{
  "best_metric": 1.0445035696029663,
  "best_model_checkpoint": "/home/stephan/code/molreactgen/checkpoints/2023-05-09_21-49-53_experiment/checkpoint-248650",
  "epoch": 49.99748655305886,
  "global_step": 248650,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1.0054293183189222e-05,
      "loss": 4.306,
      "step": 100
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.0108586366378444e-05,
      "loss": 3.8922,
      "step": 200
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.016287954956767e-05,
      "loss": 3.6512,
      "step": 300
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.021717273275689e-05,
      "loss": 3.44,
      "step": 400
    },
    {
      "epoch": 0.1,
      "learning_rate": 5.027146591594611e-05,
      "loss": 3.2164,
      "step": 500
    },
    {
      "epoch": 0.12,
      "learning_rate": 6.032575909913534e-05,
      "loss": 3.0072,
      "step": 600
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.038005228232456e-05,
      "loss": 2.8143,
      "step": 700
    },
    {
      "epoch": 0.16,
      "learning_rate": 8.043434546551378e-05,
      "loss": 2.6472,
      "step": 800
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.048863864870299e-05,
      "loss": 2.51,
      "step": 900
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00010054293183189222,
      "loss": 2.3891,
      "step": 1000
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00011059722501508144,
      "loss": 2.2905,
      "step": 1100
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00012065151819827068,
      "loss": 2.2056,
      "step": 1200
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00013070581138145988,
      "loss": 2.137,
      "step": 1300
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00014076010456464912,
      "loss": 2.0792,
      "step": 1400
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00015081439774783834,
      "loss": 2.0301,
      "step": 1500
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00016086869093102755,
      "loss": 1.9879,
      "step": 1600
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00017092298411421677,
      "loss": 1.9499,
      "step": 1700
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018097727729740598,
      "loss": 1.9279,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0001910315704805952,
      "loss": 1.8925,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00020108586366378444,
      "loss": 1.8673,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00021114015684697366,
      "loss": 1.8431,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00022119445003016287,
      "loss": 1.8236,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002312487432133521,
      "loss": 1.8023,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00024130303639654136,
      "loss": 1.7889,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00025135732957973054,
      "loss": 1.7671,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00026141162276291976,
      "loss": 1.7506,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00027146591594610903,
      "loss": 1.737,
      "step": 2700
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00028152020912929824,
      "loss": 1.7272,
      "step": 2800
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00029157450231248746,
      "loss": 1.7112,
      "step": 2900
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0003016287954956767,
      "loss": 1.699,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0003116830886788659,
      "loss": 1.6933,
      "step": 3100
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0003217373818620551,
      "loss": 1.6788,
      "step": 3200
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0003317916750452443,
      "loss": 1.6694,
      "step": 3300
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00034184596822843354,
      "loss": 1.6607,
      "step": 3400
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00035190026141162275,
      "loss": 1.6534,
      "step": 3500
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00036195455459481197,
      "loss": 1.645,
      "step": 3600
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0003720088477780012,
      "loss": 1.632,
      "step": 3700
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0003820631409611904,
      "loss": 1.6229,
      "step": 3800
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00039211743414437967,
      "loss": 1.6161,
      "step": 3900
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0004021717273275689,
      "loss": 1.6061,
      "step": 4000
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0004122260205107581,
      "loss": 1.6006,
      "step": 4100
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0004222803136939473,
      "loss": 1.5875,
      "step": 4200
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00043233460687713653,
      "loss": 1.5828,
      "step": 4300
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00044238890006032574,
      "loss": 1.5849,
      "step": 4400
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00045244319324351496,
      "loss": 1.572,
      "step": 4500
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0004624974864267042,
      "loss": 1.565,
      "step": 4600
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0004725517796098934,
      "loss": 1.5589,
      "step": 4700
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0004826060727930827,
      "loss": 1.5519,
      "step": 4800
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0004926603659762719,
      "loss": 1.5489,
      "step": 4900
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.38639571407117423,
      "eval_loss": 1.490514874458313,
      "eval_runtime": 19.6663,
      "eval_samples_per_second": 4045.908,
      "eval_steps_per_second": 15.814,
      "step": 4973
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0005027146591594611,
      "loss": 1.5386,
      "step": 5000
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0005127689523426503,
      "loss": 1.5361,
      "step": 5100
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0005228232455258395,
      "loss": 1.5279,
      "step": 5200
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0005328775387090288,
      "loss": 1.5273,
      "step": 5300
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0005429318318922181,
      "loss": 1.5192,
      "step": 5400
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0005529861250754073,
      "loss": 1.5175,
      "step": 5500
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0005630404182585965,
      "loss": 1.5165,
      "step": 5600
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0005730947114417857,
      "loss": 1.5102,
      "step": 5700
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0005831490046249749,
      "loss": 1.5006,
      "step": 5800
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0005932032978081641,
      "loss": 1.4952,
      "step": 5900
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0006032575909913533,
      "loss": 1.4956,
      "step": 6000
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0006133118841745426,
      "loss": 1.4906,
      "step": 6100
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0006233661773577318,
      "loss": 1.4845,
      "step": 6200
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0006334204705409209,
      "loss": 1.4826,
      "step": 6300
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0006434747637241102,
      "loss": 1.4844,
      "step": 6400
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0006535290569072995,
      "loss": 1.4789,
      "step": 6500
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0006635833500904886,
      "loss": 1.474,
      "step": 6600
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.000673637643273678,
      "loss": 1.4716,
      "step": 6700
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0006836919364568671,
      "loss": 1.471,
      "step": 6800
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0006937462296400564,
      "loss": 1.4675,
      "step": 6900
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0007038005228232455,
      "loss": 1.4627,
      "step": 7000
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0007138548160064348,
      "loss": 1.4575,
      "step": 7100
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0007239091091896239,
      "loss": 1.4615,
      "step": 7200
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0007339634023728133,
      "loss": 1.451,
      "step": 7300
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0007440176955560024,
      "loss": 1.4533,
      "step": 7400
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0007540719887391917,
      "loss": 1.4469,
      "step": 7500
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0007641262819223808,
      "loss": 1.4471,
      "step": 7600
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0007741805751055701,
      "loss": 1.4433,
      "step": 7700
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0007842348682887593,
      "loss": 1.4447,
      "step": 7800
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0007942891614719485,
      "loss": 1.4389,
      "step": 7900
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0008043434546551378,
      "loss": 1.4352,
      "step": 8000
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.000814397747838327,
      "loss": 1.432,
      "step": 8100
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0008244520410215162,
      "loss": 1.4366,
      "step": 8200
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0008345063342047054,
      "loss": 1.4331,
      "step": 8300
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0008445606273878946,
      "loss": 1.4331,
      "step": 8400
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0008546149205710838,
      "loss": 1.4229,
      "step": 8500
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0008646692137542731,
      "loss": 1.4255,
      "step": 8600
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0008747235069374624,
      "loss": 1.4223,
      "step": 8700
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0008847778001206515,
      "loss": 1.4239,
      "step": 8800
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0008948320933038408,
      "loss": 1.4214,
      "step": 8900
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0009048863864870299,
      "loss": 1.4187,
      "step": 9000
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0009149406796702192,
      "loss": 1.4181,
      "step": 9100
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0009249949728534083,
      "loss": 1.4188,
      "step": 9200
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0009350492660365977,
      "loss": 1.4166,
      "step": 9300
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0009451035592197868,
      "loss": 1.4089,
      "step": 9400
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0009551578524029761,
      "loss": 1.4085,
      "step": 9500
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0009652121455861654,
      "loss": 1.4111,
      "step": 9600
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0009752664387693545,
      "loss": 1.4065,
      "step": 9700
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0009853207319525437,
      "loss": 1.4065,
      "step": 9800
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.000995375025135733,
      "loss": 1.4079,
      "step": 9900
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.39677088556687123,
      "eval_loss": 1.357865333557129,
      "eval_runtime": 19.5955,
      "eval_samples_per_second": 4060.518,
      "eval_steps_per_second": 15.871,
      "step": 9946
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0010054293183189222,
      "loss": 1.4024,
      "step": 10000
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0010154836115021115,
      "loss": 1.3983,
      "step": 10100
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0010255379046853006,
      "loss": 1.3965,
      "step": 10200
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.00103559219786849,
      "loss": 1.3945,
      "step": 10300
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0010455459481198471,
      "loss": 1.3989,
      "step": 10400
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0010556002413030364,
      "loss": 1.3975,
      "step": 10500
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0010655539915543938,
      "loss": 1.3924,
      "step": 10600
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.001075608284737583,
      "loss": 1.3933,
      "step": 10700
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0010856625779207723,
      "loss": 1.394,
      "step": 10800
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0010957168711039614,
      "loss": 1.3945,
      "step": 10900
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0011057711642871507,
      "loss": 1.3929,
      "step": 11000
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0011158254574703398,
      "loss": 1.3943,
      "step": 11100
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0011258797506535291,
      "loss": 1.3893,
      "step": 11200
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0011359340438367182,
      "loss": 1.3908,
      "step": 11300
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0011459883370199076,
      "loss": 1.3887,
      "step": 11400
    },
    {
      "epoch": 2.31,
      "learning_rate": 0.0011560426302030967,
      "loss": 1.3902,
      "step": 11500
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.001166096923386286,
      "loss": 1.3934,
      "step": 11600
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.001176151216569475,
      "loss": 1.3857,
      "step": 11700
    },
    {
      "epoch": 2.37,
      "learning_rate": 0.0011862055097526644,
      "loss": 1.3885,
      "step": 11800
    },
    {
      "epoch": 2.39,
      "learning_rate": 0.0011962598029358535,
      "loss": 1.3847,
      "step": 11900
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.0012063140961190428,
      "loss": 1.3863,
      "step": 12000
    },
    {
      "epoch": 2.43,
      "learning_rate": 0.001216368389302232,
      "loss": 1.3846,
      "step": 12100
    },
    {
      "epoch": 2.45,
      "learning_rate": 0.0012264226824854213,
      "loss": 1.3815,
      "step": 12200
    },
    {
      "epoch": 2.47,
      "learning_rate": 0.0012364769756686104,
      "loss": 1.3876,
      "step": 12300
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0012465312688517997,
      "loss": 1.3812,
      "step": 12400
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.0012565855620349888,
      "loss": 1.3838,
      "step": 12500
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.0012666398552181781,
      "loss": 1.3891,
      "step": 12600
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0012766941484013675,
      "loss": 1.3852,
      "step": 12700
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.0012867484415845566,
      "loss": 1.3856,
      "step": 12800
    },
    {
      "epoch": 2.59,
      "learning_rate": 0.0012968027347677459,
      "loss": 1.3816,
      "step": 12900
    },
    {
      "epoch": 2.61,
      "learning_rate": 0.001306857027950935,
      "loss": 1.3808,
      "step": 13000
    },
    {
      "epoch": 2.63,
      "learning_rate": 0.0013169113211341243,
      "loss": 1.3808,
      "step": 13100
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.0013269656143173134,
      "loss": 1.3802,
      "step": 13200
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0013370199075005028,
      "loss": 1.3781,
      "step": 13300
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.0013470742006836919,
      "loss": 1.3813,
      "step": 13400
    },
    {
      "epoch": 2.71,
      "learning_rate": 0.0013571284938668812,
      "loss": 1.3789,
      "step": 13500
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.0013671827870500705,
      "loss": 1.3808,
      "step": 13600
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.0013772370802332596,
      "loss": 1.3793,
      "step": 13700
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.001387291373416449,
      "loss": 1.3839,
      "step": 13800
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.001397345666599638,
      "loss": 1.3746,
      "step": 13900
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.0014072994168509952,
      "loss": 1.3719,
      "step": 14000
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0014173537100341845,
      "loss": 1.3761,
      "step": 14100
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.0014274080032173739,
      "loss": 1.3775,
      "step": 14200
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.0014374622964005632,
      "loss": 1.3781,
      "step": 14300
    },
    {
      "epoch": 2.9,
      "learning_rate": 0.0014475165895837523,
      "loss": 1.376,
      "step": 14400
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.0014575708827669414,
      "loss": 1.3759,
      "step": 14500
    },
    {
      "epoch": 2.94,
      "learning_rate": 0.0014676251759501307,
      "loss": 1.3709,
      "step": 14600
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.00147767946913332,
      "loss": 1.3724,
      "step": 14700
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0014877337623165092,
      "loss": 1.3792,
      "step": 14800
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0014977880554996985,
      "loss": 1.3705,
      "step": 14900
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3990531464846882,
      "eval_loss": 1.3305182456970215,
      "eval_runtime": 19.8107,
      "eval_samples_per_second": 4016.413,
      "eval_steps_per_second": 15.699,
      "step": 14919
    },
    {
      "epoch": 3.02,
      "learning_rate": 0.0015078423486828876,
      "loss": 1.3671,
      "step": 15000
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.001517896641866077,
      "loss": 1.3701,
      "step": 15100
    },
    {
      "epoch": 3.06,
      "learning_rate": 0.001527950935049266,
      "loss": 1.3686,
      "step": 15200
    },
    {
      "epoch": 3.08,
      "learning_rate": 0.0015380052282324553,
      "loss": 1.3682,
      "step": 15300
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.0015480595214156445,
      "loss": 1.3661,
      "step": 15400
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.0015581138145988338,
      "loss": 1.3605,
      "step": 15500
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.0015681681077820229,
      "loss": 1.3689,
      "step": 15600
    },
    {
      "epoch": 3.16,
      "learning_rate": 0.0015782224009652122,
      "loss": 1.3741,
      "step": 15700
    },
    {
      "epoch": 3.18,
      "learning_rate": 0.0015882766941484015,
      "loss": 1.3674,
      "step": 15800
    },
    {
      "epoch": 3.2,
      "learning_rate": 0.0015983309873315906,
      "loss": 1.367,
      "step": 15900
    },
    {
      "epoch": 3.22,
      "learning_rate": 0.0016083852805147797,
      "loss": 1.3694,
      "step": 16000
    },
    {
      "epoch": 3.24,
      "learning_rate": 0.001618439573697969,
      "loss": 1.3683,
      "step": 16100
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.0016284938668811584,
      "loss": 1.3669,
      "step": 16200
    },
    {
      "epoch": 3.28,
      "learning_rate": 0.0016385481600643475,
      "loss": 1.3713,
      "step": 16300
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.0016486024532475366,
      "loss": 1.3663,
      "step": 16400
    },
    {
      "epoch": 3.32,
      "learning_rate": 0.001658656746430726,
      "loss": 1.3708,
      "step": 16500
    },
    {
      "epoch": 3.34,
      "learning_rate": 0.0016687110396139153,
      "loss": 1.3693,
      "step": 16600
    },
    {
      "epoch": 3.36,
      "learning_rate": 0.0016787653327971046,
      "loss": 1.3685,
      "step": 16700
    },
    {
      "epoch": 3.38,
      "learning_rate": 0.0016888196259802935,
      "loss": 1.3692,
      "step": 16800
    },
    {
      "epoch": 3.4,
      "learning_rate": 0.0016988739191634828,
      "loss": 1.3682,
      "step": 16900
    },
    {
      "epoch": 3.42,
      "learning_rate": 0.0017089282123466721,
      "loss": 1.3657,
      "step": 17000
    },
    {
      "epoch": 3.44,
      "learning_rate": 0.0017189825055298614,
      "loss": 1.3685,
      "step": 17100
    },
    {
      "epoch": 3.46,
      "learning_rate": 0.0017290367987130505,
      "loss": 1.3683,
      "step": 17200
    },
    {
      "epoch": 3.48,
      "learning_rate": 0.0017390910918962397,
      "loss": 1.368,
      "step": 17300
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.001749145385079429,
      "loss": 1.368,
      "step": 17400
    },
    {
      "epoch": 3.52,
      "learning_rate": 0.0017591996782626183,
      "loss": 1.3668,
      "step": 17500
    },
    {
      "epoch": 3.54,
      "learning_rate": 0.0017692539714458076,
      "loss": 1.363,
      "step": 17600
    },
    {
      "epoch": 3.56,
      "learning_rate": 0.0017793082646289965,
      "loss": 1.3651,
      "step": 17700
    },
    {
      "epoch": 3.58,
      "learning_rate": 0.0017893625578121858,
      "loss": 1.363,
      "step": 17800
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.0017994168509953752,
      "loss": 1.3652,
      "step": 17900
    },
    {
      "epoch": 3.62,
      "learning_rate": 0.0018094711441785645,
      "loss": 1.3666,
      "step": 18000
    },
    {
      "epoch": 3.64,
      "learning_rate": 0.0018194248944299217,
      "loss": 1.3653,
      "step": 18100
    },
    {
      "epoch": 3.66,
      "learning_rate": 0.001829479187613111,
      "loss": 1.3685,
      "step": 18200
    },
    {
      "epoch": 3.68,
      "learning_rate": 0.0018395334807963,
      "loss": 1.3641,
      "step": 18300
    },
    {
      "epoch": 3.7,
      "learning_rate": 0.0018495877739794892,
      "loss": 1.3642,
      "step": 18400
    },
    {
      "epoch": 3.72,
      "learning_rate": 0.0018596420671626785,
      "loss": 1.3668,
      "step": 18500
    },
    {
      "epoch": 3.74,
      "learning_rate": 0.0018696963603458678,
      "loss": 1.3662,
      "step": 18600
    },
    {
      "epoch": 3.76,
      "learning_rate": 0.001879650110597225,
      "loss": 1.3659,
      "step": 18700
    },
    {
      "epoch": 3.78,
      "learning_rate": 0.0018897044037804143,
      "loss": 1.368,
      "step": 18800
    },
    {
      "epoch": 3.8,
      "learning_rate": 0.0018997586969636035,
      "loss": 1.3631,
      "step": 18900
    },
    {
      "epoch": 3.82,
      "learning_rate": 0.0019098129901467928,
      "loss": 1.3629,
      "step": 19000
    },
    {
      "epoch": 3.84,
      "learning_rate": 0.0019198672833299819,
      "loss": 1.3609,
      "step": 19100
    },
    {
      "epoch": 3.86,
      "learning_rate": 0.0019299215765131712,
      "loss": 1.368,
      "step": 19200
    },
    {
      "epoch": 3.88,
      "learning_rate": 0.0019399758696963603,
      "loss": 1.3663,
      "step": 19300
    },
    {
      "epoch": 3.9,
      "learning_rate": 0.0019500301628795496,
      "loss": 1.3684,
      "step": 19400
    },
    {
      "epoch": 3.92,
      "learning_rate": 0.0019600844560627387,
      "loss": 1.3653,
      "step": 19500
    },
    {
      "epoch": 3.94,
      "learning_rate": 0.001970138749245928,
      "loss": 1.3657,
      "step": 19600
    },
    {
      "epoch": 3.96,
      "learning_rate": 0.0019801930424291174,
      "loss": 1.3619,
      "step": 19700
    },
    {
      "epoch": 3.98,
      "learning_rate": 0.0019902473356123067,
      "loss": 1.3648,
      "step": 19800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.39993003357370305,
      "eval_loss": 1.3229814767837524,
      "eval_runtime": 19.7901,
      "eval_samples_per_second": 4020.592,
      "eval_steps_per_second": 15.715,
      "step": 19893
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.0020003016287954956,
      "loss": 1.3611,
      "step": 19900
    },
    {
      "epoch": 4.02,
      "learning_rate": 0.002010355921978685,
      "loss": 1.3592,
      "step": 20000
    },
    {
      "epoch": 4.04,
      "learning_rate": 0.0020204102151618743,
      "loss": 1.3543,
      "step": 20100
    },
    {
      "epoch": 4.06,
      "learning_rate": 0.0020304645083450636,
      "loss": 1.3601,
      "step": 20200
    },
    {
      "epoch": 4.08,
      "learning_rate": 0.0020405188015282525,
      "loss": 1.3539,
      "step": 20300
    },
    {
      "epoch": 4.1,
      "learning_rate": 0.002050573094711442,
      "loss": 1.3579,
      "step": 20400
    },
    {
      "epoch": 4.12,
      "learning_rate": 0.002060627387894631,
      "loss": 1.3596,
      "step": 20500
    },
    {
      "epoch": 4.14,
      "learning_rate": 0.0020706816810778204,
      "loss": 1.3603,
      "step": 20600
    },
    {
      "epoch": 4.16,
      "learning_rate": 0.0020807359742610093,
      "loss": 1.3594,
      "step": 20700
    },
    {
      "epoch": 4.18,
      "learning_rate": 0.0020907902674441987,
      "loss": 1.3609,
      "step": 20800
    },
    {
      "epoch": 4.2,
      "learning_rate": 0.002100844560627388,
      "loss": 1.3648,
      "step": 20900
    },
    {
      "epoch": 4.22,
      "learning_rate": 0.0021108988538105773,
      "loss": 1.3616,
      "step": 21000
    },
    {
      "epoch": 4.24,
      "learning_rate": 0.0021209531469937666,
      "loss": 1.3631,
      "step": 21100
    },
    {
      "epoch": 4.26,
      "learning_rate": 0.0021310074401769555,
      "loss": 1.3639,
      "step": 21200
    },
    {
      "epoch": 4.28,
      "learning_rate": 0.002140961190428313,
      "loss": 1.3646,
      "step": 21300
    },
    {
      "epoch": 4.3,
