{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1624,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006157635467980296,
      "grad_norm": 30.824097187840486,
      "learning_rate": 1.226993865030675e-07,
      "loss": 2.8545,
      "step": 1
    },
    {
      "epoch": 0.003078817733990148,
      "grad_norm": 28.957796650466456,
      "learning_rate": 6.134969325153375e-07,
      "loss": 3.0839,
      "step": 5
    },
    {
      "epoch": 0.006157635467980296,
      "grad_norm": 22.518967134774034,
      "learning_rate": 1.226993865030675e-06,
      "loss": 2.4164,
      "step": 10
    },
    {
      "epoch": 0.009236453201970444,
      "grad_norm": 26.505709981925232,
      "learning_rate": 1.8404907975460124e-06,
      "loss": 2.0071,
      "step": 15
    },
    {
      "epoch": 0.012315270935960592,
      "grad_norm": 23.700055015543413,
      "learning_rate": 2.45398773006135e-06,
      "loss": 1.7303,
      "step": 20
    },
    {
      "epoch": 0.01539408866995074,
      "grad_norm": 11.507447830970856,
      "learning_rate": 3.0674846625766875e-06,
      "loss": 1.6109,
      "step": 25
    },
    {
      "epoch": 0.01847290640394089,
      "grad_norm": 15.955773713752077,
      "learning_rate": 3.680981595092025e-06,
      "loss": 1.4732,
      "step": 30
    },
    {
      "epoch": 0.021551724137931036,
      "grad_norm": 17.608595737910424,
      "learning_rate": 4.294478527607362e-06,
      "loss": 1.3481,
      "step": 35
    },
    {
      "epoch": 0.024630541871921183,
      "grad_norm": 15.205660456590286,
      "learning_rate": 4.9079754601227e-06,
      "loss": 1.4631,
      "step": 40
    },
    {
      "epoch": 0.02770935960591133,
      "grad_norm": 14.926475675775917,
      "learning_rate": 5.521472392638038e-06,
      "loss": 1.2939,
      "step": 45
    },
    {
      "epoch": 0.03078817733990148,
      "grad_norm": 11.878509548852609,
      "learning_rate": 6.134969325153375e-06,
      "loss": 1.1734,
      "step": 50
    },
    {
      "epoch": 0.033866995073891626,
      "grad_norm": 11.287096380726847,
      "learning_rate": 6.748466257668712e-06,
      "loss": 1.1496,
      "step": 55
    },
    {
      "epoch": 0.03694581280788178,
      "grad_norm": 11.043680477312911,
      "learning_rate": 7.36196319018405e-06,
      "loss": 1.307,
      "step": 60
    },
    {
      "epoch": 0.04002463054187192,
      "grad_norm": 19.7956889039234,
      "learning_rate": 7.975460122699386e-06,
      "loss": 1.2151,
      "step": 65
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 11.377885840797095,
      "learning_rate": 8.588957055214725e-06,
      "loss": 1.2581,
      "step": 70
    },
    {
      "epoch": 0.046182266009852216,
      "grad_norm": 11.664969491446978,
      "learning_rate": 9.202453987730062e-06,
      "loss": 1.2859,
      "step": 75
    },
    {
      "epoch": 0.04926108374384237,
      "grad_norm": 14.738756950790513,
      "learning_rate": 9.8159509202454e-06,
      "loss": 1.2863,
      "step": 80
    },
    {
      "epoch": 0.05233990147783251,
      "grad_norm": 28.65525103550205,
      "learning_rate": 1.0429447852760737e-05,
      "loss": 1.2419,
      "step": 85
    },
    {
      "epoch": 0.05541871921182266,
      "grad_norm": 14.288459529090195,
      "learning_rate": 1.1042944785276076e-05,
      "loss": 1.2857,
      "step": 90
    },
    {
      "epoch": 0.058497536945812806,
      "grad_norm": 25.32489993232227,
      "learning_rate": 1.1656441717791411e-05,
      "loss": 1.2733,
      "step": 95
    },
    {
      "epoch": 0.06157635467980296,
      "grad_norm": 18.317851808388852,
      "learning_rate": 1.226993865030675e-05,
      "loss": 1.2374,
      "step": 100
    },
    {
      "epoch": 0.06465517241379311,
      "grad_norm": 9.54576913856584,
      "learning_rate": 1.2883435582822085e-05,
      "loss": 1.2803,
      "step": 105
    },
    {
      "epoch": 0.06773399014778325,
      "grad_norm": 26.096411464584367,
      "learning_rate": 1.3496932515337424e-05,
      "loss": 1.2756,
      "step": 110
    },
    {
      "epoch": 0.0708128078817734,
      "grad_norm": 47.039712003079835,
      "learning_rate": 1.4110429447852763e-05,
      "loss": 1.3109,
      "step": 115
    },
    {
      "epoch": 0.07389162561576355,
      "grad_norm": 26.480731126203683,
      "learning_rate": 1.47239263803681e-05,
      "loss": 1.2405,
      "step": 120
    },
    {
      "epoch": 0.0769704433497537,
      "grad_norm": 22.417192506804327,
      "learning_rate": 1.5337423312883436e-05,
      "loss": 1.3085,
      "step": 125
    },
    {
      "epoch": 0.08004926108374384,
      "grad_norm": 229.896969186069,
      "learning_rate": 1.5950920245398772e-05,
      "loss": 1.2753,
      "step": 130
    },
    {
      "epoch": 0.08312807881773399,
      "grad_norm": 28.154903593032678,
      "learning_rate": 1.656441717791411e-05,
      "loss": 1.3574,
      "step": 135
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 47.5704177966797,
      "learning_rate": 1.717791411042945e-05,
      "loss": 1.3427,
      "step": 140
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 46.686158390206586,
      "learning_rate": 1.7791411042944788e-05,
      "loss": 1.3606,
      "step": 145
    },
    {
      "epoch": 0.09236453201970443,
      "grad_norm": 24.242903960555118,
      "learning_rate": 1.8404907975460123e-05,
      "loss": 1.3456,
      "step": 150
    },
    {
      "epoch": 0.09544334975369458,
      "grad_norm": 83.57525649609255,
      "learning_rate": 1.9018404907975462e-05,
      "loss": 1.2728,
      "step": 155
    },
    {
      "epoch": 0.09852216748768473,
      "grad_norm": 51.36060094160884,
      "learning_rate": 1.96319018404908e-05,
      "loss": 1.3103,
      "step": 160
    },
    {
      "epoch": 0.10160098522167488,
      "grad_norm": 23.34010834199715,
      "learning_rate": 1.999990752408443e-05,
      "loss": 1.3648,
      "step": 165
    },
    {
      "epoch": 0.10467980295566502,
      "grad_norm": 62.696412256468925,
      "learning_rate": 1.9998867189676517e-05,
      "loss": 1.3233,
      "step": 170
    },
    {
      "epoch": 0.10775862068965517,
      "grad_norm": 50.97473664105039,
      "learning_rate": 1.999667104662347e-05,
      "loss": 1.3172,
      "step": 175
    },
    {
      "epoch": 0.11083743842364532,
      "grad_norm": 29.4465582030845,
      "learning_rate": 1.9993319348786157e-05,
      "loss": 1.3562,
      "step": 180
    },
    {
      "epoch": 0.11391625615763547,
      "grad_norm": 82.46904606677337,
      "learning_rate": 1.9988812483600597e-05,
      "loss": 1.2674,
      "step": 185
    },
    {
      "epoch": 0.11699507389162561,
      "grad_norm": 75.96337504796978,
      "learning_rate": 1.9983150972033186e-05,
      "loss": 1.3079,
      "step": 190
    },
    {
      "epoch": 0.12007389162561577,
      "grad_norm": 32.18683626346483,
      "learning_rate": 1.9976335468520452e-05,
      "loss": 1.3401,
      "step": 195
    },
    {
      "epoch": 0.12315270935960591,
      "grad_norm": 60.965446926537616,
