
      "acc": 0.79340577,
      "epoch": 1.9243208562387903,
      "grad_norm": 5.25,
      "learning_rate": 3.746262408532375e-08,
      "loss": 0.72227745,
      "memory(GiB)": 138.1,
      "step": 82480,
      "train_speed(iter/s)": 0.201019
    },
    {
      "acc": 0.76795883,
      "epoch": 1.9245541638110792,
      "grad_norm": 6.40625,
      "learning_rate": 3.723215921516254e-08,
      "loss": 0.83940735,
      "memory(GiB)": 138.1,
      "step": 82490,
      "train_speed(iter/s)": 0.201032
    },
    {
      "acc": 0.78825803,
      "epoch": 1.9247874713833681,
      "grad_norm": 6.46875,
      "learning_rate": 3.7002402778668625e-08,
      "loss": 0.78783998,
      "memory(GiB)": 138.1,
      "step": 82500,
      "train_speed(iter/s)": 0.201046
    },
    {
      "epoch": 1.9247874713833681,
      "eval_acc": 0.7447004670286247,
      "eval_loss": 0.8044062852859497,
      "eval_runtime": 1271.4855,
      "eval_samples_per_second": 28.306,
      "eval_steps_per_second": 14.154,
      "step": 82500
    },
    {
      "acc": 0.79530158,
      "epoch": 1.9250207789556568,
      "grad_norm": 4.71875,
      "learning_rate": 3.6773354808640194e-08,
      "loss": 0.71304455,
      "memory(GiB)": 138.1,
      "step": 82510,
      "train_speed(iter/s)": 0.200425
    },
    {
      "acc": 0.79802475,
      "epoch": 1.925254086527946,
      "grad_norm": 3.96875,
      "learning_rate": 3.654501533777388e-08,
      "loss": 0.71960392,
      "memory(GiB)": 138.1,
      "step": 82520,
      "train_speed(iter/s)": 0.200437
    },
    {
      "acc": 0.77804232,
      "epoch": 1.9254873941002346,
      "grad_norm": 4.375,
      "learning_rate": 3.6317384398664167e-08,
      "loss": 0.79983625,
      "memory(GiB)": 138.1,
      "step": 82530,
      "train_speed(iter/s)": 0.20045
    },
    {
      "acc": 0.78172102,
      "epoch": 1.9257207016725237,
      "grad_norm": 5.0625,
      "learning_rate": 3.6090462023806175e-08,
      "loss": 0.79019833,
      "memory(GiB)": 138.1,
      "step": 82540,
      "train_speed(iter/s)": 0.200463
    },
    {
      "acc": 0.77815161,
      "epoch": 1.9259540092448124,
      "grad_norm": 3.71875,
      "learning_rate": 3.586424824559287e-08,
      "loss": 0.80372667,
      "memory(GiB)": 138.1,
      "step": 82550,
      "train_speed(iter/s)": 0.200475
    },
    {
      "acc": 0.77129049,
      "epoch": 1.9261873168171015,
      "grad_norm": 5.28125,
      "learning_rate": 3.563874309631565e-08,
      "loss": 0.84257956,
      "memory(GiB)": 138.1,
      "step": 82560,
      "train_speed(iter/s)": 0.200487
    },
    {
      "acc": 0.77810478,
      "epoch": 1.9264206243893902,
      "grad_norm": 5.03125,
      "learning_rate": 3.5413946608165995e-08,
      "loss": 0.79976082,
      "memory(GiB)": 138.1,
      "step": 82570,
      "train_speed(iter/s)": 0.200499
    },
    {
      "acc": 0.77861633,
      "epoch": 1.9266539319616793,
      "grad_norm": 5.5625,
      "learning_rate": 3.518985881323322e-08,
      "loss": 0.78327084,
      "memory(GiB)": 138.1,
      "step": 82580,
      "train_speed(iter/s)": 0.200511
    },
    {
      "acc": 0.80329628,
      "epoch": 1.926887239533968,
      "grad_norm": 6.125,
      "learning_rate": 3.496647974350509e-08,
      "loss": 0.69266319,
      "memory(GiB)": 138.1,
      "step": 82590,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.77527142,
      "epoch": 1.9271205471062571,
      "grad_norm": 4.8125,
      "learning_rate": 3.4743809430870525e-08,
      "loss": 0.80601139,
      "memory(GiB)": 138.1,
      "step": 82600,
      "train_speed(iter/s)": 0.200538
    },
    {
      "acc": 0.7910471,
      "epoch": 1.9273538546785458,
      "grad_norm": 5.1875,
      "learning_rate": 3.452184790711466e-08,
      "loss": 0.7612885,
      "memory(GiB)": 138.1,
      "step": 82610,
      "train_speed(iter/s)": 0.200551
    },
    {
      "acc": 0.77904315,
      "epoch": 1.927587162250835,
      "grad_norm": 5.53125,
      "learning_rate": 3.430059520392215e-08,
      "loss": 0.79040279,
      "memory(GiB)": 138.1,
      "step": 82620,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.80225992,
      "epoch": 1.9278204698231236,
      "grad_norm": 4.65625,
      "learning_rate": 3.408005135287773e-08,
      "loss": 0.70399323,
      "memory(GiB)": 138.1,
      "step": 82630,
      "train_speed(iter/s)": 0.200576
    },
    {
      "acc": 0.75945187,
      "epoch": 1.9280537773954127,
      "grad_norm": 5.90625,
      "learning_rate": 3.386021638546233e-08,
      "loss": 0.87596512,
      "memory(GiB)": 138.1,
      "step": 82640,
      "train_speed(iter/s)": 0.200586
    },
    {
      "acc": 0.7902338,
      "epoch": 1.9282870849677014,
      "grad_norm": 5.5,
      "learning_rate": 3.364109033305918e-08,
      "loss": 0.74522243,
      "memory(GiB)": 138.1,
      "step": 82650,
      "train_speed(iter/s)": 0.200599
    },
    {
      "acc": 0.78515444,
      "epoch": 1.9285203925399905,
      "grad_norm": 4.71875,
      "learning_rate": 3.3422673226947145e-08,
      "loss": 0.76768475,
      "memory(GiB)": 138.1,
      "step": 82660,
      "train_speed(iter/s)": 0.200611
    },
    {
      "acc": 0.77809896,
      "epoch": 1.9287537001122792,
      "grad_norm": 6.21875,
      "learning_rate": 3.3204965098306284e-08,
      "loss": 0.79778728,
      "memory(GiB)": 138.1,
      "step": 82670,
      "train_speed(iter/s)": 0.200624
    },
    {
      "acc": 0.77451992,
      "epoch": 1.9289870076845683,
      "grad_norm": 7.25,
      "learning_rate": 3.298796597821341e-08,
      "loss": 0.80516586,
      "memory(GiB)": 138.1,
      "step": 82680,
      "train_speed(iter/s)": 0.200636
    },
    {
      "acc": 0.77414317,
      "epoch": 1.929220315256857,
      "grad_norm": 6.0,
      "learning_rate": 3.2771675897645405e-08,
      "loss": 0.80401745,
      "memory(GiB)": 138.1,
      "step": 82690,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.78727055,
      "epoch": 1.929453622829146,
      "grad_norm": 5.78125,
      "learning_rate": 3.255609488747813e-08,
      "loss": 0.74130497,
      "memory(GiB)": 138.1,
      "step": 82700,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.78334398,
      "epoch": 1.9296869304014348,
      "grad_norm": 4.40625,
      "learning_rate": 3.234122297848474e-08,
      "loss": 0.75857935,
      "memory(GiB)": 138.1,
      "step": 82710,
      "train_speed(iter/s)": 0.200673
    },
    {
      "acc": 0.78041697,
      "epoch": 1.9299202379737237,
      "grad_norm": 5.4375,
      "learning_rate": 3.212706020133904e-08,
      "loss": 0.80112228,
      "memory(GiB)": 138.1,
      "step": 82720,
      "train_speed(iter/s)": 0.200685
    },
    {
      "acc": 0.78621788,
      "epoch": 1.9301535455460126,
      "grad_norm": 5.875,
      "learning_rate": 3.1913606586612135e-08,
      "loss": 0.74973865,
      "memory(GiB)": 138.1,
      "step": 82730,
      "train_speed(iter/s)": 0.200697
    },
    {
      "acc": 0.76397328,
      "epoch": 1.9303868531183015,
      "grad_norm": 6.0,
      "learning_rate": 3.170086216477464e-08,
      "loss": 0.86834965,
      "memory(GiB)": 138.1,
      "step": 82740,
      "train_speed(iter/s)": 0.20071
    },
    {
      "acc": 0.76520152,
      "epoch": 1.9306201606905904,
      "grad_norm": 4.59375,
      "learning_rate": 3.148882696619615e-08,
      "loss": 0.86656876,
      "memory(GiB)": 138.1,
      "step": 82750,
      "train_speed(iter/s)": 0.200721
    },
    {
      "acc": 0.78473167,
      "epoch": 1.9308534682628793,
      "grad_norm": 5.59375,
      "learning_rate": 3.127750102114358e-08,
      "loss": 0.77946978,
      "memory(GiB)": 138.1,
      "step": 82760,
      "train_speed(iter/s)": 0.200733
    },
    {
      "acc": 0.78480234,
      "epoch": 1.9310867758351682,
      "grad_norm": 8.0,
      "learning_rate": 3.1066884359785e-08,
      "loss": 0.77041464,
      "memory(GiB)": 138.1,
      "step": 82770,
      "train_speed(iter/s)": 0.200747
    },
    {
      "acc": 0.81039772,
      "epoch": 1.931320083407457,
      "grad_norm": 5.5,
      "learning_rate": 3.085697701218471e-08,
      "loss": 0.65555139,
      "memory(GiB)": 138.1,
      "step": 82780,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.77078209,
      "epoch": 1.931553390979746,
      "grad_norm": 4.8125,
      "learning_rate": 3.064777900830762e-08,
      "loss": 0.82027807,
      "memory(GiB)": 138.1,
      "step": 82790,
      "train_speed(iter/s)": 0.200772
    },
    {
      "acc": 0.77345366,
      "epoch": 1.9317866985520349,
      "grad_norm": 5.875,
      "learning_rate": 3.043929037801596e-08,
      "loss": 0.79783955,
      "memory(GiB)": 138.1,
      "step": 82800,
      "train_speed(iter/s)": 0.200786
    },
    {
      "acc": 0.78663902,
      "epoch": 1.9320200061243238,
      "grad_norm": 8.375,
      "learning_rate": 3.023151115107259e-08,
      "loss": 0.77612534,
      "memory(GiB)": 138.1,
      "step": 82810,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.78559351,
      "epoch": 1.9322533136966127,
      "grad_norm": 7.21875,
      "learning_rate": 3.002444135713711e-08,
      "loss": 0.76495595,
      "memory(GiB)": 138.1,
      "step": 82820,
      "train_speed(iter/s)": 0.200811
    },
    {
      "acc": 0.79293275,
      "epoch": 1.9324866212689016,
      "grad_norm": 5.03125,
      "learning_rate": 2.9818081025768667e-08,
      "loss": 0.75699339,
      "memory(GiB)": 138.1,
      "step": 82830,
      "train_speed(iter/s)": 0.200823
    },
    {
      "acc": 0.77729588,
      "epoch": 1.9327199288411905,
      "grad_norm": 7.78125,
      "learning_rate": 2.9612430186425346e-08,
      "loss": 0.81681576,
      "memory(GiB)": 138.1,
      "step": 82840,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.78396635,
      "epoch": 1.9329532364134794,
      "grad_norm": 12.375,
      "learning_rate": 2.9407488868463675e-08,
      "loss": 0.77789707,
      "memory(GiB)": 138.1,
      "step": 82850,
      "train_speed(iter/s)": 0.200849
    },