      "learning_rate": 0.0021510154836115025,
      "loss": 1.3593,
      "step": 21400
    },
    {
      "epoch": 4.32,
      "learning_rate": 0.0021610697767946913,
      "loss": 1.3604,
      "step": 21500
    },
    {
      "epoch": 4.34,
      "learning_rate": 0.0021711240699778807,
      "loss": 1.362,
      "step": 21600
    },
    {
      "epoch": 4.36,
      "learning_rate": 0.00218117836316107,
      "loss": 1.3583,
      "step": 21700
    },
    {
      "epoch": 4.38,
      "learning_rate": 0.0021912326563442593,
      "loss": 1.364,
      "step": 21800
    },
    {
      "epoch": 4.4,
      "learning_rate": 0.002201286949527448,
      "loss": 1.3625,
      "step": 21900
    },
    {
      "epoch": 4.42,
      "learning_rate": 0.0022113412427106375,
      "loss": 1.367,
      "step": 22000
    },
    {
      "epoch": 4.44,
      "learning_rate": 0.002221395535893827,
      "loss": 1.3646,
      "step": 22100
    },
    {
      "epoch": 4.46,
      "learning_rate": 0.002231449829077016,
      "loss": 1.3625,
      "step": 22200
    },
    {
      "epoch": 4.48,
      "learning_rate": 0.002241504122260205,
      "loss": 1.3617,
      "step": 22300
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.0022515584154433944,
      "loss": 1.3616,
      "step": 22400
    },
    {
      "epoch": 4.52,
      "learning_rate": 0.0022616127086265837,
      "loss": 1.3624,
      "step": 22500
    },
    {
      "epoch": 4.54,
      "learning_rate": 0.002271667001809773,
      "loss": 1.3635,
      "step": 22600
    },
    {
      "epoch": 4.56,
      "learning_rate": 0.002281721294992962,
      "loss": 1.36,
      "step": 22700
    },
    {
      "epoch": 4.58,
      "learning_rate": 0.0022917755881761512,
      "loss": 1.3599,
      "step": 22800
    },
    {
      "epoch": 4.6,
      "learning_rate": 0.0023018298813593406,
      "loss": 1.3674,
      "step": 22900
    },
    {
      "epoch": 4.62,
      "learning_rate": 0.00231188417454253,
      "loss": 1.3667,
      "step": 23000
    },
    {
      "epoch": 4.64,
      "learning_rate": 0.002321938467725719,
      "loss": 1.3555,
      "step": 23100
    },
    {
      "epoch": 4.66,
      "learning_rate": 0.002331992760908908,
      "loss": 1.3594,
      "step": 23200
    },
    {
      "epoch": 4.69,
      "learning_rate": 0.0023420470540920974,
      "loss": 1.3629,
      "step": 23300
    },
    {
      "epoch": 4.71,
      "learning_rate": 0.0023521013472752868,
      "loss": 1.3641,
      "step": 23400
    },
    {
      "epoch": 4.73,
      "learning_rate": 0.0023621556404584756,
      "loss": 1.3594,
      "step": 23500
    },
    {
      "epoch": 4.75,
      "learning_rate": 0.002372209933641665,
      "loss": 1.3594,
      "step": 23600
    },
    {
      "epoch": 4.77,
      "learning_rate": 0.0023821636838930226,
      "loss": 1.3639,
      "step": 23700
    },
    {
      "epoch": 4.79,
      "learning_rate": 0.0023922179770762115,
      "loss": 1.3609,
      "step": 23800
    },
    {
      "epoch": 4.81,
      "learning_rate": 0.002402272270259401,
      "loss": 1.3605,
      "step": 23900
    },
    {
      "epoch": 4.83,
      "learning_rate": 0.00241232656344259,
      "loss": 1.3608,
      "step": 24000
    },
    {
      "epoch": 4.85,
      "learning_rate": 0.0024223808566257794,
      "loss": 1.3646,
      "step": 24100
    },
    {
      "epoch": 4.87,
      "learning_rate": 0.0024324351498089683,
      "loss": 1.3599,
      "step": 24200
    },
    {
      "epoch": 4.89,
      "learning_rate": 0.0024424894429921577,
      "loss": 1.3586,
      "step": 24300
    },
    {
      "epoch": 4.91,
      "learning_rate": 0.002452543736175347,
      "loss": 1.3608,
      "step": 24400
    },
    {
      "epoch": 4.93,
      "learning_rate": 0.0024625980293585363,
      "loss": 1.3615,
      "step": 24500
    },
    {
      "epoch": 4.95,
      "learning_rate": 0.002472652322541725,
      "loss": 1.3629,
      "step": 24600
    },
    {
      "epoch": 4.97,
      "learning_rate": 0.0024827066157249145,
      "loss": 1.3638,
      "step": 24700
    },
    {
      "epoch": 4.99,
      "learning_rate": 0.002492760908908104,
      "loss": 1.3652,
      "step": 24800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.39963747273250233,
      "eval_loss": 1.3215824365615845,
      "eval_runtime": 19.5099,
      "eval_samples_per_second": 4078.343,
      "eval_steps_per_second": 15.941,
      "step": 24866
    },
    {
      "epoch": 5.01,
      "learning_rate": 0.002499999903431858,
      "loss": 1.3594,
      "step": 24900
    },
    {
      "epoch": 5.03,
      "learning_rate": 0.0024999979819234353,
      "loss": 1.3548,
      "step": 25000
    },
    {
      "epoch": 5.05,
      "learning_rate": 0.0024999935969463615,
      "loss": 1.3572,
      "step": 25100
    },
    {
      "epoch": 5.07,
      "learning_rate": 0.0024999867485092793,
      "loss": 1.3579,
      "step": 25200
    },
    {
      "epoch": 5.09,
      "learning_rate": 0.002499977436625685,
      "loss": 1.3553,
      "step": 25300
    },
    {
      "epoch": 5.11,
      "learning_rate": 0.0024999656613139305,
      "loss": 1.3558,
      "step": 25400
    },
    {
      "epoch": 5.13,
      "learning_rate": 0.0024999514225972227,
      "loss": 1.3556,
      "step": 25500
    },
    {
      "epoch": 5.15,
      "learning_rate": 0.002499934720503622,
      "loss": 1.356,
      "step": 25600
    },
    {
      "epoch": 5.17,
      "learning_rate": 0.0024999155550660458,
      "loss": 1.3537,
      "step": 25700
    },
    {
      "epoch": 5.19,
      "learning_rate": 0.002499893926322264,
      "loss": 1.3582,
      "step": 25800
    },
    {
      "epoch": 5.21,
      "learning_rate": 0.0024998698343149024,
      "loss": 1.3528,
      "step": 25900
    },
    {
      "epoch": 5.23,
      "learning_rate": 0.0024998432790914404,
      "loss": 1.3531,
      "step": 26000
    },
    {
      "epoch": 5.25,
      "learning_rate": 0.0024998145630805626,
      "loss": 1.3578,
      "step": 26100
    },
    {
      "epoch": 5.27,
      "learning_rate": 0.002499783106217526,
      "loss": 1.3548,
      "step": 26200
    },
    {
      "epoch": 5.29,
      "learning_rate": 0.0024997491863093103,
      "loss": 1.3516,
      "step": 26300
    },
    {
      "epoch": 5.31,
      "learning_rate": 0.002499712803422766,
      "loss": 1.3549,
      "step": 26400
    },
    {
      "epoch": 5.33,
      "learning_rate": 0.0024996739576295945,
      "loss": 1.3539,
      "step": 26500
    },
    {
      "epoch": 5.35,
      "learning_rate": 0.0024996326490063525,
      "loss": 1.356,
      "step": 26600
    },
    {
      "epoch": 5.37,
      "learning_rate": 0.0024995888776344504,
      "loss": 1.3513,
      "step": 26700
    },
    {
      "epoch": 5.39,
      "learning_rate": 0.002499542643600152,
      "loss": 1.354,
      "step": 26800
    },
    {
      "epoch": 5.41,
      "learning_rate": 0.0024994939469945737,
      "loss": 1.3535,
      "step": 26900
    },
    {
      "epoch": 5.43,
      "learning_rate": 0.0024994427879136854,
      "loss": 1.3483,
      "step": 27000
    },
    {
      "epoch": 5.45,
      "learning_rate": 0.002499389166458312,
      "loss": 1.354,
      "step": 27100
    },
    {
      "epoch": 5.47,
      "learning_rate": 0.0024993330827341276,
      "loss": 1.3512,
      "step": 27200
    },
    {
      "epoch": 5.49,
      "learning_rate": 0.0024992745368516618,
      "loss": 1.3492,
      "step": 27300
    },
    {
      "epoch": 5.51,
      "learning_rate": 0.0024992135289262953,
      "loss": 1.3521,
      "step": 27400
    },
    {
      "epoch": 5.53,
      "learning_rate": 0.002499150059078261,
      "loss": 1.3494,
      "step": 27500
    },
    {
      "epoch": 5.55,
      "learning_rate": 0.0024990841274326442,
      "loss": 1.3497,
      "step": 27600
    },
    {
      "epoch": 5.57,
      "learning_rate": 0.0024990157341193814,
      "loss": 1.3494,
      "step": 27700
    },
    {
      "epoch": 5.59,
      "learning_rate": 0.0024989448792732604,
      "loss": 1.3487,
      "step": 27800
    },
    {
      "epoch": 5.61,
      "learning_rate": 0.0024988715630339213,
      "loss": 1.3468,
      "step": 27900
    },
    {
      "epoch": 5.63,
      "learning_rate": 0.002498795785545853,
      "loss": 1.3465,
      "step": 28000
    },
    {
      "epoch": 5.65,
      "learning_rate": 0.002498718341526211,
      "loss": 1.3505,
      "step": 28100
    },
    {
      "epoch": 5.67,
      "learning_rate": 0.00249863766660223,
      "loss": 1.3512,
      "step": 28200
    },
    {
      "epoch": 5.69,
      "learning_rate": 0.0024985545308904788,
      "loss": 1.3497,
      "step": 28300
    },
    {
      "epoch": 5.71,
      "learning_rate": 0.0024984689345547983,
      "loss": 1.3495,
      "step": 28400
    },
    {
      "epoch": 5.73,
      "learning_rate": 0.002498380877763881,
      "loss": 1.3484,
      "step": 28500
    },
    {
      "epoch": 5.75,
      "learning_rate": 0.0024982903606912666,
      "loss": 1.3465,
      "step": 28600
    },
    {
      "epoch": 5.77,
      "learning_rate": 0.0024981973835153442,
      "loss": 1.3494,
      "step": 28700
    },
    {
      "epoch": 5.79,
      "learning_rate": 0.0024981019464193513,
      "loss": 1.3448,
      "step": 28800
    },
    {
      "epoch": 5.81,
      "learning_rate": 0.002498004049591373,
      "loss": 1.3447,
      "step": 28900
    },
    {
      "epoch": 5.83,
      "learning_rate": 0.002497903693224343,
      "loss": 1.3489,
      "step": 29000
    },
    {
      "epoch": 5.85,
      "learning_rate": 0.0024978008775160404,
      "loss": 1.3491,
      "step": 29100
    },
    {
      "epoch": 5.87,
      "learning_rate": 0.002497695602669093,
      "loss": 1.3511,
      "step": 29200
    },
    {
      "epoch": 5.89,
      "learning_rate": 0.002497587868890974,
      "loss": 1.344,
      "step": 29300
    },
    {
      "epoch": 5.91,
      "learning_rate": 0.002497477676394002,
      "loss": 1.3442,
      "step": 29400
    },
    {
      "epoch": 5.93,
      "learning_rate": 0.002497365025395343,
      "loss": 1.3518,
      "step": 29500
    },
    {
      "epoch": 5.95,
      "learning_rate": 0.0024972499161170065,
      "loss": 1.3396,
      "step": 29600
    },
    {
      "epoch": 5.97,
      "learning_rate": 0.0024971323487858474,
      "loss": 1.3427,
      "step": 29700
    },
    {
      "epoch": 5.99,
      "learning_rate": 0.002497012323633565,
      "loss": 1.3434,
      "step": 29800
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4013078853657854,
      "eval_loss": 1.3029085397720337,
      "eval_runtime": 20.0428,
      "eval_samples_per_second": 3969.911,
      "eval_steps_per_second": 15.517,
      "step": 29839
    },
    {
      "epoch": 6.01,
      "learning_rate": 0.002496889840896702,
      "loss": 1.3447,
      "step": 29900
    },
    {
      "epoch": 6.03,
      "learning_rate": 0.0024967649008166455,
      "loss": 1.3366,
      "step": 30000
    },
    {
      "epoch": 6.05,
      "learning_rate": 0.0024966375036396234,
      "loss": 1.3399,
      "step": 30100
    },
    {
      "epoch": 6.07,
      "learning_rate": 0.0024965076496167083,
      "loss": 1.3375,
      "step": 30200
    },
    {
      "epoch": 6.09,
      "learning_rate": 0.00249637667426921,
      "loss": 1.3349,
      "step": 30300
    },
    {
      "epoch": 6.11,
      "learning_rate": 0.0024962419318890756,
      "loss": 1.3401,
      "step": 30400
    },
    {
      "epoch": 6.13,
      "learning_rate": 0.0024961047334426316,
      "loss": 1.339,
      "step": 30500
    },
    {
      "epoch": 6.15,
      "learning_rate": 0.0024959650792002663,
      "loss": 1.3457,
      "step": 30600
    },
    {
      "epoch": 6.17,
      "learning_rate": 0.0024958229694372063,
      "loss": 1.34,
      "step": 30700
    },
    {
      "epoch": 6.19,
      "learning_rate": 0.00249567840443352,
      "loss": 1.3389,
      "step": 30800
    },
    {
      "epoch": 6.21,
      "learning_rate": 0.0024955313844741115,
      "loss": 1.3385,
      "step": 30900
    },
    {
      "epoch": 6.23,
      "learning_rate": 0.002495381909848725,
      "loss": 1.3409,
      "step": 31000
    },
    {
      "epoch": 6.25,
      "learning_rate": 0.002495229980851942,
      "loss": 1.3394,
      "step": 31100
    },
    {
      "epoch": 6.27,
      "learning_rate": 0.002495075597783181,
      "loss": 1.3392,
      "step": 31200
    },
    {
      "epoch": 6.29,
      "learning_rate": 0.0024949187609466963,
      "loss": 1.3401,
      "step": 31300
    },
    {
      "epoch": 6.31,
      "learning_rate": 0.002494759470651578,
      "loss": 1.3375,
      "step": 31400
    },
    {
      "epoch": 6.33,
      "learning_rate": 0.0024945977272117534,
      "loss": 1.342,
      "step": 31500
    },
    {
      "epoch": 6.35,
      "learning_rate": 0.002494433530945982,
      "loss": 1.335,
      "step": 31600
    },
    {
      "epoch": 6.37,
      "learning_rate": 0.002494266882177858,
      "loss": 1.3347,
      "step": 31700
    },
    {
      "epoch": 6.39,
      "learning_rate": 0.0024940977812358094,
      "loss": 1.3392,
      "step": 31800
    },
    {
      "epoch": 6.41,
      "learning_rate": 0.002493926228453097,
      "loss": 1.3365,
      "step": 31900
    },
    {
      "epoch": 6.43,
      "learning_rate": 0.0024937522241678133,
      "loss": 1.3374,
      "step": 32000
    },
    {
      "epoch": 6.45,
      "learning_rate": 0.0024935757687228814,
      "loss": 1.3372,
      "step": 32100
    },
    {
      "epoch": 6.47,
      "learning_rate": 0.002493396862466056,
      "loss": 1.3375,
      "step": 32200
    },
    {
      "epoch": 6.49,
      "learning_rate": 0.0024932173314456876,
      "loss": 1.3388,
      "step": 32300
    },
    {
      "epoch": 6.51,
      "learning_rate": 0.002493033549126893,
      "loss": 1.3394,
      "step": 32400
    },
    {
      "epoch": 6.53,
      "learning_rate": 0.002492847317064799,
      "loss": 1.3353,
      "step": 32500
    },
    {
      "epoch": 6.56,
      "learning_rate": 0.002492658635626429,
      "loss": 1.334,
      "step": 32600
    },
    {
      "epoch": 6.58,
      "learning_rate": 0.0024924675051836314,
      "loss": 1.3385,
      "step": 32700
    },
    {
      "epoch": 6.6,
      "learning_rate": 0.0024922739261130818,
      "loss": 1.3378,
      "step": 32800
    },
    {
      "epoch": 6.62,
      "learning_rate": 0.002492077898796282,
      "loss": 1.3403,
      "step": 32900
    },
    {
      "epoch": 6.64,
      "learning_rate": 0.002491879423619558,
      "loss": 1.3408,
      "step": 33000
    },
    {
      "epoch": 6.66,
      "learning_rate": 0.002491678500974061,
      "loss": 1.3345,
      "step": 33100
    },
    {
      "epoch": 6.68,
      "learning_rate": 0.002491475131255764,
      "loss": 1.3331,
      "step": 33200
    },
    {
      "epoch": 6.7,
      "learning_rate": 0.002491269314865464,
      "loss": 1.3382,
      "step": 33300
    },
    {
      "epoch": 6.72,
      "learning_rate": 0.00249106105220878,
      "loss": 1.3372,
      "step": 33400
    },
    {
      "epoch": 6.74,
      "learning_rate": 0.0024908503436961503,
      "loss": 1.3308,
      "step": 33500
    },
    {
      "epoch": 6.76,
      "learning_rate": 0.002490637189742836,
      "loss": 1.3346,
      "step": 33600
    },
    {
      "epoch": 6.78,
      "learning_rate": 0.002490421590768915,
      "loss": 1.3356,
      "step": 33700
    },
    {
      "epoch": 6.8,
      "learning_rate": 0.0024902035471992857,
      "loss": 1.3314,
      "step": 33800
    },
    {
      "epoch": 6.82,
      "learning_rate": 0.002489983059463664,
      "loss": 1.3342,
      "step": 33900
    },
    {
      "epoch": 6.84,
      "learning_rate": 0.002489760127996581,
      "loss": 1.3349,
      "step": 34000
    },
    {
      "epoch": 6.86,
      "learning_rate": 0.0024895347532373864,
      "loss": 1.336,
      "step": 34100
    },
    {
      "epoch": 6.88,
      "learning_rate": 0.002489306935630243,
      "loss": 1.3345,
      "step": 34200
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.0024890766756241293,
      "loss": 1.3351,
      "step": 34300
    },
    {
      "epoch": 6.92,
      "learning_rate": 0.0024888463127784766,
      "loss": 1.3308,
      "step": 34400
    },
    {
      "epoch": 6.94,
      "learning_rate": 0.0024886111937531884,
      "loss": 1.3369,
      "step": 34500
    },
    {
      "epoch": 6.96,
      "learning_rate": 0.0024883736337000827,
      "loss": 1.3351,
      "step": 34600
    },
    {
      "epoch": 6.98,
      "learning_rate": 0.0024881336330873373,
      "loss": 1.333,
      "step": 34700
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.0024878911923879405,
      "loss": 1.3314,
      "step": 34800
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4019466193393693,
      "eval_loss": 1.2937954664230347,
      "eval_runtime": 19.8606,
      "eval_samples_per_second": 4006.315,
      "eval_steps_per_second": 15.659,
      "step": 34812
    },
    {
      "epoch": 7.02,
      "learning_rate": 0.0024876463120796894,
      "loss": 1.3277,
      "step": 34900
    },
    {
      "epoch": 7.04,
      "learning_rate": 0.002487398992645188,
      "loss": 1.3304,
      "step": 35000
    },
    {
      "epoch": 7.06,
      "learning_rate": 0.002487149234571848,
      "loss": 1.327,
      "step": 35100
    },
    {
      "epoch": 7.08,
      "learning_rate": 0.002486897038351888,
      "loss": 1.3317,
      "step": 35200
    },
    {
      "epoch": 7.1,
      "learning_rate": 0.0024866424044823303,
      "loss": 1.3264,
      "step": 35300
    },
    {
      "epoch": 7.12,
      "learning_rate": 0.0024863853334650015,
      "loss": 1.3287,
      "step": 35400
    },
    {
      "epoch": 7.14,
      "learning_rate": 0.002486125825806532,
      "loss": 1.3266,
      "step": 35500
    },
    {
      "epoch": 7.16,
      "learning_rate": 0.002485863882018354,
      "loss": 1.3311,
      "step": 35600
    },
    {
      "epoch": 7.18,
      "learning_rate": 0.0024855995026166994,
      "loss": 1.3303,
      "step": 35700
    },
    {
      "epoch": 7.2,
      "learning_rate": 0.0024853326881226026,
      "loss": 1.328,
      "step": 35800
    },
    {
      "epoch": 7.22,
      "learning_rate": 0.002485063439061895,
      "loss": 1.3339,
      "step": 35900
    },
    {
      "epoch": 7.24,
      "learning_rate": 0.0024847917559652067,
      "loss": 1.3296,
      "step": 36000
    },
    {
      "epoch": 7.26,
      "learning_rate": 0.002484517639367966,
      "loss": 1.3294,
      "step": 36100
    },
    {
      "epoch": 7.28,
      "learning_rate": 0.0024842410898103947,
      "loss": 1.3311,
      "step": 36200
    },
    {
      "epoch": 7.3,
      "learning_rate": 0.0024839621078375103,
      "loss": 1.3282,
      "step": 36300
    },
    {
      "epoch": 7.32,
      "learning_rate": 0.002483680693999126,
      "loss": 1.3248,
      "step": 36400
    },
    {
      "epoch": 7.34,
      "learning_rate": 0.0024833996993344934,
      "loss": 1.3279,
      "step": 36500
    },
    {
      "epoch": 7.36,
      "learning_rate": 0.002483113447738441,
      "loss": 1.3267,
      "step": 36600
    },
    {
      "epoch": 7.38,
      "learning_rate": 0.0024828247659494087,
      "loss": 1.3265,
      "step": 36700
    },
    {
      "epoch": 7.4,
      "learning_rate": 0.0024825336545363243,
      "loss": 1.3287,
      "step": 36800
    },
    {
      "epoch": 7.42,
      "learning_rate": 0.0024822401140729027,
      "loss": 1.3274,
      "step": 36900
    },
    {
      "epoch": 7.44,
      "learning_rate": 0.0024819441451376482,
      "loss": 1.3317,
      "step": 37000
    },
    {
      "epoch": 7.46,
      "learning_rate": 0.00248164574831385,
      "loss": 1.3274,
      "step": 37100
    },
    {
      "epoch": 7.48,
      "learning_rate": 0.002481344924189581,
      "loss": 1.3275,
      "step": 37200
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.0024810416733576997,
      "loss": 1.3314,
      "step": 37300
    },
    {
      "epoch": 7.52,
      "learning_rate": 0.002480735996415845,
      "loss": 1.3308,
      "step": 37400
    },
    {
      "epoch": 7.54,
      "learning_rate": 0.00248042789396644,
      "loss": 1.3268,
      "step": 37500
    },
    {
      "epoch": 7.56,
      "learning_rate": 0.002480117366616685,
      "loss": 1.3238,
      "step": 37600
    },
    {
      "epoch": 7.58,
      "learning_rate": 0.002479804414978561,
      "loss": 1.3264,
      "step": 37700
    },
    {
      "epoch": 7.6,
      "learning_rate": 0.0024794890396688256,
      "loss": 1.3302,
      "step": 37800
    },
    {
      "epoch": 7.62,
      "learning_rate": 0.002479171241309015,
      "loss": 1.3292,
      "step": 37900
    },
    {
      "epoch": 7.64,
      "learning_rate": 0.002478851020525438,
      "loss": 1.3292,
      "step": 38000
    },
    {
      "epoch": 7.66,
      "learning_rate": 0.0024785283779491787,
      "loss": 1.3286,
      "step": 38100
    },
    {
      "epoch": 7.68,
      "learning_rate": 0.0024782033142160946,
      "loss": 1.3253,
      "step": 38200
    },
    {
      "epoch": 7.7,
      "learning_rate": 0.002477875829966814,
      "loss": 1.3284,
      "step": 38300
    },
    {
      "epoch": 7.72,
      "learning_rate": 0.0024775459258467355,
      "loss": 1.3302,
      "step": 38400
    },
    {
      "epoch": 7.74,
      "learning_rate": 0.00247721693771243,
      "loss": 1.3224,
      "step": 38500
    },
    {
      "epoch": 7.76,
      "learning_rate": 0.0024768822199884265,
      "loss": 1.3232,
      "step": 38600
    },
    {
      "epoch": 7.78,
      "learning_rate": 0.00247654508435181,
      "loss": 1.3267,
      "step": 38700
    },
    {
      "epoch": 7.8,
      "learning_rate": 0.002476205531467,
      "loss": 1.3268,
      "step": 38800
    },
    {
      "epoch": 7.82,
      "learning_rate": 0.0024758635620031806,
      "loss": 1.319,
      "step": 38900
    },
    {
      "epoch": 7.84,
      "learning_rate": 0.0024755191766342964,
      "loss": 1.3272,
      "step": 39000
    },
    {
      "epoch": 7.86,
      "learning_rate": 0.0024751723760390552,
      "loss": 1.3269,
      "step": 39100
    },
    {
      "epoch": 7.88,
      "learning_rate": 0.0024748231609009247,
      "loss": 1.3272,
      "step": 39200
    },
    {
      "epoch": 7.9,
      "learning_rate": 0.0024744715319081293,
      "loss": 1.3293,
      "step": 39300
    },
    {
      "epoch": 7.92,
      "learning_rate": 0.002474117489753652,
      "loss": 1.3235,
      "step": 39400
    },
    {
      "epoch": 7.94,
      "learning_rate": 0.002473761035135232,
      "loss": 1.3279,
      "step": 39500
    },
    {
      "epoch": 7.96,
      "learning_rate": 0.0024734021687553617,
      "loss": 1.3226,
      "step": 39600
    },
    {
      "epoch": 7.98,
      "learning_rate": 0.002473040891321287,
      "loss": 1.3244,
      "step": 39700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4026586923819002,
      "eval_loss": 1.2859399318695068,
      "eval_runtime": 19.8804,
      "eval_samples_per_second": 4002.332,
      "eval_steps_per_second": 15.644,
      "step": 39786
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.002472677203545006,
      "loss": 1.3265,
      "step": 39800
    },
    {
      "epoch": 8.02,
      "learning_rate": 0.002472311106143266,
      "loss": 1.3222,
      "step": 39900
    },
    {
      "epoch": 8.04,
      "learning_rate": 0.0024719425998375646,
      "loss": 1.3229,
      "step": 40000
    },
    {
      "epoch": 8.06,
      "learning_rate": 0.002471571685354145,
      "loss": 1.3203,
      "step": 40100
    },
    {
      "epoch": 8.08,
      "learning_rate": 0.002471198363423998,
      "loss": 1.324,
      "step": 40200
    },
    {
      "epoch": 8.1,
      "learning_rate": 0.002470822634782858,
      "loss": 1.3205,
      "step": 40300
    },
    {
      "epoch": 8.12,
      "learning_rate": 0.0024704445001712027,
      "loss": 1.3261,
      "step": 40400
    },
    {
      "epoch": 8.14,
      "learning_rate": 0.002470067777636028,
      "loss": 1.3198,
      "step": 40500
    },
    {
      "epoch": 8.16,
      "learning_rate": 0.0024696848573647666,
      "loss": 1.319,
      "step": 40600
    },
    {
      "epoch": 8.18,
      "learning_rate": 0.0024692995333652967,
      "loss": 1.3251,
      "step": 40700
    },
    {
      "epoch": 8.2,
      "learning_rate": 0.0024689118063970067,
      "loss": 1.3224,
      "step": 40800
    },
    {
      "epoch": 8.22,
      "learning_rate": 0.0024685216772240203,
      "loss": 1.3196,
      "step": 40900
    },
    {
      "epoch": 8.24,
      "learning_rate": 0.0024681291466151956,
      "loss": 1.3172,
      "step": 41000
    },
    {
      "epoch": 8.26,
      "learning_rate": 0.0024677342153441232,
      "loss": 1.3211,
      "step": 41100
    },
    {
      "epoch": 8.28,
      "learning_rate": 0.002467336884189125,
      "loss": 1.3166,
      "step": 41200
    },
    {
      "epoch": 8.3,
      "learning_rate": 0.0024669371539332526,
      "loss": 1.3225,
      "step": 41300
    },
    {
      "epoch": 8.32,
      "learning_rate": 0.0024665350253642855,
      "loss": 1.3187,
      "step": 41400
    },
    {
      "epoch": 8.34,
      "learning_rate": 0.00246613049927473,
      "loss": 1.3188,
      "step": 41500
    },
    {
      "epoch": 8.36,
      "learning_rate": 0.002465723576461818,
      "loss": 1.3194,
      "step": 41600
    },
    {
      "epoch": 8.38,
      "learning_rate": 0.0024653142577275022,
      "loss": 1.322,
      "step": 41700
    },
    {
      "epoch": 8.4,
      "learning_rate": 0.002464902543878461,
      "loss": 1.3192,
      "step": 41800
    },
    {
      "epoch": 8.43,
      "learning_rate": 0.00246448843572609,
      "loss": 1.3229,
      "step": 41900
    },
    {
      "epoch": 8.45,
      "learning_rate": 0.0024640719340865043,
      "loss": 1.3235,
      "step": 42000
    },
    {
      "epoch": 8.47,
      "learning_rate": 0.0024636530397805366,