      "learning_rate": 1.9968366760893437e-05,
      "loss": 1.339,
      "step": 200
    },
    {
      "epoch": 0.12623152709359606,
      "grad_norm": 44.43427299110902,
      "learning_rate": 1.9959245770286602e-05,
      "loss": 1.314,
      "step": 205
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 29.816950647227433,
      "learning_rate": 1.994897355103136e-05,
      "loss": 1.2893,
      "step": 210
    },
    {
      "epoch": 0.13238916256157635,
      "grad_norm": 65.76774435974697,
      "learning_rate": 1.9937551290534208e-05,
      "loss": 1.2433,
      "step": 215
    },
    {
      "epoch": 0.1354679802955665,
      "grad_norm": 59.89081212072458,
      "learning_rate": 1.9924980309139455e-05,
      "loss": 1.2838,
      "step": 220
    },
    {
      "epoch": 0.13854679802955666,
      "grad_norm": 28.15079173742487,
      "learning_rate": 1.9911262059976614e-05,
      "loss": 1.3156,
      "step": 225
    },
    {
      "epoch": 0.1416256157635468,
      "grad_norm": 45.241325512315925,
      "learning_rate": 1.9896398128792413e-05,
      "loss": 1.3083,
      "step": 230
    },
    {
      "epoch": 0.14470443349753695,
      "grad_norm": 44.86580327646113,
      "learning_rate": 1.988039023376751e-05,
      "loss": 1.3469,
      "step": 235
    },
    {
      "epoch": 0.1477832512315271,
      "grad_norm": 21.115693175077368,
      "learning_rate": 1.9863240225317868e-05,
      "loss": 1.4001,
      "step": 240
    },
    {
      "epoch": 0.15086206896551724,
      "grad_norm": 75.7892769694395,
      "learning_rate": 1.984495008588086e-05,
      "loss": 1.2874,
      "step": 245
    },
    {
      "epoch": 0.1539408866995074,
      "grad_norm": 53.7846306002013,
      "learning_rate": 1.982552192968612e-05,
      "loss": 1.3804,
      "step": 250
    },
    {
      "epoch": 0.15701970443349753,
      "grad_norm": 37.234004867589846,
      "learning_rate": 1.9804958002511137e-05,
      "loss": 1.3289,
      "step": 255
    },
    {
      "epoch": 0.16009852216748768,
      "grad_norm": 62.77047280372136,
      "learning_rate": 1.9783260681421667e-05,
      "loss": 1.3782,
      "step": 260
    },
    {
      "epoch": 0.16317733990147784,
      "grad_norm": 12.191885254525328,
      "learning_rate": 1.9760432474496963e-05,
      "loss": 1.3038,
      "step": 265
    },
    {
      "epoch": 0.16625615763546797,
      "grad_norm": 16.393270767950437,
      "learning_rate": 1.973647602053984e-05,
      "loss": 1.3244,
      "step": 270
    },
    {
      "epoch": 0.16933497536945813,
      "grad_norm": 40.63770181450056,
      "learning_rate": 1.9711394088771658e-05,
      "loss": 1.4164,
      "step": 275
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 31.879401330035773,
      "learning_rate": 1.9685189578512206e-05,
      "loss": 1.4478,
      "step": 280
    },
    {
      "epoch": 0.17549261083743842,
      "grad_norm": 36.847483784788245,
      "learning_rate": 1.9657865518844578e-05,
      "loss": 1.4149,
      "step": 285
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 37.64073996560774,
      "learning_rate": 1.962942506826501e-05,
      "loss": 1.4085,
      "step": 290
    },
    {
      "epoch": 0.1816502463054187,
      "grad_norm": 20.26824397019875,
      "learning_rate": 1.9599871514317785e-05,
      "loss": 1.4283,
      "step": 295
    },
    {
      "epoch": 0.18472906403940886,
      "grad_norm": 34.37390352199536,
      "learning_rate": 1.9569208273215204e-05,
      "loss": 1.3624,
      "step": 300
    },
    {
      "epoch": 0.18780788177339902,
      "grad_norm": 21.987807875880346,
      "learning_rate": 1.953743888944271e-05,
      "loss": 1.3447,
      "step": 305
    },
    {
      "epoch": 0.19088669950738915,
      "grad_norm": 71.06224034941917,
      "learning_rate": 1.950456703534915e-05,
      "loss": 1.4409,
      "step": 310
    },
    {
      "epoch": 0.1939655172413793,
      "grad_norm": 26.76193289078792,
      "learning_rate": 1.9470596510722285e-05,
      "loss": 1.3622,
      "step": 315
    },
    {
      "epoch": 0.19704433497536947,
      "grad_norm": 40.072539824672745,
      "learning_rate": 1.9435531242349545e-05,
      "loss": 1.5042,
      "step": 320
    },
    {
      "epoch": 0.2001231527093596,
      "grad_norm": 27.51647995399636,
      "learning_rate": 1.9399375283564134e-05,
      "loss": 1.4148,
      "step": 325
    },
    {
      "epoch": 0.20320197044334976,
      "grad_norm": 55.02644275886537,
      "learning_rate": 1.9362132813776472e-05,
      "loss": 1.4064,
      "step": 330
    },
    {
      "epoch": 0.2062807881773399,
      "grad_norm": 59.041970278330794,
      "learning_rate": 1.9323808137991084e-05,
      "loss": 1.4128,
      "step": 335
    },
    {
      "epoch": 0.20935960591133004,
      "grad_norm": 14.92203181577856,
      "learning_rate": 1.9284405686308982e-05,
      "loss": 1.3831,
      "step": 340
    },
    {
      "epoch": 0.2124384236453202,
      "grad_norm": 37.0357475469201,
      "learning_rate": 1.924393001341555e-05,
      "loss": 1.412,
      "step": 345
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 19.900833922344106,
      "learning_rate": 1.9202385798054073e-05,
      "loss": 1.3703,
      "step": 350
    },
    {
      "epoch": 0.2185960591133005,
      "grad_norm": 28.8884233173782,
      "learning_rate": 1.9159777842484878e-05,
      "loss": 1.411,
      "step": 355
    },
    {
      "epoch": 0.22167487684729065,
      "grad_norm": 23.69174182203317,
      "learning_rate": 1.911611107193024e-05,
      "loss": 1.4163,
      "step": 360
    },
    {
      "epoch": 0.22475369458128078,
      "grad_norm": 18.40122277597206,
      "learning_rate": 1.9071390534005045e-05,
      "loss": 1.4395,
      "step": 365
    },
    {
      "epoch": 0.22783251231527094,
      "grad_norm": 52.548048906777154,
      "learning_rate": 1.9025621398133333e-05,
      "loss": 1.4706,
      "step": 370
    },
    {
      "epoch": 0.2309113300492611,
      "grad_norm": 50.694263768607094,
      "learning_rate": 1.8978808954950722e-05,
      "loss": 1.4759,
      "step": 375
    },
    {
      "epoch": 0.23399014778325122,
      "grad_norm": 29.15179740458261,
      "learning_rate": 1.8930958615692854e-05,
      "loss": 1.5095,
      "step": 380
    },
    {
      "epoch": 0.23706896551724138,
      "grad_norm": 22.170887953050535,
      "learning_rate": 1.8882075911569887e-05,
      "loss": 1.3458,
      "step": 385
    },
    {
      "epoch": 0.24014778325123154,
      "grad_norm": 26.55797739898958,
      "learning_rate": 1.8832166493127128e-05,
      "loss": 1.4092,
      "step": 390
    },
    {
      "epoch": 0.24322660098522167,
      "grad_norm": 23.773293163951188,
      "learning_rate": 1.8781236129591847e-05,
      "loss": 1.3733,
      "step": 395
    },
    {
      "epoch": 0.24630541871921183,
      "grad_norm": 38.277657106378115,
      "learning_rate": 1.8729290708206412e-05,
      "loss": 1.4268,
      "step": 400
    },
    {
      "epoch": 0.24938423645320196,
      "grad_norm": 18.30009002302841,