    {
      "acc": 0.77509494,
      "epoch": 1.9331865439857683,
      "grad_norm": 6.46875,
      "learning_rate": 2.9203257101139694e-08,
      "loss": 0.80295143,
      "memory(GiB)": 138.1,
      "step": 82860,
      "train_speed(iter/s)": 0.200862
    },
    {
      "acc": 0.78139973,
      "epoch": 1.9334198515580572,
      "grad_norm": 5.03125,
      "learning_rate": 2.8999734913606193e-08,
      "loss": 0.7943119,
      "memory(GiB)": 138.1,
      "step": 82870,
      "train_speed(iter/s)": 0.200875
    },
    {
      "acc": 0.77936926,
      "epoch": 1.933653159130346,
      "grad_norm": 8.5,
      "learning_rate": 2.8796922334916044e-08,
      "loss": 0.78870902,
      "memory(GiB)": 138.1,
      "step": 82880,
      "train_speed(iter/s)": 0.200887
    },
    {
      "acc": 0.78523445,
      "epoch": 1.933886466702635,
      "grad_norm": 4.71875,
      "learning_rate": 2.8594819394021646e-08,
      "loss": 0.76724324,
      "memory(GiB)": 138.1,
      "step": 82890,
      "train_speed(iter/s)": 0.2009
    },
    {
      "acc": 0.78431993,
      "epoch": 1.9341197742749237,
      "grad_norm": 4.65625,
      "learning_rate": 2.8393426119772138e-08,
      "loss": 0.77278762,
      "memory(GiB)": 138.1,
      "step": 82900,
      "train_speed(iter/s)": 0.200912
    },
    {
      "acc": 0.77910967,
      "epoch": 1.9343530818472128,
      "grad_norm": 14.5625,
      "learning_rate": 2.8192742540917305e-08,
      "loss": 0.80649271,
      "memory(GiB)": 138.1,
      "step": 82910,
      "train_speed(iter/s)": 0.200925
    },
    {
      "acc": 0.78101177,
      "epoch": 1.9345863894195015,
      "grad_norm": 5.8125,
      "learning_rate": 2.799276868610368e-08,
      "loss": 0.7846736,
      "memory(GiB)": 138.1,
      "step": 82920,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.7781353,
      "epoch": 1.9348196969917906,
      "grad_norm": 4.75,
      "learning_rate": 2.7793504583878417e-08,
      "loss": 0.79075127,
      "memory(GiB)": 138.1,
      "step": 82930,
      "train_speed(iter/s)": 0.20095
    },
    {
      "acc": 0.77793598,
      "epoch": 1.9350530045640792,
      "grad_norm": 5.53125,
      "learning_rate": 2.7594950262685438e-08,
      "loss": 0.78132844,
      "memory(GiB)": 138.1,
      "step": 82940,
      "train_speed(iter/s)": 0.200962
    },
    {
      "acc": 0.76155319,
      "epoch": 1.9352863121363684,
      "grad_norm": 8.5625,
      "learning_rate": 2.739710575086929e-08,
      "loss": 0.8619318,
      "memory(GiB)": 138.1,
      "step": 82950,
      "train_speed(iter/s)": 0.200975
    },
    {
      "acc": 0.80455065,
      "epoch": 1.935519619708657,
      "grad_norm": 5.09375,
      "learning_rate": 2.719997107667127e-08,
      "loss": 0.71163216,
      "memory(GiB)": 138.1,
      "step": 82960,
      "train_speed(iter/s)": 0.200987
    },
    {
      "acc": 0.78168869,
      "epoch": 1.9357529272809462,
      "grad_norm": 8.0,
      "learning_rate": 2.7003546268233317e-08,
      "loss": 0.76260276,
      "memory(GiB)": 138.1,
      "step": 82970,
      "train_speed(iter/s)": 0.200999
    },
    {
      "acc": 0.75511971,
      "epoch": 1.9359862348532348,
      "grad_norm": 4.59375,
      "learning_rate": 2.6807831353594106e-08,
      "loss": 0.87710342,
      "memory(GiB)": 138.1,
      "step": 82980,
      "train_speed(iter/s)": 0.20101
    },
    {
      "acc": 0.80496244,
      "epoch": 1.936219542425524,
      "grad_norm": 5.28125,
      "learning_rate": 2.66128263606924e-08,
      "loss": 0.68820372,
      "memory(GiB)": 138.1,
      "step": 82990,
      "train_speed(iter/s)": 0.201023
    },
    {
      "acc": 0.78904171,
      "epoch": 1.9364528499978126,
      "grad_norm": 5.65625,
      "learning_rate": 2.6418531317364825e-08,
      "loss": 0.74845371,
      "memory(GiB)": 138.1,
      "step": 83000,
      "train_speed(iter/s)": 0.201035
    },
    {
      "epoch": 1.9364528499978126,
      "eval_acc": 0.7447185633268747,
      "eval_loss": 0.8043991327285767,
      "eval_runtime": 1272.2124,
      "eval_samples_per_second": 28.29,
      "eval_steps_per_second": 14.145,
      "step": 83000
    },
    {
      "acc": 0.76983991,
      "epoch": 1.9366861575701018,
      "grad_norm": 5.0625,
      "learning_rate": 2.6224946251346973e-08,
      "loss": 0.84707108,
      "memory(GiB)": 138.1,
      "step": 83010,
      "train_speed(iter/s)": 0.200418
    },
    {
      "acc": 0.76536455,
      "epoch": 1.9369194651423904,
      "grad_norm": 4.5625,
      "learning_rate": 2.60320711902734e-08,
      "loss": 0.84149084,
      "memory(GiB)": 138.1,
      "step": 83020,
      "train_speed(iter/s)": 0.200431
    },
    {
      "acc": 0.78333254,
      "epoch": 1.9371527727146796,
      "grad_norm": 5.375,
      "learning_rate": 2.5839906161676532e-08,
      "loss": 0.77945223,
      "memory(GiB)": 138.1,
      "step": 83030,
      "train_speed(iter/s)": 0.200443
    },
    {
      "acc": 0.79502096,
      "epoch": 1.9373860802869682,
      "grad_norm": 5.1875,
      "learning_rate": 2.564845119298831e-08,
      "loss": 0.73158178,
      "memory(GiB)": 138.1,
      "step": 83040,
      "train_speed(iter/s)": 0.200455
    },
    {
      "acc": 0.77735152,
      "epoch": 1.9376193878592574,
      "grad_norm": 5.25,
      "learning_rate": 2.5457706311538544e-08,
      "loss": 0.81248274,
      "memory(GiB)": 138.1,
      "step": 83050,
      "train_speed(iter/s)": 0.200467
    },
    {
      "acc": 0.78180962,
      "epoch": 1.937852695431546,
      "grad_norm": 4.5625,
      "learning_rate": 2.526767154455545e-08,
      "loss": 0.7838274,
      "memory(GiB)": 138.1,
      "step": 83060,
      "train_speed(iter/s)": 0.200479
    },
    {
      "acc": 0.80066261,
      "epoch": 1.9380860030038352,
      "grad_norm": 5.5,
      "learning_rate": 2.5078346919167883e-08,
      "loss": 0.71001968,
      "memory(GiB)": 138.1,
      "step": 83070,
      "train_speed(iter/s)": 0.200491
    },
    {
      "acc": 0.76299033,
      "epoch": 1.9383193105761238,
      "grad_norm": 6.84375,
      "learning_rate": 2.4889732462400895e-08,
      "loss": 0.84023495,
      "memory(GiB)": 138.1,
      "step": 83080,
      "train_speed(iter/s)": 0.200504
    },
    {
      "acc": 0.79042563,
      "epoch": 1.9385526181484127,
      "grad_norm": 5.71875,
      "learning_rate": 2.4701828201179057e-08,
      "loss": 0.74814739,
      "memory(GiB)": 138.1,
      "step": 83090,
      "train_speed(iter/s)": 0.200516
    },
    {
      "acc": 0.78727136,
      "epoch": 1.9387859257207016,
      "grad_norm": 5.53125,
      "learning_rate": 2.451463416232591e-08,
      "loss": 0.76182685,
      "memory(GiB)": 138.1,
      "step": 83100,
      "train_speed(iter/s)": 0.200529
    },
    {
      "acc": 0.78427858,
      "epoch": 1.9390192332929905,
      "grad_norm": 6.0625,
      "learning_rate": 2.4328150372563974e-08,
      "loss": 0.77642002,
      "memory(GiB)": 138.1,
      "step": 83110,
      "train_speed(iter/s)": 0.200541
    },
    {
      "acc": 0.75748053,
      "epoch": 1.9392525408652794,
      "grad_norm": 5.46875,
      "learning_rate": 2.4142376858512506e-08,
      "loss": 0.87451792,
      "memory(GiB)": 138.1,
      "step": 83120,
      "train_speed(iter/s)": 0.200554
    },
    {
      "acc": 0.79059162,
      "epoch": 1.9394858484375683,
      "grad_norm": 5.4375,
      "learning_rate": 2.3957313646691406e-08,
      "loss": 0.74902925,
      "memory(GiB)": 138.1,
      "step": 83130,
      "train_speed(iter/s)": 0.200567
    },
    {
      "acc": 0.78442965,
      "epoch": 1.9397191560098572,
      "grad_norm": 4.5625,
      "learning_rate": 2.3772960763518428e-08,
      "loss": 0.77176137,
      "memory(GiB)": 138.1,
      "step": 83140,
      "train_speed(iter/s)": 0.20058
    },
    {
      "acc": 0.78968287,
      "epoch": 1.9399524635821461,
      "grad_norm": 4.34375,
      "learning_rate": 2.358931823530919e-08,
      "loss": 0.73917861,
      "memory(GiB)": 138.1,
      "step": 83150,
      "train_speed(iter/s)": 0.200592
    },
    {
      "acc": 0.79481835,
      "epoch": 1.940185771154435,
      "grad_norm": 4.625,
      "learning_rate": 2.3406386088279387e-08,
      "loss": 0.73219633,
      "memory(GiB)": 138.1,
      "step": 83160,
      "train_speed(iter/s)": 0.200605
    },
    {
      "acc": 0.78026934,
      "epoch": 1.940419078726724,
      "grad_norm": 9.0,
      "learning_rate": 2.3224164348542576e-08,
      "loss": 0.78946395,
      "memory(GiB)": 138.1,
      "step": 83170,
      "train_speed(iter/s)": 0.200617
    },
    {
      "acc": 0.77109671,
      "epoch": 1.9406523862990128,
      "grad_norm": 10.5625,
      "learning_rate": 2.3042653042110175e-08,
      "loss": 0.8287899,
      "memory(GiB)": 138.1,
      "step": 83180,
      "train_speed(iter/s)": 0.20063
    },
    {
      "acc": 0.77708349,
      "epoch": 1.9408856938713017,
      "grad_norm": 4.84375,
      "learning_rate": 2.2861852194893118e-08,
      "loss": 0.80719547,
      "memory(GiB)": 138.1,
      "step": 83190,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.78134727,
      "epoch": 1.9411190014435906,
      "grad_norm": 6.28125,
      "learning_rate": 2.2681761832701323e-08,
      "loss": 0.78660898,
      "memory(GiB)": 138.1,
      "step": 83200,
      "train_speed(iter/s)": 0.200654
    },
    {
      "acc": 0.76660013,
      "epoch": 1.9413523090158795,
      "grad_norm": 7.09375,
      "learning_rate": 2.2502381981241993e-08,
      "loss": 0.84043627,
      "memory(GiB)": 138.1,
      "step": 83210,
      "train_speed(iter/s)": 0.200666
    },
    {
      "acc": 0.78337593,
      "epoch": 1.9415856165881684,
      "grad_norm": 4.5,
      "learning_rate": 2.2323712666121324e-08,
      "loss": 0.77146144,
      "memory(GiB)": 138.1,
      "step": 83220,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.78208766,
      "epoch": 1.9418189241604573,
      "grad_norm": 4.625,
      "learning_rate": 2.2145753912845014e-08,
      "loss": 0.78833818,
      "memory(GiB)": 138.1,