      "loss": 1.3213,
      "step": 42100
    },
    {
      "epoch": 8.49,
      "learning_rate": 0.002463231753633735,
      "loss": 1.3187,
      "step": 42200
    },
    {
      "epoch": 8.51,
      "learning_rate": 0.0024628080764763603,
      "loss": 1.3226,
      "step": 42300
    },
    {
      "epoch": 8.53,
      "learning_rate": 0.002462382009143387,
      "loss": 1.3192,
      "step": 42400
    },
    {
      "epoch": 8.55,
      "learning_rate": 0.002461953552474498,
      "loss": 1.3222,
      "step": 42500
    },
    {
      "epoch": 8.57,
      "learning_rate": 0.0024615270275859403,
      "loss": 1.3172,
      "step": 42600
    },
    {
      "epoch": 8.59,
      "learning_rate": 0.002461093818655314,
      "loss": 1.3191,
      "step": 42700
    },
    {
      "epoch": 8.61,
      "learning_rate": 0.0024606582229275095,
      "loss": 1.3203,
      "step": 42800
    },
    {
      "epoch": 8.63,
      "learning_rate": 0.0024602202412609907,
      "loss": 1.3171,
      "step": 42900
    },
    {
      "epoch": 8.65,
      "learning_rate": 0.0024597798745189217,
      "loss": 1.3239,
      "step": 43000
    },
    {
      "epoch": 8.67,
      "learning_rate": 0.0024593371235691673,
      "loss": 1.3189,
      "step": 43100
    },
    {
      "epoch": 8.69,
      "learning_rate": 0.0024588919892842924,
      "loss": 1.3168,
      "step": 43200
    },
    {
      "epoch": 8.71,
      "learning_rate": 0.002458444472541557,
      "loss": 1.3214,
      "step": 43300
    },
    {
      "epoch": 8.73,
      "learning_rate": 0.0024579945742229177,
      "loss": 1.321,
      "step": 43400
    },
    {
      "epoch": 8.75,
      "learning_rate": 0.0024575422952150235,
      "loss": 1.3202,
      "step": 43500
    },
    {
      "epoch": 8.77,
      "learning_rate": 0.002457087636409217,
      "loss": 1.3188,
      "step": 43600
    },
    {
      "epoch": 8.79,
      "learning_rate": 0.0024566305987015298,
      "loss": 1.3166,
      "step": 43700
    },
    {
      "epoch": 8.81,
      "learning_rate": 0.002456171182992681,
      "loss": 1.3244,
      "step": 43800
    },
    {
      "epoch": 8.83,
      "learning_rate": 0.002455709390188077,
      "loss": 1.3186,
      "step": 43900
    },
    {
      "epoch": 8.85,
      "learning_rate": 0.00245524522119781,
      "loss": 1.3191,
      "step": 44000
    },
    {
      "epoch": 8.87,
      "learning_rate": 0.002454778676936653,
      "loss": 1.3218,
      "step": 44100
    },
    {
      "epoch": 8.89,
      "learning_rate": 0.0024543097583240615,
      "loss": 1.3217,
      "step": 44200
    },
    {
      "epoch": 8.91,
      "learning_rate": 0.0024538384662841704,
      "loss": 1.3169,
      "step": 44300
    },
    {
      "epoch": 8.93,
      "learning_rate": 0.0024533648017457917,
      "loss": 1.3201,
      "step": 44400
    },
    {
      "epoch": 8.95,
      "learning_rate": 0.002452888765642413,
      "loss": 1.3189,
      "step": 44500
    },
    {
      "epoch": 8.97,
      "learning_rate": 0.0024524103589121955,
      "loss": 1.3206,
      "step": 44600
    },
    {
      "epoch": 8.99,
      "learning_rate": 0.002451934401988944,
      "loss": 1.3163,
      "step": 44700
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4035565789440192,
      "eval_loss": 1.2766544818878174,
      "eval_runtime": 19.7806,
      "eval_samples_per_second": 4022.534,
      "eval_steps_per_second": 15.723,
      "step": 44759
    },
    {
      "epoch": 9.01,
      "learning_rate": 0.0024514512805208794,
      "loss": 1.3158,
      "step": 44800
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0024509657912589406,
      "loss": 1.3109,
      "step": 44900
    },
    {
      "epoch": 9.05,
      "learning_rate": 0.0024504779351599195,
      "loss": 1.312,
      "step": 45000
    },
    {
      "epoch": 9.07,
      "learning_rate": 0.002449987713185271,
      "loss": 1.3134,
      "step": 45100
    },
    {
      "epoch": 9.09,
      "learning_rate": 0.002449495126301115,
      "loss": 1.3156,
      "step": 45200
    },
    {
      "epoch": 9.11,
      "learning_rate": 0.002449000175478231,
      "loss": 1.3188,
      "step": 45300
    },
    {
      "epoch": 9.13,
      "learning_rate": 0.0024485028616920562,
      "loss": 1.3159,
      "step": 45400
    },
    {
      "epoch": 9.15,
      "learning_rate": 0.0024480031859226863,
      "loss": 1.3154,
      "step": 45500
    },
    {
      "epoch": 9.17,
      "learning_rate": 0.0024475011491548715,
      "loss": 1.3151,
      "step": 45600
    },
    {
      "epoch": 9.19,
      "learning_rate": 0.002446996752378015,
      "loss": 1.3152,
      "step": 45700
    },
    {
      "epoch": 9.21,
      "learning_rate": 0.0024464899965861704,
      "loss": 1.3159,
      "step": 45800
    },
    {
      "epoch": 9.23,
      "learning_rate": 0.002445980882778041,
      "loss": 1.3189,
      "step": 45900
    },
    {
      "epoch": 9.25,
      "learning_rate": 0.0024454694119569777,
      "loss": 1.3132,
      "step": 46000
    },
    {
      "epoch": 9.27,
      "learning_rate": 0.0024449555851309753,
      "loss": 1.3136,
      "step": 46100
    },
    {
      "epoch": 9.29,
      "learning_rate": 0.0024444394033126733,
      "loss": 1.3154,
      "step": 46200
    },
    {
      "epoch": 9.31,
      "learning_rate": 0.0024439208675193502,
      "loss": 1.3148,
      "step": 46300
    },
    {
      "epoch": 9.33,
      "learning_rate": 0.0024433999787729266,
      "loss": 1.3119,
      "step": 46400
    },
    {
      "epoch": 9.35,
      "learning_rate": 0.002442876738099957,
      "loss": 1.3143,
      "step": 46500
    },
    {
      "epoch": 9.37,
      "learning_rate": 0.002442351146531634,
      "loss": 1.316,
      "step": 46600
    },
    {
      "epoch": 9.39,
      "learning_rate": 0.0024418284961464525,
      "loss": 1.3157,
      "step": 46700
    },
    {
      "epoch": 9.41,
      "learning_rate": 0.002441298229382551,
      "loss": 1.3142,
      "step": 46800
    },
    {
      "epoch": 9.43,
      "learning_rate": 0.002440765614834186,
      "loss": 1.3125,
      "step": 46900
    },
    {
      "epoch": 9.45,
      "learning_rate": 0.0024402306535510226,
      "loss": 1.3187,
      "step": 47000
    },
    {
      "epoch": 9.47,
      "learning_rate": 0.0024396933465873506,
      "loss": 1.316,
      "step": 47100
    },
    {
      "epoch": 9.49,
      "learning_rate": 0.002439153695002083,
      "loss": 1.3129,
      "step": 47200
    },
    {
      "epoch": 9.51,
      "learning_rate": 0.0024386116998587537,
      "loss": 1.3164,
      "step": 47300
    },
    {
      "epoch": 9.53,
      "learning_rate": 0.002438067362225514,
      "loss": 1.3156,
      "step": 47400
    },
    {
      "epoch": 9.55,
      "learning_rate": 0.0024375206831751335,
      "loss": 1.3151,
      "step": 47500
    },
    {
      "epoch": 9.57,
      "learning_rate": 0.0024369716637849944,
      "loss": 1.3131,
      "step": 47600
    },
    {
      "epoch": 9.59,
      "learning_rate": 0.0024364203051370926,
      "loss": 1.3154,
      "step": 47700
    },
    {
      "epoch": 9.61,
      "learning_rate": 0.0024358666083180336,
      "loss": 1.3141,
      "step": 47800
    },
    {
      "epoch": 9.63,
      "learning_rate": 0.0024353105744190314,
      "loss": 1.3131,
      "step": 47900
    },
    {
      "epoch": 9.65,
      "learning_rate": 0.0024347522045359044,
      "loss": 1.3147,
      "step": 48000
    },
    {
      "epoch": 9.67,
      "learning_rate": 0.0024341914997690764,
      "loss": 1.3159,
      "step": 48100
    },
    {
      "epoch": 9.69,
      "learning_rate": 0.002433628461223572,
      "loss": 1.3135,
      "step": 48200
    },
    {
      "epoch": 9.71,
      "learning_rate": 0.0024330630900090155,
      "loss": 1.3133,
      "step": 48300
    },
    {
      "epoch": 9.73,
      "learning_rate": 0.0024324953872396277,
      "loss": 1.3126,
      "step": 48400
    },
    {
      "epoch": 9.75,
      "learning_rate": 0.0024319253540342253,
      "loss": 1.3144,
      "step": 48500
    },
    {
      "epoch": 9.77,
      "learning_rate": 0.0024313529915162163,
      "loss": 1.315,
      "step": 48600
    },
    {
      "epoch": 9.79,
      "learning_rate": 0.002430784059241426,
      "loss": 1.315,
      "step": 48700
    },
    {
      "epoch": 9.81,
      "learning_rate": 0.0024302070647516908,
      "loss": 1.3129,
      "step": 48800
    },
    {
      "epoch": 9.83,
      "learning_rate": 0.002429627744335717,
      "loss": 1.3163,
      "step": 48900
    },
    {
      "epoch": 9.85,
      "learning_rate": 0.0024290460991352155,
      "loss": 1.3101,
      "step": 49000
    },
    {
      "epoch": 9.87,
      "learning_rate": 0.0024284621302964804,
      "loss": 1.3152,
      "step": 49100
    },
    {
      "epoch": 9.89,
      "learning_rate": 0.0024278758389703853,
      "loss": 1.3163,
      "step": 49200
    },
    {
      "epoch": 9.91,
      "learning_rate": 0.002427287226312379,
      "loss": 1.3089,
      "step": 49300
    },
    {
      "epoch": 9.93,
      "learning_rate": 0.002426696293482488,
      "loss": 1.3141,
      "step": 49400
    },
    {
      "epoch": 9.95,
      "learning_rate": 0.002426103041645309,
      "loss": 1.3113,
      "step": 49500
    },
    {
      "epoch": 9.97,
      "learning_rate": 0.0024255074719700094,
      "loss": 1.3098,
      "step": 49600
    },
    {
      "epoch": 9.99,
      "learning_rate": 0.0024249095856303253,
      "loss": 1.3112,
      "step": 49700
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.4034808535713104,
      "eval_loss": 1.2766634225845337,
      "eval_runtime": 19.8931,
      "eval_samples_per_second": 3999.773,
      "eval_steps_per_second": 15.634,
      "step": 49732
    },
    {
      "epoch": 10.01,
      "learning_rate": 0.0024243093838045584,
      "loss": 1.3131,
      "step": 49800
    },
    {
      "epoch": 10.03,
      "learning_rate": 0.002423706867675572,
      "loss": 1.3102,
      "step": 49900
    },
    {
      "epoch": 10.05,
      "learning_rate": 0.002423102038430793,
      "loss": 1.3043,
      "step": 50000
    },
    {
      "epoch": 10.07,
      "learning_rate": 0.0024224948972622054,
      "loss": 1.3073,
      "step": 50100
    },
    {
      "epoch": 10.09,
      "learning_rate": 0.002421885445366349,
      "loss": 1.3097,
      "step": 50200
    },
    {
      "epoch": 10.11,
      "learning_rate": 0.002421273683944319,
      "loss": 1.3065,
      "step": 50300
    },
    {
      "epoch": 10.13,
      "learning_rate": 0.002420659614201761,
      "loss": 1.3096,
      "step": 50400
    },
    {
      "epoch": 10.15,
      "learning_rate": 0.002420043237348871,
      "loss": 1.3118,
      "step": 50500
    },
    {
      "epoch": 10.17,
      "learning_rate": 0.0024194245546003894,
      "loss": 1.3081,
      "step": 50600
    },
    {
      "epoch": 10.19,
      "learning_rate": 0.002418803567175604,
      "loss": 1.3087,
      "step": 50700
    },
    {
      "epoch": 10.21,
      "learning_rate": 0.002418186520605176,
      "loss": 1.3084,
      "step": 50800
    },
    {
      "epoch": 10.23,
      "learning_rate": 0.0024175609505199523,
      "loss": 1.3071,
      "step": 50900
    },
    {
      "epoch": 10.25,
      "learning_rate": 0.002416933079431175,
      "loss": 1.3071,
      "step": 51000
    },
    {
      "epoch": 10.27,
      "learning_rate": 0.0024163029085762376,
      "loss": 1.3126,
      "step": 51100
    },
    {
      "epoch": 10.3,
      "learning_rate": 0.0024156704391970684,
      "loss": 1.3117,
      "step": 51200
    },
    {
      "epoch": 10.32,
      "learning_rate": 0.0024150356725401233,
      "loss": 1.3088,
      "step": 51300
    },
    {
      "epoch": 10.34,
      "learning_rate": 0.002414404991844438,
      "loss": 1.3101,
      "step": 51400
    },
    {
      "epoch": 10.36,
      "learning_rate": 0.0024137656573309015,
      "loss": 1.3126,
      "step": 51500
    },
    {
      "epoch": 10.38,
      "learning_rate": 0.002413124029293492,
      "loss": 1.3098,
      "step": 51600
    },
    {
      "epoch": 10.4,
      "learning_rate": 0.0024124801089967156,
      "loss": 1.3148,
      "step": 51700
    },
    {
      "epoch": 10.42,
      "learning_rate": 0.0024118338977095963,
      "loss": 1.3099,
      "step": 51800
    },
    {
      "epoch": 10.44,
      "learning_rate": 0.0024111853967056732,
      "loss": 1.3077,
      "step": 51900
    },
    {
      "epoch": 10.46,
      "learning_rate": 0.002410534607262998,
      "loss": 1.3099,
      "step": 52000
    },
    {
      "epoch": 10.48,
      "learning_rate": 0.0024098815306641316,
      "loss": 1.3082,
      "step": 52100
    },
    {
      "epoch": 10.5,
      "learning_rate": 0.002409226168196144,
      "loss": 1.3068,
      "step": 52200
    },
    {
      "epoch": 10.52,
      "learning_rate": 0.0024085685211506086,
      "loss": 1.3094,
      "step": 52300
    },
    {
      "epoch": 10.54,
      "learning_rate": 0.002407908590823602,
      "loss": 1.3076,
      "step": 52400
    },
    {
      "epoch": 10.56,
      "learning_rate": 0.0024072463785157006,
      "loss": 1.3075,
      "step": 52500
    },
    {
      "epoch": 10.58,
      "learning_rate": 0.0024065818855319773,
      "loss": 1.3118,
      "step": 52600
    },
    {
      "epoch": 10.6,
      "learning_rate": 0.0024059151131820013,
      "loss": 1.3101,
      "step": 52700
    },
    {
      "epoch": 10.62,
      "learning_rate": 0.002405246062779832,
      "loss": 1.3058,
      "step": 52800
    },
    {
      "epoch": 10.64,
      "learning_rate": 0.0024045747356440197,
      "loss": 1.3079,
      "step": 52900
    },
    {
      "epoch": 10.66,
      "learning_rate": 0.0024039011330976015,
      "loss": 1.309,
      "step": 53000
    },
    {
      "epoch": 10.68,
      "learning_rate": 0.0024032252564680978,
      "loss": 1.3076,
      "step": 53100
    },
    {
      "epoch": 10.7,
      "learning_rate": 0.0024025471070875117,
      "loss": 1.3102,
      "step": 53200
    },
    {
      "epoch": 10.72,
      "learning_rate": 0.002401866686292326,
      "loss": 1.3088,
      "step": 53300
    },
    {
      "epoch": 10.74,
      "learning_rate": 0.0024011839954234983,
      "loss": 1.3089,
      "step": 53400
    },
    {
      "epoch": 10.76,
      "learning_rate": 0.0024004990358264607,
      "loss": 1.3063,
      "step": 53500
    },
    {
      "epoch": 10.78,
      "learning_rate": 0.0023998118088511167,
      "loss": 1.3066,
      "step": 53600
    },
    {
      "epoch": 10.8,
      "learning_rate": 0.0023991223158518387,
      "loss": 1.3071,
      "step": 53700
    },
    {
      "epoch": 10.82,
      "learning_rate": 0.0023984305581874637,
      "loss": 1.3097,
      "step": 53800
    },
    {
      "epoch": 10.84,
      "learning_rate": 0.002397736537221293,
      "loss": 1.3097,
      "step": 53900
    },
    {
      "epoch": 10.86,
      "learning_rate": 0.002397040254321088,
      "loss": 1.3088,
      "step": 54000
    },
    {
      "epoch": 10.88,
      "learning_rate": 0.0023963417108590675,
      "loss": 1.3071,
      "step": 54100
    },
    {
      "epoch": 10.9,
      "learning_rate": 0.0023956409082119055,
      "loss": 1.3055,
      "step": 54200
    },
    {
      "epoch": 10.92,
      "learning_rate": 0.002394937847760729,
      "loss": 1.3059,
      "step": 54300
    },
    {
      "epoch": 10.94,
      "learning_rate": 0.002394232530891114,
      "loss": 1.3016,
      "step": 54400
    },
    {
      "epoch": 10.96,
      "learning_rate": 0.0023935249589930835,
      "loss": 1.308,
      "step": 54500
    },
    {
      "epoch": 10.98,
      "learning_rate": 0.0023928151334611045,
      "loss": 1.3063,
      "step": 54600
    },
    {
      "epoch": 11.0,
      "learning_rate": 0.002392103055694086,
      "loss": 1.3058,
      "step": 54700
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.4042048581326505,
      "eval_loss": 1.2673618793487549,
      "eval_runtime": 19.7081,
      "eval_samples_per_second": 4037.327,
      "eval_steps_per_second": 15.78,
      "step": 54705
    },
    {
      "epoch": 11.02,
      "learning_rate": 0.0023913887270953743,
      "loss": 1.2989,
      "step": 54800
    },
    {
      "epoch": 11.04,
      "learning_rate": 0.002390672149072754,
      "loss": 1.302,
      "step": 54900
    },
    {
      "epoch": 11.06,
      "learning_rate": 0.0023899605224217933,
      "loss": 1.2996,
      "step": 55000
    },
    {
      "epoch": 11.08,
      "learning_rate": 0.002389239472251357,
      "loss": 1.2986,
      "step": 55100
    },
    {
      "epoch": 11.1,
      "learning_rate": 0.0023885161768927156,
      "loss": 1.3003,
      "step": 55200
    },
    {
      "epoch": 11.12,
      "learning_rate": 0.002387790637771324,
      "loss": 1.3006,
      "step": 55300
    },
    {
      "epoch": 11.14,
      "learning_rate": 0.0023870628563170586,
      "loss": 1.3063,
      "step": 55400
    },
    {
      "epoch": 11.16,
      "learning_rate": 0.0023863328339642155,
      "loss": 1.3082,
      "step": 55500
    },
    {
      "epoch": 11.18,
      "learning_rate": 0.002385600572151506,
      "loss": 1.3049,
      "step": 55600
    },
    {
      "epoch": 11.2,
      "learning_rate": 0.002384866072322057,
      "loss": 1.3041,
      "step": 55700
    },
    {
      "epoch": 11.22,
      "learning_rate": 0.0023841293359234033,
      "loss": 1.305,
      "step": 55800
    },
    {
      "epoch": 11.24,
      "learning_rate": 0.00238339036440749,
      "loss": 1.3048,
      "step": 55900
    },
    {
      "epoch": 11.26,
      "learning_rate": 0.002382649159230665,
      "loss": 1.3019,
      "step": 56000
    },
    {
      "epoch": 11.28,
      "learning_rate": 0.00238190572185368,
      "loss": 1.3049,
      "step": 56100
    },
    {
      "epoch": 11.3,
      "learning_rate": 0.002381160053741684,
      "loss": 1.306,
      "step": 56200
    },
    {
      "epoch": 11.32,
      "learning_rate": 0.0023804121563642247,
      "loss": 1.3058,
      "step": 56300
    },
    {
      "epoch": 11.34,
      "learning_rate": 0.0023796620311952415,
      "loss": 1.305,
      "step": 56400
    },
    {
      "epoch": 11.36,
      "learning_rate": 0.0023789096797130643,
      "loss": 1.3037,
      "step": 56500
    },
    {
      "epoch": 11.38,
      "learning_rate": 0.002378155103400411,
      "loss": 1.3064,
      "step": 56600
    },
    {
      "epoch": 11.4,
      "learning_rate": 0.002377398303744384,
      "loss": 1.3038,
      "step": 56700
    },
    {
      "epoch": 11.42,
      "learning_rate": 0.002376639282236468,
      "loss": 1.3057,
      "step": 56800
    },
    {
      "epoch": 11.44,
      "learning_rate": 0.0023758780403725257,
      "loss": 1.3044,
      "step": 56900
    },
    {
      "epoch": 11.46,
      "learning_rate": 0.0023751145796527956,
      "loss": 1.3057,
      "step": 57000
    },
    {
      "epoch": 11.48,
      "learning_rate": 0.00237434890158189,
      "loss": 1.3034,
      "step": 57100
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.0023735810076687893,
      "loss": 1.3055,
      "step": 57200
    },
    {
      "epoch": 11.52,
      "learning_rate": 0.0023728108994268433,
      "loss": 1.3021,
      "step": 57300
    },
    {
      "epoch": 11.54,
      "learning_rate": 0.0023720385783737637,
      "loss": 1.3051,
      "step": 57400
    },
    {
      "epoch": 11.56,
      "learning_rate": 0.0023712640460316244,
      "loss": 1.3066,
      "step": 57500
    },
    {
      "epoch": 11.58,
      "learning_rate": 0.0023704873039268565,
      "loss": 1.306,
      "step": 57600
    },
    {
      "epoch": 11.6,
      "learning_rate": 0.0023697083535902464,
      "loss": 1.2988,
      "step": 57700
    },
    {
      "epoch": 11.62,
      "learning_rate": 0.0023689271965569325,
      "loss": 1.3044,
      "step": 57800
    },
    {
      "epoch": 11.64,
      "learning_rate": 0.0023681438343664016,
      "loss": 1.3054,
      "step": 57900
    },
    {
      "epoch": 11.66,
      "learning_rate": 0.002367358268562487,
      "loss": 1.3035,
      "step": 58000
    },
    {
      "epoch": 11.68,
      "learning_rate": 0.0023665705006933645,
      "loss": 1.2999,
      "step": 58100
    },
    {
      "epoch": 11.7,
      "learning_rate": 0.00236578053231155,
      "loss": 1.3027,
      "step": 58200
    },
    {
      "epoch": 11.72,
      "learning_rate": 0.0023649883649738954,
      "loss": 1.3035,
      "step": 58300
    },
    {
      "epoch": 11.74,
      "learning_rate": 0.0023641940002415876,
      "loss": 1.3023,
      "step": 58400
    },
    {
      "epoch": 11.76,
      "learning_rate": 0.002363397439680142,
      "loss": 1.2999,
      "step": 58500
    },
    {
      "epoch": 11.78,
      "learning_rate": 0.002362598684859403,
      "loss": 1.2994,
      "step": 58600
    },
    {
      "epoch": 11.8,
      "learning_rate": 0.0023617977373535403,
      "loss": 1.3024,
      "step": 58700
    },
    {
      "epoch": 11.82,
      "learning_rate": 0.0023609945987410432,
      "loss": 1.302,
      "step": 58800
    },
    {
      "epoch": 11.84,
      "learning_rate": 0.0023601892706047187,
      "loss": 1.3044,
      "step": 58900
    },
    {
      "epoch": 11.86,
      "learning_rate": 0.002359381754531691,
      "loss": 1.3026,
      "step": 59000
    },
    {
      "epoch": 11.88,
      "learning_rate": 0.0023585801599547503,
      "loss": 1.3061,
      "step": 59100
    },
    {
      "epoch": 11.9,
      "learning_rate": 0.002357768294626511,
      "loss": 1.2997,
      "step": 59200
    },
    {
      "epoch": 11.92,
      "learning_rate": 0.0023569542461327744,
      "loss": 1.2999,
      "step": 59300
    },
    {
      "epoch": 11.94,
      "learning_rate": 0.0023561380160778494,
      "loss": 1.3,
      "step": 59400
    },
    {
      "epoch": 11.96,
      "learning_rate": 0.0023553196060703448,
      "loss": 1.3009,
      "step": 59500
    },
    {
      "epoch": 11.98,
      "learning_rate": 0.0023544990177231644,
      "loss": 1.3043,
      "step": 59600
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.40439273978216544,
      "eval_loss": 1.2642629146575928,
      "eval_runtime": 19.7374,
      "eval_samples_per_second": 4031.338,
      "eval_steps_per_second": 15.757,
      "step": 59679
    },
    {
      "epoch": 12.0,
      "learning_rate": 0.0023536762526535065,
      "loss": 1.3005,
      "step": 59700
    },
    {
      "epoch": 12.02,
      "learning_rate": 0.002352851312482858,
      "loss": 1.2972,
      "step": 59800
    },
    {
      "epoch": 12.04,
      "learning_rate": 0.0023520241988369927,
      "loss": 1.2979,
      "step": 59900
    },
    {
      "epoch": 12.06,
      "learning_rate": 0.0023511949133459688,
      "loss": 1.2933,
      "step": 60000
    },
    {
      "epoch": 12.08,
      "learning_rate": 0.0023503634576441234,
      "loss": 1.2949,
      "step": 60100
    },
    {
      "epoch": 12.1,
      "learning_rate": 0.0023495298333700712,
      "loss": 1.3001,
      "step": 60200
    },
    {
      "epoch": 12.12,
      "learning_rate": 0.0023486940421667012,
      "loss": 1.2958,
      "step": 60300
    },
    {
      "epoch": 12.14,
      "learning_rate": 0.0023478560856811715,
      "loss": 1.2996,
      "step": 60400
    },
    {
      "epoch": 12.17,
      "learning_rate": 0.0023470159655649093,
      "loss": 1.2979,
      "step": 60500
    },
    {
      "epoch": 12.19,
      "learning_rate": 0.0023461736834736045,
      "loss": 1.2997,
      "step": 60600
    },
    {
      "epoch": 12.21,
      "learning_rate": 0.0023453292410672085,
      "loss": 1.3016,
      "step": 60700
    },
    {
      "epoch": 12.23,
      "learning_rate": 0.002344482640009929,
      "loss": 1.299,
      "step": 60800
    },
    {
      "epoch": 12.25,
      "learning_rate": 0.00234363388197023,
      "loss": 1.3009,
      "step": 60900
    },
    {
      "epoch": 12.27,
      "learning_rate": 0.0023427829686208245,
      "loss": 1.3003,
      "step": 61000
    },
    {
      "epoch": 12.29,
      "learning_rate": 0.002341938442963461,
      "loss": 1.303,
      "step": 61100
    },
    {
      "epoch": 12.31,
      "learning_rate": 0.0023410832455409507,
      "loss": 1.2995,
      "step": 61200
    },
    {
      "epoch": 12.33,
      "learning_rate": 0.002340225897835473,
      "loss": 1.2984,
      "step": 61300
    },
    {
      "epoch": 12.35,
      "learning_rate": 0.0023393664015366696,
      "loss": 1.3007,
      "step": 61400
    },
    {
      "epoch": 12.37,
      "learning_rate": 0.0023385047583384167,
      "loss": 1.3039,
      "step": 61500
    },
    {
      "epoch": 12.39,
      "learning_rate": 0.0023376582667187836,
      "loss": 1.2987,
      "step": 61600
    },
    {
      "epoch": 12.41,
      "learning_rate": 0.0023367923776734523,
      "loss": 1.295,
      "step": 61700
    },
    {
      "epoch": 12.43,
      "learning_rate": 0.0023359243468015017,
      "loss": 1.3033,
      "step": 61800
    },
    {
      "epoch": 12.45,
      "learning_rate": 0.002335054175813628,
      "loss": 1.3007,
      "step": 61900
    },
    {
      "epoch": 12.47,
      "learning_rate": 0.002334181866424745,
      "loss": 1.3005,
      "step": 62000
    },
    {
      "epoch": 12.49,
      "learning_rate": 0.002333307420353981,
      "loss": 1.3006,
      "step": 62100
    },
    {
      "epoch": 12.51,
      "learning_rate": 0.0023324308393246752,
      "loss": 1.2948,
      "step": 62200
    },
    {
      "epoch": 12.53,
      "learning_rate": 0.002331552125064374,
      "loss": 1.2972,
      "step": 62300
    },
    {
      "epoch": 12.55,
      "learning_rate": 0.0023306712793048287,
      "loss": 1.3009,
      "step": 62400
    },
    {
      "epoch": 12.57,
      "learning_rate": 0.0023297883037819906,
      "loss": 1.2992,
      "step": 62500
    },
    {
      "epoch": 12.59,
      "learning_rate": 0.002328903200236008,
      "loss": 1.3011,
      "step": 62600
    },
    {