      "learning_rate": 1.867633623354773e-05,
      "loss": 1.4443,
      "step": 405
    },
    {
      "epoch": 0.2524630541871921,
      "grad_norm": 46.84980763430157,
      "learning_rate": 1.8622378826833186e-05,
      "loss": 1.4581,
      "step": 410
    },
    {
      "epoch": 0.2555418719211823,
      "grad_norm": 14.0324980387513,
      "learning_rate": 1.856742472521304e-05,
      "loss": 1.4517,
      "step": 415
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 15.706776811258118,
      "learning_rate": 1.8511480281049475e-05,
      "loss": 1.4459,
      "step": 420
    },
    {
      "epoch": 0.2616995073891626,
      "grad_norm": 13.632481739237138,
      "learning_rate": 1.8454551961182276e-05,
      "loss": 1.4012,
      "step": 425
    },
    {
      "epoch": 0.2647783251231527,
      "grad_norm": 45.128284427780166,
      "learning_rate": 1.8396646346181327e-05,
      "loss": 1.4216,
      "step": 430
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 18.61287164945975,
      "learning_rate": 1.8337770129585918e-05,
      "loss": 1.4912,
      "step": 435
    },
    {
      "epoch": 0.270935960591133,
      "grad_norm": 33.25836863651266,
      "learning_rate": 1.8277930117131025e-05,
      "loss": 1.4829,
      "step": 440
    },
    {
      "epoch": 0.27401477832512317,
      "grad_norm": 31.56004722863592,
      "learning_rate": 1.8217133225960597e-05,
      "loss": 1.4024,
      "step": 445
    },
    {
      "epoch": 0.2770935960591133,
      "grad_norm": 23.203442638131012,
      "learning_rate": 1.8155386483827995e-05,
      "loss": 1.4356,
      "step": 450
    },
    {
      "epoch": 0.2801724137931034,
      "grad_norm": 17.965500461906817,
      "learning_rate": 1.8092697028283598e-05,
      "loss": 1.4272,
      "step": 455
    },
    {
      "epoch": 0.2832512315270936,
      "grad_norm": 17.880734027839548,
      "learning_rate": 1.8029072105849767e-05,
      "loss": 1.4096,
      "step": 460
    },
    {
      "epoch": 0.28633004926108374,
      "grad_norm": 34.46926823265854,
      "learning_rate": 1.7964519071183188e-05,
      "loss": 1.4568,
      "step": 465
    },
    {
      "epoch": 0.2894088669950739,
      "grad_norm": 33.69072735250714,
      "learning_rate": 1.789904538622471e-05,
      "loss": 1.4619,
      "step": 470
    },
    {
      "epoch": 0.29248768472906406,
      "grad_norm": 40.54364819482425,
      "learning_rate": 1.7832658619336794e-05,
      "loss": 1.4764,
      "step": 475
    },
    {
      "epoch": 0.2955665024630542,
      "grad_norm": 16.132335947901723,
      "learning_rate": 1.7765366444428655e-05,
      "loss": 1.4647,
      "step": 480
    },
    {
      "epoch": 0.2986453201970443,
      "grad_norm": 14.936210207581604,
      "learning_rate": 1.7697176640069217e-05,
      "loss": 1.3972,
      "step": 485
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 36.71541735028421,
      "learning_rate": 1.762809708858793e-05,
      "loss": 1.4476,
      "step": 490
    },
    {
      "epoch": 0.30480295566502463,
      "grad_norm": 27.189619322043708,
      "learning_rate": 1.7558135775163645e-05,
      "loss": 1.4563,
      "step": 495
    },
    {
      "epoch": 0.3078817733990148,
      "grad_norm": 28.56013806002953,
      "learning_rate": 1.7487300786901568e-05,
      "loss": 1.373,
      "step": 500
    },
    {
      "epoch": 0.31096059113300495,
      "grad_norm": 39.40623446814172,
      "learning_rate": 1.7415600311898436e-05,
      "loss": 1.5221,
      "step": 505
    },
    {
      "epoch": 0.31403940886699505,
      "grad_norm": 92.53539153545226,
      "learning_rate": 1.734304263829602e-05,
      "loss": 1.5007,
      "step": 510
    },
    {
      "epoch": 0.3171182266009852,
      "grad_norm": 41.59882820578932,
      "learning_rate": 1.726963615332308e-05,
      "loss": 1.5236,
      "step": 515
    },
    {
      "epoch": 0.32019704433497537,
      "grad_norm": 45.960026820732985,
      "learning_rate": 1.7195389342325843e-05,
      "loss": 1.4876,
      "step": 520
    },
    {
      "epoch": 0.3232758620689655,
      "grad_norm": 22.864306475782886,
      "learning_rate": 1.7120310787787136e-05,
      "loss": 1.4331,
      "step": 525
    },
    {
      "epoch": 0.3263546798029557,
      "grad_norm": 12.642507080193463,
      "learning_rate": 1.7044409168334327e-05,
      "loss": 1.4261,
      "step": 530
    },
    {
      "epoch": 0.3294334975369458,
      "grad_norm": 20.36555056001907,
      "learning_rate": 1.696769325773611e-05,
      "loss": 1.3857,
      "step": 535
    },
    {
      "epoch": 0.33251231527093594,
      "grad_norm": 34.44303190420096,
      "learning_rate": 1.6890171923888323e-05,
      "loss": 1.4474,
      "step": 540
    },
    {
      "epoch": 0.3355911330049261,
      "grad_norm": 23.295684871020136,
      "learning_rate": 1.6811854127788857e-05,
      "loss": 1.5049,
      "step": 545
    },
    {
      "epoch": 0.33866995073891626,
      "grad_norm": 32.38754659254767,
      "learning_rate": 1.6732748922501832e-05,
      "loss": 1.5208,
      "step": 550
    },
    {
      "epoch": 0.3417487684729064,
      "grad_norm": 13.296570283727492,
      "learning_rate": 1.6652865452111115e-05,
      "loss": 1.3801,
      "step": 555
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 16.805850521256307,
      "learning_rate": 1.657221295066332e-05,
      "loss": 1.4474,
      "step": 560
    },
    {
      "epoch": 0.3479064039408867,
      "grad_norm": 27.42875404689886,
      "learning_rate": 1.6490800741100396e-05,
      "loss": 1.4329,
      "step": 565
    },
    {
      "epoch": 0.35098522167487683,
      "grad_norm": 28.230443570138807,
      "learning_rate": 1.6408638234181975e-05,
      "loss": 1.4483,
      "step": 570
    },
    {
      "epoch": 0.354064039408867,
      "grad_norm": 13.32519726088819,
      "learning_rate": 1.6325734927397514e-05,
      "loss": 1.3896,
      "step": 575
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 27.4675811711817,
      "learning_rate": 1.6242100403868472e-05,
      "loss": 1.4182,
      "step": 580
    },
    {
      "epoch": 0.3602216748768473,
      "grad_norm": 17.582212129726077,
      "learning_rate": 1.615774433124054e-05,
      "loss": 1.4333,
      "step": 585
    },
    {
      "epoch": 0.3633004926108374,
      "grad_norm": 32.55456987218266,
      "learning_rate": 1.6072676460566136e-05,
      "loss": 1.4552,
      "step": 590
    },
    {
      "epoch": 0.36637931034482757,
      "grad_norm": 14.52780288644456,
      "learning_rate": 1.5986906625177215e-05,
      "loss": 1.4584,
      "step": 595
    },
    {
      "epoch": 0.3694581280788177,
      "grad_norm": 38.095789663333,
      "learning_rate": 1.590044473954863e-05,
      "loss": 1.5388,
      "step": 600
    },
    {
      "epoch": 0.3725369458128079,
      "grad_norm": 13.377997762493838,
      "learning_rate": 1.5813300798152048e-05,
      "loss": 1.4395,
      "step": 605
    },
    {
      "epoch": 0.37561576354679804,
      "grad_norm": 23.274433382861258,
      "learning_rate": 1.5725484874300673e-05,