      "step": 83230,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.77869391,
      "epoch": 1.9420522317327462,
      "grad_norm": 7.71875,
      "learning_rate": 2.1968505746815527e-08,
      "loss": 0.79904451,
      "memory(GiB)": 138.1,
      "step": 83240,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.78262606,
      "epoch": 1.9422855393050351,
      "grad_norm": 5.78125,
      "learning_rate": 2.179196819333651e-08,
      "loss": 0.77370358,
      "memory(GiB)": 138.1,
      "step": 83250,
      "train_speed(iter/s)": 0.200716
    },
    {
      "acc": 0.77597246,
      "epoch": 1.942518846877324,
      "grad_norm": 7.6875,
      "learning_rate": 2.1616141277607804e-08,
      "loss": 0.80707121,
      "memory(GiB)": 138.1,
      "step": 83260,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.79199247,
      "epoch": 1.9427521544496127,
      "grad_norm": 5.6875,
      "learning_rate": 2.1441025024728225e-08,
      "loss": 0.74981995,
      "memory(GiB)": 138.1,
      "step": 83270,
      "train_speed(iter/s)": 0.200741
    },
    {
      "acc": 0.77484131,
      "epoch": 1.9429854620219018,
      "grad_norm": 5.78125,
      "learning_rate": 2.1266619459696102e-08,
      "loss": 0.81854687,
      "memory(GiB)": 138.1,
      "step": 83280,
      "train_speed(iter/s)": 0.200753
    },
    {
      "acc": 0.78900671,
      "epoch": 1.9432187695941905,
      "grad_norm": 5.96875,
      "learning_rate": 2.1092924607408195e-08,
      "loss": 0.76061816,
      "memory(GiB)": 138.1,
      "step": 83290,
      "train_speed(iter/s)": 0.200766
    },
    {
      "acc": 0.7663866,
      "epoch": 1.9434520771664796,
      "grad_norm": 5.0,
      "learning_rate": 2.091994049265855e-08,
      "loss": 0.85271311,
      "memory(GiB)": 138.1,
      "step": 83300,
      "train_speed(iter/s)": 0.200778
    },
    {
      "acc": 0.79749975,
      "epoch": 1.9436853847387683,
      "grad_norm": 6.40625,
      "learning_rate": 2.074766714014076e-08,
      "loss": 0.72964478,
      "memory(GiB)": 138.1,
      "step": 83310,
      "train_speed(iter/s)": 0.200791
    },
    {
      "acc": 0.77364941,
      "epoch": 1.9439186923110574,
      "grad_norm": 5.90625,
      "learning_rate": 2.057610457444792e-08,
      "loss": 0.82341738,
      "memory(GiB)": 138.1,
      "step": 83320,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.78957624,
      "epoch": 1.944151999883346,
      "grad_norm": 5.875,
      "learning_rate": 2.0405252820068776e-08,
      "loss": 0.76410146,
      "memory(GiB)": 138.1,
      "step": 83330,
      "train_speed(iter/s)": 0.200817
    },
    {
      "acc": 0.77884073,
      "epoch": 1.9443853074556352,
      "grad_norm": 3.71875,
      "learning_rate": 2.0235111901393266e-08,
      "loss": 0.77160797,
      "memory(GiB)": 138.1,
      "step": 83340,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.76247535,
      "epoch": 1.9446186150279239,
      "grad_norm": 5.84375,
      "learning_rate": 2.0065681842709185e-08,
      "loss": 0.86031971,
      "memory(GiB)": 138.1,
      "step": 83350,
      "train_speed(iter/s)": 0.20084
    },
    {
      "acc": 0.77901812,
      "epoch": 1.944851922600213,
      "grad_norm": 4.34375,
      "learning_rate": 1.989696266820218e-08,
      "loss": 0.79860239,
      "memory(GiB)": 138.1,
      "step": 83360,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.77776785,
      "epoch": 1.9450852301725017,
      "grad_norm": 5.84375,
      "learning_rate": 1.972895440195688e-08,
      "loss": 0.7847723,
      "memory(GiB)": 138.1,
      "step": 83370,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.79139214,
      "epoch": 1.9453185377447908,
      "grad_norm": 4.375,
      "learning_rate": 1.9561657067956874e-08,
      "loss": 0.75544043,
      "memory(GiB)": 138.1,
      "step": 83380,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.76193299,
      "epoch": 1.9455518453170795,
      "grad_norm": 5.375,
      "learning_rate": 1.9395070690083617e-08,
      "loss": 0.84404163,
      "memory(GiB)": 138.1,
      "step": 83390,
      "train_speed(iter/s)": 0.200889
    },
    {
      "acc": 0.78581266,
      "epoch": 1.9457851528893686,
      "grad_norm": 5.4375,
      "learning_rate": 1.9229195292116976e-08,
      "loss": 0.76360941,
      "memory(GiB)": 138.1,
      "step": 83400,
      "train_speed(iter/s)": 0.200901
    },
    {
      "acc": 0.79550982,
      "epoch": 1.9460184604616573,
      "grad_norm": 6.375,
      "learning_rate": 1.9064030897735232e-08,
      "loss": 0.73615227,
      "memory(GiB)": 138.1,
      "step": 83410,
      "train_speed(iter/s)": 0.200913
    },
    {
      "acc": 0.78520379,
      "epoch": 1.9462517680339464,
      "grad_norm": 4.21875,
      "learning_rate": 1.8899577530516744e-08,
      "loss": 0.77724113,
      "memory(GiB)": 138.1,
      "step": 83420,
      "train_speed(iter/s)": 0.200926
    },
    {
      "acc": 0.78606596,
      "epoch": 1.946485075606235,
      "grad_norm": 5.90625,
      "learning_rate": 1.8735835213936627e-08,
      "loss": 0.76647534,
      "memory(GiB)": 138.1,
      "step": 83430,
      "train_speed(iter/s)": 0.200938
    },
    {
      "acc": 0.78347244,
      "epoch": 1.9467183831785242,
      "grad_norm": 5.59375,
      "learning_rate": 1.8572803971368404e-08,
      "loss": 0.80083179,
      "memory(GiB)": 138.1,
      "step": 83440,
      "train_speed(iter/s)": 0.200951
    },
    {
      "acc": 0.77931328,
      "epoch": 1.9469516907508129,
      "grad_norm": 8.6875,
      "learning_rate": 1.841048382608568e-08,
      "loss": 0.79017577,
      "memory(GiB)": 138.1,
      "step": 83450,
      "train_speed(iter/s)": 0.200963
    },
    {
      "acc": 0.77811432,
      "epoch": 1.947184998323102,
      "grad_norm": 5.8125,
      "learning_rate": 1.8248874801259363e-08,
      "loss": 0.79184465,
      "memory(GiB)": 138.1,
      "step": 83460,
      "train_speed(iter/s)": 0.200976
    },
    {
      "acc": 0.79206524,
      "epoch": 1.9474183058953907,
      "grad_norm": 5.625,
      "learning_rate": 1.8087976919958783e-08,
      "loss": 0.73837852,
      "memory(GiB)": 138.1,
      "step": 83470,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.78631954,
      "epoch": 1.9476516134676796,
      "grad_norm": 7.09375,
      "learning_rate": 1.792779020515223e-08,
      "loss": 0.76524954,
      "memory(GiB)": 138.1,
      "step": 83480,
      "train_speed(iter/s)": 0.201002
    },
    {
      "acc": 0.78290014,
      "epoch": 1.9478849210399685,
      "grad_norm": 6.65625,
      "learning_rate": 1.776831467970641e-08,
      "loss": 0.77942362,
      "memory(GiB)": 138.1,
      "step": 83490,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.79167795,
      "epoch": 1.9481182286122574,
      "grad_norm": 5.4375,
      "learning_rate": 1.7609550366385897e-08,
      "loss": 0.74010186,
      "memory(GiB)": 138.1,
      "step": 83500,
      "train_speed(iter/s)": 0.201027
    },
    {
      "epoch": 1.9481182286122574,
      "eval_acc": 0.7446927801054742,
      "eval_loss": 0.8044261932373047,
      "eval_runtime": 1271.6845,
      "eval_samples_per_second": 28.302,
      "eval_steps_per_second": 14.151,
      "step": 83500
    },
    {
      "acc": 0.77565384,
      "epoch": 1.9483515361845463,
      "grad_norm": 5.25,
      "learning_rate": 1.7451497287855334e-08,
      "loss": 0.77604108,
      "memory(GiB)": 138.1,
      "step": 83510,
      "train_speed(iter/s)": 0.200414
    },
    {
      "acc": 0.7824501,
      "epoch": 1.9485848437568352,
      "grad_norm": 3.953125,
      "learning_rate": 1.7294155466675567e-08,
      "loss": 0.80296421,
      "memory(GiB)": 138.1,
      "step": 83520,
      "train_speed(iter/s)": 0.200427
    },
    {
      "acc": 0.7742733,
      "epoch": 1.948818151329124,
      "grad_norm": 4.5625,
      "learning_rate": 1.7137524925307515e-08,
      "loss": 0.83972244,
      "memory(GiB)": 138.1,
      "step": 83530,
      "train_speed(iter/s)": 0.20044
    },
    {
      "acc": 0.76697764,
      "epoch": 1.949051458901413,
      "grad_norm": 6.34375,
      "learning_rate": 1.698160568611107e-08,
      "loss": 0.82968264,
      "memory(GiB)": 138.1,
      "step": 83540,
      "train_speed(iter/s)": 0.200453
    },
    {
      "acc": 0.77781119,
      "epoch": 1.9492847664737019,
      "grad_norm": 5.5625,
      "learning_rate": 1.6826397771342317e-08,
      "loss": 0.80852528,
      "memory(GiB)": 138.1,
      "step": 83550,
      "train_speed(iter/s)": 0.200466
    },
    {
      "acc": 0.77753439,
      "epoch": 1.9495180740459908,
      "grad_norm": 7.4375,
      "learning_rate": 1.6671901203157425e-08,
      "loss": 0.83437614,
      "memory(GiB)": 138.1,
      "step": 83560,
      "train_speed(iter/s)": 0.200479
    },
    {
      "acc": 0.79008942,
      "epoch": 1.9497513816182797,
      "grad_norm": 5.875,
      "learning_rate": 1.6518116003611527e-08,
      "loss": 0.74177456,
      "memory(GiB)": 138.1,
      "step": 83570,
      "train_speed(iter/s)": 0.200492
    },
    {
      "acc": 0.80940037,
      "epoch": 1.9499846891905686,
      "grad_norm": 10.3125,
      "learning_rate": 1.636504219465651e-08,
      "loss": 0.66792626,
      "memory(GiB)": 138.1,
      "step": 83580,
      "train_speed(iter/s)": 0.200505
    },
    {
      "acc": 0.76242118,
      "epoch": 1.9502179967628575,
      "grad_norm": 5.65625,
      "learning_rate": 1.6212679798143782e-08,
      "loss": 0.85964575,
      "memory(GiB)": 138.1,
      "step": 83590,
      "train_speed(iter/s)": 0.200517
    },
    {
      "acc": 0.7804348,
      "epoch": 1.9504513043351464,
      "grad_norm": 5.96875,
      "learning_rate": 1.6061028835823723e-08,
      "loss": 0.78724127,
      "memory(GiB)": 138.1,
      "step": 83600,
      "train_speed(iter/s)": 0.20053
    },
    {
      "acc": 0.76076965,
      "epoch": 1.9506846119074353,
      "grad_norm": 6.0625,
      "learning_rate": 1.591008932934346e-08,
      "loss": 0.86802006,
      "memory(GiB)": 138.1,