      "epoch": 12.61,
      "learning_rate": 0.0023280159704112255,
      "loss": 1.3016,
      "step": 62700
    },
    {
      "epoch": 12.63,
      "learning_rate": 0.0023271266160561736,
      "loss": 1.2949,
      "step": 62800
    },
    {
      "epoch": 12.65,
      "learning_rate": 0.0023262351389235743,
      "loss": 1.2994,
      "step": 62900
    },
    {
      "epoch": 12.67,
      "learning_rate": 0.00232534154077033,
      "loss": 1.3009,
      "step": 63000
    },
    {
      "epoch": 12.69,
      "learning_rate": 0.002324445823357525,
      "loss": 1.2988,
      "step": 63100
    },
    {
      "epoch": 12.71,
      "learning_rate": 0.0023235479884504196,
      "loss": 1.2966,
      "step": 63200
    },
    {
      "epoch": 12.73,
      "learning_rate": 0.0023226480378184472,
      "loss": 1.2976,
      "step": 63300
    },
    {
      "epoch": 12.75,
      "learning_rate": 0.0023217459732352104,
      "loss": 1.2981,
      "step": 63400
    },
    {
      "epoch": 12.77,
      "learning_rate": 0.002320841796478479,
      "loss": 1.2969,
      "step": 63500
    },
    {
      "epoch": 12.79,
      "learning_rate": 0.0023199355093301835,
      "loss": 1.2999,
      "step": 63600
    },
    {
      "epoch": 12.81,
      "learning_rate": 0.0023190271135764165,
      "loss": 1.295,
      "step": 63700
    },
    {
      "epoch": 12.83,
      "learning_rate": 0.0023181166110074224,
      "loss": 1.2908,
      "step": 63800
    },
    {
      "epoch": 12.85,
      "learning_rate": 0.002317204003417602,
      "loss": 1.2966,
      "step": 63900
    },
    {
      "epoch": 12.87,
      "learning_rate": 0.0023162892926055006,
      "loss": 1.2965,
      "step": 64000
    },
    {
      "epoch": 12.89,
      "learning_rate": 0.002315372480373811,
      "loss": 1.2946,
      "step": 64100
    },
    {
      "epoch": 12.91,
      "learning_rate": 0.0023144535685293665,
      "loss": 1.2988,
      "step": 64200
    },
    {
      "epoch": 12.93,
      "learning_rate": 0.0023135325588831386,
      "loss": 1.3009,
      "step": 64300
    },
    {
      "epoch": 12.95,
      "learning_rate": 0.0023126094532502327,
      "loss": 1.2955,
      "step": 64400
    },
    {
      "epoch": 12.97,
      "learning_rate": 0.0023116842534498857,
      "loss": 1.3004,
      "step": 64500
    },
    {
      "epoch": 12.99,
      "learning_rate": 0.00231075696130546,
      "loss": 1.2943,
      "step": 64600
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.4050755408318846,
      "eval_loss": 1.25895094871521,
      "eval_runtime": 19.7319,
      "eval_samples_per_second": 4032.464,
      "eval_steps_per_second": 15.761,
      "step": 64652
    },
    {
      "epoch": 13.01,
      "learning_rate": 0.0023098275786444435,
      "loss": 1.2935,
      "step": 64700
    },
    {
      "epoch": 13.03,
      "learning_rate": 0.002308896107298443,
      "loss": 1.2893,
      "step": 64800
    },
    {
      "epoch": 13.05,
      "learning_rate": 0.0023079625491031824,
      "loss": 1.2912,
      "step": 64900
    },
    {
      "epoch": 13.07,
      "learning_rate": 0.002307026905898497,
      "loss": 1.2944,
      "step": 65000
    },
    {
      "epoch": 13.09,
      "learning_rate": 0.002306089179528332,
      "loss": 1.2949,
      "step": 65100
    },
    {
      "epoch": 13.11,
      "learning_rate": 0.002305149371840738,
      "loss": 1.2898,
      "step": 65200
    },
    {
      "epoch": 13.13,
      "learning_rate": 0.002304207484687868,
      "loss": 1.2934,
      "step": 65300
    },
    {
      "epoch": 13.15,
      "learning_rate": 0.002303263519925972,
      "loss": 1.2922,
      "step": 65400
    },
    {
      "epoch": 13.17,
      "learning_rate": 0.002302317479415395,
      "loss": 1.2913,
      "step": 65500
    },
    {
      "epoch": 13.19,
      "learning_rate": 0.002301369365020573,
      "loss": 1.2921,
      "step": 65600
    },
    {
      "epoch": 13.21,
      "learning_rate": 0.0023004191786100297,
      "loss": 1.2935,
      "step": 65700
    },
    {
      "epoch": 13.23,
      "learning_rate": 0.002299466922056371,
      "loss": 1.298,
      "step": 65800
    },
    {
      "epoch": 13.25,
      "learning_rate": 0.0022985125972362834,
      "loss": 1.2949,
      "step": 65900
    },
    {
      "epoch": 13.27,
      "learning_rate": 0.0022975562060305295,
      "loss": 1.2926,
      "step": 66000
    },
    {
      "epoch": 13.29,
      "learning_rate": 0.002296597750323944,
      "loss": 1.293,
      "step": 66100
    },
    {
      "epoch": 13.31,
      "learning_rate": 0.0022956372320054306,
      "loss": 1.2942,
      "step": 66200
    },
    {
      "epoch": 13.33,
      "learning_rate": 0.0022946746529679575,
      "loss": 1.2944,
      "step": 66300
    },
    {
      "epoch": 13.35,
      "learning_rate": 0.002293710015108555,
      "loss": 1.297,
      "step": 66400
    },
    {
      "epoch": 13.37,
      "learning_rate": 0.0022927433203283093,
      "loss": 1.2982,
      "step": 66500
    },
    {
      "epoch": 13.39,
      "learning_rate": 0.0022917745705323617,
      "loss": 1.2904,
      "step": 66600
    },
    {
      "epoch": 13.41,
      "learning_rate": 0.0022908037676299038,
      "loss": 1.2948,
      "step": 66700
    },
    {
      "epoch": 13.43,
      "learning_rate": 0.0022898309135341716,
      "loss": 1.2965,
      "step": 66800
    },
    {
      "epoch": 13.45,
      "learning_rate": 0.0022888560101624452,
      "loss": 1.2892,
      "step": 66900
    },
    {
      "epoch": 13.47,
      "learning_rate": 0.002287888839071394,
      "loss": 1.2942,
      "step": 67000
    },
    {
      "epoch": 13.49,
      "learning_rate": 0.0022869098633604185,
      "loss": 1.2886,
      "step": 67100
    },
    {
      "epoch": 13.51,
      "learning_rate": 0.0022859288441301902,
      "loss": 1.2901,
      "step": 67200
    },
    {
      "epoch": 13.53,
      "learning_rate": 0.0022849457833140803,
      "loss": 1.2941,
      "step": 67300
    },
    {
      "epoch": 13.55,
      "learning_rate": 0.0022839606828494842,
      "loss": 1.2947,
      "step": 67400
    },
    {
      "epoch": 13.57,
      "learning_rate": 0.002282973544677816,
      "loss": 1.2952,
      "step": 67500
    },
    {
      "epoch": 13.59,
      "learning_rate": 0.0022819843707445058,
      "loss": 1.2945,
      "step": 67600
    },
    {
      "epoch": 13.61,
      "learning_rate": 0.002280993162998996,
      "loss": 1.2935,
      "step": 67700
    },
    {
      "epoch": 13.63,
      "learning_rate": 0.002279999923394737,
      "loss": 1.2948,
      "step": 67800
    },
    {
      "epoch": 13.65,
      "learning_rate": 0.0022790046538891844,
      "loss": 1.2895,
      "step": 67900
    },
    {
      "epoch": 13.67,
      "learning_rate": 0.0022780073564437927,
      "loss": 1.2952,
      "step": 68000
    },
    {
      "epoch": 13.69,
      "learning_rate": 0.002277008033024014,
      "loss": 1.2946,
      "step": 68100
    },
    {
      "epoch": 13.71,
      "learning_rate": 0.002276006685599293,
      "loss": 1.298,
      "step": 68200
    },
    {
      "epoch": 13.73,
      "learning_rate": 0.002275003316143064,
      "loss": 1.2942,
      "step": 68300
    },
    {
      "epoch": 13.75,
      "learning_rate": 0.0022739979266327448,
      "loss": 1.2936,
      "step": 68400
    },
    {
      "epoch": 13.77,
      "learning_rate": 0.002272990519049735,
      "loss": 1.2937,
      "step": 68500
    },
    {
      "epoch": 13.79,
      "learning_rate": 0.0022719810953794116,
      "loss": 1.2991,
      "step": 68600
    },
    {
      "epoch": 13.81,
      "learning_rate": 0.002270969657611125,
      "loss": 1.2924,
      "step": 68700
    },
    {
      "epoch": 13.83,
      "learning_rate": 0.0022699562077381943,
      "loss": 1.2924,
      "step": 68800
    },
    {
      "epoch": 13.85,
      "learning_rate": 0.002268940747757904,
      "loss": 1.2936,
      "step": 68900
    },
    {
      "epoch": 13.87,
      "learning_rate": 0.002267923279671501,
      "loss": 1.2963,
      "step": 69000
    },
    {
      "epoch": 13.89,
      "learning_rate": 0.0022669038054841887,
      "loss": 1.2875,
      "step": 69100
    },
    {
      "epoch": 13.91,
      "learning_rate": 0.0022658823272051244,
      "loss": 1.2912,
      "step": 69200
    },
    {
      "epoch": 13.93,
      "learning_rate": 0.0022648588468474158,
      "loss": 1.2943,
      "step": 69300
    },
    {
      "epoch": 13.95,
      "learning_rate": 0.002263833366428115,
      "loss": 1.2974,
      "step": 69400
    },
    {
      "epoch": 13.97,
      "learning_rate": 0.002262816172636471,
      "loss": 1.2929,
      "step": 69500
    },
    {
      "epoch": 13.99,
      "learning_rate": 0.002261786718111027,
      "loss": 1.2926,
      "step": 69600
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.40531926198942647,
      "eval_loss": 1.2532228231430054,
      "eval_runtime": 19.5692,
      "eval_samples_per_second": 4065.991,
      "eval_steps_per_second": 15.892,
      "step": 69625
    },
    {
      "epoch": 14.01,
      "learning_rate": 0.002260755269578475,
      "loss": 1.2829,
      "step": 69700
    },
    {
      "epoch": 14.04,
      "learning_rate": 0.0022597218290715715,
      "loss": 1.2825,
      "step": 69800
    },
    {
      "epoch": 14.06,
      "learning_rate": 0.002258686398626998,
      "loss": 1.2912,
      "step": 69900
    },
    {
      "epoch": 14.08,
      "learning_rate": 0.0022576489802853578,
      "loss": 1.2877,
      "step": 70000
    },
    {
      "epoch": 14.1,
      "learning_rate": 0.002256609576091173,
      "loss": 1.2876,
      "step": 70100
    },
    {
      "epoch": 14.12,
      "learning_rate": 0.0022555681880928784,
      "loss": 1.2921,
      "step": 70200
    },
    {
      "epoch": 14.14,
      "learning_rate": 0.0022545248183428184,
      "loss": 1.2898,
      "step": 70300
    },
    {
      "epoch": 14.16,
      "learning_rate": 0.0022534794688972436,
      "loss": 1.2881,
      "step": 70400
    },
    {
      "epoch": 14.18,
      "learning_rate": 0.0022524321418163056,
      "loss": 1.2887,
      "step": 70500
    },
    {
      "epoch": 14.2,
      "learning_rate": 0.0022513828391640535,
      "loss": 1.2835,
      "step": 70600
    },
    {
      "epoch": 14.22,
      "learning_rate": 0.00225033156300843,
      "loss": 1.2945,
      "step": 70700
    },
    {
      "epoch": 14.24,
      "learning_rate": 0.0022492783154212676,
      "loss": 1.2912,
      "step": 70800
    },
    {
      "epoch": 14.26,
      "learning_rate": 0.002248223098478284,
      "loss": 1.2864,
      "step": 70900
    },
    {
      "epoch": 14.28,
      "learning_rate": 0.0022471659142590758,
      "loss": 1.2914,
      "step": 71000
    },
    {
      "epoch": 14.3,
      "learning_rate": 0.00224610676484712,
      "loss": 1.2877,
      "step": 71100
    },
    {
      "epoch": 14.32,
      "learning_rate": 0.0022450456523297646,
      "loss": 1.288,
      "step": 71200
    },
    {
      "epoch": 14.34,
      "learning_rate": 0.0022439825787982275,
      "loss": 1.2901,
      "step": 71300
    },
    {
      "epoch": 14.36,
      "learning_rate": 0.0022429175463475897,
      "loss": 1.2885,
      "step": 71400
    },
    {
      "epoch": 14.38,
      "learning_rate": 0.002241861236648861,
      "loss": 1.2915,
      "step": 71500
    },
    {
      "epoch": 14.4,
      "learning_rate": 0.00224079231219746,
      "loss": 1.2862,
      "step": 71600
    },
    {
      "epoch": 14.42,
      "learning_rate": 0.002239721435114266,
      "loss": 1.2887,
      "step": 71700
    },
    {
      "epoch": 14.44,
      "learning_rate": 0.0022386486075097406,
      "loss": 1.2953,
      "step": 71800
    },
    {
      "epoch": 14.46,
      "learning_rate": 0.0022375738314981885,
      "loss": 1.2835,
      "step": 71900
    },
    {
      "epoch": 14.48,
      "learning_rate": 0.0022364971091977555,
      "loss": 1.2931,
      "step": 72000
    },
    {
      "epoch": 14.5,
      "learning_rate": 0.002235418442730422,
      "loss": 1.2878,
      "step": 72100
    },
    {
      "epoch": 14.52,
      "learning_rate": 0.0022343378342220004,
      "loss": 1.2877,
      "step": 72200
    },
    {
      "epoch": 14.54,
      "learning_rate": 0.0022332552858021298,
      "loss": 1.2856,
      "step": 72300
    },
    {
      "epoch": 14.56,
      "learning_rate": 0.002232170799604273,
      "loss": 1.2855,
      "step": 72400
    },
    {
      "epoch": 14.58,
      "learning_rate": 0.002231084377765712,
      "loss": 1.2872,
      "step": 72500
    },
    {
      "epoch": 14.6,
      "learning_rate": 0.002229996022427543,
      "loss": 1.2863,
      "step": 72600
    },
    {
      "epoch": 14.62,
      "learning_rate": 0.002228905735734673,
      "loss": 1.2894,
      "step": 72700
    },
    {
      "epoch": 14.64,
      "learning_rate": 0.002227813519835815,
      "loss": 1.2877,
      "step": 72800
    },
    {
      "epoch": 14.66,
      "learning_rate": 0.0022267193768834843,
      "loss": 1.2847,
      "step": 72900
    },
    {
      "epoch": 14.68,
      "learning_rate": 0.002225623309033993,
      "loss": 1.2925,
      "step": 73000
    },
    {
      "epoch": 14.7,
      "learning_rate": 0.0022245253184474496,
      "loss": 1.2866,
      "step": 73100
    },
    {
      "epoch": 14.72,
      "learning_rate": 0.002223425407287748,
      "loss": 1.2889,
      "step": 73200
    },
    {
      "epoch": 14.74,
      "learning_rate": 0.00222232357772257,
      "loss": 1.2886,
      "step": 73300
    },
    {
      "epoch": 14.76,
      "learning_rate": 0.0022212198319233765,
      "loss": 1.2896,
      "step": 73400
    },
    {
      "epoch": 14.78,
      "learning_rate": 0.0022201141720654062,
      "loss": 1.2915,
      "step": 73500
    },
    {
      "epoch": 14.8,
      "learning_rate": 0.002219006600327669,
      "loss": 1.287,
      "step": 73600
    },
    {
      "epoch": 14.82,
      "learning_rate": 0.002217897118892943,
      "loss": 1.2942,
      "step": 73700
    },
    {
      "epoch": 14.84,
      "learning_rate": 0.0022167857299477704,
      "loss": 1.2878,
      "step": 73800
    },
    {
      "epoch": 14.86,
      "learning_rate": 0.0022156724356824516,
      "loss": 1.285,
      "step": 73900
    },
    {
      "epoch": 14.88,
      "learning_rate": 0.0022145683996782177,
      "loss": 1.2891,
      "step": 74000
    },
    {
      "epoch": 14.9,
      "learning_rate": 0.0022134513203569167,
      "loss": 1.29,
      "step": 74100
    },
    {
      "epoch": 14.92,
      "learning_rate": 0.0022123323422868512,
      "loss": 1.2884,
      "step": 74200
    },
    {
      "epoch": 14.94,
      "learning_rate": 0.002211211467673279,
      "loss": 1.2901,
      "step": 74300
    },
    {
      "epoch": 14.96,
      "learning_rate": 0.0022100886987251943,
      "loss": 1.2901,
      "step": 74400
    },
    {
      "epoch": 14.98,
      "learning_rate": 0.0022089640376553267,
      "loss": 1.2917,
      "step": 74500
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.40564348112720083,
      "eval_loss": 1.2492839097976685,
      "eval_runtime": 19.7618,
      "eval_samples_per_second": 4026.36,
      "eval_steps_per_second": 15.737,
      "step": 74598
    },
    {
      "epoch": 15.0,
      "learning_rate": 0.0022078374866801326,
      "loss": 1.2901,
      "step": 74600
    },
    {
      "epoch": 15.02,
      "learning_rate": 0.002206709048019794,
      "loss": 1.2815,
      "step": 74700
    },
    {
      "epoch": 15.04,
      "learning_rate": 0.0022055787238982145,
      "loss": 1.2824,
      "step": 74800
    },
    {
      "epoch": 15.06,
      "learning_rate": 0.0022044465165430107,
      "loss": 1.2836,
      "step": 74900
    },
    {
      "epoch": 15.08,
      "learning_rate": 0.002203312428185512,
      "loss": 1.2848,
      "step": 75000
    },
    {
      "epoch": 15.1,
      "learning_rate": 0.002202176461060756,
      "loss": 1.2856,
      "step": 75100
    },
    {
      "epoch": 15.12,
      "learning_rate": 0.0022010386174074813,
      "loss": 1.2858,
      "step": 75200
    },
    {
      "epoch": 15.14,
      "learning_rate": 0.002199898899468125,
      "loss": 1.2827,
      "step": 75300
    },
    {
      "epoch": 15.16,
      "learning_rate": 0.0021987573094888182,
      "loss": 1.2816,
      "step": 75400
    },
    {
      "epoch": 15.18,
      "learning_rate": 0.0021976138497193823,
      "loss": 1.2845,
      "step": 75500
    },
    {
      "epoch": 15.2,
      "learning_rate": 0.0021964685224133214,
      "loss": 1.2876,
      "step": 75600
    },
    {
      "epoch": 15.22,
      "learning_rate": 0.0021953213298278223,
      "loss": 1.2815,
      "step": 75700
    },
    {
      "epoch": 15.24,
      "learning_rate": 0.0021941837739942976,
      "loss": 1.2838,
      "step": 75800
    },
    {
      "epoch": 15.26,
      "learning_rate": 0.002193032876232499,
      "loss": 1.2826,
      "step": 75900
    },
    {
      "epoch": 15.28,
      "learning_rate": 0.002191880119962158,
      "loss": 1.2794,
      "step": 76000
    },
    {
      "epoch": 15.3,
      "learning_rate": 0.0021907255074551024,
      "loss": 1.2845,
      "step": 76100
    },
    {
      "epoch": 15.32,
      "learning_rate": 0.0021895690409868165,
      "loss": 1.2861,
      "step": 76200
    },
    {
      "epoch": 15.34,
      "learning_rate": 0.0021884107228364387,
      "loss": 1.2845,
      "step": 76300
    },
    {
      "epoch": 15.36,
      "learning_rate": 0.002187250555286758,
      "loss": 1.2845,
      "step": 76400
    },
    {
      "epoch": 15.38,
      "learning_rate": 0.0021860885406242065,
      "loss": 1.2879,
      "step": 76500
    },
    {
      "epoch": 15.4,
      "learning_rate": 0.002184924681138858,
      "loss": 1.2798,
      "step": 76600
    },
    {
      "epoch": 15.42,
      "learning_rate": 0.0021837589791244205,
      "loss": 1.2823,
      "step": 76700
    },
    {
      "epoch": 15.44,
      "learning_rate": 0.0021825914368782343,
      "loss": 1.282,
      "step": 76800
    },
    {
      "epoch": 15.46,
      "learning_rate": 0.002181422056701266,
      "loss": 1.2844,
      "step": 76900
    },
    {
      "epoch": 15.48,
      "learning_rate": 0.002180250840898105,
      "loss": 1.2854,
      "step": 77000
    },
    {
      "epoch": 15.5,
      "learning_rate": 0.002179077791776957,
      "loss": 1.287,
      "step": 77100
    },
    {
      "epoch": 15.52,
      "learning_rate": 0.0021779029116496423,
      "loss": 1.284,
      "step": 77200
    },
    {
      "epoch": 15.54,
      "learning_rate": 0.002176726202831588,
      "loss": 1.2866,
      "step": 77300
    },
    {
      "epoch": 15.56,
      "learning_rate": 0.0021755476676418267,
      "loss": 1.29,
      "step": 77400
    },
    {
      "epoch": 15.58,
      "learning_rate": 0.0021743673084029897,
      "loss": 1.2852,
      "step": 77500
    },
    {
      "epoch": 15.6,
      "learning_rate": 0.002173185127441303,
      "loss": 1.28,
      "step": 77600
    },
    {
      "epoch": 15.62,
      "learning_rate": 0.0021720129760884666,
      "loss": 1.2861,
      "step": 77700
    },
    {
      "epoch": 15.64,
      "learning_rate": 0.0021708271768331494,
      "loss": 1.2843,
      "step": 77800
    },
    {
      "epoch": 15.66,
      "learning_rate": 0.002169639562831796,
      "loss": 1.2825,
      "step": 77900
    },
    {
      "epoch": 15.68,
      "learning_rate": 0.002168450136424931,
      "loss": 1.2862,
      "step": 78000
    },
    {
      "epoch": 15.7,
      "learning_rate": 0.0021672588999566487,
      "loss": 1.2842,
      "step": 78100
    },
    {
      "epoch": 15.72,
      "learning_rate": 0.0021660658557746126,
      "loss": 1.2826,
      "step": 78200
    },
    {
      "epoch": 15.74,
      "learning_rate": 0.0021648710062300482,
      "loss": 1.2851,
      "step": 78300
    },
    {
      "epoch": 15.76,
      "learning_rate": 0.002163674353677738,
      "loss": 1.28,
      "step": 78400
    },
    {
      "epoch": 15.78,
      "learning_rate": 0.0021624759004760198,
      "loss": 1.2835,
      "step": 78500
    },
    {
      "epoch": 15.8,
      "learning_rate": 0.0021612756489867773,
      "loss": 1.282,
      "step": 78600
    },
    {
      "epoch": 15.82,
      "learning_rate": 0.002160073601575442,
      "loss": 1.2815,
      "step": 78700
    },
    {
      "epoch": 15.84,
      "learning_rate": 0.0021588697606109808,
      "loss": 1.2854,
      "step": 78800
    },
    {
      "epoch": 15.86,
      "learning_rate": 0.0021576641284658978,
      "loss": 1.2859,
      "step": 78900
    },
    {
      "epoch": 15.88,
      "learning_rate": 0.0021564567075162263,
      "loss": 1.2845,
      "step": 79000
    },
    {
      "epoch": 15.91,
      "learning_rate": 0.002155247500141525,
      "loss": 1.286,
      "step": 79100
    },
    {
      "epoch": 15.93,
      "learning_rate": 0.0021540365087248737,
      "loss": 1.2849,
      "step": 79200
    },
    {
      "epoch": 15.95,
      "learning_rate": 0.0021528237356528675,
      "loss": 1.2842,
      "step": 79300
    },
    {
      "epoch": 15.97,
      "learning_rate": 0.002151609183315613,
      "loss": 1.283,
      "step": 79400
    },
    {
      "epoch": 15.99,
      "learning_rate": 0.0021503928541067234,
      "loss": 1.2873,
      "step": 79500
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.40597676821927015,
      "eval_loss": 1.2457048892974854,
      "eval_runtime": 19.9932,
      "eval_samples_per_second": 3979.752,
      "eval_steps_per_second": 15.555,
      "step": 79572
    },
    {
      "epoch": 16.01,
      "learning_rate": 0.002149174750423314,
      "loss": 1.2817,
      "step": 79600
    },
    {
      "epoch": 16.03,
      "learning_rate": 0.0021479548746659964,
      "loss": 1.2792,
      "step": 79700
    },
    {
      "epoch": 16.05,
      "learning_rate": 0.002146733229238875,
      "loss": 1.2774,
      "step": 79800
    },
    {
      "epoch": 16.07,
      "learning_rate": 0.002145509816549542,
      "loss": 1.2808,
      "step": 79900
    },
    {
      "epoch": 16.09,
      "learning_rate": 0.0021442846390090724,
      "loss": 1.2771,
      "step": 80000
    },
    {
      "epoch": 16.11,
      "learning_rate": 0.0021430576990320196,
      "loss": 1.2769,
      "step": 80100
    },
    {
      "epoch": 16.13,
      "learning_rate": 0.002141828999036409,
      "loss": 1.2819,
      "step": 80200
    },
    {
      "epoch": 16.15,
      "learning_rate": 0.0021405985414437367,
      "loss": 1.2789,
      "step": 80300
    },
    {
      "epoch": 16.17,
      "learning_rate": 0.002139366328678961,
      "loss": 1.2777,
      "step": 80400
    },
    {
      "epoch": 16.19,
      "learning_rate": 0.0021381323631705002,
      "loss": 1.2804,
      "step": 80500
    },
    {
      "epoch": 16.21,
      "learning_rate": 0.002136896647350226,
      "loss": 1.2802,
      "step": 80600
    },
    {
      "epoch": 16.23,
      "learning_rate": 0.0021356591836534607,
      "loss": 1.2759,
      "step": 80700
    },
    {
      "epoch": 16.25,
      "learning_rate": 0.00213441997451897,
      "loss": 1.2834,
      "step": 80800
    },
    {
      "epoch": 16.27,
      "learning_rate": 0.002133179022388961,
      "loss": 1.2772,
      "step": 80900
    },
    {
      "epoch": 16.29,
      "learning_rate": 0.0021319363297090744,
      "loss": 1.2805,
      "step": 81000
    },
    {
      "epoch": 16.31,
      "learning_rate": 0.0021306918989283815,
      "loss": 1.2752,
      "step": 81100
    },
    {
      "epoch": 16.33,
      "learning_rate": 0.0021294457324993795,
      "loss": 1.279,
      "step": 81200
    },
    {
      "epoch": 16.35,
      "learning_rate": 0.0021281978328779863,
      "loss": 1.2825,
      "step": 81300
    },
    {
      "epoch": 16.37,
      "learning_rate": 0.0021269482025235344,
      "loss": 1.2824,
      "step": 81400
    },
    {
      "epoch": 16.39,
      "learning_rate": 0.002125696843898769,
      "loss": 1.2801,
      "step": 81500
    },
    {
      "epoch": 16.41,
      "learning_rate": 0.0021244437594698383,
      "loss": 1.2782,
      "step": 81600
    },
    {
      "epoch": 16.43,
      "learning_rate": 0.0021231889517062965,
      "loss": 1.2817,
      "step": 81700
    },
    {
      "epoch": 16.45,
      "learning_rate": 0.0021219324230810884,
      "loss": 1.2788,
      "step": 81800
    },
    {
      "epoch": 16.47,
      "learning_rate": 0.002120674176070555,
      "loss": 1.2796,
      "step": 81900
    },
    {
      "epoch": 16.49,
      "learning_rate": 0.0021194142131544212,
      "loss": 1.2799,
      "step": 82000
    },
    {
      "epoch": 16.51,
      "learning_rate": 0.002118152536815795,
      "loss": 1.2786,
      "step": 82100
    },
    {
      "epoch": 16.53,
      "learning_rate": 0.0021168891495411592,
      "loss": 1.2794,
      "step": 82200
    },
    {
      "epoch": 16.55,
      "learning_rate": 0.0021156240538203713,
      "loss": 1.2814,
      "step": 82300
    },
    {
      "epoch": 16.57,
      "learning_rate": 0.0021143572521466533,
      "loss": 1.2811,
      "step": 82400
    },
    {
      "epoch": 16.59,
      "learning_rate": 0.00211308874701659,
      "loss": 1.2821,
      "step": 82500
    },
    {
      "epoch": 16.61,
      "learning_rate": 0.002111818540930124,
      "loss": 1.2811,
      "step": 82600
    },
    {
      "epoch": 16.63,
      "learning_rate": 0.00211054663639055,
      "loss": 1.2759,
      "step": 82700
    },
    {
      "epoch": 16.65,
      "learning_rate": 0.0021092730359045086,
      "loss": 1.2797,
      "step": 82800
    },
    {
      "epoch": 16.67,
      "learning_rate": 0.0021079977419819853,
      "loss": 1.2784,
      "step": 82900
    },
    {
      "epoch": 16.69,
      "learning_rate": 0.0021067207571362997,
      "loss": 1.2816,
      "step": 83000
    },
    {
      "epoch": 16.71,
      "learning_rate": 0.0021054420838841066,
      "loss": 1.2801,
      "step": 83100
    },
    {
      "epoch": 16.73,