      "loss": 1.4929,
      "step": 610
    },
    {
      "epoch": 0.3786945812807882,
      "grad_norm": 28.742993330560854,
      "learning_rate": 1.5637007118984814e-05,
      "loss": 1.4463,
      "step": 615
    },
    {
      "epoch": 0.3817733990147783,
      "grad_norm": 19.366450983318067,
      "learning_rate": 1.5547877759698498e-05,
      "loss": 1.4487,
      "step": 620
    },
    {
      "epoch": 0.38485221674876846,
      "grad_norm": 20.41604184202308,
      "learning_rate": 1.5458107099257245e-05,
      "loss": 1.4614,
      "step": 625
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 13.577917003534072,
      "learning_rate": 1.5367705514607107e-05,
      "loss": 1.4096,
      "step": 630
    },
    {
      "epoch": 0.3910098522167488,
      "grad_norm": 14.857276479488586,
      "learning_rate": 1.527668345562516e-05,
      "loss": 1.4112,
      "step": 635
    },
    {
      "epoch": 0.39408866995073893,
      "grad_norm": 13.564931885711946,
      "learning_rate": 1.518505144391157e-05,
      "loss": 1.4325,
      "step": 640
    },
    {
      "epoch": 0.39716748768472904,
      "grad_norm": 19.557479990947304,
      "learning_rate": 1.5092820071573358e-05,
      "loss": 1.4823,
      "step": 645
    },
    {
      "epoch": 0.4002463054187192,
      "grad_norm": 26.631608175080384,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.453,
      "step": 650
    },
    {
      "epoch": 0.40332512315270935,
      "grad_norm": 30.709105530238208,
      "learning_rate": 1.490660195863106e-05,
      "loss": 1.4173,
      "step": 655
    },
    {
      "epoch": 0.4064039408866995,
      "grad_norm": 18.57808903837633,
      "learning_rate": 1.4812636743715912e-05,
      "loss": 1.4274,
      "step": 660
    },
    {
      "epoch": 0.40948275862068967,
      "grad_norm": 14.798272723989436,
      "learning_rate": 1.4718115217065766e-05,
      "loss": 1.4599,
      "step": 665
    },
    {
      "epoch": 0.4125615763546798,
      "grad_norm": 14.5356657578583,
      "learning_rate": 1.462304830479811e-05,
      "loss": 1.4103,
      "step": 670
    },
    {
      "epoch": 0.41564039408866993,
      "grad_norm": 36.27561493072455,
      "learning_rate": 1.4527446996073714e-05,
      "loss": 1.4217,
      "step": 675
    },
    {
      "epoch": 0.4187192118226601,
      "grad_norm": 32.24737869090583,
      "learning_rate": 1.4431322341826348e-05,
      "loss": 1.4406,
      "step": 680
    },
    {
      "epoch": 0.42179802955665024,
      "grad_norm": 16.536051884815873,
      "learning_rate": 1.433468545348537e-05,
      "loss": 1.4618,
      "step": 685
    },
    {
      "epoch": 0.4248768472906404,
      "grad_norm": 19.30327522321086,
      "learning_rate": 1.4237547501691298e-05,
      "loss": 1.4777,
      "step": 690
    },
    {
      "epoch": 0.42795566502463056,
      "grad_norm": 13.035000719900685,
      "learning_rate": 1.4139919715004558e-05,
      "loss": 1.4576,
      "step": 695
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 24.61662490249352,
      "learning_rate": 1.4041813378607534e-05,
      "loss": 1.5325,
      "step": 700
    },
    {
      "epoch": 0.4341133004926108,
      "grad_norm": 31.28228825862423,
      "learning_rate": 1.3943239833000068e-05,
      "loss": 1.4268,
      "step": 705
    },
    {
      "epoch": 0.437192118226601,
      "grad_norm": 13.659381666322691,
      "learning_rate": 1.3844210472688557e-05,
      "loss": 1.394,
      "step": 710
    },
    {
      "epoch": 0.44027093596059114,
      "grad_norm": 11.196127515614847,
      "learning_rate": 1.3744736744868832e-05,
      "loss": 1.4252,
      "step": 715
    },
    {
      "epoch": 0.4433497536945813,
      "grad_norm": 22.383901587181395,
      "learning_rate": 1.3644830148102915e-05,
      "loss": 1.4811,
      "step": 720
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 17.83615025815634,
      "learning_rate": 1.3544502230989868e-05,
      "loss": 1.4489,
      "step": 725
    },
    {
      "epoch": 0.44950738916256155,
      "grad_norm": 20.615332188902258,
      "learning_rate": 1.344376459083085e-05,
      "loss": 1.4591,
      "step": 730
    },
    {
      "epoch": 0.4525862068965517,
      "grad_norm": 35.365894961811776,
      "learning_rate": 1.3342628872288505e-05,
      "loss": 1.4503,
      "step": 735
    },
    {
      "epoch": 0.45566502463054187,
      "grad_norm": 12.761046766989047,
      "learning_rate": 1.3241106766040957e-05,
      "loss": 1.4035,
      "step": 740
    },
    {
      "epoch": 0.45874384236453203,
      "grad_norm": 30.14161857999299,
      "learning_rate": 1.3139210007430404e-05,
      "loss": 1.4156,
      "step": 745
    },
    {
      "epoch": 0.4618226600985222,
      "grad_norm": 16.885037039184166,
      "learning_rate": 1.3036950375106588e-05,
      "loss": 1.4376,
      "step": 750
    },
    {
      "epoch": 0.4649014778325123,
      "grad_norm": 22.489148333724426,
      "learning_rate": 1.2934339689665274e-05,
      "loss": 1.4184,
      "step": 755
    },
    {
      "epoch": 0.46798029556650245,
      "grad_norm": 10.494163494195895,
      "learning_rate": 1.283138981228184e-05,
      "loss": 1.4479,
      "step": 760
    },
    {
      "epoch": 0.4710591133004926,
      "grad_norm": 23.592356007614587,
      "learning_rate": 1.2728112643340201e-05,
      "loss": 1.4353,
      "step": 765
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 17.52804287104563,
      "learning_rate": 1.26245201210572e-05,
      "loss": 1.4191,
      "step": 770
    },
    {
      "epoch": 0.4772167487684729,
      "grad_norm": 18.009386689354745,
      "learning_rate": 1.2520624220102623e-05,
      "loss": 1.4509,
      "step": 775
    },
    {
      "epoch": 0.4802955665024631,
      "grad_norm": 14.48974652665118,
      "learning_rate": 1.2416436950215001e-05,
      "loss": 1.4469,
      "step": 780
    },
    {
      "epoch": 0.4833743842364532,
      "grad_norm": 10.958672298212598,
      "learning_rate": 1.2311970354813345e-05,
      "loss": 1.5327,
      "step": 785
    },
    {
      "epoch": 0.48645320197044334,
      "grad_norm": 7.7667441337425,
      "learning_rate": 1.220723650960502e-05,
      "loss": 1.42,
      "step": 790
    },
    {
      "epoch": 0.4895320197044335,
      "grad_norm": 9.14412567269568,
      "learning_rate": 1.2102247521189838e-05,
      "loss": 1.3924,
      "step": 795
    },
    {
      "epoch": 0.49261083743842365,
      "grad_norm": 15.758001094856118,
      "learning_rate": 1.199701552566064e-05,
      "loss": 1.3854,
      "step": 800
    },
    {
      "epoch": 0.4956896551724138,
      "grad_norm": 9.118909852310507,
      "learning_rate": 1.1891552687200414e-05,
      "loss": 1.4841,
      "step": 805
    },
    {
      "epoch": 0.4987684729064039,
      "grad_norm": 17.14996406716137,
      "learning_rate": 1.1785871196676196e-05,
      "loss": 1.5099,
      "step": 810
    },
    {
      "epoch": 0.5018472906403941,
      "grad_norm": 21.72121993201183,
      "learning_rate": 1.167998327022988e-05,
      "loss": 1.5371,