      "step": 83610,
      "train_speed(iter/s)": 0.200542
    },
    {
      "acc": 0.77423449,
      "epoch": 1.9509179194797241,
      "grad_norm": 5.6875,
      "learning_rate": 1.57598613002502e-08,
      "loss": 0.81039219,
      "memory(GiB)": 138.1,
      "step": 83620,
      "train_speed(iter/s)": 0.200555
    },
    {
      "acc": 0.76987734,
      "epoch": 1.951151227052013,
      "grad_norm": 6.78125,
      "learning_rate": 1.5610344769989017e-08,
      "loss": 0.83137999,
      "memory(GiB)": 138.1,
      "step": 83630,
      "train_speed(iter/s)": 0.200567
    },
    {
      "acc": 0.76419888,
      "epoch": 1.951384534624302,
      "grad_norm": 5.84375,
      "learning_rate": 1.5461539759902832e-08,
      "loss": 0.84437933,
      "memory(GiB)": 138.1,
      "step": 83640,
      "train_speed(iter/s)": 0.200579
    },
    {
      "acc": 0.79972954,
      "epoch": 1.9516178421965908,
      "grad_norm": 4.40625,
      "learning_rate": 1.5313446291234104e-08,
      "loss": 0.706955,
      "memory(GiB)": 138.1,
      "step": 83650,
      "train_speed(iter/s)": 0.200591
    },
    {
      "acc": 0.79801159,
      "epoch": 1.9518511497688795,
      "grad_norm": 4.5,
      "learning_rate": 1.516606438512258e-08,
      "loss": 0.73285971,
      "memory(GiB)": 138.1,
      "step": 83660,
      "train_speed(iter/s)": 0.200604
    },
    {
      "acc": 0.77286844,
      "epoch": 1.9520844573411686,
      "grad_norm": 6.5,
      "learning_rate": 1.501939406260755e-08,
      "loss": 0.82769775,
      "memory(GiB)": 138.1,
      "step": 83670,
      "train_speed(iter/s)": 0.200617
    },
    {
      "acc": 0.78321648,
      "epoch": 1.9523177649134573,
      "grad_norm": 5.28125,
      "learning_rate": 1.4873435344625597e-08,
      "loss": 0.78056812,
      "memory(GiB)": 138.1,
      "step": 83680,
      "train_speed(iter/s)": 0.200629
    },
    {
      "acc": 0.77889185,
      "epoch": 1.9525510724857464,
      "grad_norm": 4.09375,
      "learning_rate": 1.4728188252012832e-08,
      "loss": 0.79473181,
      "memory(GiB)": 138.1,
      "step": 83690,
      "train_speed(iter/s)": 0.200642
    },
    {
      "acc": 0.78766642,
      "epoch": 1.9527843800580351,
      "grad_norm": 4.9375,
      "learning_rate": 1.4583652805503223e-08,
      "loss": 0.76746063,
      "memory(GiB)": 138.1,
      "step": 83700,
      "train_speed(iter/s)": 0.200654
    },
    {
      "acc": 0.79879084,
      "epoch": 1.9530176876303242,
      "grad_norm": 7.625,
      "learning_rate": 1.4439829025728047e-08,
      "loss": 0.72144575,
      "memory(GiB)": 138.1,
      "step": 83710,
      "train_speed(iter/s)": 0.200667
    },
    {
      "acc": 0.78694019,
      "epoch": 1.953250995202613,
      "grad_norm": 8.3125,
      "learning_rate": 1.4296716933219768e-08,
      "loss": 0.7669951,
      "memory(GiB)": 138.1,
      "step": 83720,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.77646313,
      "epoch": 1.953484302774902,
      "grad_norm": 5.09375,
      "learning_rate": 1.4154316548406488e-08,
      "loss": 0.8079896,
      "memory(GiB)": 138.1,
      "step": 83730,
      "train_speed(iter/s)": 0.200692
    },
    {
      "acc": 0.77782555,
      "epoch": 1.9537176103471907,
      "grad_norm": 5.625,
      "learning_rate": 1.4012627891615838e-08,
      "loss": 0.79529219,
      "memory(GiB)": 138.1,
      "step": 83740,
      "train_speed(iter/s)": 0.200704
    },
    {
      "acc": 0.77345643,
      "epoch": 1.9539509179194798,
      "grad_norm": 7.25,
      "learning_rate": 1.3871650983074414e-08,
      "loss": 0.81681423,
      "memory(GiB)": 138.1,
      "step": 83750,
      "train_speed(iter/s)": 0.200717
    },
    {
      "acc": 0.77701006,
      "epoch": 1.9541842254917685,
      "grad_norm": 5.625,
      "learning_rate": 1.3731385842906675e-08,
      "loss": 0.79945378,
      "memory(GiB)": 138.1,
      "step": 83760,
      "train_speed(iter/s)": 0.200729
    },
    {
      "acc": 0.78949852,
      "epoch": 1.9544175330640576,
      "grad_norm": 6.1875,
      "learning_rate": 1.3591832491134382e-08,
      "loss": 0.75286026,
      "memory(GiB)": 138.1,
      "step": 83770,
      "train_speed(iter/s)": 0.200742
    },
    {
      "acc": 0.77981601,
      "epoch": 1.9546508406363463,
      "grad_norm": 4.1875,
      "learning_rate": 1.3452990947679933e-08,
      "loss": 0.78779631,
      "memory(GiB)": 138.1,
      "step": 83780,
      "train_speed(iter/s)": 0.200753
    },
    {
      "acc": 0.77697887,
      "epoch": 1.9548841482086354,
      "grad_norm": 6.28125,
      "learning_rate": 1.3314861232362475e-08,
      "loss": 0.8031744,
      "memory(GiB)": 138.1,
      "step": 83790,
      "train_speed(iter/s)": 0.200766
    },
    {
      "acc": 0.7862566,
      "epoch": 1.955117455780924,
      "grad_norm": 7.625,
      "learning_rate": 1.3177443364899567e-08,
      "loss": 0.76605148,
      "memory(GiB)": 138.1,
      "step": 83800,
      "train_speed(iter/s)": 0.200779
    },
    {
      "acc": 0.78782101,
      "epoch": 1.9553507633532132,
      "grad_norm": 5.6875,
      "learning_rate": 1.3040737364908295e-08,
      "loss": 0.75679483,
      "memory(GiB)": 138.1,
      "step": 83810,
      "train_speed(iter/s)": 0.200791
    },
    {
      "acc": 0.77958326,
      "epoch": 1.955584070925502,
      "grad_norm": 5.5,
      "learning_rate": 1.2904743251902496e-08,
      "loss": 0.79298873,
      "memory(GiB)": 138.1,
      "step": 83820,
      "train_speed(iter/s)": 0.200804
    },
    {
      "acc": 0.77609186,
      "epoch": 1.955817378497791,
      "grad_norm": 6.21875,
      "learning_rate": 1.2769461045296083e-08,
      "loss": 0.81230192,
      "memory(GiB)": 138.1,
      "step": 83830,
      "train_speed(iter/s)": 0.200816
    },
    {
      "acc": 0.78773966,
      "epoch": 1.9560506860700797,
      "grad_norm": 5.75,
      "learning_rate": 1.2634890764400832e-08,
      "loss": 0.78044167,
      "memory(GiB)": 138.1,
      "step": 83840,
      "train_speed(iter/s)": 0.200828
    },
    {
      "acc": 0.77997303,
      "epoch": 1.9562839936423686,
      "grad_norm": 6.375,
      "learning_rate": 1.2501032428425264e-08,
      "loss": 0.80062265,
      "memory(GiB)": 138.1,
      "step": 83850,
      "train_speed(iter/s)": 0.200841
    },
    {
      "acc": 0.77704172,
      "epoch": 1.9565173012146575,
      "grad_norm": 6.25,
      "learning_rate": 1.2367886056479095e-08,
      "loss": 0.79110088,
      "memory(GiB)": 138.1,
      "step": 83860,
      "train_speed(iter/s)": 0.200853
    },
    {
      "acc": 0.77524996,
      "epoch": 1.9567506087869464,
      "grad_norm": 4.84375,
      "learning_rate": 1.2235451667567678e-08,
      "loss": 0.80665646,
      "memory(GiB)": 138.1,
      "step": 83870,
      "train_speed(iter/s)": 0.200865
    },
    {
      "acc": 0.80717812,
      "epoch": 1.9569839163592353,
      "grad_norm": 4.59375,
      "learning_rate": 1.2103729280596998e-08,
      "loss": 0.69145899,
      "memory(GiB)": 138.1,
      "step": 83880,
      "train_speed(iter/s)": 0.200877
    },
    {
      "acc": 0.7651278,
      "epoch": 1.9572172239315242,
      "grad_norm": 5.46875,
      "learning_rate": 1.1972718914370351e-08,
      "loss": 0.84759274,
      "memory(GiB)": 138.1,
      "step": 83890,
      "train_speed(iter/s)": 0.20089
    },
    {
      "acc": 0.78047352,
      "epoch": 1.957450531503813,
      "grad_norm": 4.625,
      "learning_rate": 1.1842420587588333e-08,
      "loss": 0.80353117,
      "memory(GiB)": 138.1,
      "step": 83900,
      "train_speed(iter/s)": 0.200902
    },
    {
      "acc": 0.77811413,
      "epoch": 1.957683839076102,
      "grad_norm": 5.375,
      "learning_rate": 1.1712834318852173e-08,
      "loss": 0.79200544,
      "memory(GiB)": 138.1,
      "step": 83910,
      "train_speed(iter/s)": 0.200914
    },
    {
      "acc": 0.79376502,
      "epoch": 1.957917146648391,
      "grad_norm": 6.0,
      "learning_rate": 1.15839601266593e-08,
      "loss": 0.74129534,
      "memory(GiB)": 138.1,
      "step": 83920,
      "train_speed(iter/s)": 0.200927
    },
    {
      "acc": 0.78470793,
      "epoch": 1.9581504542206798,
      "grad_norm": 5.5,
      "learning_rate": 1.1455798029407772e-08,
      "loss": 0.76864552,
      "memory(GiB)": 138.1,
      "step": 83930,
      "train_speed(iter/s)": 0.200939
    },
    {
      "acc": 0.78365679,
      "epoch": 1.9583837617929687,
      "grad_norm": 3.96875,
      "learning_rate": 1.132834804539129e-08,
      "loss": 0.75253015,
      "memory(GiB)": 138.1,
      "step": 83940,
      "train_speed(iter/s)": 0.200952
    },
    {
      "acc": 0.78898249,
      "epoch": 1.9586170693652576,
      "grad_norm": 4.96875,
      "learning_rate": 1.120161019280419e-08,
      "loss": 0.74453573,
      "memory(GiB)": 138.1,
      "step": 83950,
      "train_speed(iter/s)": 0.200964
    },
    {
      "acc": 0.77433839,
      "epoch": 1.9588503769375465,
      "grad_norm": 6.5,
      "learning_rate": 1.1075584489737557e-08,
      "loss": 0.82852116,
      "memory(GiB)": 138.1,
      "step": 83960,
      "train_speed(iter/s)": 0.200977
    },
    {
      "acc": 0.77746682,
      "epoch": 1.9590836845098354,
      "grad_norm": 4.78125,
      "learning_rate": 1.0950270954181997e-08,
      "loss": 0.79795814,
      "memory(GiB)": 138.1,
      "step": 83970,
      "train_speed(iter/s)": 0.200989
    },
    {
      "acc": 0.77838507,
      "epoch": 1.9593169920821243,
      "grad_norm": 6.46875,
      "learning_rate": 1.0825669604026534e-08,
      "loss": 0.79831772,
      "memory(GiB)": 138.1,
      "step": 83980,
      "train_speed(iter/s)": 0.201001
    },
    {
      "acc": 0.78435926,
      "epoch": 1.9595502996544132,
      "grad_norm": 6.28125,
      "learning_rate": 1.0701780457056943e-08,
      "loss": 0.79196739,
      "memory(GiB)": 138.1,
      "step": 83990,
      "train_speed(iter/s)": 0.201014
    },
    {
      "acc": 0.78845625,
      "epoch": 1.959783607226702,
      "grad_norm": 4.9375,