      "learning_rate": 0.0021041617247453863,
      "loss": 1.284,
      "step": 83200
    },
    {
      "epoch": 16.75,
      "learning_rate": 0.0021028796822434442,
      "loss": 1.2809,
      "step": 83300
    },
    {
      "epoch": 16.77,
      "learning_rate": 0.0021015959589049003,
      "loss": 1.2786,
      "step": 83400
    },
    {
      "epoch": 16.79,
      "learning_rate": 0.0021003105572596887,
      "loss": 1.2824,
      "step": 83500
    },
    {
      "epoch": 16.81,
      "learning_rate": 0.0020990234798410508,
      "loss": 1.2771,
      "step": 83600
    },
    {
      "epoch": 16.83,
      "learning_rate": 0.0020977347291855314,
      "loss": 1.2829,
      "step": 83700
    },
    {
      "epoch": 16.85,
      "learning_rate": 0.0020964443078329703,
      "loss": 1.2805,
      "step": 83800
    },
    {
      "epoch": 16.87,
      "learning_rate": 0.0020951522183265034,
      "loss": 1.2761,
      "step": 83900
    },
    {
      "epoch": 16.89,
      "learning_rate": 0.0020938584632125513,
      "loss": 1.2776,
      "step": 84000
    },
    {
      "epoch": 16.91,
      "learning_rate": 0.002092563045040819,
      "loss": 1.2783,
      "step": 84100
    },
    {
      "epoch": 16.93,
      "learning_rate": 0.0020912659663642878,
      "loss": 1.2802,
      "step": 84200
    },
    {
      "epoch": 16.95,
      "learning_rate": 0.002089967229739212,
      "loss": 1.2749,
      "step": 84300
    },
    {
      "epoch": 16.97,
      "learning_rate": 0.0020886668377251135,
      "loss": 1.2798,
      "step": 84400
    },
    {
      "epoch": 16.99,
      "learning_rate": 0.002087364792884776,
      "loss": 1.2791,
      "step": 84500
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.40644209745282756,
      "eval_loss": 1.2400243282318115,
      "eval_runtime": 19.8275,
      "eval_samples_per_second": 4013.004,
      "eval_steps_per_second": 15.685,
      "step": 84545
    },
    {
      "epoch": 17.01,
      "learning_rate": 0.0020860610977842414,
      "loss": 1.2757,
      "step": 84600
    },
    {
      "epoch": 17.03,
      "learning_rate": 0.0020847557549928037,
      "loss": 1.2693,
      "step": 84700
    },
    {
      "epoch": 17.05,
      "learning_rate": 0.002083448767083003,
      "loss": 1.2748,
      "step": 84800
    },
    {
      "epoch": 17.07,
      "learning_rate": 0.002082140136630623,
      "loss": 1.2713,
      "step": 84900
    },
    {
      "epoch": 17.09,
      "learning_rate": 0.002080829866214684,
      "loss": 1.2743,
      "step": 85000
    },
    {
      "epoch": 17.11,
      "learning_rate": 0.0020795310855919614,
      "loss": 1.2762,
      "step": 85100
    },
    {
      "epoch": 17.13,
      "learning_rate": 0.0020782175593340372,
      "loss": 1.2779,
      "step": 85200
    },
    {
      "epoch": 17.15,
      "learning_rate": 0.0020769024008430834,
      "loss": 1.272,
      "step": 85300
    },
    {
      "epoch": 17.17,
      "learning_rate": 0.0020755856127109857,
      "loss": 1.2736,
      "step": 85400
    },
    {
      "epoch": 17.19,
      "learning_rate": 0.0020742671975328406,
      "loss": 1.2722,
      "step": 85500
    },
    {
      "epoch": 17.21,
      "learning_rate": 0.0020729471579069526,
      "loss": 1.2711,
      "step": 85600
    },
    {
      "epoch": 17.23,
      "learning_rate": 0.002071625496434827,
      "loss": 1.2761,
      "step": 85700
    },
    {
      "epoch": 17.25,
      "learning_rate": 0.0020703022157211644,
      "loss": 1.2755,
      "step": 85800
    },
    {
      "epoch": 17.27,
      "learning_rate": 0.002068977318373858,
      "loss": 1.28,
      "step": 85900
    },
    {
      "epoch": 17.29,
      "learning_rate": 0.002067650807003987,
      "loss": 1.2775,
      "step": 86000
    },
    {
      "epoch": 17.31,
      "learning_rate": 0.00206632268422581,
      "loss": 1.2762,
      "step": 86100
    },
    {
      "epoch": 17.33,
      "learning_rate": 0.002064992952656763,
      "loss": 1.2791,
      "step": 86200
    },
    {
      "epoch": 17.35,
      "learning_rate": 0.0020636616149174508,
      "loss": 1.2801,
      "step": 86300
    },
    {
      "epoch": 17.37,
      "learning_rate": 0.002062328673631646,
      "loss": 1.2765,
      "step": 86400
    },
    {
      "epoch": 17.39,
      "learning_rate": 0.002060994131426279,
      "loss": 1.271,
      "step": 86500
    },
    {
      "epoch": 17.41,
      "learning_rate": 0.0020596713602392754,
      "loss": 1.2774,
      "step": 86600
    },
    {
      "epoch": 17.43,
      "learning_rate": 0.002058333640031713,
      "loss": 1.2742,
      "step": 86700
    },
    {
      "epoch": 17.45,
      "learning_rate": 0.002056994326777913,
      "loss": 1.2761,
      "step": 86800
    },
    {
      "epoch": 17.47,
      "learning_rate": 0.0020556534231173655,
      "loss": 1.2749,
      "step": 86900
    },
    {
      "epoch": 17.49,
      "learning_rate": 0.002054310931692694,
      "loss": 1.2757,
      "step": 87000
    },
    {
      "epoch": 17.51,
      "learning_rate": 0.0020529668551496506,
      "loss": 1.2755,
      "step": 87100
    },
    {
      "epoch": 17.53,
      "learning_rate": 0.0020516211961371126,
      "loss": 1.2766,
      "step": 87200
    },
    {
      "epoch": 17.55,
      "learning_rate": 0.0020502739573070757,
      "loss": 1.2719,
      "step": 87300
    },
    {
      "epoch": 17.57,
      "learning_rate": 0.002048925141314647,
      "loss": 1.2771,
      "step": 87400
    },
    {
      "epoch": 17.59,
      "learning_rate": 0.0020475747508180457,
      "loss": 1.2732,
      "step": 87500
    },
    {
      "epoch": 17.61,
      "learning_rate": 0.00204622278847859,
      "loss": 1.2741,
      "step": 87600
    },
    {
      "epoch": 17.63,
      "learning_rate": 0.0020448692569606988,
      "loss": 1.275,
      "step": 87700
    },
    {
      "epoch": 17.65,
      "learning_rate": 0.0020435141589318817,
      "loss": 1.2755,
      "step": 87800
    },
    {
      "epoch": 17.67,
      "learning_rate": 0.0020421574970627366,
      "loss": 1.2727,
      "step": 87900
    },
    {
      "epoch": 17.69,
      "learning_rate": 0.002040799274026943,
      "loss": 1.2753,
      "step": 88000
    },
    {
      "epoch": 17.71,
      "learning_rate": 0.0020394394925012565,
      "loss": 1.2749,
      "step": 88100
    },
    {
      "epoch": 17.73,
      "learning_rate": 0.002038078155165506,
      "loss": 1.2776,
      "step": 88200
    },
    {
      "epoch": 17.75,
      "learning_rate": 0.002036715264702584,
      "loss": 1.2752,
      "step": 88300
    },
    {
      "epoch": 17.78,
      "learning_rate": 0.0020353508237984466,
      "loss": 1.2771,
      "step": 88400
    },
    {
      "epoch": 17.8,
      "learning_rate": 0.002033984835142102,
      "loss": 1.2761,
      "step": 88500
    },
    {
      "epoch": 17.82,
      "learning_rate": 0.002032617301425613,
      "loss": 1.2707,
      "step": 88600
    },
    {
      "epoch": 17.84,
      "learning_rate": 0.0020312482253440835,
      "loss": 1.2749,
      "step": 88700
    },
    {
      "epoch": 17.86,
      "learning_rate": 0.0020298776095956594,
      "loss": 1.2735,
      "step": 88800
    },
    {
      "epoch": 17.88,
      "learning_rate": 0.00202850545688152,
      "loss": 1.2736,
      "step": 88900
    },
    {
      "epoch": 17.9,
      "learning_rate": 0.002027131769905874,
      "loss": 1.2763,
      "step": 89000
    },
    {
      "epoch": 17.92,
      "learning_rate": 0.002025756551375953,
      "loss": 1.2728,
      "step": 89100
    },
    {
      "epoch": 17.94,
      "learning_rate": 0.0020243798040020084,
      "loss": 1.2773,
      "step": 89200
    },
    {
      "epoch": 17.96,
      "learning_rate": 0.002023001530497303,
      "loss": 1.2736,
      "step": 89300
    },
    {
      "epoch": 17.98,
      "learning_rate": 0.0020216217335781088,
      "loss": 1.2755,
      "step": 89400
    },
    {
      "epoch": 18.0,
      "learning_rate": 0.002020240415963699,
      "loss": 1.2724,
      "step": 89500
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.40673402194635844,
      "eval_loss": 1.2363520860671997,
      "eval_runtime": 19.8757,
      "eval_samples_per_second": 4003.284,
      "eval_steps_per_second": 15.647,
      "step": 89518
    },
    {
      "epoch": 18.02,
      "learning_rate": 0.0020188575803763435,
      "loss": 1.2653,
      "step": 89600
    },
    {
      "epoch": 18.04,
      "learning_rate": 0.0020174732295413058,
      "loss": 1.265,
      "step": 89700
    },
    {
      "epoch": 18.06,
      "learning_rate": 0.0020160873661868328,
      "loss": 1.2724,
      "step": 89800
    },
    {
      "epoch": 18.08,
      "learning_rate": 0.0020146999930441547,
      "loss": 1.2692,
      "step": 89900
    },
    {
      "epoch": 18.1,
      "learning_rate": 0.002013311112847475,
      "loss": 1.2678,
      "step": 90000
    },
    {
      "epoch": 18.12,
      "learning_rate": 0.0020119207283339695,
      "loss": 1.2709,
      "step": 90100
    },
    {
      "epoch": 18.14,
      "learning_rate": 0.0020105288422437775,
      "loss": 1.2674,
      "step": 90200
    },
    {
      "epoch": 18.16,
      "learning_rate": 0.0020091354573199965,
      "loss": 1.2694,
      "step": 90300
    },
    {
      "epoch": 18.18,
      "learning_rate": 0.00200774057630868,
      "loss": 1.2691,
      "step": 90400
    },
    {
      "epoch": 18.2,
      "learning_rate": 0.0020063442019588283,
      "loss": 1.2665,
      "step": 90500
    },
    {
      "epoch": 18.22,
      "learning_rate": 0.002004946337022386,
      "loss": 1.2669,
      "step": 90600
    },
    {
      "epoch": 18.24,
      "learning_rate": 0.0020035469842542347,
      "loss": 1.272,
      "step": 90700
    },
    {
      "epoch": 18.26,
      "learning_rate": 0.002002146146412188,
      "loss": 1.2658,
      "step": 90800
    },
    {
      "epoch": 18.28,
      "learning_rate": 0.002000743826256986,
      "loss": 1.2719,
      "step": 90900
    },
    {
      "epoch": 18.3,
      "learning_rate": 0.0019993400265522917,
      "loss": 1.2716,
      "step": 91000
    },
    {
      "epoch": 18.32,
      "learning_rate": 0.001997934750064681,
      "loss": 1.2676,
      "step": 91100
    },
    {
      "epoch": 18.34,
      "learning_rate": 0.0019965279995636438,
      "loss": 1.2705,
      "step": 91200
    },
    {
      "epoch": 18.36,
      "learning_rate": 0.001995119777821572,
      "loss": 1.2741,
      "step": 91300
    },
    {
      "epoch": 18.38,
      "learning_rate": 0.0019937100876137592,
      "loss": 1.2688,
      "step": 91400
    },
    {
      "epoch": 18.4,
      "learning_rate": 0.001992298931718391,
      "loss": 1.2709,
      "step": 91500
    },
    {
      "epoch": 18.42,
      "learning_rate": 0.0019908863129165432,
      "loss": 1.2719,
      "step": 91600
    },
    {
      "epoch": 18.44,
      "learning_rate": 0.0019894722339921737,
      "loss": 1.2726,
      "step": 91700
    },
    {
      "epoch": 18.46,
      "learning_rate": 0.0019880566977321184,
      "loss": 1.2734,
      "step": 91800
    },
    {
      "epoch": 18.48,
      "learning_rate": 0.001986639706926085,
      "loss": 1.2703,
      "step": 91900
    },
    {
      "epoch": 18.5,
      "learning_rate": 0.0019852354559692456,
      "loss": 1.2692,
      "step": 92000
    },
    {
      "epoch": 18.52,
      "learning_rate": 0.0019838155789275737,
      "loss": 1.2722,
      "step": 92100
    },
    {
      "epoch": 18.54,
      "learning_rate": 0.0019823942556982275,
      "loss": 1.2754,
      "step": 92200
    },
    {
      "epoch": 18.56,
      "learning_rate": 0.001980971489082321,
      "loss": 1.2689,
      "step": 92300
    },
    {
      "epoch": 18.58,
      "learning_rate": 0.00197954728188381,
      "loss": 1.2756,
      "step": 92400
    },
    {
      "epoch": 18.6,
      "learning_rate": 0.0019781216369094915,
      "loss": 1.2725,
      "step": 92500
    },
    {
      "epoch": 18.62,
      "learning_rate": 0.001976694556968995,
      "loss": 1.2689,
      "step": 92600
    },
    {
      "epoch": 18.64,
      "learning_rate": 0.0019752660448747795,
      "loss": 1.2701,
      "step": 92700
    },
    {
      "epoch": 18.66,
      "learning_rate": 0.001973836103442124,
      "loss": 1.2694,
      "step": 92800
    },
    {
      "epoch": 18.68,
      "learning_rate": 0.0019724047354891263,
      "loss": 1.2722,
      "step": 92900
    },
    {
      "epoch": 18.7,
      "learning_rate": 0.001970971943836695,
      "loss": 1.2724,
      "step": 93000
    },
    {
      "epoch": 18.72,
      "learning_rate": 0.0019695377313085453,
      "loss": 1.2683,
      "step": 93100
    },
    {
      "epoch": 18.74,
      "learning_rate": 0.0019681021007311905,
      "loss": 1.2739,
      "step": 93200
    },
    {
      "epoch": 18.76,
      "learning_rate": 0.001966665054933941,
      "loss": 1.2722,
      "step": 93300
    },
    {
      "epoch": 18.78,
      "learning_rate": 0.0019652265967488943,
      "loss": 1.2705,
      "step": 93400
    },
    {
      "epoch": 18.8,
      "learning_rate": 0.0019637867290109327,
      "loss": 1.2687,
      "step": 93500
    },
    {
      "epoch": 18.82,
      "learning_rate": 0.001962345454557716,
      "loss": 1.2698,
      "step": 93600
    },
    {
      "epoch": 18.84,
      "learning_rate": 0.0019609027762296765,
      "loss": 1.2722,
      "step": 93700
    },
    {
      "epoch": 18.86,
      "learning_rate": 0.001959458696870013,
      "loss": 1.274,
      "step": 93800
    },
    {
      "epoch": 18.88,
      "learning_rate": 0.0019580132193246854,
      "loss": 1.2683,
      "step": 93900
    },
    {
      "epoch": 18.9,
      "learning_rate": 0.001956566346442409,
      "loss": 1.268,
      "step": 94000
    },
    {
      "epoch": 18.92,
      "learning_rate": 0.001955118081074649,
      "loss": 1.2674,
      "step": 94100
    },
    {
      "epoch": 18.94,
      "learning_rate": 0.001953668426075616,
      "loss": 1.267,
      "step": 94200
    },
    {
      "epoch": 18.96,
      "learning_rate": 0.0019522173843022578,
      "loss": 1.2677,
      "step": 94300
    },
    {
      "epoch": 18.98,
      "learning_rate": 0.0019507649586142553,
      "loss": 1.2669,
      "step": 94400
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.4069155401191752,
      "eval_loss": 1.2321081161499023,
      "eval_runtime": 19.6805,
      "eval_samples_per_second": 4042.992,
      "eval_steps_per_second": 15.802,
      "step": 94491
    },
    {
      "epoch": 19.0,
      "learning_rate": 0.0019493111518740181,
      "loss": 1.2724,
      "step": 94500
    },
    {
      "epoch": 19.02,
      "learning_rate": 0.001947855966946676,
      "loss": 1.2629,
      "step": 94600
    },
    {
      "epoch": 19.04,
      "learning_rate": 0.0019463994067000763,
      "loss": 1.2614,
      "step": 94700
    },
    {
      "epoch": 19.06,
      "learning_rate": 0.001944941474004775,
      "loss": 1.2603,
      "step": 94800
    },
    {
      "epoch": 19.08,
      "learning_rate": 0.0019434821717340346,
      "loss": 1.2643,
      "step": 94900
    },
    {
      "epoch": 19.1,
      "learning_rate": 0.001942021502763816,
      "loss": 1.2619,
      "step": 95000
    },
    {
      "epoch": 19.12,
      "learning_rate": 0.001940559469972774,
      "loss": 1.2676,
      "step": 95100
    },
    {
      "epoch": 19.14,
      "learning_rate": 0.0019390960762422499,
      "loss": 1.2606,
      "step": 95200
    },
    {
      "epoch": 19.16,
      "learning_rate": 0.0019376313244562687,
      "loss": 1.2628,
      "step": 95300
    },
    {
      "epoch": 19.18,
      "learning_rate": 0.0019361798852696786,
      "loss": 1.2691,
      "step": 95400
    },
    {
      "epoch": 19.2,
      "learning_rate": 0.0019347124395440384,
      "loss": 1.2652,
      "step": 95500
    },
    {
      "epoch": 19.22,
      "learning_rate": 0.0019332436444021162,
      "loss": 1.2672,
      "step": 95600
    },
    {
      "epoch": 19.24,
      "learning_rate": 0.0019317735027385814,
      "loss": 1.2668,
      "step": 95700
    },
    {
      "epoch": 19.26,
      "learning_rate": 0.0019303020174507568,
      "loss": 1.2604,
      "step": 95800
    },
    {
      "epoch": 19.28,
      "learning_rate": 0.001928829191438613,
      "loss": 1.2699,
      "step": 95900
    },
    {
      "epoch": 19.3,
      "learning_rate": 0.0019273550276047641,
      "loss": 1.2617,
      "step": 96000
    },
    {
      "epoch": 19.32,
      "learning_rate": 0.0019258795288544595,
      "loss": 1.2651,
      "step": 96100
    },
    {
      "epoch": 19.34,
      "learning_rate": 0.0019244026980955796,
      "loss": 1.2651,
      "step": 96200
    },
    {
      "epoch": 19.36,
      "learning_rate": 0.0019229245382386302,
      "loss": 1.2654,
      "step": 96300
    },
    {
      "epoch": 19.38,
      "learning_rate": 0.0019214450521967369,
      "loss": 1.2645,
      "step": 96400
    },
    {
      "epoch": 19.4,
      "learning_rate": 0.0019199642428856373,
      "loss": 1.2674,
      "step": 96500
    },
    {
      "epoch": 19.42,
      "learning_rate": 0.0019184821132236796,
      "loss": 1.2707,
      "step": 96600
    },
    {
      "epoch": 19.44,
      "learning_rate": 0.0019169986661318106,
      "loss": 1.2653,
      "step": 96700
    },
    {
      "epoch": 19.46,
      "learning_rate": 0.0019155139045335771,
      "loss": 1.2664,
      "step": 96800
    },
    {
      "epoch": 19.48,
      "learning_rate": 0.0019140278313551134,
      "loss": 1.2657,
      "step": 96900
    },
    {
      "epoch": 19.5,
      "learning_rate": 0.0019125404495251408,
      "loss": 1.2665,
      "step": 97000
    },
    {
      "epoch": 19.52,
      "learning_rate": 0.001911066655304144,
      "loss": 1.2674,
      "step": 97100
    },
    {
      "epoch": 19.54,
      "learning_rate": 0.0019095766779809568,
      "loss": 1.2644,
      "step": 97200
    },
    {
      "epoch": 19.56,
      "learning_rate": 0.0019080854007784964,
      "loss": 1.2648,
      "step": 97300
    },
    {
      "epoch": 19.58,
      "learning_rate": 0.0019065928266357385,
      "loss": 1.2669,
      "step": 97400
    },
    {
      "epoch": 19.6,
      "learning_rate": 0.001905098958494216,
      "loss": 1.2676,
      "step": 97500
    },
    {
      "epoch": 19.62,
      "learning_rate": 0.001903618757271021,
      "loss": 1.2662,
      "step": 97600
    },
    {
      "epoch": 19.65,
      "learning_rate": 0.001902122322833248,
      "loss": 1.2677,
      "step": 97700
    },
    {
      "epoch": 19.67,
      "learning_rate": 0.0019006246032070807,
      "loss": 1.2659,
      "step": 97800
    },
    {
      "epoch": 19.69,
      "learning_rate": 0.0018991256013441932,
      "loss": 1.2669,
      "step": 97900
    },
    {
      "epoch": 19.71,
      "learning_rate": 0.0018976253201987848,
      "loss": 1.2649,
      "step": 98000
    },
    {
      "epoch": 19.73,
      "learning_rate": 0.0018961237627275773,
      "loss": 1.2679,
      "step": 98100
    },
    {
      "epoch": 19.75,
      "learning_rate": 0.001894620931889807,
      "loss": 1.2692,
      "step": 98200
    },
    {
      "epoch": 19.77,
      "learning_rate": 0.0018931168306472199,
      "loss": 1.2655,
      "step": 98300
    },
    {
      "epoch": 19.79,
      "learning_rate": 0.0018916114619640656,
      "loss": 1.2646,
      "step": 98400
    },
    {
      "epoch": 19.81,
      "learning_rate": 0.0018901048288070927,
      "loss": 1.2647,
      "step": 98500
    },
    {
      "epoch": 19.83,
      "learning_rate": 0.0018885969341455395,
      "loss": 1.2651,
      "step": 98600
    },
    {
      "epoch": 19.85,
      "learning_rate": 0.0018870877809511327,
      "loss": 1.2642,
      "step": 98700
    },
    {
      "epoch": 19.87,
      "learning_rate": 0.001885577372198078,
      "loss": 1.2656,
      "step": 98800
    },
    {
      "epoch": 19.89,
      "learning_rate": 0.001884065710863056,
      "loss": 1.2627,
      "step": 98900
    },
    {
      "epoch": 19.91,
      "learning_rate": 0.0018825527999252157,
      "loss": 1.266,
      "step": 99000
    },
    {
      "epoch": 19.93,
      "learning_rate": 0.0018810386423661694,
      "loss": 1.2654,
      "step": 99100
    },
    {
      "epoch": 19.95,
      "learning_rate": 0.0018795232411699847,
      "loss": 1.264,
      "step": 99200
    },
    {
      "epoch": 19.97,
      "learning_rate": 0.0018780065993231816,
      "loss": 1.2633,
      "step": 99300
    },
    {
      "epoch": 19.99,
      "learning_rate": 0.0018764887198147245,
      "loss": 1.2683,
      "step": 99400
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.4075025708445861,
      "eval_loss": 1.2268297672271729,
      "eval_runtime": 19.801,
      "eval_samples_per_second": 4018.383,
      "eval_steps_per_second": 15.706,
      "step": 99465
    },
    {
      "epoch": 20.01,
      "learning_rate": 0.0018749696056360177,
      "loss": 1.2603,
      "step": 99500
    },
    {
      "epoch": 20.03,
      "learning_rate": 0.0018734492597808972,
      "loss": 1.2609,
      "step": 99600
    },
    {
      "epoch": 20.05,
      "learning_rate": 0.001871927685245628,
      "loss": 1.2587,
      "step": 99700
    },
    {
      "epoch": 20.07,
      "learning_rate": 0.0018704048850288952,
      "loss": 1.2575,
      "step": 99800
    },
    {
      "epoch": 20.09,
      "learning_rate": 0.001868880862131801,
      "loss": 1.2582,
      "step": 99900
    },
    {
      "epoch": 20.11,
      "learning_rate": 0.0018673556195578558,
      "loss": 1.2595,
      "step": 100000
    },
    {
      "epoch": 20.13,
      "learning_rate": 0.0018658291603129745,
      "loss": 1.2615,
      "step": 100100
    },
    {
      "epoch": 20.15,
      "learning_rate": 0.0018643014874054691,
      "loss": 1.2603,
      "step": 100200
    },
    {
      "epoch": 20.17,
      "learning_rate": 0.0018627726038460447,
      "loss": 1.2634,
      "step": 100300
    },
    {
      "epoch": 20.19,
      "learning_rate": 0.001861242512647791,
      "loss": 1.2634,
      "step": 100400
    },
    {
      "epoch": 20.21,
      "learning_rate": 0.0018597112168261781,
      "loss": 1.2646,
      "step": 100500
    },
    {
      "epoch": 20.23,
      "learning_rate": 0.0018581787193990508,
      "loss": 1.2591,
      "step": 100600
    },
    {
      "epoch": 20.25,
      "learning_rate": 0.0018566450233866208,
      "loss": 1.2609,
      "step": 100700
    },
    {
      "epoch": 20.27,
      "learning_rate": 0.0018551101318114629,
      "loss": 1.2647,
      "step": 100800
    },
    {
      "epoch": 20.29,
      "learning_rate": 0.0018535740476985083,
      "loss": 1.2634,
      "step": 100900
    },
    {
      "epoch": 20.31,
      "learning_rate": 0.0018520367740750374,
      "loss": 1.2589,
      "step": 101000
    },
    {
      "epoch": 20.33,
      "learning_rate": 0.0018504983139706755,
      "loss": 1.2615,
      "step": 101100
    },
    {
      "epoch": 20.35,
      "learning_rate": 0.001848958670417386,
      "loss": 1.2561,
      "step": 101200
    },
    {
      "epoch": 20.37,
      "learning_rate": 0.0018474178464494648,
      "loss": 1.2627,
      "step": 101300
    },
    {
      "epoch": 20.39,
      "learning_rate": 0.001845875845103534,
      "loss": 1.2583,
      "step": 101400
    },
    {
      "epoch": 20.41,
      "learning_rate": 0.0018443326694185364,
      "loss": 1.2633,
      "step": 101500
    },
    {
      "epoch": 20.43,
      "learning_rate": 0.0018427883224357284,
      "loss": 1.2611,
      "step": 101600
    },
    {
      "epoch": 20.45,
      "learning_rate": 0.0018412428071986754,
      "loss": 1.2625,
      "step": 101700
    },
    {
      "epoch": 20.47,
      "learning_rate": 0.001839711599315474,
      "loss": 1.2607,
      "step": 101800
    },
    {
      "epoch": 20.49,
      "learning_rate": 0.001838163768316336,
      "loss": 1.2623,
      "step": 101900
    },
    {
      "epoch": 20.51,
      "learning_rate": 0.0018366147781769232,
      "loss": 1.2569,
      "step": 102000
    },
    {
      "epoch": 20.53,
      "learning_rate": 0.001835064631949951,
      "loss": 1.2585,
      "step": 102100
    },
    {
      "epoch": 20.55,
      "learning_rate": 0.0018335133326904144,
      "loss": 1.2631,
      "step": 102200
    },
    {
      "epoch": 20.57,
      "learning_rate": 0.001831960883455579,
      "loss": 1.2592,
      "step": 102300
    },
    {
      "epoch": 20.59,
      "learning_rate": 0.0018304072873049785,
      "loss": 1.2613,
      "step": 102400
    },
    {
      "epoch": 20.61,
      "learning_rate": 0.0018288525473004055,
      "loss": 1.2601,
      "step": 102500
    },
    {
      "epoch": 20.63,
      "learning_rate": 0.0018272966665059086,
      "loss": 1.2608,
      "step": 102600
    },
    {
      "epoch": 20.65,
      "learning_rate": 0.0018257396479877822,
      "loss": 1.2608,
      "step": 102700
    },
    {
      "epoch": 20.67,
      "learning_rate": 0.0018241814948145656,
      "loss": 1.2619,
      "step": 102800
    },
    {
      "epoch": 20.69,
      "learning_rate": 0.0018226222100570322,
      "loss": 1.2599,
      "step": 102900
    },
    {
      "epoch": 20.71,
      "learning_rate": 0.0018210617967881865,
      "loss": 1.2589,
      "step": 103000
    },
    {
      "epoch": 20.73,
      "learning_rate": 0.0018195002580832567,
      "loss": 1.2573,
      "step": 103100
    },
    {
      "epoch": 20.75,
      "learning_rate": 0.0018179375970196893,
      "loss": 1.2582,
      "step": 103200
    },
    {
      "epoch": 20.77,
      "learning_rate": 0.0018163738166771422,
      "loss": 1.2581,
      "step": 103300
    },
    {
      "epoch": 20.79,
      "learning_rate": 0.0018148089201374795,
      "loss": 1.2628,
      "step": 103400
    },
    {
      "epoch": 20.81,
      "learning_rate": 0.0018132429104847653,
      "loss": 1.2598,
      "step": 103500
    },
    {
      "epoch": 20.83,
      "learning_rate": 0.0018116757908052572,