      "step": 815
    },
    {
      "epoch": 0.5049261083743842,
      "grad_norm": 15.280975846414641,
      "learning_rate": 1.1573901147866108e-05,
      "loss": 1.4375,
      "step": 820
    },
    {
      "epoch": 0.5080049261083743,
      "grad_norm": 16.226129759030037,
      "learning_rate": 1.1467637092037399e-05,
      "loss": 1.4342,
      "step": 825
    },
    {
      "epoch": 0.5110837438423645,
      "grad_norm": 8.038464949776358,
      "learning_rate": 1.1361203386226672e-05,
      "loss": 1.4387,
      "step": 830
    },
    {
      "epoch": 0.5141625615763546,
      "grad_norm": 19.365038241185463,
      "learning_rate": 1.1254612333527368e-05,
      "loss": 1.384,
      "step": 835
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 14.789904589004234,
      "learning_rate": 1.1147876255221274e-05,
      "loss": 1.4382,
      "step": 840
    },
    {
      "epoch": 0.520320197044335,
      "grad_norm": 21.23210928827207,
      "learning_rate": 1.1041007489354263e-05,
      "loss": 1.4166,
      "step": 845
    },
    {
      "epoch": 0.5233990147783252,
      "grad_norm": 10.650809601674084,
      "learning_rate": 1.093401838931009e-05,
      "loss": 1.4269,
      "step": 850
    },
    {
      "epoch": 0.5264778325123153,
      "grad_norm": 17.153586268388544,
      "learning_rate": 1.0826921322382407e-05,
      "loss": 1.4112,
      "step": 855
    },
    {
      "epoch": 0.5295566502463054,
      "grad_norm": 12.23318853282514,
      "learning_rate": 1.071972866834519e-05,
      "loss": 1.3409,
      "step": 860
    },
    {
      "epoch": 0.5326354679802956,
      "grad_norm": 9.79980745523444,
      "learning_rate": 1.061245281802171e-05,
      "loss": 1.4392,
      "step": 865
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 13.39047711234012,
      "learning_rate": 1.0505106171852226e-05,
      "loss": 1.3005,
      "step": 870
    },
    {
      "epoch": 0.5387931034482759,
      "grad_norm": 27.935567545937946,
      "learning_rate": 1.039770113846056e-05,
      "loss": 1.3531,
      "step": 875
    },
    {
      "epoch": 0.541871921182266,
      "grad_norm": 21.112232870571965,
      "learning_rate": 1.0290250133219755e-05,
      "loss": 1.4231,
      "step": 880
    },
    {
      "epoch": 0.5449507389162561,
      "grad_norm": 14.018625193344477,
      "learning_rate": 1.0182765576816916e-05,
      "loss": 1.493,
      "step": 885
    },
    {
      "epoch": 0.5480295566502463,
      "grad_norm": 47.019213796032155,
      "learning_rate": 1.0075259893817465e-05,
      "loss": 1.4117,
      "step": 890
    },
    {
      "epoch": 0.5511083743842364,
      "grad_norm": 21.905968001772294,
      "learning_rate": 9.967745511228922e-06,
      "loss": 1.3714,
      "step": 895
    },
    {
      "epoch": 0.5541871921182266,
      "grad_norm": 10.600750962435368,
      "learning_rate": 9.860234857064439e-06,
      "loss": 1.4149,
      "step": 900
    },
    {
      "epoch": 0.5572660098522167,
      "grad_norm": 19.368746856411462,
      "learning_rate": 9.752740358906168e-06,
      "loss": 1.4068,
      "step": 905
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 10.340212723912037,
      "learning_rate": 9.645274442468746e-06,
      "loss": 1.381,
      "step": 910
    },
    {
      "epoch": 0.5634236453201971,
      "grad_norm": 17.209489283241304,
      "learning_rate": 9.537849530162915e-06,
      "loss": 1.4018,
      "step": 915
    },
    {
      "epoch": 0.5665024630541872,
      "grad_norm": 16.192710445420914,
      "learning_rate": 9.430478039659604e-06,
      "loss": 1.3763,
      "step": 920
    },
    {
      "epoch": 0.5695812807881774,
      "grad_norm": 10.665049867143576,
      "learning_rate": 9.323172382454494e-06,
      "loss": 1.3917,
      "step": 925
    },
    {
      "epoch": 0.5726600985221675,
      "grad_norm": 23.488462682807576,
      "learning_rate": 9.215944962433329e-06,
      "loss": 1.3135,
      "step": 930
    },
    {
      "epoch": 0.5757389162561576,
      "grad_norm": 21.67168819889345,
      "learning_rate": 9.108808174438128e-06,
      "loss": 1.407,
      "step": 935
    },
    {
      "epoch": 0.5788177339901478,
      "grad_norm": 21.02636023745943,
      "learning_rate": 9.001774402834375e-06,
      "loss": 1.3427,
      "step": 940
    },
    {
      "epoch": 0.5818965517241379,
      "grad_norm": 8.044561556098252,
      "learning_rate": 8.89485602007949e-06,
      "loss": 1.378,
      "step": 945
    },
    {
      "epoch": 0.5849753694581281,
      "grad_norm": 16.484248656670655,
      "learning_rate": 8.788065385292637e-06,
      "loss": 1.3633,
      "step": 950
    },
    {
      "epoch": 0.5880541871921182,
      "grad_norm": 12.965003760380185,
      "learning_rate": 8.68141484282609e-06,
      "loss": 1.4083,
      "step": 955
    },
    {
      "epoch": 0.5911330049261084,
      "grad_norm": 18.259361751180567,
      "learning_rate": 8.574916720838293e-06,
      "loss": 1.431,
      "step": 960
    },
    {
      "epoch": 0.5942118226600985,
      "grad_norm": 14.931940073877918,
      "learning_rate": 8.468583329868805e-06,
      "loss": 1.4134,
      "step": 965
    },
    {
      "epoch": 0.5972906403940886,
      "grad_norm": 7.9389796967423845,
      "learning_rate": 8.362426961415274e-06,
      "loss": 1.3798,
      "step": 970
    },
    {
      "epoch": 0.6003694581280788,
      "grad_norm": 10.911398304396144,
      "learning_rate": 8.256459886512618e-06,
      "loss": 1.3546,
      "step": 975
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 19.695152731292648,
      "learning_rate": 8.150694354314556e-06,
      "loss": 1.4425,
      "step": 980
    },
    {
      "epoch": 0.6065270935960592,
      "grad_norm": 15.555720674983748,
      "learning_rate": 8.0451425906777e-06,
      "loss": 1.3925,
      "step": 985
    },
    {
      "epoch": 0.6096059113300493,
      "grad_norm": 26.83456471123147,
      "learning_rate": 7.939816796748296e-06,
      "loss": 1.3358,
      "step": 990
    },
    {
      "epoch": 0.6126847290640394,
      "grad_norm": 17.412398845782374,
      "learning_rate": 7.834729147551858e-06,
      "loss": 1.3782,
      "step": 995
    },
    {
      "epoch": 0.6157635467980296,
      "grad_norm": 14.175946894851235,
      "learning_rate": 7.729891790585817e-06,
      "loss": 1.3165,
      "step": 1000
    },
    {
      "epoch": 0.6188423645320197,
      "grad_norm": 8.623021255345796,
      "learning_rate": 7.625316844415327e-06,
      "loss": 1.3017,
      "step": 1005
    },
    {
      "epoch": 0.6219211822660099,
      "grad_norm": 10.20752311555882,
      "learning_rate": 7.521016397272436e-06,
      "loss": 1.3772,
      "step": 1010
    },
    {
      "epoch": 0.625,
      "grad_norm": 8.29466538031788,
      "learning_rate": 7.417002505658773e-06,
      "loss": 1.4227,
      "step": 1015
    },
    {
      "epoch": 0.6280788177339901,
      "grad_norm": 9.706478317484397,
      "learning_rate": 7.313287192951866e-06,
      "loss": 1.3848,
      "step": 1020
    },
    {
      "epoch": 0.6311576354679803,