      "learning_rate": 1.0578603530958519e-08,
      "loss": 0.76660895,
      "memory(GiB)": 138.1,
      "step": 84000,
      "train_speed(iter/s)": 0.201026
    },
    {
      "epoch": 1.959783607226702,
      "eval_acc": 0.7447569979426271,
      "eval_loss": 0.8044305443763733,
      "eval_runtime": 1271.2957,
      "eval_samples_per_second": 28.31,
      "eval_steps_per_second": 14.156,
      "step": 84000
    },
    {
      "acc": 0.77818713,
      "epoch": 1.960016914798991,
      "grad_norm": 6.59375,
      "learning_rate": 1.0456138843315534e-08,
      "loss": 0.80791035,
      "memory(GiB)": 138.1,
      "step": 84010,
      "train_speed(iter/s)": 0.200416
    },
    {
      "acc": 0.7748589,
      "epoch": 1.9602502223712799,
      "grad_norm": 6.375,
      "learning_rate": 1.0334386411609e-08,
      "loss": 0.7919076,
      "memory(GiB)": 138.1,
      "step": 84020,
      "train_speed(iter/s)": 0.200429
    },
    {
      "acc": 0.77099795,
      "epoch": 1.9604835299435688,
      "grad_norm": 5.40625,
      "learning_rate": 1.0213346253219459e-08,
      "loss": 0.7921875,
      "memory(GiB)": 138.1,
      "step": 84030,
      "train_speed(iter/s)": 0.200441
    },
    {
      "acc": 0.79554253,
      "epoch": 1.9607168375158577,
      "grad_norm": 5.4375,
      "learning_rate": 1.0093018385424757e-08,
      "loss": 0.73114367,
      "memory(GiB)": 138.1,
      "step": 84040,
      "train_speed(iter/s)": 0.200454
    },
    {
      "acc": 0.80636587,
      "epoch": 1.9609501450881464,
      "grad_norm": 6.21875,
      "learning_rate": 9.973402825402823e-09,
      "loss": 0.70025764,
      "memory(GiB)": 138.1,
      "step": 84050,
      "train_speed(iter/s)": 0.200466
    },
    {
      "acc": 0.79211783,
      "epoch": 1.9611834526604355,
      "grad_norm": 7.6875,
      "learning_rate": 9.854499590227217e-09,
      "loss": 0.72471952,
      "memory(GiB)": 138.1,
      "step": 84060,
      "train_speed(iter/s)": 0.200478
    },
    {
      "acc": 0.77074847,
      "epoch": 1.9614167602327242,
      "grad_norm": 4.5625,
      "learning_rate": 9.736308696872698e-09,
      "loss": 0.8087081,
      "memory(GiB)": 138.1,
      "step": 84070,
      "train_speed(iter/s)": 0.200491
    },
    {
      "acc": 0.77772079,
      "epoch": 1.9616500678050133,
      "grad_norm": 6.1875,
      "learning_rate": 9.618830162210213e-09,
      "loss": 0.80092773,
      "memory(GiB)": 138.1,
      "step": 84080,
      "train_speed(iter/s)": 0.200502
    },
    {
      "acc": 0.78463898,
      "epoch": 1.961883375377302,
      "grad_norm": 4.46875,
      "learning_rate": 9.502064003010236e-09,
      "loss": 0.78809996,
      "memory(GiB)": 138.1,
      "step": 84090,
      "train_speed(iter/s)": 0.200515
    },
    {
      "acc": 0.7911252,
      "epoch": 1.962116682949591,
      "grad_norm": 5.3125,
      "learning_rate": 9.386010235940546e-09,
      "loss": 0.76758842,
      "memory(GiB)": 138.1,
      "step": 84100,
      "train_speed(iter/s)": 0.200527
    },
    {
      "acc": 0.76237383,
      "epoch": 1.9623499905218798,
      "grad_norm": 7.28125,
      "learning_rate": 9.270668877568444e-09,
      "loss": 0.84567051,
      "memory(GiB)": 138.1,
      "step": 84110,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.77878971,
      "epoch": 1.9625832980941689,
      "grad_norm": 5.25,
      "learning_rate": 9.156039944358542e-09,
      "loss": 0.79879036,
      "memory(GiB)": 138.1,
      "step": 84120,
      "train_speed(iter/s)": 0.200552
    },
    {
      "acc": 0.76202507,
      "epoch": 1.9628166056664575,
      "grad_norm": 4.875,
      "learning_rate": 9.042123452673856e-09,
      "loss": 0.8723732,
      "memory(GiB)": 138.1,
      "step": 84130,
      "train_speed(iter/s)": 0.200564
    },
    {
      "acc": 0.77010603,
      "epoch": 1.9630499132387467,
      "grad_norm": 4.96875,
      "learning_rate": 8.928919418776382e-09,
      "loss": 0.83259068,
      "memory(GiB)": 138.1,
      "step": 84140,
      "train_speed(iter/s)": 0.200577
    },
    {
      "acc": 0.77653313,
      "epoch": 1.9632832208110353,
      "grad_norm": 5.65625,
      "learning_rate": 8.816427858825416e-09,
      "loss": 0.79427109,
      "memory(GiB)": 138.1,
      "step": 84150,
      "train_speed(iter/s)": 0.200589
    },
    {
      "acc": 0.79803519,
      "epoch": 1.9635165283833245,
      "grad_norm": 5.78125,
      "learning_rate": 8.704648788879777e-09,
      "loss": 0.73506384,
      "memory(GiB)": 138.1,
      "step": 84160,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.76183743,
      "epoch": 1.9637498359556131,
      "grad_norm": 4.375,
      "learning_rate": 8.593582224895036e-09,
      "loss": 0.85956383,
      "memory(GiB)": 138.1,
      "step": 84170,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.76150913,
      "epoch": 1.9639831435279023,
      "grad_norm": 8.1875,
      "learning_rate": 8.483228182726843e-09,
      "loss": 0.85972614,
      "memory(GiB)": 138.1,
      "step": 84180,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.79041133,
      "epoch": 1.964216451100191,
      "grad_norm": 4.71875,
      "learning_rate": 8.3735866781276e-09,
      "loss": 0.75700512,
      "memory(GiB)": 138.1,
      "step": 84190,
      "train_speed(iter/s)": 0.200637
    },
    {
      "acc": 0.76137877,
      "epoch": 1.96444975867248,
      "grad_norm": 5.78125,
      "learning_rate": 8.264657726749226e-09,
      "loss": 0.84933109,
      "memory(GiB)": 138.1,
      "step": 84200,
      "train_speed(iter/s)": 0.200649
    },
    {
      "acc": 0.782689,
      "epoch": 1.9646830662447687,
      "grad_norm": 4.1875,
      "learning_rate": 8.156441344140398e-09,
      "loss": 0.7834971,
      "memory(GiB)": 138.1,
      "step": 84210,
      "train_speed(iter/s)": 0.200661
    },
    {
      "acc": 0.78736172,
      "epoch": 1.9649163738170579,
      "grad_norm": 7.0625,
      "learning_rate": 8.048937545749313e-09,
      "loss": 0.74641953,
      "memory(GiB)": 138.1,
      "step": 84220,
      "train_speed(iter/s)": 0.200674
    },
    {
      "acc": 0.77960968,
      "epoch": 1.9651496813893465,
      "grad_norm": 6.65625,
      "learning_rate": 7.942146346922586e-09,
      "loss": 0.81022434,
      "memory(GiB)": 138.1,
      "step": 84230,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.75209432,
      "epoch": 1.9653829889616354,
      "grad_norm": 4.125,
      "learning_rate": 7.83606776290413e-09,
      "loss": 0.89543324,
      "memory(GiB)": 138.1,
      "step": 84240,
      "train_speed(iter/s)": 0.200699
    },
    {
      "acc": 0.79684596,
      "epoch": 1.9656162965339243,
      "grad_norm": 4.875,
      "learning_rate": 7.730701808836837e-09,
      "loss": 0.72995892,
      "memory(GiB)": 138.1,
      "step": 84250,
      "train_speed(iter/s)": 0.200712
    },
    {
      "acc": 0.78651543,
      "epoch": 1.9658496041062132,
      "grad_norm": 5.9375,
      "learning_rate": 7.626048499761452e-09,
      "loss": 0.74779282,
      "memory(GiB)": 138.1,
      "step": 84260,
      "train_speed(iter/s)": 0.200723
    },
    {
      "acc": 0.7833818,
      "epoch": 1.9660829116785021,
      "grad_norm": 6.28125,
      "learning_rate": 7.522107850617689e-09,
      "loss": 0.78160305,
      "memory(GiB)": 138.1,
      "step": 84270,
      "train_speed(iter/s)": 0.200736
    },
    {
      "acc": 0.80099115,
      "epoch": 1.966316219250791,
      "grad_norm": 3.59375,
      "learning_rate": 7.418879876242014e-09,
      "loss": 0.70863304,
      "memory(GiB)": 138.1,
      "step": 84280,
      "train_speed(iter/s)": 0.200749
    },
    {
      "acc": 0.76215916,
      "epoch": 1.96654952682308,
      "grad_norm": 6.40625,
      "learning_rate": 7.316364591371527e-09,
      "loss": 0.85770187,
      "memory(GiB)": 138.1,
      "step": 84290,
      "train_speed(iter/s)": 0.20076
    },
    {
      "acc": 0.79879122,
      "epoch": 1.9667828343953688,
      "grad_norm": 4.9375,
      "learning_rate": 7.214562010639525e-09,
      "loss": 0.73097854,
      "memory(GiB)": 138.1,
      "step": 84300,
      "train_speed(iter/s)": 0.200772
    },
    {
      "acc": 0.77161856,
      "epoch": 1.9670161419676577,
      "grad_norm": 6.625,
      "learning_rate": 7.113472148578271e-09,
      "loss": 0.85240879,
      "memory(GiB)": 138.1,
      "step": 84310,
      "train_speed(iter/s)": 0.200784
    },
    {
      "acc": 0.78391027,
      "epoch": 1.9672494495399466,
      "grad_norm": 4.625,
      "learning_rate": 7.013095019618443e-09,
      "loss": 0.76578526,
      "memory(GiB)": 138.1,
      "step": 84320,
      "train_speed(iter/s)": 0.200796
    },
    {
      "acc": 0.76423545,
      "epoch": 1.9674827571122355,
      "grad_norm": 6.28125,
      "learning_rate": 6.9134306380885805e-09,
      "loss": 0.87942724,
      "memory(GiB)": 138.1,
      "step": 84330,
      "train_speed(iter/s)": 0.200808
    },
    {
      "acc": 0.76918364,
      "epoch": 1.9677160646845244,
      "grad_norm": 8.1875,
      "learning_rate": 6.814479018216192e-09,
      "loss": 0.83055038,
      "memory(GiB)": 138.1,
      "step": 84340,
      "train_speed(iter/s)": 0.200821
    },
    {
      "acc": 0.76374469,
      "epoch": 1.9679493722568133,
      "grad_norm": 4.65625,
      "learning_rate": 6.7162401741266425e-09,
      "loss": 0.86087818,
      "memory(GiB)": 138.1,
      "step": 84350,
      "train_speed(iter/s)": 0.200833
    },
    {
      "acc": 0.76832781,
      "epoch": 1.9681826798291022,
      "grad_norm": 6.25,
      "learning_rate": 6.6187141198431615e-09,
      "loss": 0.83810291,
      "memory(GiB)": 138.1,
      "step": 84360,
      "train_speed(iter/s)": 0.200845
    },
    {
      "acc": 0.76127386,
      "epoch": 1.9684159874013911,
      "grad_norm": 7.15625,
      "learning_rate": 6.52190086928739e-09,
      "loss": 0.86279221,
      "memory(GiB)": 138.1,
      "step": 84370,
      "train_speed(iter/s)": 0.200857
    },
    {
      "acc": 0.78920145,
      "epoch": 1.96864929497368,
      "grad_norm": 5.59375,