      "loss": 1.2618,
      "step": 103600
    },
    {
      "epoch": 20.85,
      "learning_rate": 0.0018101075641874,
      "loss": 1.266,
      "step": 103700
    },
    {
      "epoch": 20.87,
      "learning_rate": 0.0018085382337218203,
      "loss": 1.2608,
      "step": 103800
    },
    {
      "epoch": 20.89,
      "learning_rate": 0.0018069678025013202,
      "loss": 1.2583,
      "step": 103900
    },
    {
      "epoch": 20.91,
      "learning_rate": 0.0018053962736208717,
      "loss": 1.2578,
      "step": 104000
    },
    {
      "epoch": 20.93,
      "learning_rate": 0.0018038236501776092,
      "loss": 1.2621,
      "step": 104100
    },
    {
      "epoch": 20.95,
      "learning_rate": 0.0018022499352708247,
      "loss": 1.2574,
      "step": 104200
    },
    {
      "epoch": 20.97,
      "learning_rate": 0.0018006751320019604,
      "loss": 1.2589,
      "step": 104300
    },
    {
      "epoch": 20.99,
      "learning_rate": 0.001799099243474605,
      "loss": 1.2632,
      "step": 104400
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.40797378629408937,
      "eval_loss": 1.2228056192398071,
      "eval_runtime": 19.5365,
      "eval_samples_per_second": 4072.794,
      "eval_steps_per_second": 15.919,
      "step": 104438
    },
    {
      "epoch": 21.01,
      "learning_rate": 0.0017975222727944844,
      "loss": 1.2558,
      "step": 104500
    },
    {
      "epoch": 21.03,
      "learning_rate": 0.0017959442230694584,
      "loss": 1.2493,
      "step": 104600
    },
    {
      "epoch": 21.05,
      "learning_rate": 0.0017943650974095123,
      "loss": 1.2564,
      "step": 104700
    },
    {
      "epoch": 21.07,
      "learning_rate": 0.001792784898926753,
      "loss": 1.2552,
      "step": 104800
    },
    {
      "epoch": 21.09,
      "learning_rate": 0.0017912036307354006,
      "loss": 1.2495,
      "step": 104900
    },
    {
      "epoch": 21.11,
      "learning_rate": 0.0017896212959517843,
      "loss": 1.2554,
      "step": 105000
    },
    {
      "epoch": 21.13,
      "learning_rate": 0.0017880378976943344,
      "loss": 1.2587,
      "step": 105100
    },
    {
      "epoch": 21.15,
      "learning_rate": 0.0017864534390835776,
      "loss": 1.2574,
      "step": 105200
    },
    {
      "epoch": 21.17,
      "learning_rate": 0.00178486792324213,
      "loss": 1.2543,
      "step": 105300
    },
    {
      "epoch": 21.19,
      "learning_rate": 0.0017832813532946924,
      "loss": 1.2566,
      "step": 105400
    },
    {
      "epoch": 21.21,
      "learning_rate": 0.001781693732368041,
      "loss": 1.2563,
      "step": 105500
    },
    {
      "epoch": 21.23,
      "learning_rate": 0.001780105063591025,
      "loss": 1.259,
      "step": 105600
    },
    {
      "epoch": 21.25,
      "learning_rate": 0.0017785153500945576,
      "loss": 1.2568,
      "step": 105700
    },
    {
      "epoch": 21.27,
      "learning_rate": 0.001776924595011612,
      "loss": 1.2594,
      "step": 105800
    },
    {
      "epoch": 21.29,
      "learning_rate": 0.0017753328014772126,
      "loss": 1.2547,
      "step": 105900
    },
    {
      "epoch": 21.31,
      "learning_rate": 0.0017737399726284325,
      "loss": 1.2585,
      "step": 106000
    },
    {
      "epoch": 21.33,
      "learning_rate": 0.0017721461116043825,
      "loss": 1.2554,
      "step": 106100
    },
    {
      "epoch": 21.35,
      "learning_rate": 0.00177055122154621,
      "loss": 1.2586,
      "step": 106200
    },
    {
      "epoch": 21.37,
      "learning_rate": 0.001768955305597089,
      "loss": 1.2551,
      "step": 106300
    },
    {
      "epoch": 21.39,
      "learning_rate": 0.0017673583669022158,
      "loss": 1.256,
      "step": 106400
    },
    {
      "epoch": 21.41,
      "learning_rate": 0.0017657604086088023,
      "loss": 1.2537,
      "step": 106500
    },
    {
      "epoch": 21.43,
      "learning_rate": 0.0017641614338660694,
      "loss": 1.2552,
      "step": 106600
    },
    {
      "epoch": 21.45,
      "learning_rate": 0.0017625614458252417,
      "loss": 1.2555,
      "step": 106700
    },
    {
      "epoch": 21.47,
      "learning_rate": 0.0017609604476395407,
      "loss": 1.2555,
      "step": 106800
    },
    {
      "epoch": 21.49,
      "learning_rate": 0.0017593584424641785,
      "loss": 1.253,
      "step": 106900
    },
    {
      "epoch": 21.52,
      "learning_rate": 0.0017577714685050292,
      "loss": 1.2582,
      "step": 107000
    },
    {
      "epoch": 21.54,
      "learning_rate": 0.0017561674688150015,
      "loss": 1.2545,
      "step": 107100
    },
    {
      "epoch": 21.56,
      "learning_rate": 0.0017545624715812104,
      "loss": 1.2584,
      "step": 107200
    },
    {
      "epoch": 21.58,
      "learning_rate": 0.0017529564799667488,
      "loss": 1.2594,
      "step": 107300
    },
    {
      "epoch": 21.6,
      "learning_rate": 0.001751365571861091,
      "loss": 1.2576,
      "step": 107400
    },
    {
      "epoch": 21.62,
      "learning_rate": 0.0017497576108472055,
      "loss": 1.256,
      "step": 107500
    },
    {
      "epoch": 21.64,
      "learning_rate": 0.0017481486649219638,
      "loss": 1.2531,
      "step": 107600
    },
    {
      "epoch": 21.66,
      "learning_rate": 0.001746538737256242,
      "loss": 1.255,
      "step": 107700
    },
    {
      "epoch": 21.68,
      "learning_rate": 0.0017449278310228496,
      "loss": 1.258,
      "step": 107800
    },
    {
      "epoch": 21.7,
      "learning_rate": 0.0017433159493965259,
      "loss": 1.2543,
      "step": 107900
    },
    {
      "epoch": 21.72,
      "learning_rate": 0.0017417030955539316,
      "loss": 1.2588,
      "step": 108000
    },
    {
      "epoch": 21.74,
      "learning_rate": 0.0017400892726736443,
      "loss": 1.2568,
      "step": 108100
    },
    {
      "epoch": 21.76,
      "learning_rate": 0.0017384744839361499,
      "loss": 1.254,
      "step": 108200
    },
    {
      "epoch": 21.78,
      "learning_rate": 0.0017368587325238393,
      "loss": 1.2562,
      "step": 108300
    },
    {
      "epoch": 21.8,
      "learning_rate": 0.0017352420216209996,
      "loss": 1.2515,
      "step": 108400
    },
    {
      "epoch": 21.82,
      "learning_rate": 0.0017336243544138097,
      "loss": 1.2532,
      "step": 108500
    },
    {
      "epoch": 21.84,
      "learning_rate": 0.0017320057340903326,
      "loss": 1.2579,
      "step": 108600
    },
    {
      "epoch": 21.86,
      "learning_rate": 0.0017303861638405097,
      "loss": 1.2541,
      "step": 108700
    },
    {
      "epoch": 21.88,
      "learning_rate": 0.001728765646856154,
      "loss": 1.2553,
      "step": 108800
    },
    {
      "epoch": 21.9,
      "learning_rate": 0.0017271441863309462,
      "loss": 1.2541,
      "step": 108900
    },
    {
      "epoch": 21.92,
      "learning_rate": 0.0017255217854604242,
      "loss": 1.2558,
      "step": 109000
    },
    {
      "epoch": 21.94,
      "learning_rate": 0.0017238984474419804,
      "loss": 1.2548,
      "step": 109100
    },
    {
      "epoch": 21.96,
      "learning_rate": 0.0017222741754748536,
      "loss": 1.2568,
      "step": 109200
    },
    {
      "epoch": 21.98,
      "learning_rate": 0.0017206489727601237,
      "loss": 1.2524,
      "step": 109300
    },
    {
      "epoch": 22.0,
      "learning_rate": 0.0017190228425007044,
      "loss": 1.2563,
      "step": 109400
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.40835639033056986,
      "eval_loss": 1.2162704467773438,
      "eval_runtime": 19.8119,
      "eval_samples_per_second": 4016.173,
      "eval_steps_per_second": 15.698,
      "step": 109411
    },
    {
      "epoch": 22.02,
      "learning_rate": 0.001717395787901338,
      "loss": 1.2459,
      "step": 109500
    },
    {
      "epoch": 22.04,
      "learning_rate": 0.0017157678121685874,
      "loss": 1.2485,
      "step": 109600
    },
    {
      "epoch": 22.06,
      "learning_rate": 0.0017141389185108325,
      "loss": 1.25,
      "step": 109700
    },
    {
      "epoch": 22.08,
      "learning_rate": 0.0017125091101382601,
      "loss": 1.2473,
      "step": 109800
    },
    {
      "epoch": 22.1,
      "learning_rate": 0.0017108783902628616,
      "loss": 1.2489,
      "step": 109900
    },
    {
      "epoch": 22.12,
      "learning_rate": 0.0017092467620984239,
      "loss": 1.2498,
      "step": 110000
    },
    {
      "epoch": 22.14,
      "learning_rate": 0.0017076142288605242,
      "loss": 1.2485,
      "step": 110100
    },
    {
      "epoch": 22.16,
      "learning_rate": 0.0017059807937665227,
      "loss": 1.2542,
      "step": 110200
    },
    {
      "epoch": 22.18,
      "learning_rate": 0.0017043464600355585,
      "loss": 1.2515,
      "step": 110300
    },
    {
      "epoch": 22.2,
      "learning_rate": 0.0017027112308885396,
      "loss": 1.2534,
      "step": 110400
    },
    {
      "epoch": 22.22,
      "learning_rate": 0.0017010751095481403,
      "loss": 1.2504,
      "step": 110500
    },
    {
      "epoch": 22.24,
      "learning_rate": 0.001699438099238793,
      "loss": 1.25,
      "step": 110600
    },
    {
      "epoch": 22.26,
      "learning_rate": 0.0016978002031866812,
      "loss": 1.2516,
      "step": 110700
    },
    {
      "epoch": 22.28,
      "learning_rate": 0.0016961614246197348,
      "loss": 1.2541,
      "step": 110800
    },
    {
      "epoch": 22.3,
      "learning_rate": 0.001694521766767623,
      "loss": 1.2501,
      "step": 110900
    },
    {
      "epoch": 22.32,
      "learning_rate": 0.001692881232861747,
      "loss": 1.2477,
      "step": 111000
    },
    {
      "epoch": 22.34,
      "learning_rate": 0.001691239826135236,
      "loss": 1.2528,
      "step": 111100
    },
    {
      "epoch": 22.36,
      "learning_rate": 0.0016895975498229378,
      "loss": 1.2535,
      "step": 111200
    },
    {
      "epoch": 22.38,
      "learning_rate": 0.0016879544071614144,
      "loss": 1.2523,
      "step": 111300
    },
    {
      "epoch": 22.4,
      "learning_rate": 0.0016863104013889359,
      "loss": 1.2528,
      "step": 111400
    },
    {
      "epoch": 22.42,
      "learning_rate": 0.001684681988647626,
      "loss": 1.248,
      "step": 111500
    },
    {
      "epoch": 22.44,
      "learning_rate": 0.0016830362749250844,
      "loss": 1.2526,
      "step": 111600
    },
    {
      "epoch": 22.46,
      "learning_rate": 0.0016813897077841348,
      "loss": 1.2497,
      "step": 111700
    },
    {
      "epoch": 22.48,
      "learning_rate": 0.0016797422904697957,
      "loss": 1.2518,
      "step": 111800
    },
    {
      "epoch": 22.5,
      "learning_rate": 0.001678094026228761,
      "loss": 1.2476,
      "step": 111900
    },
    {
      "epoch": 22.52,
      "learning_rate": 0.0016764614135541252,
      "loss": 1.2516,
      "step": 112000
    },
    {
      "epoch": 22.54,
      "learning_rate": 0.001674811473594641,
      "loss": 1.2452,
      "step": 112100
    },
    {
      "epoch": 22.56,
      "learning_rate": 0.0016731606964260073,
      "loss": 1.2497,
      "step": 112200
    },
    {
      "epoch": 22.58,
      "learning_rate": 0.0016715090853015398,
      "loss": 1.2523,
      "step": 112300
    },
    {
      "epoch": 22.6,
      "learning_rate": 0.0016698566434761963,
      "loss": 1.2479,
      "step": 112400
    },
    {
      "epoch": 22.62,
      "learning_rate": 0.0016682033742065746,
      "loss": 1.2541,
      "step": 112500
    },
    {
      "epoch": 22.64,
      "learning_rate": 0.0016665492807509006,
      "loss": 1.2487,
      "step": 112600
    },
    {
      "epoch": 22.66,
      "learning_rate": 0.0016648943663690257,
      "loss": 1.2482,
      "step": 112700
    },
    {
      "epoch": 22.68,
      "learning_rate": 0.0016632386343224186,
      "loss": 1.2489,
      "step": 112800
    },
    {
      "epoch": 22.7,
      "learning_rate": 0.0016615820878741606,
      "loss": 1.2516,
      "step": 112900
    },
    {
      "epoch": 22.72,
      "learning_rate": 0.0016599247302889358,
      "loss": 1.2487,
      "step": 113000
    },
    {
      "epoch": 22.74,
      "learning_rate": 0.0016582665648330298,
      "loss": 1.2538,
      "step": 113100
    },
    {
      "epoch": 22.76,
      "learning_rate": 0.0016566075947743175,
      "loss": 1.2529,
      "step": 113200
    },
    {
      "epoch": 22.78,
      "learning_rate": 0.0016549478233822618,
      "loss": 1.2498,
      "step": 113300
    },
    {
      "epoch": 22.8,
      "learning_rate": 0.0016532872539279028,
      "loss": 1.2516,
      "step": 113400
    },
    {
      "epoch": 22.82,
      "learning_rate": 0.0016516258896838553,
      "loss": 1.2495,
      "step": 113500
    },
    {
      "epoch": 22.84,
      "learning_rate": 0.0016499637339242989,
      "loss": 1.2488,
      "step": 113600
    },
    {
      "epoch": 22.86,
      "learning_rate": 0.0016483007899249745,
      "loss": 1.2488,
      "step": 113700
    },
    {
      "epoch": 22.88,
      "learning_rate": 0.0016466370609631749,
      "loss": 1.2482,
      "step": 113800
    },
    {
      "epoch": 22.9,
      "learning_rate": 0.0016449725503177412,
      "loss": 1.2518,
      "step": 113900
    },
    {
      "epoch": 22.92,
      "learning_rate": 0.0016433072612690542,
      "loss": 1.2484,
      "step": 114000
    },
    {
      "epoch": 22.94,
      "learning_rate": 0.0016416411970990297,
      "loss": 1.2538,
      "step": 114100
    },
    {
      "epoch": 22.96,
      "learning_rate": 0.0016399743610911097,
      "loss": 1.2527,
      "step": 114200
    },
    {
      "epoch": 22.98,
      "learning_rate": 0.0016383067565302588,
      "loss": 1.2523,
      "step": 114300
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.4088067653938865,
      "eval_loss": 1.212856650352478,
      "eval_runtime": 19.7647,
      "eval_samples_per_second": 4025.761,
      "eval_steps_per_second": 15.735,
      "step": 114384
    },
    {
      "epoch": 23.0,
      "learning_rate": 0.001636638386702955,
      "loss": 1.2473,
      "step": 114400
    },
    {
      "epoch": 23.02,
      "learning_rate": 0.0016349692548971854,
      "loss": 1.2428,
      "step": 114500
    },
    {
      "epoch": 23.04,
      "learning_rate": 0.001633299364402438,
      "loss": 1.2422,
      "step": 114600
    },
    {
      "epoch": 23.06,
      "learning_rate": 0.0016316287185096973,
      "loss": 1.2432,
      "step": 114700
    },
    {
      "epoch": 23.08,
      "learning_rate": 0.0016299573205114343,
      "loss": 1.2439,
      "step": 114800
    },
    {
      "epoch": 23.1,
      "learning_rate": 0.001628301898865501,
      "loss": 1.2479,
      "step": 114900
    },
    {
      "epoch": 23.12,
      "learning_rate": 0.0016266290139783787,
      "loss": 1.2494,
      "step": 115000
    },
    {
      "epoch": 23.14,
      "learning_rate": 0.0016249553868390434,
      "loss": 1.2441,
      "step": 115100
    },
    {
      "epoch": 23.16,
      "learning_rate": 0.0016232810207458424,
      "loss": 1.245,
      "step": 115200
    },
    {
      "epoch": 23.18,
      "learning_rate": 0.0016216059189985796,
      "loss": 1.2475,
      "step": 115300
    },
    {
      "epoch": 23.2,
      "learning_rate": 0.0016199300848985091,
      "loss": 1.2451,
      "step": 115400
    },
    {
      "epoch": 23.22,
      "learning_rate": 0.0016182535217483282,
      "loss": 1.247,
      "step": 115500
    },
    {
      "epoch": 23.24,
      "learning_rate": 0.0016165762328521703,
      "loss": 1.2434,
      "step": 115600
    },
    {
      "epoch": 23.26,
      "learning_rate": 0.0016148982215156002,
      "loss": 1.2469,
      "step": 115700
    },
    {
      "epoch": 23.28,
      "learning_rate": 0.0016132194910456056,
      "loss": 1.245,
      "step": 115800
    },
    {
      "epoch": 23.3,
      "learning_rate": 0.0016115400447505918,
      "loss": 1.2507,
      "step": 115900
    },
    {
      "epoch": 23.32,
      "learning_rate": 0.0016098598859403746,
      "loss": 1.2434,
      "step": 116000
    },
    {
      "epoch": 23.34,
      "learning_rate": 0.0016081790179261746,
      "loss": 1.2418,
      "step": 116100
    },
    {
      "epoch": 23.37,
      "learning_rate": 0.0016064974440206093,
      "loss": 1.2476,
      "step": 116200
    },
    {
      "epoch": 23.39,
      "learning_rate": 0.0016048151675376878,
      "loss": 1.2492,
      "step": 116300
    },
    {
      "epoch": 23.41,
      "learning_rate": 0.001603132191792804,
      "loss": 1.2383,
      "step": 116400
    },
    {
      "epoch": 23.43,
      "learning_rate": 0.0016014485201027297,
      "loss": 1.2473,
      "step": 116500
    },
    {
      "epoch": 23.45,
      "learning_rate": 0.0015997641557856073,
      "loss": 1.2489,
      "step": 116600
    },
    {
      "epoch": 23.47,
      "learning_rate": 0.0015980791021609464,
      "loss": 1.2463,
      "step": 116700
    },
    {
      "epoch": 23.49,
      "learning_rate": 0.001596393362549613,
      "loss": 1.2469,
      "step": 116800
    },
    {
      "epoch": 23.51,
      "learning_rate": 0.0015947069402738262,
      "loss": 1.2443,
      "step": 116900
    },
    {
      "epoch": 23.53,
      "learning_rate": 0.0015930367130251378,
      "loss": 1.2448,
      "step": 117000
    },
    {
      "epoch": 23.55,
      "learning_rate": 0.0015913489421361716,
      "loss": 1.2469,
      "step": 117100
    },
    {
      "epoch": 23.57,
      "learning_rate": 0.0015896604985241852,
      "loss": 1.2453,
      "step": 117200
    },
    {
      "epoch": 23.59,
      "learning_rate": 0.0015879713855167263,
      "loss": 1.2467,
      "step": 117300
    },
    {
      "epoch": 23.61,
      "learning_rate": 0.0015862816064426619,
      "loss": 1.2437,
      "step": 117400
    },
    {
      "epoch": 23.63,
      "learning_rate": 0.0015845911646321712,
      "loss": 1.2453,
      "step": 117500
    },
    {
      "epoch": 23.65,
      "learning_rate": 0.00158290006341674,
      "loss": 1.2459,
      "step": 117600
    },
    {
      "epoch": 23.67,
      "learning_rate": 0.0015812083061291539,
      "loss": 1.2456,
      "step": 117700
    },
    {
      "epoch": 23.69,
      "learning_rate": 0.0015795158961034905,
      "loss": 1.2425,
      "step": 117800
    },
    {
      "epoch": 23.71,
      "learning_rate": 0.0015778228366751152,
      "loss": 1.2451,
      "step": 117900
    },
    {
      "epoch": 23.73,
      "learning_rate": 0.001576129131180672,
      "loss": 1.246,
      "step": 118000
    },
    {
      "epoch": 23.75,
      "learning_rate": 0.001574434782958078,
      "loss": 1.2433,
      "step": 118100
    },
    {
      "epoch": 23.77,
      "learning_rate": 0.001572739795346519,
      "loss": 1.2463,
      "step": 118200
    },
    {
      "epoch": 23.79,
      "learning_rate": 0.001571044171686438,
      "loss": 1.2468,
      "step": 118300
    },
    {
      "epoch": 23.81,
      "learning_rate": 0.001569347915319534,
      "loss": 1.2445,
      "step": 118400
    },
    {
      "epoch": 23.83,
      "learning_rate": 0.0015676510295887514,
      "loss": 1.246,
      "step": 118500
    },
    {
      "epoch": 23.85,
      "learning_rate": 0.0015659535178382759,
      "loss": 1.2459,
      "step": 118600
    },
    {
      "epoch": 23.87,
      "learning_rate": 0.0015642553834135256,
      "loss": 1.2422,
      "step": 118700
    },
    {
      "epoch": 23.89,
      "learning_rate": 0.0015625566296611475,
      "loss": 1.2426,
      "step": 118800
    },
    {
      "epoch": 23.91,
      "learning_rate": 0.0015608572599290084,
      "loss": 1.2443,
      "step": 118900
    },
    {
      "epoch": 23.93,
      "learning_rate": 0.001559157277566188,
      "loss": 1.245,
      "step": 119000
    },
    {
      "epoch": 23.95,
      "learning_rate": 0.001557456685922975,
      "loss": 1.2447,
      "step": 119100
    },
    {
      "epoch": 23.97,
      "learning_rate": 0.0015557554883508585,
      "loss": 1.245,
      "step": 119200
    },
    {
      "epoch": 23.99,
      "learning_rate": 0.0015540536882025203,
      "loss": 1.2394,
      "step": 119300
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.408813924305172,
      "eval_loss": 1.208786129951477,
      "eval_runtime": 19.9606,
      "eval_samples_per_second": 3986.246,
      "eval_steps_per_second": 15.581,
      "step": 119358
    },
    {
      "epoch": 24.01,
      "learning_rate": 0.001552351288831832,
      "loss": 1.239,
      "step": 119400
    },
    {
      "epoch": 24.03,
      "learning_rate": 0.0015506482935938443,
      "loss": 1.2356,
      "step": 119500
    },
    {
      "epoch": 24.05,
      "learning_rate": 0.0015489447058447835,
      "loss": 1.2393,
      "step": 119600
    },
    {
      "epoch": 24.07,
      "learning_rate": 0.0015472405289420421,
      "loss": 1.2373,
      "step": 119700
    },
    {
      "epoch": 24.09,
      "learning_rate": 0.0015455357662441758,
      "loss": 1.2384,
      "step": 119800
    },
    {
      "epoch": 24.11,
      "learning_rate": 0.0015438304211108924,
      "loss": 1.2399,
      "step": 119900
    },
    {
      "epoch": 24.13,
      "learning_rate": 0.00154212449690305,
      "loss": 1.238,
      "step": 120000
    },
    {
      "epoch": 24.15,
      "learning_rate": 0.0015404179969826454,
      "loss": 1.2407,
      "step": 120100
    },
    {
      "epoch": 24.17,
      "learning_rate": 0.0015387109247128126,
      "loss": 1.2427,
      "step": 120200
    },
    {
      "epoch": 24.19,
      "learning_rate": 0.001537003283457811,
      "loss": 1.2432,
      "step": 120300
    },
    {
      "epoch": 24.21,
      "learning_rate": 0.0015352950765830234,
      "loss": 1.2396,
      "step": 120400
    },
    {
      "epoch": 24.23,
      "learning_rate": 0.001533586307454946,
      "loss": 1.2404,
      "step": 120500
    },
    {
      "epoch": 24.25,
      "learning_rate": 0.0015318769794411841,
      "loss": 1.2383,
      "step": 120600
    },
    {
      "epoch": 24.27,
      "learning_rate": 0.0015301670959104435,
      "loss": 1.2408,
      "step": 120700
    },
    {
      "epoch": 24.29,
      "learning_rate": 0.0015284566602325254,
      "loss": 1.2389,
      "step": 120800
    },
    {
      "epoch": 24.31,
      "learning_rate": 0.0015267456757783189,
      "loss": 1.2397,
      "step": 120900
    },
    {
      "epoch": 24.33,
      "learning_rate": 0.0015250341459197947,
      "loss": 1.2415,
      "step": 121000
    },
    {
      "epoch": 24.35,
      "learning_rate": 0.0015233563207578788,
      "loss": 1.2356,
      "step": 121100
    },
    {
      "epoch": 24.37,
      "learning_rate": 0.0015216437209509943,
      "loss": 1.2364,
      "step": 121200
    },
    {
      "epoch": 24.39,
      "learning_rate": 0.001519930585794613,
      "loss": 1.2443,
      "step": 121300
    },
    {
      "epoch": 24.41,
      "learning_rate": 0.0015182169186649438,
      "loss": 1.2386,
      "step": 121400
    },
    {
      "epoch": 24.43,
      "learning_rate": 0.001516502722939245,
      "loss": 1.2404,
      "step": 121500
    },
    {
      "epoch": 24.45,
      "learning_rate": 0.0015147880019958154,
      "loss": 1.2416,
      "step": 121600
    },
    {
      "epoch": 24.47,
      "learning_rate": 0.0015130727592139904,
      "loss": 1.2418,
      "step": 121700
    },
    {
      "epoch": 24.49,
      "learning_rate": 0.0015113741581417957,
      "loss": 1.2432,
      "step": 121800
    },
    {
      "epoch": 24.51,
      "learning_rate": 0.001509657886959314,
      "loss": 1.2408,
      "step": 121900
    },
    {
      "epoch": 24.53,
      "learning_rate": 0.0015079411040487545,
      "loss": 1.242,
      "step": 122000
    },
    {
      "epoch": 24.55,
      "learning_rate": 0.0015062238127935158,
      "loss": 1.2398,
      "step": 122100
    },
    {
      "epoch": 24.57,
      "learning_rate": 0.0015045060165779975,
      "loss": 1.2379,
      "step": 122200
    },
    {
      "epoch": 24.59,
      "learning_rate": 0.001502787718787595,
      "loss": 1.2364,
      "step": 122300
    },
    {
      "epoch": 24.61,
      "learning_rate": 0.0015010689228086916,
      "loss": 1.2375,
      "step": 122400
    },
    {
      "epoch": 24.63,
      "learning_rate": 0.0014993496320286532,
      "loss": 1.2353,
      "step": 122500
    },
    {
      "epoch": 24.65,
      "learning_rate": 0.00149762984983582,
      "loss": 1.2391,
      "step": 122600
    },
    {
      "epoch": 24.67,
      "learning_rate": 0.0014959095796195018,
      "loss": 1.2424,
      "step": 122700
    },
    {
      "epoch": 24.69,
      "learning_rate": 0.0014941888247699687,
      "loss": 1.2423,
      "step": 122800
    },
    {
      "epoch": 24.71,
      "learning_rate": 0.0014924675886784473,
      "loss": 1.2402,
      "step": 122900
    },
    {
      "epoch": 24.73,
      "learning_rate": 0.001490745874737111,
      "loss": 1.2396,
      "step": 123000
    },
    {
      "epoch": 24.75,
      "learning_rate": 0.001489023686339077,
      "loss": 1.2386,
      "step": 123100
    },
    {
      "epoch": 24.77,
      "learning_rate": 0.001487301026878396,
      "loss": 1.2402,
      "step": 123200
    },
    {
      "epoch": 24.79,
      "learning_rate": 0.001485577899750048,
      "loss": 1.2379,
      "step": 123300
    },
    {
      "epoch": 24.81,
      "learning_rate": 0.0014838543083499334,
      "loss": 1.2432,
      "step": 123400
    },
    {
      "epoch": 24.83,
      "learning_rate": 0.001482130256074869,
      "loss": 1.24,
      "step": 123500
    },
    {
      "epoch": 24.85,
      "learning_rate": 0.001480405746322579,
      "loss": 1.2389,
      "step": 123600
    },
    {
      "epoch": 24.87,
      "learning_rate": 0.0014786807824916897,
      "loss": 1.2393,
      "step": 123700
    },
    {
      "epoch": 24.89,
      "learning_rate": 0.0014769553679817215,
      "loss": 1.237,
      "step": 123800
    },
    {
      "epoch": 24.91,
      "learning_rate": 0.0014752295061930846,