      "grad_norm": 10.802336009782953,
      "learning_rate": 7.209882448015338e-06,
      "loss": 1.3446,
      "step": 1025
    },
    {
      "epoch": 0.6342364532019704,
      "grad_norm": 18.98546241888069,
      "learning_rate": 7.1068002238130465e-06,
      "loss": 1.3682,
      "step": 1030
    },
    {
      "epoch": 0.6373152709359606,
      "grad_norm": 9.536444897433284,
      "learning_rate": 7.004052436027397e-06,
      "loss": 1.3615,
      "step": 1035
    },
    {
      "epoch": 0.6403940886699507,
      "grad_norm": 8.159252929197551,
      "learning_rate": 6.901650961681976e-06,
      "loss": 1.3406,
      "step": 1040
    },
    {
      "epoch": 0.6434729064039408,
      "grad_norm": 8.417979006236454,
      "learning_rate": 6.799607637768621e-06,
      "loss": 1.3694,
      "step": 1045
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 11.129778045558496,
      "learning_rate": 6.6979342598791395e-06,
      "loss": 1.3766,
      "step": 1050
    },
    {
      "epoch": 0.6496305418719212,
      "grad_norm": 11.957608169763887,
      "learning_rate": 6.596642580841827e-06,
      "loss": 1.3597,
      "step": 1055
    },
    {
      "epoch": 0.6527093596059114,
      "grad_norm": 16.8182044633009,
      "learning_rate": 6.495744309362879e-06,
      "loss": 1.3064,
      "step": 1060
    },
    {
      "epoch": 0.6557881773399015,
      "grad_norm": 17.103923571365467,
      "learning_rate": 6.395251108672975e-06,
      "loss": 1.331,
      "step": 1065
    },
    {
      "epoch": 0.6588669950738916,
      "grad_norm": 9.356263358149233,
      "learning_rate": 6.295174595179041e-06,
      "loss": 1.36,
      "step": 1070
    },
    {
      "epoch": 0.6619458128078818,
      "grad_norm": 14.770806738027815,
      "learning_rate": 6.195526337121483e-06,
      "loss": 1.3919,
      "step": 1075
    },
    {
      "epoch": 0.6650246305418719,
      "grad_norm": 22.999125215792372,
      "learning_rate": 6.096317853236975e-06,
      "loss": 1.2782,
      "step": 1080
    },
    {
      "epoch": 0.6681034482758621,
      "grad_norm": 8.809923298393809,
      "learning_rate": 5.997560611426947e-06,
      "loss": 1.3444,
      "step": 1085
    },
    {
      "epoch": 0.6711822660098522,
      "grad_norm": 18.77979667174953,
      "learning_rate": 5.899266027431965e-06,
      "loss": 1.3848,
      "step": 1090
    },
    {
      "epoch": 0.6742610837438424,
      "grad_norm": 11.03776076734672,
      "learning_rate": 5.80144546351216e-06,
      "loss": 1.3011,
      "step": 1095
    },
    {
      "epoch": 0.6773399014778325,
      "grad_norm": 8.735851907944854,
      "learning_rate": 5.704110227133792e-06,
      "loss": 1.3439,
      "step": 1100
    },
    {
      "epoch": 0.6804187192118226,
      "grad_norm": 14.77561891944379,
      "learning_rate": 5.607271569662203e-06,
      "loss": 1.3166,
      "step": 1105
    },
    {
      "epoch": 0.6834975369458128,
      "grad_norm": 21.95252112441737,
      "learning_rate": 5.510940685061202e-06,
      "loss": 1.3284,
      "step": 1110
    },
    {
      "epoch": 0.6865763546798029,
      "grad_norm": 22.872979512272977,
      "learning_rate": 5.41512870859912e-06,
      "loss": 1.3419,
      "step": 1115
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 18.730673975715575,
      "learning_rate": 5.319846715561656e-06,
      "loss": 1.3474,
      "step": 1120
    },
    {
      "epoch": 0.6927339901477833,
      "grad_norm": 12.94402628113151,
      "learning_rate": 5.225105719971615e-06,
      "loss": 1.284,
      "step": 1125
    },
    {
      "epoch": 0.6958128078817734,
      "grad_norm": 11.076020901601611,
      "learning_rate": 5.130916673315762e-06,
      "loss": 1.3774,
      "step": 1130
    },
    {
      "epoch": 0.6988916256157636,
      "grad_norm": 15.519595404751641,
      "learning_rate": 5.037290463278914e-06,
      "loss": 1.3096,
      "step": 1135
    },
    {
      "epoch": 0.7019704433497537,
      "grad_norm": 9.993229180399556,
      "learning_rate": 4.94423791248536e-06,
      "loss": 1.3125,
      "step": 1140
    },
    {
      "epoch": 0.7050492610837439,
      "grad_norm": 19.385051432139623,
      "learning_rate": 4.851769777247857e-06,
      "loss": 1.4055,
      "step": 1145
    },
    {
      "epoch": 0.708128078817734,
      "grad_norm": 8.020441232215644,
      "learning_rate": 4.759896746324247e-06,
      "loss": 1.322,
      "step": 1150
    },
    {
      "epoch": 0.7112068965517241,
      "grad_norm": 8.425717242898326,
      "learning_rate": 4.668629439681907e-06,
      "loss": 1.3147,
      "step": 1155
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 12.858189319574025,
      "learning_rate": 4.577978407270156e-06,
      "loss": 1.33,
      "step": 1160
    },
    {
      "epoch": 0.7173645320197044,
      "grad_norm": 17.66515297002037,
      "learning_rate": 4.487954127800726e-06,
      "loss": 1.3114,
      "step": 1165
    },
    {
      "epoch": 0.7204433497536946,
      "grad_norm": 18.24711748661215,
      "learning_rate": 4.398567007536493e-06,
      "loss": 1.4077,
      "step": 1170
    },
    {
      "epoch": 0.7235221674876847,
      "grad_norm": 13.478769959260893,
      "learning_rate": 4.309827379088589e-06,
      "loss": 1.3231,
      "step": 1175
    },
    {
      "epoch": 0.7266009852216748,
      "grad_norm": 7.302618771342409,
      "learning_rate": 4.221745500221992e-06,
      "loss": 1.2599,
      "step": 1180
    },
    {
      "epoch": 0.729679802955665,
      "grad_norm": 7.6843380691337835,
      "learning_rate": 4.134331552669812e-06,
      "loss": 1.3517,
      "step": 1185
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 8.515558984433651,
      "learning_rate": 4.047595640956326e-06,
      "loss": 1.3627,
      "step": 1190
    },
    {
      "epoch": 0.7358374384236454,
      "grad_norm": 8.423972687424229,
      "learning_rate": 3.961547791228963e-06,
      "loss": 1.277,
      "step": 1195
    },
    {
      "epoch": 0.7389162561576355,
      "grad_norm": 9.232139997585016,
      "learning_rate": 3.876197950099351e-06,
      "loss": 1.292,
      "step": 1200
    },
    {
      "epoch": 0.7419950738916257,
      "grad_norm": 10.31580251619475,
      "learning_rate": 3.7915559834935355e-06,
      "loss": 1.3203,
      "step": 1205
    },
    {
      "epoch": 0.7450738916256158,
      "grad_norm": 7.4684234026975895,
      "learning_rate": 3.7076316755115407e-06,
      "loss": 1.2874,
      "step": 1210
    },
    {
      "epoch": 0.7481527093596059,
      "grad_norm": 8.002591900900384,
      "learning_rate": 3.6244347272963974e-06,
      "loss": 1.3066,
      "step": 1215
    },
    {
      "epoch": 0.7512315270935961,
      "grad_norm": 11.054705747477877,
      "learning_rate": 3.5419747559127294e-06,
      "loss": 1.3229,
      "step": 1220
    },
    {
      "epoch": 0.7543103448275862,
      "grad_norm": 7.81922171969261,
      "learning_rate": 3.4602612932351065e-06,
      "loss": 1.3395,
      "step": 1225
    },
    {
      "epoch": 0.7573891625615764,