      "learning_rate": 6.425800436279383e-09,
      "loss": 0.74230909,
      "memory(GiB)": 138.1,
      "step": 84380,
      "train_speed(iter/s)": 0.20087
    },
    {
      "acc": 0.77091441,
      "epoch": 1.968882602545969,
      "grad_norm": 5.125,
      "learning_rate": 6.330412834538169e-09,
      "loss": 0.83819075,
      "memory(GiB)": 138.1,
      "step": 84390,
      "train_speed(iter/s)": 0.200882
    },
    {
      "acc": 0.77305069,
      "epoch": 1.9691159101182578,
      "grad_norm": 5.1875,
      "learning_rate": 6.235738077680076e-09,
      "loss": 0.84126186,
      "memory(GiB)": 138.1,
      "step": 84400,
      "train_speed(iter/s)": 0.200894
    },
    {
      "acc": 0.7800487,
      "epoch": 1.9693492176905467,
      "grad_norm": 4.78125,
      "learning_rate": 6.141776179219294e-09,
      "loss": 0.7732388,
      "memory(GiB)": 138.1,
      "step": 84410,
      "train_speed(iter/s)": 0.200906
    },
    {
      "acc": 0.79444857,
      "epoch": 1.9695825252628356,
      "grad_norm": 10.0625,
      "learning_rate": 6.048527152569539e-09,
      "loss": 0.73840342,
      "memory(GiB)": 138.1,
      "step": 84420,
      "train_speed(iter/s)": 0.200918
    },
    {
      "acc": 0.78184214,
      "epoch": 1.9698158328351245,
      "grad_norm": 5.1875,
      "learning_rate": 5.955991011041273e-09,
      "loss": 0.80885363,
      "memory(GiB)": 138.1,
      "step": 84430,
      "train_speed(iter/s)": 0.200931
    },
    {
      "acc": 0.76946917,
      "epoch": 1.9700491404074132,
      "grad_norm": 11.5,
      "learning_rate": 5.864167767845041e-09,
      "loss": 0.82882366,
      "memory(GiB)": 138.1,
      "step": 84440,
      "train_speed(iter/s)": 0.200943
    },
    {
      "acc": 0.79174566,
      "epoch": 1.9702824479797023,
      "grad_norm": 8.375,
      "learning_rate": 5.773057436087581e-09,
      "loss": 0.72687259,
      "memory(GiB)": 138.1,
      "step": 84450,
      "train_speed(iter/s)": 0.200956
    },
    {
      "acc": 0.77029505,
      "epoch": 1.970515755551991,
      "grad_norm": 5.375,
      "learning_rate": 5.6826600287757105e-09,
      "loss": 0.83553743,
      "memory(GiB)": 138.1,
      "step": 84460,
      "train_speed(iter/s)": 0.200968
    },
    {
      "acc": 0.81181126,
      "epoch": 1.97074906312428,
      "grad_norm": 6.03125,
      "learning_rate": 5.592975558813551e-09,
      "loss": 0.68757429,
      "memory(GiB)": 138.1,
      "step": 84470,
      "train_speed(iter/s)": 0.20098
    },
    {
      "acc": 0.7760201,
      "epoch": 1.9709823706965688,
      "grad_norm": 4.9375,
      "learning_rate": 5.504004039002531e-09,
      "loss": 0.78468761,
      "memory(GiB)": 138.1,
      "step": 84480,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.78242598,
      "epoch": 1.971215678268858,
      "grad_norm": 6.6875,
      "learning_rate": 5.41574548204471e-09,
      "loss": 0.77616792,
      "memory(GiB)": 138.1,
      "step": 84490,
      "train_speed(iter/s)": 0.201005
    },
    {
      "acc": 0.77654591,
      "epoch": 1.9714489858411466,
      "grad_norm": 8.1875,
      "learning_rate": 5.32819990053779e-09,
      "loss": 0.82183952,
      "memory(GiB)": 138.1,
      "step": 84500,
      "train_speed(iter/s)": 0.201017
    },
    {
      "epoch": 1.9714489858411466,
      "eval_acc": 0.7447488305867797,
      "eval_loss": 0.8044179677963257,
      "eval_runtime": 1271.4413,
      "eval_samples_per_second": 28.307,
      "eval_steps_per_second": 14.154,
      "step": 84500
    },
    {
      "acc": 0.77603683,
      "epoch": 1.9716822934134357,
      "grad_norm": 4.75,
      "learning_rate": 5.24136730697955e-09,
      "loss": 0.79193287,
      "memory(GiB)": 138.1,
      "step": 84510,
      "train_speed(iter/s)": 0.200412
    },
    {
      "acc": 0.7718574,
      "epoch": 1.9719156009857244,
      "grad_norm": 5.78125,
      "learning_rate": 5.155247713765077e-09,
      "loss": 0.82882061,
      "memory(GiB)": 138.1,
      "step": 84520,
      "train_speed(iter/s)": 0.200425
    },
    {
      "acc": 0.76282473,
      "epoch": 1.9721489085580135,
      "grad_norm": 4.0625,
      "learning_rate": 5.069841133187869e-09,
      "loss": 0.83978472,
      "memory(GiB)": 138.1,
      "step": 84530,
      "train_speed(iter/s)": 0.200437
    },
    {
      "acc": 0.79579945,
      "epoch": 1.9723822161303022,
      "grad_norm": 4.125,
      "learning_rate": 4.985147577439842e-09,
      "loss": 0.73812609,
      "memory(GiB)": 138.1,
      "step": 84540,
      "train_speed(iter/s)": 0.200449
    },
    {
      "acc": 0.7669486,
      "epoch": 1.9726155237025913,
      "grad_norm": 6.875,
      "learning_rate": 4.901167058610767e-09,
      "loss": 0.84001894,
      "memory(GiB)": 138.1,
      "step": 84550,
      "train_speed(iter/s)": 0.200461
    },
    {
      "acc": 0.78317232,
      "epoch": 1.97284883127488,
      "grad_norm": 11.375,
      "learning_rate": 4.8178995886893895e-09,
      "loss": 0.76386995,
      "memory(GiB)": 138.1,
      "step": 84560,
      "train_speed(iter/s)": 0.200474
    },
    {
      "acc": 0.78392029,
      "epoch": 1.973082138847169,
      "grad_norm": 4.90625,
      "learning_rate": 4.735345179561757e-09,
      "loss": 0.77154541,
      "memory(GiB)": 138.1,
      "step": 84570,
      "train_speed(iter/s)": 0.200485
    },
    {
      "acc": 0.77048054,
      "epoch": 1.9733154464194578,
      "grad_norm": 5.96875,
      "learning_rate": 4.65350384301233e-09,
      "loss": 0.82347736,
      "memory(GiB)": 138.1,
      "step": 84580,
      "train_speed(iter/s)": 0.200498
    },
    {
      "acc": 0.7574182,
      "epoch": 1.973548753991747,
      "grad_norm": 4.5625,
      "learning_rate": 4.572375590723988e-09,
      "loss": 0.88755779,
      "memory(GiB)": 138.1,
      "step": 84590,
      "train_speed(iter/s)": 0.20051
    },
    {
      "acc": 0.786063,
      "epoch": 1.9737820615640356,
      "grad_norm": 5.375,
      "learning_rate": 4.49196043427802e-09,
      "loss": 0.76739092,
      "memory(GiB)": 138.1,
      "step": 84600,
      "train_speed(iter/s)": 0.200524
    },
    {
      "acc": 0.75360541,
      "epoch": 1.9740153691363247,
      "grad_norm": 6.4375,
      "learning_rate": 4.4122583851535785e-09,
      "loss": 0.87514687,
      "memory(GiB)": 138.1,
      "step": 84610,
      "train_speed(iter/s)": 0.200535
    },
    {
      "acc": 0.78751106,
      "epoch": 1.9742486767086134,
      "grad_norm": 5.6875,
      "learning_rate": 4.3332694547276736e-09,
      "loss": 0.75879245,
      "memory(GiB)": 138.1,
      "step": 84620,
      "train_speed(iter/s)": 0.200548
    },
    {
      "acc": 0.77941394,
      "epoch": 1.9744819842809023,
      "grad_norm": 5.25,
      "learning_rate": 4.254993654276285e-09,
      "loss": 0.7969101,
      "memory(GiB)": 138.1,
      "step": 84630,
      "train_speed(iter/s)": 0.20056
    },
    {
      "acc": 0.78160868,
      "epoch": 1.9747152918531912,
      "grad_norm": 5.25,
      "learning_rate": 4.177430994973808e-09,
      "loss": 0.76313763,
      "memory(GiB)": 138.1,
      "step": 84640,
      "train_speed(iter/s)": 0.200572
    },
    {
      "acc": 0.78432083,
      "epoch": 1.97494859942548,
      "grad_norm": 7.34375,
      "learning_rate": 4.1005814878913865e-09,
      "loss": 0.78618231,
      "memory(GiB)": 138.1,
      "step": 84650,
      "train_speed(iter/s)": 0.200584
    },
    {
      "acc": 0.77199898,
      "epoch": 1.975181906997769,
      "grad_norm": 4.75,
      "learning_rate": 4.024445143999689e-09,
      "loss": 0.81763344,
      "memory(GiB)": 138.1,
      "step": 84660,
      "train_speed(iter/s)": 0.200596
    },
    {
      "acc": 0.77098942,
      "epoch": 1.9754152145700579,
      "grad_norm": 4.75,
      "learning_rate": 3.9490219741672445e-09,
      "loss": 0.82059212,
      "memory(GiB)": 138.1,
      "step": 84670,
      "train_speed(iter/s)": 0.200608
    },
    {
      "acc": 0.78812075,
      "epoch": 1.9756485221423468,
      "grad_norm": 6.84375,
      "learning_rate": 3.87431198916044e-09,
      "loss": 0.75257215,
      "memory(GiB)": 138.1,
      "step": 84680,
      "train_speed(iter/s)": 0.20062
    },
    {
      "acc": 0.78079576,
      "epoch": 1.9758818297146357,
      "grad_norm": 4.8125,
      "learning_rate": 3.800315199644078e-09,
      "loss": 0.78688898,
      "memory(GiB)": 138.1,
      "step": 84690,
      "train_speed(iter/s)": 0.200632
    },
    {
      "acc": 0.78553157,
      "epoch": 1.9761151372869246,
      "grad_norm": 4.34375,
      "learning_rate": 3.727031616181376e-09,
      "loss": 0.76703348,
      "memory(GiB)": 138.1,
      "step": 84700,
      "train_speed(iter/s)": 0.200644
    },
    {
      "acc": 0.78337941,
      "epoch": 1.9763484448592135,
      "grad_norm": 6.03125,
      "learning_rate": 3.6544612492334097e-09,
      "loss": 0.76780548,
      "memory(GiB)": 138.1,
      "step": 84710,
      "train_speed(iter/s)": 0.200655
    },
    {
      "acc": 0.77466536,
      "epoch": 1.9765817524315024,
      "grad_norm": 4.71875,
      "learning_rate": 3.582604109159671e-09,
      "loss": 0.80691957,
      "memory(GiB)": 138.1,
      "step": 84720,
      "train_speed(iter/s)": 0.200668
    },
    {
      "acc": 0.77310543,
      "epoch": 1.9768150600037913,
      "grad_norm": 4.375,
      "learning_rate": 3.5114602062180646e-09,
      "loss": 0.8129674,
      "memory(GiB)": 138.1,
      "step": 84730,
      "train_speed(iter/s)": 0.200679
    },
    {
      "acc": 0.77795973,
      "epoch": 1.9770483675760802,
      "grad_norm": 5.40625,
      "learning_rate": 3.4410295505638013e-09,
      "loss": 0.82256088,
      "memory(GiB)": 138.1,
      "step": 84740,
      "train_speed(iter/s)": 0.200691
    },
    {
      "acc": 0.78295822,
      "epoch": 1.977281675148369,
      "grad_norm": 5.90625,
      "learning_rate": 3.3713121522510607e-09,
      "loss": 0.79164433,
      "memory(GiB)": 138.1,
      "step": 84750,
      "train_speed(iter/s)": 0.200703
    },
    {
      "acc": 0.76906929,