      "loss": 1.2358,
      "step": 123900
    },
    {
      "epoch": 24.93,
      "learning_rate": 0.0014735032005270684,
      "loss": 1.2392,
      "step": 124000
    },
    {
      "epoch": 24.95,
      "learning_rate": 0.0014717764543858392,
      "loss": 1.2389,
      "step": 124100
    },
    {
      "epoch": 24.97,
      "learning_rate": 0.00147004927117243,
      "loss": 1.2392,
      "step": 124200
    },
    {
      "epoch": 24.99,
      "learning_rate": 0.001468321654290736,
      "loss": 1.2387,
      "step": 124300
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.40966806196499067,
      "eval_loss": 1.2022136449813843,
      "eval_runtime": 19.7693,
      "eval_samples_per_second": 4024.827,
      "eval_steps_per_second": 15.731,
      "step": 124331
    },
    {
      "epoch": 25.01,
      "learning_rate": 0.0014665936071455062,
      "loss": 1.2317,
      "step": 124400
    },
    {
      "epoch": 25.03,
      "learning_rate": 0.0014648651331423384,
      "loss": 1.23,
      "step": 124500
    },
    {
      "epoch": 25.05,
      "learning_rate": 0.0014631362356876715,
      "loss": 1.2321,
      "step": 124600
    },
    {
      "epoch": 25.07,
      "learning_rate": 0.0014614069181887784,
      "loss": 1.2331,
      "step": 124700
    },
    {
      "epoch": 25.09,
      "learning_rate": 0.0014596771840537605,
      "loss": 1.2315,
      "step": 124800
    },
    {
      "epoch": 25.11,
      "learning_rate": 0.0014579470366915396,
      "loss": 1.2357,
      "step": 124900
    },
    {
      "epoch": 25.13,
      "learning_rate": 0.0014562164795118523,
      "loss": 1.2365,
      "step": 125000
    },
    {
      "epoch": 25.15,
      "learning_rate": 0.0014544855159252432,
      "loss": 1.2316,
      "step": 125100
    },
    {
      "epoch": 25.17,
      "learning_rate": 0.001452754149343057,
      "loss": 1.2337,
      "step": 125200
    },
    {
      "epoch": 25.19,
      "learning_rate": 0.001451022383177433,
      "loss": 1.235,
      "step": 125300
    },
    {
      "epoch": 25.21,
      "learning_rate": 0.0014492902208412984,
      "loss": 1.2352,
      "step": 125400
    },
    {
      "epoch": 25.24,
      "learning_rate": 0.0014475576657483604,
      "loss": 1.231,
      "step": 125500
    },
    {
      "epoch": 25.26,
      "learning_rate": 0.001445824721313101,
      "loss": 1.2361,
      "step": 125600
    },
    {
      "epoch": 25.28,
      "learning_rate": 0.0014441087261535155,
      "loss": 1.2323,
      "step": 125700
    },
    {
      "epoch": 25.3,
      "learning_rate": 0.0014423750170883193,
      "loss": 1.2378,
      "step": 125800
    },
    {
      "epoch": 25.32,
      "learning_rate": 0.001440640928894652,
      "loss": 1.2323,
      "step": 125900
    },
    {
      "epoch": 25.34,
      "learning_rate": 0.0014389064649900165,
      "loss": 1.2387,
      "step": 126000
    },
    {
      "epoch": 25.36,
      "learning_rate": 0.001437171628792657,
      "loss": 1.2401,
      "step": 126100
    },
    {
      "epoch": 25.38,
      "learning_rate": 0.0014354364237215494,
      "loss": 1.2334,
      "step": 126200
    },
    {
      "epoch": 25.4,
      "learning_rate": 0.0014337008531963994,
      "loss": 1.2348,
      "step": 126300
    },
    {
      "epoch": 25.42,
      "learning_rate": 0.0014319649206376301,
      "loss": 1.236,
      "step": 126400
    },
    {
      "epoch": 25.44,
      "learning_rate": 0.00143022862946638,
      "loss": 1.2343,
      "step": 126500
    },
    {
      "epoch": 25.46,
      "learning_rate": 0.0014284919831044935,
      "loss": 1.2348,
      "step": 126600
    },
    {
      "epoch": 25.48,
      "learning_rate": 0.0014267549849745155,
      "loss": 1.2337,
      "step": 126700
    },
    {
      "epoch": 25.5,
      "learning_rate": 0.0014250176384996832,
      "loss": 1.2328,
      "step": 126800
    },
    {
      "epoch": 25.52,
      "learning_rate": 0.0014232799471039221,
      "loss": 1.2345,
      "step": 126900
    },
    {
      "epoch": 25.54,
      "learning_rate": 0.0014215419142118353,
      "loss": 1.2391,
      "step": 127000
    },
    {
      "epoch": 25.56,
      "learning_rate": 0.0014198035432487004,
      "loss": 1.2323,
      "step": 127100
    },
    {
      "epoch": 25.58,
      "learning_rate": 0.0014180648376404608,
      "loss": 1.2352,
      "step": 127200
    },
    {
      "epoch": 25.6,
      "learning_rate": 0.0014163258008137198,
      "loss": 1.2319,
      "step": 127300
    },
    {
      "epoch": 25.62,
      "learning_rate": 0.0014145864361957325,
      "loss": 1.2329,
      "step": 127400
    },
    {
      "epoch": 25.64,
      "learning_rate": 0.001412846747214401,
      "loss": 1.2351,
      "step": 127500
    },
    {
      "epoch": 25.66,
      "learning_rate": 0.001411106737298266,
      "loss": 1.2341,
      "step": 127600
    },
    {
      "epoch": 25.68,
      "learning_rate": 0.001409366409876501,
      "loss": 1.2328,
      "step": 127700
    },
    {
      "epoch": 25.7,
      "learning_rate": 0.001407625768378905,
      "loss": 1.235,
      "step": 127800
    },
    {
      "epoch": 25.72,
      "learning_rate": 0.0014058848162358966,
      "loss": 1.2304,
      "step": 127900
    },
    {
      "epoch": 25.74,
      "learning_rate": 0.001404160970981524,
      "loss": 1.2324,
      "step": 128000
    },
    {
      "epoch": 25.76,
      "learning_rate": 0.0014024194108622253,
      "loss": 1.2356,
      "step": 128100
    },
    {
      "epoch": 25.78,
      "learning_rate": 0.0014006775503580902,
      "loss": 1.2296,
      "step": 128200
    },
    {
      "epoch": 25.8,
      "learning_rate": 0.0013989353929019378,
      "loss": 1.2354,
      "step": 128300
    },
    {
      "epoch": 25.82,
      "learning_rate": 0.0013971929419271745,
      "loss": 1.2314,
      "step": 128400
    },
    {
      "epoch": 25.84,
      "learning_rate": 0.0013954502008677843,
      "loss": 1.2336,
      "step": 128500
    },
    {
      "epoch": 25.86,
      "learning_rate": 0.0013937071731583237,
      "loss": 1.2356,
      "step": 128600
    },
    {
      "epoch": 25.88,
      "learning_rate": 0.001391963862233913,
      "loss": 1.2314,
      "step": 128700
    },
    {
      "epoch": 25.9,
      "learning_rate": 0.0013902202715302314,
      "loss": 1.2336,
      "step": 128800
    },
    {
      "epoch": 25.92,
      "learning_rate": 0.0013884764044835088,
      "loss": 1.234,
      "step": 128900
    },
    {
      "epoch": 25.94,
      "learning_rate": 0.001386732264530521,
      "loss": 1.2355,
      "step": 129000
    },
    {
      "epoch": 25.96,
      "learning_rate": 0.0013849878551085804,
      "loss": 1.2334,
      "step": 129100
    },
    {
      "epoch": 25.98,
      "learning_rate": 0.0013832431796555308,
      "loss": 1.2339,
      "step": 129200
    },
    {
      "epoch": 26.0,
      "learning_rate": 0.0013814982416097406,
      "loss": 1.234,
      "step": 129300
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.4099652363267976,
      "eval_loss": 1.1980103254318237,
      "eval_runtime": 19.8186,
      "eval_samples_per_second": 4014.806,
      "eval_steps_per_second": 15.692,
      "step": 129304
    },
    {
      "epoch": 26.02,
      "learning_rate": 0.0013797530444100952,
      "loss": 1.2258,
      "step": 129400
    },
    {
      "epoch": 26.04,
      "learning_rate": 0.0013780075914959912,
      "loss": 1.2242,
      "step": 129500
    },
    {
      "epoch": 26.06,
      "learning_rate": 0.001376261886307329,
      "loss": 1.2273,
      "step": 129600
    },
    {
      "epoch": 26.08,
      "learning_rate": 0.001374515932284506,
      "loss": 1.2251,
      "step": 129700
    },
    {
      "epoch": 26.1,
      "learning_rate": 0.001372787196065969,
      "loss": 1.23,
      "step": 129800
    },
    {
      "epoch": 26.12,
      "learning_rate": 0.0013710407571004548,
      "loss": 1.2325,
      "step": 129900
    },
    {
      "epoch": 26.14,
      "learning_rate": 0.0013692940795904673,
      "loss": 1.2274,
      "step": 130000
    },
    {
      "epoch": 26.16,
      "learning_rate": 0.0013675471669783198,
      "loss": 1.2278,
      "step": 130100
    },
    {
      "epoch": 26.18,
      "learning_rate": 0.0013658000227067901,
      "loss": 1.2245,
      "step": 130200
    },
    {
      "epoch": 26.2,
      "learning_rate": 0.0013640526502191111,
      "loss": 1.2266,
      "step": 130300
    },
    {
      "epoch": 26.22,
      "learning_rate": 0.0013623050529589667,
      "loss": 1.2315,
      "step": 130400
    },
    {
      "epoch": 26.24,
      "learning_rate": 0.001360557234370483,
      "loss": 1.2302,
      "step": 130500
    },
    {
      "epoch": 26.26,
      "learning_rate": 0.001358809197898223,
      "loss": 1.2324,
      "step": 130600
    },
    {
      "epoch": 26.28,
      "learning_rate": 0.0013570609469871781,
      "loss": 1.2305,
      "step": 130700
    },
    {
      "epoch": 26.3,
      "learning_rate": 0.001355312485082764,
      "loss": 1.2294,
      "step": 130800
    },
    {
      "epoch": 26.32,
      "learning_rate": 0.0013535638156308098,
      "loss": 1.2278,
      "step": 130900
    },
    {
      "epoch": 26.34,
      "learning_rate": 0.0013518149420775557,
      "loss": 1.224,
      "step": 131000
    },
    {
      "epoch": 26.36,
      "learning_rate": 0.0013500658678696432,
      "loss": 1.2296,
      "step": 131100
    },
    {
      "epoch": 26.38,
      "learning_rate": 0.00134831659645411,
      "loss": 1.228,
      "step": 131200
    },
    {
      "epoch": 26.4,
      "learning_rate": 0.0013465671312783809,
      "loss": 1.2245,
      "step": 131300
    },
    {
      "epoch": 26.42,
      "learning_rate": 0.0013448174757902645,
      "loss": 1.2305,
      "step": 131400
    },
    {
      "epoch": 26.44,
      "learning_rate": 0.0013430676334379426,
      "loss": 1.2305,
      "step": 131500
    },
    {
      "epoch": 26.46,
      "learning_rate": 0.0013413176076699674,
      "loss": 1.2294,
      "step": 131600
    },
    {
      "epoch": 26.48,
      "learning_rate": 0.0013395674019352503,
      "loss": 1.2312,
      "step": 131700
    },
    {
      "epoch": 26.5,
      "learning_rate": 0.0013378170196830588,
      "loss": 1.2278,
      "step": 131800
    },
    {
      "epoch": 26.52,
      "learning_rate": 0.001336066464363008,
      "loss": 1.2289,
      "step": 131900
    },
    {
      "epoch": 26.54,
      "learning_rate": 0.0013343157394250537,
      "loss": 1.2242,
      "step": 132000
    },
    {
      "epoch": 26.56,
      "learning_rate": 0.0013325648483194865,
      "loss": 1.2275,
      "step": 132100
    },
    {
      "epoch": 26.58,
      "learning_rate": 0.0013308137944969248,
      "loss": 1.2262,
      "step": 132200
    },
    {
      "epoch": 26.6,
      "learning_rate": 0.001329062581408306,
      "loss": 1.2289,
      "step": 132300
    },
    {
      "epoch": 26.62,
      "learning_rate": 0.0013273112125048833,
      "loss": 1.2341,
      "step": 132400
    },
    {
      "epoch": 26.64,
      "learning_rate": 0.0013255596912382156,
      "loss": 1.2288,
      "step": 132500
    },
    {
      "epoch": 26.66,
      "learning_rate": 0.0013238080210601635,
      "loss": 1.2267,
      "step": 132600
    },
    {
      "epoch": 26.68,
      "learning_rate": 0.0013220562054228793,
      "loss": 1.2267,
      "step": 132700
    },
    {
      "epoch": 26.7,
      "learning_rate": 0.0013203042477788038,
      "loss": 1.2287,
      "step": 132800
    },
    {
      "epoch": 26.72,
      "learning_rate": 0.001318552151580656,
      "loss": 1.226,
      "step": 132900
    },
    {
      "epoch": 26.74,
      "learning_rate": 0.001316799920281429,
      "loss": 1.2294,
      "step": 133000
    },
    {
      "epoch": 26.76,
      "learning_rate": 0.0013150475573343817,
      "loss": 1.2272,
      "step": 133100
    },
    {
      "epoch": 26.78,
      "learning_rate": 0.0013132950661930332,
      "loss": 1.23,
      "step": 133200
    },
    {
      "epoch": 26.8,
      "learning_rate": 0.001311542450311154,
      "loss": 1.2282,
      "step": 133300
    },
    {
      "epoch": 26.82,
      "learning_rate": 0.0013097897131427616,
      "loss": 1.2297,
      "step": 133400
    },
    {
      "epoch": 26.84,
      "learning_rate": 0.0013080368581421117,
      "loss": 1.2269,
      "step": 133500
    },
    {
      "epoch": 26.86,
      "learning_rate": 0.0013062838887636927,
      "loss": 1.2283,
      "step": 133600
    },
    {
      "epoch": 26.88,
      "learning_rate": 0.001304530808462218,
      "loss": 1.2292,
      "step": 133700
    },
    {
      "epoch": 26.9,
      "learning_rate": 0.0013027776206926205,
      "loss": 1.2306,
      "step": 133800
    },
    {
      "epoch": 26.92,
      "learning_rate": 0.0013010243289100437,
      "loss": 1.2264,
      "step": 133900
    },
    {
      "epoch": 26.94,
      "learning_rate": 0.0012992709365698368,
      "loss": 1.2249,
      "step": 134000
    },
    {
      "epoch": 26.96,
      "learning_rate": 0.001297517447127547,
      "loss": 1.2279,
      "step": 134100
    },
    {
      "epoch": 26.98,
      "learning_rate": 0.001295763864038913,
      "loss": 1.2272,
      "step": 134200
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.41072249005388595,
      "eval_loss": 1.1898874044418335,
      "eval_runtime": 19.6468,
      "eval_samples_per_second": 4049.93,
      "eval_steps_per_second": 15.83,
      "step": 134277
    },
    {
      "epoch": 27.0,
      "learning_rate": 0.0012940101907598575,
      "loss": 1.2222,
      "step": 134300
    },
    {
      "epoch": 27.02,
      "learning_rate": 0.0012922564307464824,
      "loss": 1.2168,
      "step": 134400
    },
    {
      "epoch": 27.04,
      "learning_rate": 0.0012905025874550586,
      "loss": 1.2195,
      "step": 134500
    },
    {
      "epoch": 27.06,
      "learning_rate": 0.0012887486643420223,
      "loss": 1.2188,
      "step": 134600
    },
    {
      "epoch": 27.08,
      "learning_rate": 0.001286994664863967,
      "loss": 1.2162,
      "step": 134700
    },
    {
      "epoch": 27.11,
      "learning_rate": 0.0012852405924776362,
      "loss": 1.2236,
      "step": 134800
    },
    {
      "epoch": 27.13,
      "learning_rate": 0.0012834864506399174,
      "loss": 1.2245,
      "step": 134900
    },
    {
      "epoch": 27.15,
      "learning_rate": 0.001281732242807835,
      "loss": 1.2193,
      "step": 135000
    },
    {
      "epoch": 27.17,
      "learning_rate": 0.0012799779724385432,
      "loss": 1.2235,
      "step": 135100
    },
    {
      "epoch": 27.19,
      "learning_rate": 0.0012782236429893202,
      "loss": 1.2252,
      "step": 135200
    },
    {
      "epoch": 27.21,
      "learning_rate": 0.0012764692579175594,
      "loss": 1.2253,
      "step": 135300
    },
    {
      "epoch": 27.23,
      "learning_rate": 0.0012747148206807646,
      "loss": 1.2226,
      "step": 135400
    },
    {
      "epoch": 27.25,
      "learning_rate": 0.0012729603347365424,
      "loss": 1.2217,
      "step": 135500
    },
    {
      "epoch": 27.27,
      "learning_rate": 0.0012712058035425956,
      "loss": 1.2222,
      "step": 135600
    },
    {
      "epoch": 27.29,
      "learning_rate": 0.0012694512305567152,
      "loss": 1.2209,
      "step": 135700
    },
    {
      "epoch": 27.31,
      "learning_rate": 0.0012677141655283745,
      "loss": 1.2256,
      "step": 135800
    },
    {
      "epoch": 27.33,
      "learning_rate": 0.0012659595196639695,
      "loss": 1.2215,
      "step": 135900
    },
    {
      "epoch": 27.35,
      "learning_rate": 0.0012642048423468924,
      "loss": 1.2238,
      "step": 136000
    },
    {
      "epoch": 27.37,
      "learning_rate": 0.0012624501370352233,
      "loss": 1.222,
      "step": 136100
    },
    {
      "epoch": 27.39,
      "learning_rate": 0.0012606954071870964,
      "loss": 1.2196,
      "step": 136200
    },
    {
      "epoch": 27.41,
      "learning_rate": 0.0012589406562606954,
      "loss": 1.2271,
      "step": 136300
    },
    {
      "epoch": 27.43,
      "learning_rate": 0.0012571858877142449,
      "loss": 1.2228,
      "step": 136400
    },
    {
      "epoch": 27.45,
      "learning_rate": 0.001255431105006004,
      "loss": 1.2225,
      "step": 136500
    },
    {
      "epoch": 27.47,
      "learning_rate": 0.0012536763115942604,
      "loss": 1.2253,
      "step": 136600
    },
    {
      "epoch": 27.49,
      "learning_rate": 0.0012519215109373229,
      "loss": 1.222,
      "step": 136700
    },
    {
      "epoch": 27.51,
      "learning_rate": 0.001250166706493513,
      "loss": 1.2234,
      "step": 136800
    },
    {
      "epoch": 27.53,
      "learning_rate": 0.001248411901721162,
      "loss": 1.2235,
      "step": 136900
    },
    {
      "epoch": 27.55,
      "learning_rate": 0.0012466571000786,
      "loss": 1.2218,
      "step": 137000
    },
    {
      "epoch": 27.57,
      "learning_rate": 0.001244902305024152,
      "loss": 1.2255,
      "step": 137100
    },
    {
      "epoch": 27.59,
      "learning_rate": 0.0012431475200161302,
      "loss": 1.2253,
      "step": 137200
    },
    {
      "epoch": 27.61,
      "learning_rate": 0.0012413927485128253,
      "loss": 1.2223,
      "step": 137300
    },
    {
      "epoch": 27.63,
      "learning_rate": 0.001239637993972503,
      "loss": 1.2251,
      "step": 137400
    },
    {
      "epoch": 27.65,
      "learning_rate": 0.0012378832598533957,
      "loss": 1.2225,
      "step": 137500
    },
    {
      "epoch": 27.67,
      "learning_rate": 0.0012361285496136948,
      "loss": 1.2206,
      "step": 137600
    },
    {
      "epoch": 27.69,
      "learning_rate": 0.001234373866711544,
      "loss": 1.2238,
      "step": 137700
    },
    {
      "epoch": 27.71,
      "learning_rate": 0.0012326192146050346,
      "loss": 1.2259,
      "step": 137800
    },
    {
      "epoch": 27.73,
      "learning_rate": 0.0012308645967521966,
      "loss": 1.2215,
      "step": 137900
    },
    {
      "epoch": 27.75,
      "learning_rate": 0.0012291100166109926,
      "loss": 1.2229,
      "step": 138000
    },
    {
      "epoch": 27.77,
      "learning_rate": 0.0012273554776393101,
      "loss": 1.2194,
      "step": 138100
    },
    {
      "epoch": 27.79,
      "learning_rate": 0.0012256009832949562,
      "loss": 1.2188,
      "step": 138200
    },
    {
      "epoch": 27.81,
      "learning_rate": 0.0012238465370356501,
      "loss": 1.2199,
      "step": 138300
    },
    {
      "epoch": 27.83,
      "learning_rate": 0.0012220921423190167,
      "loss": 1.2211,
      "step": 138400
    },
    {
      "epoch": 27.85,
      "learning_rate": 0.0012203378026025777,
      "loss": 1.2184,
      "step": 138500
    },
    {
      "epoch": 27.87,
      "learning_rate": 0.0012185835213437478,
      "loss": 1.2187,
      "step": 138600
    },
    {
      "epoch": 27.89,
      "learning_rate": 0.0012168293019998258,
      "loss": 1.2232,
      "step": 138700
    },
    {
      "epoch": 27.91,
      "learning_rate": 0.0012150751480279897,
      "loss": 1.2241,
      "step": 138800
    },
    {
      "epoch": 27.93,
      "learning_rate": 0.0012133210628852867,
      "loss": 1.2215,
      "step": 138900
    },
    {
      "epoch": 27.95,
      "learning_rate": 0.0012115670500286294,
      "loss": 1.2231,
      "step": 139000
    },
    {
      "epoch": 27.97,
      "learning_rate": 0.0012098131129147888,
      "loss": 1.2214,
      "step": 139100
    },
    {
      "epoch": 27.99,
      "learning_rate": 0.0012080767931761424,
      "loss": 1.2187,
      "step": 139200
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.4111801831154055,
      "eval_loss": 1.1840488910675049,
      "eval_runtime": 19.7999,
      "eval_samples_per_second": 4018.599,
      "eval_steps_per_second": 15.707,
      "step": 139251
    },
    {
      "epoch": 28.01,
      "learning_rate": 0.0012063230170739731,
      "loss": 1.2143,
      "step": 139300
    },
    {
      "epoch": 28.03,
      "learning_rate": 0.0012045693270494448,
      "loss": 1.211,
      "step": 139400
    },
    {
      "epoch": 28.05,
      "learning_rate": 0.0012028157265586918,
      "loss": 1.2165,
      "step": 139500
    },
    {
      "epoch": 28.07,
      "learning_rate": 0.0012010622190576717,
      "loss": 1.217,
      "step": 139600
    },
    {
      "epoch": 28.09,
      "learning_rate": 0.001199308808002159,
      "loss": 1.2124,
      "step": 139700
    },
    {
      "epoch": 28.11,
      "learning_rate": 0.001197555496847737,
      "loss": 1.2182,
      "step": 139800
    },
    {
      "epoch": 28.13,
      "learning_rate": 0.0011958022890497934,
      "loss": 1.2156,
      "step": 139900
    },
    {
      "epoch": 28.15,
      "learning_rate": 0.0011940491880635118,
      "loss": 1.2185,
      "step": 140000
    },
    {
      "epoch": 28.17,
      "learning_rate": 0.0011922961973438657,
      "loss": 1.2143,
      "step": 140100
    },
    {
      "epoch": 28.19,
      "learning_rate": 0.0011905433203456097,
      "loss": 1.2148,
      "step": 140200
    },
    {
      "epoch": 28.21,
      "learning_rate": 0.001188790560523276,
      "loss": 1.2149,
      "step": 140300
    },
    {
      "epoch": 28.23,
      "learning_rate": 0.0011870379213311652,
      "loss": 1.2188,
      "step": 140400
    },
    {
      "epoch": 28.25,
      "learning_rate": 0.0011852854062233409,
      "loss": 1.2153,
      "step": 140500
    },
    {
      "epoch": 28.27,
      "learning_rate": 0.0011835330186536204,
      "loss": 1.2163,
      "step": 140600
    },
    {
      "epoch": 28.29,
      "learning_rate": 0.0011817807620755712,
      "loss": 1.2149,
      "step": 140700
    },
    {
      "epoch": 28.31,
      "learning_rate": 0.001180028639942502,
      "loss": 1.2188,
      "step": 140800
    },
    {
      "epoch": 28.33,
      "learning_rate": 0.0011782766557074578,
      "loss": 1.2193,
      "step": 140900
    },
    {
      "epoch": 28.35,
      "learning_rate": 0.0011765248128232095,
      "loss": 1.2181,
      "step": 141000
    },
    {
      "epoch": 28.37,
      "learning_rate": 0.001174773114742251,
      "loss": 1.2133,
      "step": 141100
    },
    {
      "epoch": 28.39,
      "learning_rate": 0.0011730215649167904,
      "loss": 1.2198,
      "step": 141200
    },
    {
      "epoch": 28.41,
      "learning_rate": 0.001171270166798745,
      "loss": 1.2163,
      "step": 141300
    },
    {
      "epoch": 28.43,
      "learning_rate": 0.00116951892383973,
      "loss": 1.219,
      "step": 141400
    },
    {
      "epoch": 28.45,
      "learning_rate": 0.0011677678394910577,
      "loss": 1.2176,
      "step": 141500
    },
    {
      "epoch": 28.47,
      "learning_rate": 0.0011660169172037266,
      "loss": 1.2186,
      "step": 141600
    },
    {
      "epoch": 28.49,
      "learning_rate": 0.0011642661604284164,
      "loss": 1.2152,
      "step": 141700
    },
    {
      "epoch": 28.51,
      "learning_rate": 0.0011625155726154794,
      "loss": 1.2113,
      "step": 141800
    },
    {
      "epoch": 28.53,
      "learning_rate": 0.0011607651572149362,
      "loss": 1.2138,
      "step": 141900
    },
    {
      "epoch": 28.55,
      "learning_rate": 0.001159014917676467,
      "loss": 1.2151,
      "step": 142000
    },
    {
      "epoch": 28.57,
      "learning_rate": 0.0011572648574494063,
      "loss": 1.217,
      "step": 142100
    },
    {
      "epoch": 28.59,
      "learning_rate": 0.001155514979982733,
      "loss": 1.2148,
      "step": 142200
    },
    {
      "epoch": 28.61,
      "learning_rate": 0.0011537652887250683,
      "loss": 1.2149,
      "step": 142300
    },
    {
      "epoch": 28.63,
      "learning_rate": 0.0011520157871246655,
      "loss": 1.2161,
      "step": 142400
    },
    {
      "epoch": 28.65,
      "learning_rate": 0.001150266478629404,
      "loss": 1.2111,
      "step": 142500
    },
    {
      "epoch": 28.67,
      "learning_rate": 0.001148517366686782,
      "loss": 1.2201,
      "step": 142600
    },
    {
      "epoch": 28.69,
      "learning_rate": 0.0011467684547439116,
      "loss": 1.2148,
      "step": 142700
    },
    {
      "epoch": 28.71,
      "learning_rate": 0.00114501974624751,
      "loss": 1.2185,
      "step": 142800
    },
    {
      "epoch": 28.73,
      "learning_rate": 0.0011432887286244955,
      "loss": 1.2136,
      "step": 142900
    },
    {
      "epoch": 28.75,
      "learning_rate": 0.0011415404352391302,
      "loss": 1.2153,
      "step": 143000
    },
    {
      "epoch": 28.77,
      "learning_rate": 0.0011397923556035006,
      "loss": 1.2159,
      "step": 143100
    },
    {
      "epoch": 28.79,
      "learning_rate": 0.0011380444931626827,
      "loss": 1.2155,
      "step": 143200
    },
    {
      "epoch": 28.81,
      "learning_rate": 0.0011362968513613262,
      "loss": 1.2182,
      "step": 143300
    },
    {
      "epoch": 28.83,
      "learning_rate": 0.001134549433643645,
      "loss": 1.2113,
      "step": 143400
    },
    {
      "epoch": 28.85,
      "learning_rate": 0.0011328022434534126,
      "loss": 1.2165,
      "step": 143500
    },
    {
      "epoch": 28.87,
      "learning_rate": 0.0011310552842339516,
      "loss": 1.2119,
      "step": 143600
    },
    {
      "epoch": 28.89,
      "learning_rate": 0.001129308559428132,
      "loss": 1.2145,
      "step": 143700
    },
    {
      "epoch": 28.91,
      "learning_rate": 0.0011275620724783605,
      "loss": 1.2153,
      "step": 143800
    },
    {
      "epoch": 28.93,
      "learning_rate": 0.001125815826826576,
      "loss": 1.2151,
      "step": 143900
    },
    {
      "epoch": 28.95,
      "learning_rate": 0.0011240698259142399,
      "loss": 1.219,
      "step": 144000
    },
    {
      "epoch": 28.98,
      "learning_rate": 0.0011223240731823335,
      "loss": 1.2149,
      "step": 144100
    },
    {
      "epoch": 29.0,
      "learning_rate": 0.0011205785720713479,
      "loss": 1.2162,