      "grad_norm": 9.80272601474901,
      "learning_rate": 3.3793037848461873e-06,
      "loss": 1.301,
      "step": 1230
    },
    {
      "epoch": 0.7604679802955665,
      "grad_norm": 8.64305287763975,
      "learning_rate": 3.2991115889448877e-06,
      "loss": 1.2926,
      "step": 1235
    },
    {
      "epoch": 0.7635467980295566,
      "grad_norm": 7.806886192716367,
      "learning_rate": 3.2196939752646183e-06,
      "loss": 1.2926,
      "step": 1240
    },
    {
      "epoch": 0.7666256157635468,
      "grad_norm": 7.547666757027036,
      "learning_rate": 3.141060124001776e-06,
      "loss": 1.3068,
      "step": 1245
    },
    {
      "epoch": 0.7697044334975369,
      "grad_norm": 9.555552308692176,
      "learning_rate": 3.063219124754543e-06,
      "loss": 1.3023,
      "step": 1250
    },
    {
      "epoch": 0.7727832512315271,
      "grad_norm": 8.411419674044241,
      "learning_rate": 2.9861799754722033e-06,
      "loss": 1.2665,
      "step": 1255
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 10.709948989400203,
      "learning_rate": 2.9099515814150335e-06,
      "loss": 1.2541,
      "step": 1260
    },
    {
      "epoch": 0.7789408866995073,
      "grad_norm": 14.980457406276473,
      "learning_rate": 2.8345427541248993e-06,
      "loss": 1.2431,
      "step": 1265
    },
    {
      "epoch": 0.7820197044334976,
      "grad_norm": 7.561020286323297,
      "learning_rate": 2.7599622104066937e-06,
      "loss": 1.3489,
      "step": 1270
    },
    {
      "epoch": 0.7850985221674877,
      "grad_norm": 6.762857047686262,
      "learning_rate": 2.6862185713207467e-06,
      "loss": 1.2933,
      "step": 1275
    },
    {
      "epoch": 0.7881773399014779,
      "grad_norm": 10.436824156417362,
      "learning_rate": 2.6133203611862554e-06,
      "loss": 1.279,
      "step": 1280
    },
    {
      "epoch": 0.791256157635468,
      "grad_norm": 9.19121896336613,
      "learning_rate": 2.5412760065959386e-06,
      "loss": 1.2856,
      "step": 1285
    },
    {
      "epoch": 0.7943349753694581,
      "grad_norm": 7.299874225629273,
      "learning_rate": 2.4700938354419823e-06,
      "loss": 1.2808,
      "step": 1290
    },
    {
      "epoch": 0.7974137931034483,
      "grad_norm": 6.93762581606662,
      "learning_rate": 2.3997820759533654e-06,
      "loss": 1.3062,
      "step": 1295
    },
    {
      "epoch": 0.8004926108374384,
      "grad_norm": 6.867928680181049,
      "learning_rate": 2.3303488557447374e-06,
      "loss": 1.2471,
      "step": 1300
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 6.6252232320657924,
      "learning_rate": 2.26180220087692e-06,
      "loss": 1.2657,
      "step": 1305
    },
    {
      "epoch": 0.8066502463054187,
      "grad_norm": 7.151355762927701,
      "learning_rate": 2.194150034929133e-06,
      "loss": 1.2972,
      "step": 1310
    },
    {
      "epoch": 0.8097290640394089,
      "grad_norm": 10.61150455539557,
      "learning_rate": 2.1274001780830776e-06,
      "loss": 1.2988,
      "step": 1315
    },
    {
      "epoch": 0.812807881773399,
      "grad_norm": 8.478966940847858,
      "learning_rate": 2.0615603462189824e-06,
      "loss": 1.317,
      "step": 1320
    },
    {
      "epoch": 0.8158866995073891,
      "grad_norm": 7.647249125843244,
      "learning_rate": 1.9966381500236786e-06,
      "loss": 1.2202,
      "step": 1325
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 6.7760137509227425,
      "learning_rate": 1.932641094110855e-06,
      "loss": 1.265,
      "step": 1330
    },
    {
      "epoch": 0.8220443349753694,
      "grad_norm": 11.005843646770034,
      "learning_rate": 1.869576576153581e-06,
      "loss": 1.2099,
      "step": 1335
    },
    {
      "epoch": 0.8251231527093597,
      "grad_norm": 10.046995706677444,
      "learning_rate": 1.8074518860291646e-06,
      "loss": 1.2691,
      "step": 1340
    },
    {
      "epoch": 0.8282019704433498,
      "grad_norm": 7.467331282453166,
      "learning_rate": 1.746274204976498e-06,
      "loss": 1.2611,
      "step": 1345
    },
    {
      "epoch": 0.8312807881773399,
      "grad_norm": 7.229696328604628,
      "learning_rate": 1.6860506047659442e-06,
      "loss": 1.2688,
      "step": 1350
    },
    {
      "epoch": 0.8343596059113301,
      "grad_norm": 6.744110485516448,
      "learning_rate": 1.6267880468818787e-06,
      "loss": 1.2412,
      "step": 1355
    },
    {
      "epoch": 0.8374384236453202,
      "grad_norm": 8.852264527144929,
      "learning_rate": 1.5684933817180014e-06,
      "loss": 1.2771,
      "step": 1360
    },
    {
      "epoch": 0.8405172413793104,
      "grad_norm": 12.735778529994441,
      "learning_rate": 1.5111733477854507e-06,
      "loss": 1.2872,
      "step": 1365
    },
    {
      "epoch": 0.8435960591133005,
      "grad_norm": 7.240920825332387,
      "learning_rate": 1.454834570933884e-06,
      "loss": 1.2302,
      "step": 1370
    },
    {
      "epoch": 0.8466748768472906,
      "grad_norm": 7.571689813788829,
      "learning_rate": 1.399483563585573e-06,
      "loss": 1.275,
      "step": 1375
    },
    {
      "epoch": 0.8497536945812808,
      "grad_norm": 7.780172978665113,
      "learning_rate": 1.345126723982594e-06,
      "loss": 1.34,
      "step": 1380
    },
    {
      "epoch": 0.8528325123152709,
      "grad_norm": 8.561225757508065,
      "learning_rate": 1.2917703354472467e-06,
      "loss": 1.2557,
      "step": 1385
    },
    {
      "epoch": 0.8559113300492611,
      "grad_norm": 10.571516988186687,
      "learning_rate": 1.2394205656557224e-06,
      "loss": 1.2487,
      "step": 1390
    },
    {
      "epoch": 0.8589901477832512,
      "grad_norm": 7.087630630952262,
      "learning_rate": 1.1880834659251706e-06,
      "loss": 1.2715,
      "step": 1395
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 9.265456612804266,
      "learning_rate": 1.1377649705142012e-06,
      "loss": 1.2785,
      "step": 1400
    },
    {
      "epoch": 0.8651477832512315,
      "grad_norm": 9.408590480859424,
      "learning_rate": 1.0884708959369116e-06,
      "loss": 1.303,
      "step": 1405
    },
    {
      "epoch": 0.8682266009852216,
      "grad_norm": 7.075930275063177,
      "learning_rate": 1.040206940290547e-06,
      "loss": 1.2274,
      "step": 1410
    },
    {
      "epoch": 0.8713054187192119,
      "grad_norm": 6.354199837846897,
      "learning_rate": 9.929786825968213e-07,
      "loss": 1.2342,
      "step": 1415
    },
    {
      "epoch": 0.874384236453202,
      "grad_norm": 7.4318395549675405,
      "learning_rate": 9.467915821570228e-07,
      "loss": 1.2917,
      "step": 1420
    },
    {
      "epoch": 0.8774630541871922,
      "grad_norm": 6.936418377158421,
      "learning_rate": 9.016509779209536e-07,
      "loss": 1.2708,
      "step": 1425
    },
    {
      "epoch": 0.8805418719211823,
      "grad_norm": 8.103125888344566,
      "learning_rate": 8.575620878697744e-07,
      "loss": 1.2316,
      "step": 1430
    },
    {
      "epoch": 0.8836206896551724,
      "grad_norm": 10.248646825430681,