      "epoch": 1.977514982720658,
      "grad_norm": 5.28125,
      "learning_rate": 3.3023080212318814e-09,
      "loss": 0.8484437,
      "memory(GiB)": 138.1,
      "step": 84760,
      "train_speed(iter/s)": 0.200715
    },
    {
      "acc": 0.80698633,
      "epoch": 1.9777482902929469,
      "grad_norm": 4.59375,
      "learning_rate": 3.234017167356718e-09,
      "loss": 0.68034239,
      "memory(GiB)": 138.1,
      "step": 84770,
      "train_speed(iter/s)": 0.200727
    },
    {
      "acc": 0.77349758,
      "epoch": 1.9779815978652358,
      "grad_norm": 7.46875,
      "learning_rate": 3.1664396003738827e-09,
      "loss": 0.82069979,
      "memory(GiB)": 138.1,
      "step": 84780,
      "train_speed(iter/s)": 0.200739
    },
    {
      "acc": 0.77547059,
      "epoch": 1.9782149054375247,
      "grad_norm": 5.28125,
      "learning_rate": 3.0995753299306598e-09,
      "loss": 0.80736742,
      "memory(GiB)": 138.1,
      "step": 84790,
      "train_speed(iter/s)": 0.20075
    },
    {
      "acc": 0.7942091,
      "epoch": 1.9784482130098135,
      "grad_norm": 5.15625,
      "learning_rate": 3.0334243655710805e-09,
      "loss": 0.7481492,
      "memory(GiB)": 138.1,
      "step": 84800,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.78032541,
      "epoch": 1.9786815205821022,
      "grad_norm": 5.5625,
      "learning_rate": 2.9679867167387024e-09,
      "loss": 0.79240317,
      "memory(GiB)": 138.1,
      "step": 84810,
      "train_speed(iter/s)": 0.200775
    },
    {
      "acc": 0.78130064,
      "epoch": 1.9789148281543913,
      "grad_norm": 5.125,
      "learning_rate": 2.9032623927743864e-09,
      "loss": 0.79106226,
      "memory(GiB)": 138.1,
      "step": 84820,
      "train_speed(iter/s)": 0.200787
    },
    {
      "acc": 0.77278428,
      "epoch": 1.97914813572668,
      "grad_norm": 5.53125,
      "learning_rate": 2.839251402917964e-09,
      "loss": 0.81999369,
      "memory(GiB)": 138.1,
      "step": 84830,
      "train_speed(iter/s)": 0.200799
    },
    {
      "acc": 0.78221178,
      "epoch": 1.9793814432989691,
      "grad_norm": 4.21875,
      "learning_rate": 2.7759537563065706e-09,
      "loss": 0.76479435,
      "memory(GiB)": 138.1,
      "step": 84840,
      "train_speed(iter/s)": 0.200812
    },
    {
      "acc": 0.7911087,
      "epoch": 1.9796147508712578,
      "grad_norm": 6.25,
      "learning_rate": 2.7133694619763117e-09,
      "loss": 0.75844345,
      "memory(GiB)": 138.1,
      "step": 84850,
      "train_speed(iter/s)": 0.200824
    },
    {
      "acc": 0.77396116,
      "epoch": 1.979848058443547,
      "grad_norm": 4.25,
      "learning_rate": 2.6514985288605964e-09,
      "loss": 0.83643398,
      "memory(GiB)": 138.1,
      "step": 84860,
      "train_speed(iter/s)": 0.200836
    },
    {
      "acc": 0.79459782,
      "epoch": 1.9800813660158356,
      "grad_norm": 10.875,
      "learning_rate": 2.590340965791804e-09,
      "loss": 0.74517059,
      "memory(GiB)": 138.1,
      "step": 84870,
      "train_speed(iter/s)": 0.200848
    },
    {
      "acc": 0.77023711,
      "epoch": 1.9803146735881247,
      "grad_norm": 5.40625,
      "learning_rate": 2.529896781500174e-09,
      "loss": 0.81559296,
      "memory(GiB)": 138.1,
      "step": 84880,
      "train_speed(iter/s)": 0.200861
    },
    {
      "acc": 0.75082717,
      "epoch": 1.9805479811604134,
      "grad_norm": 4.90625,
      "learning_rate": 2.4701659846138036e-09,
      "loss": 0.89876308,
      "memory(GiB)": 138.1,
      "step": 84890,
      "train_speed(iter/s)": 0.200873
    },
    {
      "acc": 0.7946104,
      "epoch": 1.9807812887327025,
      "grad_norm": 4.3125,
      "learning_rate": 2.4111485836592065e-09,
      "loss": 0.73043966,
      "memory(GiB)": 138.1,
      "step": 84900,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.76876898,
      "epoch": 1.9810145963049912,
      "grad_norm": 6.75,
      "learning_rate": 2.3528445870618643e-09,
      "loss": 0.83043594,
      "memory(GiB)": 138.1,
      "step": 84910,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.7821094,
      "epoch": 1.9812479038772803,
      "grad_norm": 5.71875,
      "learning_rate": 2.2952540031440096e-09,
      "loss": 0.79190273,
      "memory(GiB)": 138.1,
      "step": 84920,
      "train_speed(iter/s)": 0.200909
    },
    {
      "acc": 0.78448,
      "epoch": 1.981481211449569,
      "grad_norm": 5.21875,
      "learning_rate": 2.2383768401268435e-09,
      "loss": 0.78527408,
      "memory(GiB)": 138.1,
      "step": 84930,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.779772,
      "epoch": 1.9817145190218581,
      "grad_norm": 4.9375,
      "learning_rate": 2.182213106129427e-09,
      "loss": 0.78437147,
      "memory(GiB)": 138.1,
      "step": 84940,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.76009808,
      "epoch": 1.9819478265941468,
      "grad_norm": 4.625,
      "learning_rate": 2.126762809169236e-09,
      "loss": 0.84811478,
      "memory(GiB)": 138.1,
      "step": 84950,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.7567162,
      "epoch": 1.982181134166436,
      "grad_norm": 5.875,
      "learning_rate": 2.072025957161605e-09,
      "loss": 0.86951723,
      "memory(GiB)": 138.1,
      "step": 84960,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.78688498,
      "epoch": 1.9824144417387246,
      "grad_norm": 5.9375,
      "learning_rate": 2.0180025579202844e-09,
      "loss": 0.77867031,
      "memory(GiB)": 138.1,
      "step": 84970,
      "train_speed(iter/s)": 0.200969
    },
    {
      "acc": 0.79099741,
      "epoch": 1.9826477493110137,
      "grad_norm": 3.921875,
      "learning_rate": 1.964692619157438e-09,
      "loss": 0.73870716,
      "memory(GiB)": 138.1,
      "step": 84980,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.77988491,
      "epoch": 1.9828810568833024,
      "grad_norm": 3.890625,
      "learning_rate": 1.912096148482534e-09,
      "loss": 0.80207052,
      "memory(GiB)": 138.1,
      "step": 84990,
      "train_speed(iter/s)": 0.200994
    },
    {
      "acc": 0.78414559,
      "epoch": 1.9831143644555915,
      "grad_norm": 5.3125,
      "learning_rate": 1.8602131534045665e-09,
      "loss": 0.79055219,
      "memory(GiB)": 138.1,
      "step": 85000,
      "train_speed(iter/s)": 0.201006
    },
    {
      "epoch": 1.9831143644555915,
      "eval_acc": 0.7447177626057132,
      "eval_loss": 0.8044196963310242,
      "eval_runtime": 1271.3001,
      "eval_samples_per_second": 28.31,
      "eval_steps_per_second": 14.156,
      "step": 85000
    },
    {
      "acc": 0.78565392,
      "epoch": 1.9833476720278802,
      "grad_norm": 6.03125,
      "learning_rate": 1.8090436413287226e-09,
      "loss": 0.76658969,
      "memory(GiB)": 138.1,
      "step": 85010,
      "train_speed(iter/s)": 0.200404
    },
    {
      "acc": 0.79666553,
      "epoch": 1.983580979600169,
      "grad_norm": 5.34375,
      "learning_rate": 1.758587619559715e-09,
      "loss": 0.73290777,
      "memory(GiB)": 138.1,
      "step": 85020,
      "train_speed(iter/s)": 0.200416
    },
    {
      "acc": 0.78800335,
      "epoch": 1.983814287172458,
      "grad_norm": 5.46875,
      "learning_rate": 1.7088450953006708e-09,
      "loss": 0.73815637,
      "memory(GiB)": 138.1,
      "step": 85030,
      "train_speed(iter/s)": 0.200428
    },
    {
      "acc": 0.78558655,
      "epoch": 1.984047594744747,
      "grad_norm": 5.09375,
      "learning_rate": 1.659816075652021e-09,
      "loss": 0.76726046,
      "memory(GiB)": 138.1,
      "step": 85040,
      "train_speed(iter/s)": 0.20044
    },
    {
      "acc": 0.79275599,
      "epoch": 1.9842809023170358,
      "grad_norm": 5.59375,
      "learning_rate": 1.6115005676120565e-09,
      "loss": 0.72810202,
      "memory(GiB)": 138.1,
      "step": 85050,
      "train_speed(iter/s)": 0.200452
    },
    {
      "acc": 0.78331842,
      "epoch": 1.9845142098893247,
      "grad_norm": 5.09375,
      "learning_rate": 1.5638985780791483e-09,
      "loss": 0.78223214,
      "memory(GiB)": 138.1,
      "step": 85060,
      "train_speed(iter/s)": 0.200463
    },
    {
      "acc": 0.77881184,
      "epoch": 1.9847475174616136,
      "grad_norm": 5.625,
      "learning_rate": 1.517010113847306e-09,
      "loss": 0.79921713,
      "memory(GiB)": 138.1,
      "step": 85070,
      "train_speed(iter/s)": 0.200476
    },
    {
      "acc": 0.77289662,
      "epoch": 1.9849808250339025,
      "grad_norm": 5.25,
      "learning_rate": 1.4708351816100641e-09,
      "loss": 0.80485191,
      "memory(GiB)": 138.1,
      "step": 85080,
      "train_speed(iter/s)": 0.200488
    },
    {
      "acc": 0.77916284,
      "epoch": 1.9852141326061914,
      "grad_norm": 4.5625,
      "learning_rate": 1.425373787958817e-09,
      "loss": 0.77773075,
      "memory(GiB)": 138.1,
      "step": 85090,
      "train_speed(iter/s)": 0.200501
    },
    {
      "acc": 0.79094009,
      "epoch": 1.9854474401784803,
      "grad_norm": 4.6875,
      "learning_rate": 1.3806259393839282e-09,
      "loss": 0.7560596,
      "memory(GiB)": 138.1,
      "step": 85100,
      "train_speed(iter/s)": 0.200513
    },
    {
      "acc": 0.77891779,
      "epoch": 1.9856807477507692,
      "grad_norm": 4.53125,
      "learning_rate": 1.336591642271956e-09,
      "loss": 0.79700637,
      "memory(GiB)": 138.1,
      "step": 85110,
      "train_speed(iter/s)": 0.200526
    },
    {
      "acc": 0.76863756,
      "epoch": 1.985914055323058,
      "grad_norm": 6.71875,
      "learning_rate": 1.2932709029100933e-09,
      "loss": 0.8220705,
      "memory(GiB)": 138.1,
      "step": 85120,
      "train_speed(iter/s)": 0.200539
    },
    {
      "acc": 0.77171407,
      "epoch": 1.986147362895347,
      "grad_norm": 5.625,
      "learning_rate": 1.2506637274811717e-09,
      "loss": 0.81749229,
      "memory(GiB)": 138.1,
      "step": 85130,
      "train_speed(iter/s)": 0.200551
    },
    {
      "acc": 0.78697386,
      "epoch": 1.986380670467636,
      "grad_norm": 7.96875,
      "learning_rate": 1.2087701220681036e-09,
      "loss": 0.76459389,
      "memory(GiB)": 138.1,
      "step": 85140,
      "train_speed(iter/s)": 0.200565
    },
    {
      "acc": 0.78453169,
      "epoch": 1.9866139780399248,
      "grad_norm": 7.8125,
      "learning_rate": 1.1675900926511053e-09,
      "loss": 0.77584524,
      "memory(GiB)": 138.1,
      "step": 85150,
      "train_speed(iter/s)": 0.200577
    },
    {
      "acc": 0.79571877,
      "epoch": 1.9868472856122137,
      "grad_norm": 4.5625,
      "learning_rate": 1.1271236451082524e-09,
      "loss": 0.72640424,
      "memory(GiB)": 138.1,
      "step": 85160,
      "train_speed(iter/s)": 0.200589
    },
    {
      "acc": 0.77864351,
      "epoch": 1.9870805931845026,
      "grad_norm": 7.5,
      "learning_rate": 1.0873707852160354e-09,
      "loss": 0.80453529,
      "memory(GiB)": 138.1,
      "step": 85170,
      "train_speed(iter/s)": 0.200601
    },
    {
      "acc": 0.76175656,
      "epoch": 1.9873139007567915,
      "grad_norm": 5.25,
      "learning_rate": 1.0483315186499143e-09,
      "loss": 0.84362011,
      "memory(GiB)": 138.1,
      "step": 85180,
      "train_speed(iter/s)": 0.200613
    },
    {
      "acc": 0.77368731,
      "epoch": 1.9875472083290804,
      "grad_norm": 3.71875,
      "learning_rate": 1.0100058509815435e-09,
      "loss": 0.80538511,
      "memory(GiB)": 138.1,
      "step": 85190,
      "train_speed(iter/s)": 0.200625
    },
    {
      "acc": 0.76618557,
      "epoch": 1.987780515901369,
      "grad_norm": 7.21875,
      "learning_rate": 9.723937876832124e-10,
      "loss": 0.82723331,
      "memory(GiB)": 138.1,
      "step": 85200,
      "train_speed(iter/s)": 0.200638
    },
    {
      "acc": 0.7858901,
      "epoch": 1.9880138234736582,
      "grad_norm": 5.0,
      "learning_rate": 9.354953341234042e-10,
      "loss": 0.75257692,
      "memory(GiB)": 138.1,
      "step": 85210,
      "train_speed(iter/s)": 0.20065
    },
    {
      "acc": 0.78722382,
      "epoch": 1.9882471310459469,
      "grad_norm": 5.625,
      "learning_rate": 8.99310495569572e-10,
      "loss": 0.74956484,
      "memory(GiB)": 138.1,
      "step": 85220,
      "train_speed(iter/s)": 0.200662
    },
    {
      "acc": 0.80738592,
      "epoch": 1.988480438618236,
      "grad_norm": 4.125,
      "learning_rate": 8.638392771864734e-10,
      "loss": 0.67543035,
      "memory(GiB)": 138.1,
      "step": 85230,
      "train_speed(iter/s)": 0.200675
    },
    {
      "acc": 0.79583402,
      "epoch": 1.9887137461905247,
      "grad_norm": 5.4375,
      "learning_rate": 8.290816840383908e-10,
      "loss": 0.73210516,
      "memory(GiB)": 138.1,
      "step": 85240,
      "train_speed(iter/s)": 0.200687
    },
    {
      "acc": 0.76078939,
      "epoch": 1.9889470537628138,
      "grad_norm": 6.4375,
      "learning_rate": 7.950377210863558e-10,
      "loss": 0.84969578,
      "memory(GiB)": 138.1,
      "step": 85250,
      "train_speed(iter/s)": 0.2007
    },
    {
      "acc": 0.77894077,
      "epoch": 1.9891803613351025,
      "grad_norm": 7.0625,
      "learning_rate": 7.617073931909247e-10,
      "loss": 0.79533377,
      "memory(GiB)": 138.1,
      "step": 85260,
      "train_speed(iter/s)": 0.200713
    },
    {
      "acc": 0.78382912,
      "epoch": 1.9894136689073916,
      "grad_norm": 5.21875,
      "learning_rate": 7.29090705108848e-10,
      "loss": 0.77288427,
      "memory(GiB)": 138.1,
      "step": 85270,
      "train_speed(iter/s)": 0.200725
    },
    {
      "acc": 0.75326347,
      "epoch": 1.9896469764796803,
      "grad_norm": 4.8125,
      "learning_rate": 6.971876614969564e-10,
      "loss": 0.89215813,
      "memory(GiB)": 138.1,
      "step": 85280,
      "train_speed(iter/s)": 0.200737
    },
    {
      "acc": 0.79840131,
      "epoch": 1.9898802840519694,
      "grad_norm": 6.84375,
      "learning_rate": 6.659982669093845e-10,
      "loss": 0.70803699,
      "memory(GiB)": 138.1,
      "step": 85290,
      "train_speed(iter/s)": 0.20075
    },
    {
      "acc": 0.78066187,
      "epoch": 1.990113591624258,
      "grad_norm": 4.90625,
      "learning_rate": 6.355225257981268e-10,
      "loss": 0.79505005,
      "memory(GiB)": 138.1,
      "step": 85300,
      "train_speed(iter/s)": 0.200762
    },
    {
      "acc": 0.78745785,
      "epoch": 1.9903468991965472,
      "grad_norm": 4.65625,
      "learning_rate": 6.057604425135921e-10,
      "loss": 0.74916191,
      "memory(GiB)": 138.1,
      "step": 85310,
      "train_speed(iter/s)": 0.200774
    },
    {
      "acc": 0.75778542,
      "epoch": 1.9905802067688358,
      "grad_norm": 5.96875,
      "learning_rate": 5.76712021304604e-10,
      "loss": 0.86030426,
      "memory(GiB)": 138.1,
      "step": 85320,
      "train_speed(iter/s)": 0.200785
    },
    {
      "acc": 0.76393342,
      "epoch": 1.990813514341125,
      "grad_norm": 6.6875,
      "learning_rate": 5.483772663178455e-10,
      "loss": 0.85037289,
      "memory(GiB)": 138.1,
      "step": 85330,
      "train_speed(iter/s)": 0.200798
    },
    {
      "acc": 0.77822199,
      "epoch": 1.9910468219134136,
      "grad_norm": 6.40625,
      "learning_rate": 5.20756181597859e-10,
      "loss": 0.80846586,
      "memory(GiB)": 138.1,
      "step": 85340,
      "train_speed(iter/s)": 0.20081
    },
    {
      "acc": 0.78142519,
      "epoch": 1.9912801294857028,
      "grad_norm": 6.0625,
      "learning_rate": 4.938487710870465e-10,
      "loss": 0.78094144,
      "memory(GiB)": 138.1,
      "step": 85350,
      "train_speed(iter/s)": 0.200822
    },
    {
      "acc": 0.77061892,
      "epoch": 1.9915134370579914,
      "grad_norm": 4.9375,
      "learning_rate": 4.676550386273349e-10,
      "loss": 0.82950497,
      "memory(GiB)": 138.1,
      "step": 85360,
      "train_speed(iter/s)": 0.200834
    },
    {
      "acc": 0.78219571,
      "epoch": 1.9917467446302806,
      "grad_norm": 6.28125,
      "learning_rate": 4.421749879574e-10,
      "loss": 0.76107216,
      "memory(GiB)": 138.1,
      "step": 85370,
      "train_speed(iter/s)": 0.200847
    },
    {
      "acc": 0.76850634,
      "epoch": 1.9919800522025692,
      "grad_norm": 5.09375,
      "learning_rate": 4.174086227148877e-10,
      "loss": 0.84992552,
      "memory(GiB)": 138.1,
      "step": 85380,
      "train_speed(iter/s)": 0.200859
    },
    {
      "acc": 0.76265931,
      "epoch": 1.9922133597748581,
      "grad_norm": 6.25,
      "learning_rate": 3.9335594643419294e-10,
      "loss": 0.84543304,
      "memory(GiB)": 138.1,
      "step": 85390,
      "train_speed(iter/s)": 0.200872
    },
    {
      "acc": 0.78111825,
      "epoch": 1.992446667347147,
      "grad_norm": 5.15625,
      "learning_rate": 3.700169625503458e-10,
      "loss": 0.79352822,
      "memory(GiB)": 138.1,
      "step": 85400,
      "train_speed(iter/s)": 0.200884
    },
    {
      "acc": 0.78446589,
      "epoch": 1.992679974919436,
      "grad_norm": 5.15625,
      "learning_rate": 3.4739167439346024e-10,
      "loss": 0.77862563,
      "memory(GiB)": 138.1,
      "step": 85410,
      "train_speed(iter/s)": 0.200896
    },
    {
      "acc": 0.78342505,
      "epoch": 1.9929132824917248,
      "grad_norm": 4.03125,
      "learning_rate": 3.254800851948403e-10,
      "loss": 0.78520279,
      "memory(GiB)": 138.1,
      "step": 85420,
      "train_speed(iter/s)": 0.200908
    },
    {
      "acc": 0.77086182,
      "epoch": 1.9931465900640137,
      "grad_norm": 5.59375,
      "learning_rate": 3.042821980808741e-10,
      "loss": 0.82681541,
      "memory(GiB)": 138.1,
      "step": 85430,
      "train_speed(iter/s)": 0.20092
    },
    {
      "acc": 0.76659908,
      "epoch": 1.9933798976363026,
      "grad_norm": 6.09375,
      "learning_rate": 2.8379801607858473e-10,
      "loss": 0.83367481,
      "memory(GiB)": 138.1,
      "step": 85440,
      "train_speed(iter/s)": 0.200933
    },
    {
      "acc": 0.77113953,
      "epoch": 1.9936132052085915,
      "grad_norm": 6.15625,
      "learning_rate": 2.640275421111893e-10,
      "loss": 0.82532921,
      "memory(GiB)": 138.1,
      "step": 85450,
      "train_speed(iter/s)": 0.200945
    },
    {
      "acc": 0.80559187,
      "epoch": 1.9938465127808804,
      "grad_norm": 4.71875,
      "learning_rate": 2.449707790019851e-10,
      "loss": 0.6829802,
      "memory(GiB)": 138.1,
      "step": 85460,
      "train_speed(iter/s)": 0.200957
    },
    {
      "acc": 0.77535009,
      "epoch": 1.9940798203531693,
      "grad_norm": 5.1875,
      "learning_rate": 2.266277294704633e-10,
      "loss": 0.80879288,
      "memory(GiB)": 138.1,
      "step": 85470,
      "train_speed(iter/s)": 0.20097
    },
    {
      "acc": 0.78569078,
      "epoch": 1.9943131279254582,
      "grad_norm": 7.46875,
      "learning_rate": 2.0899839613508499e-10,
      "loss": 0.76950636,
      "memory(GiB)": 138.1,
      "step": 85480,
      "train_speed(iter/s)": 0.200981
    },
    {
      "acc": 0.78273611,
      "epoch": 1.9945464354977471,
      "grad_norm": 7.84375,
      "learning_rate": 1.9208278151328087e-10,
      "loss": 0.78465424,
      "memory(GiB)": 138.1,
      "step": 85490,
      "train_speed(iter/s)": 0.200993
    },
    {
      "acc": 0.79737329,
      "epoch": 1.994779743070036,
      "grad_norm": 5.84375,
      "learning_rate": 1.7588088801923088e-10,
      "loss": 0.73678904,
      "memory(GiB)": 138.1,
      "step": 85500,
      "train_speed(iter/s)": 0.201006
    },
    {
      "epoch": 1.994779743070036,
      "eval_acc": 0.744678847557264,
      "eval_loss": 0.8044396638870239,
      "eval_runtime": 1270.037,
      "eval_samples_per_second": 28.339,
      "eval_steps_per_second": 14.17,
      "step": 85500
    }
  ],
  "logging_steps": 10,
  "max_steps": 85722,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 2.0183064448975503e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}