      "learning_rate": 8.145300084128349e-07,
      "loss": 1.2637,
      "step": 1435
    },
    {
      "epoch": 0.8866995073891626,
      "grad_norm": 5.970922877419386,
      "learning_rate": 7.725597137985741e-07,
      "loss": 1.1892,
      "step": 1440
    },
    {
      "epoch": 0.8897783251231527,
      "grad_norm": 8.287948622274875,
      "learning_rate": 7.316560555395069e-07,
      "loss": 1.2157,
      "step": 1445
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 8.447501834857656,
      "learning_rate": 6.918237618514378e-07,
      "loss": 1.1855,
      "step": 1450
    },
    {
      "epoch": 0.895935960591133,
      "grad_norm": 7.159131346667986,
      "learning_rate": 6.530674371068946e-07,
      "loss": 1.257,
      "step": 1455
    },
    {
      "epoch": 0.8990147783251231,
      "grad_norm": 10.533560917394905,
      "learning_rate": 6.153915613028915e-07,
      "loss": 1.2347,
      "step": 1460
    },
    {
      "epoch": 0.9020935960591133,
      "grad_norm": 8.866898875480516,
      "learning_rate": 5.788004895430799e-07,
      "loss": 1.1854,
      "step": 1465
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 8.194881676375312,
      "learning_rate": 5.43298451534312e-07,
      "loss": 1.2982,
      "step": 1470
    },
    {
      "epoch": 0.9082512315270936,
      "grad_norm": 7.016734395939443,
      "learning_rate": 5.088895510977154e-07,
      "loss": 1.2356,
      "step": 1475
    },
    {
      "epoch": 0.9113300492610837,
      "grad_norm": 7.085958022802171,
      "learning_rate": 4.755777656943239e-07,
      "loss": 1.2744,
      "step": 1480
    },
    {
      "epoch": 0.9144088669950738,
      "grad_norm": 9.454871601471178,
      "learning_rate": 4.433669459652945e-07,
      "loss": 1.2263,
      "step": 1485
    },
    {
      "epoch": 0.9174876847290641,
      "grad_norm": 7.94562063795173,
      "learning_rate": 4.1226081528680907e-07,
      "loss": 1.2799,
      "step": 1490
    },
    {
      "epoch": 0.9205665024630542,
      "grad_norm": 6.748156588514909,
      "learning_rate": 3.822629693396651e-07,
      "loss": 1.2173,
      "step": 1495
    },
    {
      "epoch": 0.9236453201970444,
      "grad_norm": 7.1052859489691125,
      "learning_rate": 3.5337687569363734e-07,
      "loss": 1.2749,
      "step": 1500
    },
    {
      "epoch": 0.9267241379310345,
      "grad_norm": 7.280418494562587,
      "learning_rate": 3.2560587340665694e-07,
      "loss": 1.2514,
      "step": 1505
    },
    {
      "epoch": 0.9298029556650246,
      "grad_norm": 6.511194976095366,
      "learning_rate": 2.989531726388262e-07,
      "loss": 1.2674,
      "step": 1510
    },
    {
      "epoch": 0.9328817733990148,
      "grad_norm": 9.154084778928757,
      "learning_rate": 2.7342185428134604e-07,
      "loss": 1.2745,
      "step": 1515
    },
    {
      "epoch": 0.9359605911330049,
      "grad_norm": 7.957060401797473,
      "learning_rate": 2.4901486960039025e-07,
      "loss": 1.2681,
      "step": 1520
    },
    {
      "epoch": 0.9390394088669951,
      "grad_norm": 6.873067183698468,
      "learning_rate": 2.257350398959457e-07,
      "loss": 1.2312,
      "step": 1525
    },
    {
      "epoch": 0.9421182266009852,
      "grad_norm": 10.68387181654707,
      "learning_rate": 2.035850561756969e-07,
      "loss": 1.245,
      "step": 1530
    },
    {
      "epoch": 0.9451970443349754,
      "grad_norm": 6.894582373427124,
      "learning_rate": 1.8256747884395577e-07,
      "loss": 1.2479,
      "step": 1535
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 7.101633764094371,
      "learning_rate": 1.6268473740569723e-07,
      "loss": 1.3312,
      "step": 1540
    },
    {
      "epoch": 0.9513546798029556,
      "grad_norm": 7.043121051092255,
      "learning_rate": 1.4393913018572182e-07,
      "loss": 1.2224,
      "step": 1545
    },
    {
      "epoch": 0.9544334975369458,
      "grad_norm": 6.668479034325025,
      "learning_rate": 1.2633282406298576e-07,
      "loss": 1.2493,
      "step": 1550
    },
    {
      "epoch": 0.9575123152709359,
      "grad_norm": 7.036438292666957,
      "learning_rate": 1.0986785422011593e-07,
      "loss": 1.2326,
      "step": 1555
    },
    {
      "epoch": 0.9605911330049262,
      "grad_norm": 6.0704303828222965,
      "learning_rate": 9.454612390816686e-08,
      "loss": 1.2284,
      "step": 1560
    },
    {
      "epoch": 0.9636699507389163,
      "grad_norm": 6.331625313562051,
      "learning_rate": 8.036940422660345e-08,
      "loss": 1.2493,
      "step": 1565
    },
    {
      "epoch": 0.9667487684729064,
      "grad_norm": 7.32852887589745,
      "learning_rate": 6.733933391858238e-08,
      "loss": 1.3042,
      "step": 1570
    },
    {
      "epoch": 0.9698275862068966,
      "grad_norm": 9.963995739558282,
      "learning_rate": 5.5457419181517145e-08,
      "loss": 1.236,
      "step": 1575
    },
    {
      "epoch": 0.9729064039408867,
      "grad_norm": 6.627733686300748,
      "learning_rate": 4.472503349297497e-08,
      "loss": 1.1996,
      "step": 1580
    },
    {
      "epoch": 0.9759852216748769,
      "grad_norm": 7.234989701478282,
      "learning_rate": 3.5143417451907195e-08,
      "loss": 1.2076,
      "step": 1585
    },
    {
      "epoch": 0.979064039408867,
      "grad_norm": 12.472880054360536,
      "learning_rate": 2.671367863524732e-08,
      "loss": 1.2612,
      "step": 1590
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 7.0817535398795854,
      "learning_rate": 1.943679146988009e-08,
      "loss": 1.276,
      "step": 1595
    },
    {
      "epoch": 0.9852216748768473,
      "grad_norm": 8.107209810494863,
      "learning_rate": 1.3313597120002686e-08,
      "loss": 1.2662,
      "step": 1600
    },
    {
      "epoch": 0.9883004926108374,
      "grad_norm": 7.688683508919599,
      "learning_rate": 8.34480338989141e-09,
      "loss": 1.2943,
      "step": 1605
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 6.1964489042565765,
      "learning_rate": 4.530984642087122e-09,
      "loss": 1.1709,
      "step": 1610
    },
    {
      "epoch": 0.9944581280788177,
      "grad_norm": 6.73265107054657,
      "learning_rate": 1.872581730997247e-09,
      "loss": 1.2853,
      "step": 1615
    },
    {
      "epoch": 0.9975369458128078,
      "grad_norm": 6.616322833119161,
      "learning_rate": 3.699019519387559e-10,
      "loss": 1.1907,
      "step": 1620
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.2451802492141724,
      "eval_runtime": 910.1999,
      "eval_samples_per_second": 25.389,
      "eval_steps_per_second": 0.794,
      "step": 1624
    },
    {
      "epoch": 1.0,
      "step": 1624,
      "total_flos": 55902786895872.0,
      "train_loss": 1.3673941268415875,
      "train_runtime": 11829.2918,
      "train_samples_per_second": 4.393,
      "train_steps_per_second": 0.137
    }
  ],
  "logging_steps": 5,
  "max_steps": 1624,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 55902786895872.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}