      "step": 144200
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.41123713623185454,
      "eval_loss": 1.1828089952468872,
      "eval_runtime": 19.8923,
      "eval_samples_per_second": 3999.934,
      "eval_steps_per_second": 15.634,
      "step": 144224
    },
    {
      "epoch": 29.02,
      "learning_rate": 0.0011188333260212788,
      "loss": 1.2079,
      "step": 144300
    },
    {
      "epoch": 29.04,
      "learning_rate": 0.001117088338471618,
      "loss": 1.2047,
      "step": 144400
    },
    {
      "epoch": 29.06,
      "learning_rate": 0.001115343612861349,
      "loss": 1.2053,
      "step": 144500
    },
    {
      "epoch": 29.08,
      "learning_rate": 0.0011135991526289393,
      "loss": 1.2075,
      "step": 144600
    },
    {
      "epoch": 29.1,
      "learning_rate": 0.0011118549612123333,
      "loss": 1.2061,
      "step": 144700
    },
    {
      "epoch": 29.12,
      "learning_rate": 0.0011101110420489442,
      "loss": 1.2101,
      "step": 144800
    },
    {
      "epoch": 29.14,
      "learning_rate": 0.0011083673985756498,
      "loss": 1.2117,
      "step": 144900
    },
    {
      "epoch": 29.16,
      "learning_rate": 0.001106624034228785,
      "loss": 1.2104,
      "step": 145000
    },
    {
      "epoch": 29.18,
      "learning_rate": 0.0011048809524441346,
      "loss": 1.2105,
      "step": 145100
    },
    {
      "epoch": 29.2,
      "learning_rate": 0.0011031381566569247,
      "loss": 1.2052,
      "step": 145200
    },
    {
      "epoch": 29.22,
      "learning_rate": 0.0011013956503018196,
      "loss": 1.2104,
      "step": 145300
    },
    {
      "epoch": 29.24,
      "learning_rate": 0.0010996534368129128,
      "loss": 1.2103,
      "step": 145400
    },
    {
      "epoch": 29.26,
      "learning_rate": 0.0010979115196237208,
      "loss": 1.2142,
      "step": 145500
    },
    {
      "epoch": 29.28,
      "learning_rate": 0.001096169902167175,
      "loss": 1.2038,
      "step": 145600
    },
    {
      "epoch": 29.3,
      "learning_rate": 0.0010944285878756177,
      "loss": 1.2115,
      "step": 145700
    },
    {
      "epoch": 29.32,
      "learning_rate": 0.0010926875801807927,
      "loss": 1.2122,
      "step": 145800
    },
    {
      "epoch": 29.34,
      "learning_rate": 0.0010909468825138404,
      "loss": 1.2085,
      "step": 145900
    },
    {
      "epoch": 29.36,
      "learning_rate": 0.0010892064983052884,
      "loss": 1.2095,
      "step": 146000
    },
    {
      "epoch": 29.38,
      "learning_rate": 0.0010874664309850487,
      "loss": 1.2102,
      "step": 146100
    },
    {
      "epoch": 29.4,
      "learning_rate": 0.0010857266839824074,
      "loss": 1.2116,
      "step": 146200
    },
    {
      "epoch": 29.42,
      "learning_rate": 0.0010839872607260209,
      "loss": 1.2121,
      "step": 146300
    },
    {
      "epoch": 29.44,
      "learning_rate": 0.0010822481646439047,
      "loss": 1.2101,
      "step": 146400
    },
    {
      "epoch": 29.46,
      "learning_rate": 0.0010805093991634325,
      "loss": 1.2077,
      "step": 146500
    },
    {
      "epoch": 29.48,
      "learning_rate": 0.001078770967711325,
      "loss": 1.2103,
      "step": 146600
    },
    {
      "epoch": 29.5,
      "learning_rate": 0.001077032873713645,
      "loss": 1.2079,
      "step": 146700
    },
    {
      "epoch": 29.52,
      "learning_rate": 0.0010752951205957896,
      "loss": 1.2105,
      "step": 146800
    },
    {
      "epoch": 29.54,
      "learning_rate": 0.0010735577117824847,
      "loss": 1.2118,
      "step": 146900
    },
    {
      "epoch": 29.56,
      "learning_rate": 0.0010718206506977778,
      "loss": 1.2105,
      "step": 147000
    },
    {
      "epoch": 29.58,
      "learning_rate": 0.0010700839407650313,
      "loss": 1.2045,
      "step": 147100
    },
    {
      "epoch": 29.6,
      "learning_rate": 0.001068347585406914,
      "loss": 1.2123,
      "step": 147200
    },
    {
      "epoch": 29.62,
      "learning_rate": 0.0010666115880453974,
      "loss": 1.2078,
      "step": 147300
    },
    {
      "epoch": 29.64,
      "learning_rate": 0.0010648759521017476,
      "loss": 1.2099,
      "step": 147400
    },
    {
      "epoch": 29.66,
      "learning_rate": 0.0010631406809965178,
      "loss": 1.2085,
      "step": 147500
    },
    {
      "epoch": 29.68,
      "learning_rate": 0.0010614057781495414,
      "loss": 1.2119,
      "step": 147600
    },
    {
      "epoch": 29.7,
      "learning_rate": 0.001059671246979928,
      "loss": 1.2093,
      "step": 147700
    },
    {
      "epoch": 29.72,
      "learning_rate": 0.001057937090906053,
      "loss": 1.2063,
      "step": 147800
    },
    {
      "epoch": 29.74,
      "learning_rate": 0.001056203313345554,
      "loss": 1.2055,
      "step": 147900
    },
    {
      "epoch": 29.76,
      "learning_rate": 0.0010544699177153208,
      "loss": 1.2086,
      "step": 148000
    },
    {
      "epoch": 29.78,
      "learning_rate": 0.0010527369074314922,
      "loss": 1.2115,
      "step": 148100
    },
    {
      "epoch": 29.8,
      "learning_rate": 0.0010510042859094464,
      "loss": 1.2102,
      "step": 148200
    },
    {
      "epoch": 29.82,
      "learning_rate": 0.0010492720565637972,
      "loss": 1.2079,
      "step": 148300
    },
    {
      "epoch": 29.84,
      "learning_rate": 0.001047540222808383,
      "loss": 1.2114,
      "step": 148400
    },
    {
      "epoch": 29.86,
      "learning_rate": 0.001045808788056264,
      "loss": 1.2092,
      "step": 148500
    },
    {
      "epoch": 29.88,
      "learning_rate": 0.001044077755719714,
      "loss": 1.2106,
      "step": 148600
    },
    {
      "epoch": 29.9,
      "learning_rate": 0.0010423471292102147,
      "loss": 1.2099,
      "step": 148700
    },
    {
      "epoch": 29.92,
      "learning_rate": 0.0010406169119384452,
      "loss": 1.2035,
      "step": 148800
    },
    {
      "epoch": 29.94,
      "learning_rate": 0.0010388871073142806,
      "loss": 1.2079,
      "step": 148900
    },
    {
      "epoch": 29.96,
      "learning_rate": 0.0010371577187467818,
      "loss": 1.2114,
      "step": 149000
    },
    {
      "epoch": 29.98,
      "learning_rate": 0.001035428749644191,
      "loss": 1.2087,
      "step": 149100
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.41176053219028325,
      "eval_loss": 1.1731864213943481,
      "eval_runtime": 19.7738,
      "eval_samples_per_second": 4023.916,
      "eval_steps_per_second": 15.728,
      "step": 149197
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.001033700203413921,
      "loss": 1.2107,
      "step": 149200
    },
    {
      "epoch": 30.02,
      "learning_rate": 0.0010319893625408032,
      "loss": 1.204,
      "step": 149300
    },
    {
      "epoch": 30.04,
      "learning_rate": 0.0010302616679603773,
      "loss": 1.2,
      "step": 149400
    },
    {
      "epoch": 30.06,
      "learning_rate": 0.0010285344064354445,
      "loss": 1.2002,
      "step": 149500
    },
    {
      "epoch": 30.08,
      "learning_rate": 0.0010268075813700541,
      "loss": 1.1993,
      "step": 149600
    },
    {
      "epoch": 30.1,
      "learning_rate": 0.0010250811961673946,
      "loss": 1.2016,
      "step": 149700
    },
    {
      "epoch": 30.12,
      "learning_rate": 0.0010233552542297884,
      "loss": 1.2029,
      "step": 149800
    },
    {
      "epoch": 30.14,
      "learning_rate": 0.001021629758958684,
      "loss": 1.2027,
      "step": 149900
    },
    {
      "epoch": 30.16,
      "learning_rate": 0.0010199047137546503,
      "loss": 1.2043,
      "step": 150000
    },
    {
      "epoch": 30.18,
      "learning_rate": 0.0010181801220173676,
      "loss": 1.2028,
      "step": 150100
    },
    {
      "epoch": 30.2,
      "learning_rate": 0.0010164559871456242,
      "loss": 1.2031,
      "step": 150200
    },
    {
      "epoch": 30.22,
      "learning_rate": 0.0010147323125373072,
      "loss": 1.2036,
      "step": 150300
    },
    {
      "epoch": 30.24,
      "learning_rate": 0.0010130263313926052,
      "loss": 1.2056,
      "step": 150400
    },
    {
      "epoch": 30.26,
      "learning_rate": 0.001011303582813796,
      "loss": 1.2015,
      "step": 150500
    },
    {
      "epoch": 30.28,
      "learning_rate": 0.0010095813046526582,
      "loss": 1.2047,
      "step": 150600
    },
    {
      "epoch": 30.3,
      "learning_rate": 0.0010078595003034205,
      "loss": 1.2056,
      "step": 150700
    },
    {
      "epoch": 30.32,
      "learning_rate": 0.0010061381731593774,
      "loss": 1.2017,
      "step": 150800
    },
    {
      "epoch": 30.34,
      "learning_rate": 0.0010044345326882533,
      "loss": 1.2007,
      "step": 150900
    },
    {
      "epoch": 30.36,
      "learning_rate": 0.0010027141652740393,
      "loss": 1.1998,
      "step": 151000
    },
    {
      "epoch": 30.38,
      "learning_rate": 0.0010009942852053342,
      "loss": 1.2035,
      "step": 151100
    },
    {
      "epoch": 30.4,
      "learning_rate": 0.0009992748958716382,
      "loss": 1.1996,
      "step": 151200
    },
    {
      "epoch": 30.42,
      "learning_rate": 0.0009975560006614873,
      "loss": 1.2052,
      "step": 151300
    },
    {
      "epoch": 30.44,
      "learning_rate": 0.0009958376029624422,
      "loss": 1.2017,
      "step": 151400
    },
    {
      "epoch": 30.46,
      "learning_rate": 0.0009941197061610842,
      "loss": 1.2044,
      "step": 151500
    },
    {
      "epoch": 30.48,
      "learning_rate": 0.0009924023136430055,
      "loss": 1.2078,
      "step": 151600
    },
    {
      "epoch": 30.5,
      "learning_rate": 0.000990685428792806,
      "loss": 1.2025,
      "step": 151700
    },
    {
      "epoch": 30.52,
      "learning_rate": 0.0009889690549940852,
      "loss": 1.2001,
      "step": 151800
    },
    {
      "epoch": 30.54,
      "learning_rate": 0.0009872531956294354,
      "loss": 1.1996,
      "step": 151900
    },
    {
      "epoch": 30.56,
      "learning_rate": 0.0009855378540804332,
      "loss": 1.2012,
      "step": 152000
    },
    {
      "epoch": 30.58,
      "learning_rate": 0.0009838230337276372,
      "loss": 1.2039,
      "step": 152100
    },
    {
      "epoch": 30.6,
      "learning_rate": 0.0009821087379505776,
      "loss": 1.2026,
      "step": 152200
    },
    {
      "epoch": 30.62,
      "learning_rate": 0.0009803949701277515,
      "loss": 1.2004,
      "step": 152300
    },
    {
      "epoch": 30.64,
      "learning_rate": 0.0009786817336366138,
      "loss": 1.2015,
      "step": 152400
    },
    {
      "epoch": 30.66,
      "learning_rate": 0.0009769690318535743,
      "loss": 1.2026,
      "step": 152500
    },
    {
      "epoch": 30.68,
      "learning_rate": 0.0009752739871163907,
      "loss": 1.2037,
      "step": 152600
    },
    {
      "epoch": 30.7,
      "learning_rate": 0.0009735623594432755,
      "loss": 1.2016,
      "step": 152700
    },
    {
      "epoch": 30.72,
      "learning_rate": 0.0009718512765674095,
      "loss": 1.2024,
      "step": 152800
    },
    {
      "epoch": 30.74,
      "learning_rate": 0.0009701407418609562,
      "loss": 1.202,
      "step": 152900
    },
    {
      "epoch": 30.76,
      "learning_rate": 0.0009684307586950005,
      "loss": 1.206,
      "step": 153000
    },
    {
      "epoch": 30.78,
      "learning_rate": 0.0009667213304395399,
      "loss": 1.1989,
      "step": 153100
    },
    {
      "epoch": 30.8,
      "learning_rate": 0.0009650124604634786,
      "loss": 1.2081,
      "step": 153200
    },
    {
      "epoch": 30.82,
      "learning_rate": 0.0009633041521346189,
      "loss": 1.2015,
      "step": 153300
    },
    {
      "epoch": 30.85,
      "learning_rate": 0.0009615964088196581,
      "loss": 1.1989,
      "step": 153400
    },
    {
      "epoch": 30.87,
      "learning_rate": 0.0009598892338841794,
      "loss": 1.2006,
      "step": 153500
    },
    {
      "epoch": 30.89,
      "learning_rate": 0.0009581826306926464,
      "loss": 1.1991,
      "step": 153600
    },
    {
      "epoch": 30.91,
      "learning_rate": 0.000956493660031415,
      "loss": 1.2033,
      "step": 153700
    },
    {
      "epoch": 30.93,
      "learning_rate": 0.0009547882046153125,
      "loss": 1.2024,
      "step": 153800
    },
    {
      "epoch": 30.95,
      "learning_rate": 0.000953083330996152,
      "loss": 1.2042,
      "step": 153900
    },
    {
      "epoch": 30.97,
      "learning_rate": 0.0009513790425338609,
      "loss": 1.2005,
      "step": 154000
    },
    {
      "epoch": 30.99,
      "learning_rate": 0.000949675342587214,
      "loss": 1.2005,
      "step": 154100
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.41267003209737646,
      "eval_loss": 1.1657705307006836,
      "eval_runtime": 19.6203,
      "eval_samples_per_second": 4055.399,
      "eval_steps_per_second": 15.851,
      "step": 154170
    },
    {
      "epoch": 31.01,
      "learning_rate": 0.0009479722345138251,
      "loss": 1.1976,
      "step": 154200
    },
    {
      "epoch": 31.03,
      "learning_rate": 0.0009462697216701424,
      "loss": 1.1919,
      "step": 154300
    },
    {
      "epoch": 31.05,
      "learning_rate": 0.0009445678074114414,
      "loss": 1.1973,
      "step": 154400
    },
    {
      "epoch": 31.07,
      "learning_rate": 0.0009428664950918177,
      "loss": 1.1941,
      "step": 154500
    },
    {
      "epoch": 31.09,
      "learning_rate": 0.0009411657880641792,
      "loss": 1.1909,
      "step": 154600
    },
    {
      "epoch": 31.11,
      "learning_rate": 0.0009394656896802428,
      "loss": 1.1942,
      "step": 154700
    },
    {
      "epoch": 31.13,
      "learning_rate": 0.0009377662032905253,
      "loss": 1.1925,
      "step": 154800
    },
    {
      "epoch": 31.15,
      "learning_rate": 0.0009360673322443375,
      "loss": 1.1955,
      "step": 154900
    },
    {
      "epoch": 31.17,
      "learning_rate": 0.0009343690798897762,
      "loss": 1.1939,
      "step": 155000
    },
    {
      "epoch": 31.19,
      "learning_rate": 0.0009326714495737206,
      "loss": 1.1936,
      "step": 155100
    },
    {
      "epoch": 31.21,
      "learning_rate": 0.0009309744446418236,
      "loss": 1.195,
      "step": 155200
    },
    {
      "epoch": 31.23,
      "learning_rate": 0.0009292780684385055,
      "loss": 1.1938,
      "step": 155300
    },
    {
      "epoch": 31.25,
      "learning_rate": 0.0009275823243069464,
      "loss": 1.1969,
      "step": 155400
    },
    {
      "epoch": 31.27,
      "learning_rate": 0.0009258872155890821,
      "loss": 1.1971,
      "step": 155500
    },
    {
      "epoch": 31.29,
      "learning_rate": 0.0009241927456255962,
      "loss": 1.1924,
      "step": 155600
    },
    {
      "epoch": 31.31,
      "learning_rate": 0.0009224989177559132,
      "loss": 1.197,
      "step": 155700
    },
    {
      "epoch": 31.33,
      "learning_rate": 0.0009208057353181909,
      "loss": 1.1955,
      "step": 155800
    },
    {
      "epoch": 31.35,
      "learning_rate": 0.0009191132016493168,
      "loss": 1.1927,
      "step": 155900
    },
    {
      "epoch": 31.37,
      "learning_rate": 0.0009174213200848991,
      "loss": 1.1954,
      "step": 156000
    },
    {
      "epoch": 31.39,
      "learning_rate": 0.0009157300939592614,
      "loss": 1.1945,
      "step": 156100
    },
    {
      "epoch": 31.41,
      "learning_rate": 0.0009140395266054343,
      "loss": 1.1989,
      "step": 156200
    },
    {
      "epoch": 31.43,
      "learning_rate": 0.0009123496213551513,
      "loss": 1.1914,
      "step": 156300
    },
    {
      "epoch": 31.45,
      "learning_rate": 0.0009106603815388409,
      "loss": 1.1953,
      "step": 156400
    },
    {
      "epoch": 31.47,
      "learning_rate": 0.0009089718104856201,
      "loss": 1.1952,
      "step": 156500
    },
    {
      "epoch": 31.49,
      "learning_rate": 0.0009072839115232867,
      "loss": 1.1936,
      "step": 156600
    },
    {
      "epoch": 31.51,
      "learning_rate": 0.0009055966879783159,
      "loss": 1.197,
      "step": 156700
    },
    {
      "epoch": 31.53,
      "learning_rate": 0.0009039101431758506,
      "loss": 1.1957,
      "step": 156800
    },
    {
      "epoch": 31.55,
      "learning_rate": 0.0009022242804396972,
      "loss": 1.1971,
      "step": 156900
    },
    {
      "epoch": 31.57,
      "learning_rate": 0.0009005391030923156,
      "loss": 1.1941,
      "step": 157000
    },
    {
      "epoch": 31.59,
      "learning_rate": 0.0008988546144548173,
      "loss": 1.1935,
      "step": 157100
    },
    {
      "epoch": 31.61,
      "learning_rate": 0.0008971708178469554,
      "loss": 1.1977,
      "step": 157200
    },
    {
      "epoch": 31.63,
      "learning_rate": 0.00089548771658712,
      "loss": 1.1978,
      "step": 157300
    },
    {
      "epoch": 31.65,
      "learning_rate": 0.0008938053139923291,
      "loss": 1.1932,
      "step": 157400
    },
    {
      "epoch": 31.67,
      "learning_rate": 0.0008921236133782254,
      "loss": 1.1928,
      "step": 157500
    },
    {
      "epoch": 31.69,
      "learning_rate": 0.0008904426180590678,
      "loss": 1.1976,
      "step": 157600
    },
    {
      "epoch": 31.71,
      "learning_rate": 0.0008887623313477256,
      "loss": 1.1946,
      "step": 157700
    },
    {
      "epoch": 31.73,
      "learning_rate": 0.0008870827565556696,
      "loss": 1.1975,
      "step": 157800
    },
    {
      "epoch": 31.75,
      "learning_rate": 0.0008854038969929701,
      "loss": 1.195,
      "step": 157900
    },
    {
      "epoch": 31.77,
      "learning_rate": 0.0008837257559682865,
      "loss": 1.1947,
      "step": 158000
    },
    {
      "epoch": 31.79,
      "learning_rate": 0.0008820483367888628,
      "loss": 1.1963,
      "step": 158100
    },
    {
      "epoch": 31.81,
      "learning_rate": 0.0008803716427605191,
      "loss": 1.1963,
      "step": 158200
    },
    {
      "epoch": 31.83,
      "learning_rate": 0.0008786956771876478,
      "loss": 1.1923,
      "step": 158300
    },
    {
      "epoch": 31.85,
      "learning_rate": 0.000877020443373205,
      "loss": 1.1971,
      "step": 158400
    },
    {
      "epoch": 31.87,
      "learning_rate": 0.0008753459446187053,
      "loss": 1.1951,
      "step": 158500
    },
    {
      "epoch": 31.89,
      "learning_rate": 0.0008736721842242136,
      "loss": 1.1981,
      "step": 158600
    },
    {
      "epoch": 31.91,
      "learning_rate": 0.0008719991654883402,
      "loss": 1.1991,
      "step": 158700
    },
    {
      "epoch": 31.93,
      "learning_rate": 0.0008703268917082342,
      "loss": 1.1949,
      "step": 158800
    },
    {
      "epoch": 31.95,
      "learning_rate": 0.0008686553661795765,
      "loss": 1.1921,
      "step": 158900
    },
    {
      "epoch": 31.97,
      "learning_rate": 0.0008669845921965718,
      "loss": 1.1974,
      "step": 159000
    },
    {
      "epoch": 31.99,
      "learning_rate": 0.0008653145730519456,
      "loss": 1.1944,
      "step": 159100
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.4131173845092614,
      "eval_loss": 1.1602274179458618,
      "eval_runtime": 19.7773,
      "eval_samples_per_second": 4023.208,
      "eval_steps_per_second": 15.725,
      "step": 159144
    },
    {
      "epoch": 32.01,
      "learning_rate": 0.000863645312036935,
      "loss": 1.1882,
      "step": 159200
    },
    {
      "epoch": 32.03,
      "learning_rate": 0.0008619768124412836,
      "loss": 1.1821,
      "step": 159300
    },
    {
      "epoch": 32.05,
      "learning_rate": 0.000860309077553233,
      "loss": 1.1848,
      "step": 159400
    },
    {
      "epoch": 32.07,
      "learning_rate": 0.0008586421106595186,
      "loss": 1.1857,
      "step": 159500
    },
    {
      "epoch": 32.09,
      "learning_rate": 0.0008569759150453628,
      "loss": 1.1859,
      "step": 159600
    },
    {
      "epoch": 32.11,
      "learning_rate": 0.0008553104939944677,
      "loss": 1.189,
      "step": 159700
    },
    {
      "epoch": 32.13,
      "learning_rate": 0.0008536458507890077,
      "loss": 1.189,
      "step": 159800
    },
    {
      "epoch": 32.15,
      "learning_rate": 0.0008519819887096256,
      "loss": 1.1908,
      "step": 159900
    },
    {
      "epoch": 32.17,
      "learning_rate": 0.0008503189110354243,
      "loss": 1.1872,
      "step": 160000
    },
    {
      "epoch": 32.19,
      "learning_rate": 0.0008486566210439614,
      "loss": 1.1861,
      "step": 160100
    },
    {
      "epoch": 32.21,
      "learning_rate": 0.0008470117330755695,
      "loss": 1.1838,
      "step": 160200
    },
    {
      "epoch": 32.23,
      "learning_rate": 0.0008453510203175021,
      "loss": 1.1876,
      "step": 160300
    },
    {
      "epoch": 32.25,
      "learning_rate": 0.0008436911050327827,
      "loss": 1.1843,
      "step": 160400
    },
    {
      "epoch": 32.27,
      "learning_rate": 0.0008420319904927365,
      "loss": 1.1884,
      "step": 160500
    },
    {
      "epoch": 32.29,
      "learning_rate": 0.0008403736799671097,
      "loss": 1.191,
      "step": 160600
    },
    {
      "epoch": 32.31,
      "learning_rate": 0.000838716176724065,
      "loss": 1.1915,
      "step": 160700
    },
    {
      "epoch": 32.33,
      "learning_rate": 0.0008370594840301723,
      "loss": 1.188,
      "step": 160800
    },
    {
      "epoch": 32.35,
      "learning_rate": 0.000835403605150406,
      "loss": 1.1913,
      "step": 160900
    },
    {
      "epoch": 32.37,
      "learning_rate": 0.000833748543348136,
      "loss": 1.1905,
      "step": 161000
    },
    {
      "epoch": 32.39,
      "learning_rate": 0.0008320943018851221,
      "loss": 1.1876,
      "step": 161100
    },
    {
      "epoch": 32.41,
      "learning_rate": 0.0008304408840215062,
      "loss": 1.1882,
      "step": 161200
    },
    {
      "epoch": 32.43,
      "learning_rate": 0.0008287882930158088,
      "loss": 1.1874,
      "step": 161300
    },
    {
      "epoch": 32.45,
      "learning_rate": 0.0008271365321249197,
      "loss": 1.1908,
      "step": 161400
    },
    {
      "epoch": 32.47,
      "learning_rate": 0.0008254856046040937,
      "loss": 1.1903,
      "step": 161500
    },
    {
      "epoch": 32.49,
      "learning_rate": 0.0008238355137069418,
      "loss": 1.192,
      "step": 161600
    },
    {
      "epoch": 32.51,
      "learning_rate": 0.0008221862626854274,
      "loss": 1.1883,
      "step": 161700
    },
    {
      "epoch": 32.53,
      "learning_rate": 0.0008205378547898581,
      "loss": 1.1918,
      "step": 161800
    },
    {
      "epoch": 32.55,
      "learning_rate": 0.0008188902932688807,
      "loss": 1.1885,
      "step": 161900
    },
    {
      "epoch": 32.57,
      "learning_rate": 0.0008172435813694726,
      "loss": 1.1887,
      "step": 162000
    },
    {
      "epoch": 32.59,
      "learning_rate": 0.0008155977223369379,
      "loss": 1.1873,
      "step": 162100
    },
    {
      "epoch": 32.61,
      "learning_rate": 0.0008139527194148993,
      "loss": 1.1863,
      "step": 162200
    },
    {
      "epoch": 32.63,
      "learning_rate": 0.0008123085758452935,
      "loss": 1.1897,
      "step": 162300
    },
    {
      "epoch": 32.65,
      "learning_rate": 0.0008106652948683613,
      "loss": 1.1856,
      "step": 162400
    },
    {
      "epoch": 32.67,
      "learning_rate": 0.0008090392995776086,
      "loss": 1.1922,
      "step": 162500
    },
    {
      "epoch": 32.69,
      "learning_rate": 0.0008073977447932476,
      "loss": 1.1861,
      "step": 162600
    },
    {
      "epoch": 32.72,
      "learning_rate": 0.0008057570622797192,
      "loss": 1.1881,
      "step": 162700
    },
    {
      "epoch": 32.74,
      "learning_rate": 0.0008041336489961784,
      "loss": 1.1852,
      "step": 162800
    },
    {
      "epoch": 32.76,
      "learning_rate": 0.0008024947119195025,
      "loss": 1.1851,
      "step": 162900
    },
    {
      "epoch": 32.78,
      "learning_rate": 0.000800856656776449,
      "loss": 1.1819,
      "step": 163000
    },
    {
      "epoch": 32.8,
      "learning_rate": 0.0007992194867952607,
      "loss": 1.1882,
      "step": 163100
    },
    {
      "epoch": 32.82,
      "learning_rate": 0.0007975832052024367,
      "loss": 1.1864,
      "step": 163200
    },
    {
      "epoch": 32.84,
      "learning_rate": 0.0007959478152227251,
      "loss": 1.1914,
      "step": 163300
    },
    {
      "epoch": 32.86,
      "learning_rate": 0.0007943133200791164,
      "loss": 1.1888,
      "step": 163400
    },
    {
      "epoch": 32.88,
      "learning_rate": 0.0007926797229928376,
      "loss": 1.1831,
      "step": 163500
    },
    {
      "epoch": 32.9,
      "learning_rate": 0.0007910470271833464,
      "loss": 1.1878,
      "step": 163600
    },
    {
      "epoch": 32.92,
      "learning_rate": 0.0007894152358683243,
      "loss": 1.1889,
      "step": 163700
    },
    {
      "epoch": 32.94,
      "learning_rate": 0.0007877843522636694,
      "loss": 1.1872,
      "step": 163800
    },
    {
      "epoch": 32.96,
      "learning_rate": 0.0007861543795834913,
      "loss": 1.1851,
      "step": 163900
    },
    {
      "epoch": 32.98,
      "learning_rate": 0.0007845253210401045,
      "loss": 1.1905,
      "step": 164000
    },
    {
      "epoch": 33.0,
      "learning_rate": 0.0007828971798440226,
      "loss": 1.1887,
      "step": 164100
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.41392475061534817,
      "eval_loss": 1.1511569023132324,
      "eval_runtime": 19.7992,
      "eval_samples_per_second": 4018.746,
      "eval_steps_per_second": 15.708,
      "step": 164117
    },
    {
      "epoch": 33.02,
      "learning_rate": 0.0007812699592039499,
      "loss": 1.1783,
      "step": 164200
    },
    {
      "epoch": 33.04,
      "learning_rate": 0.0007796436623267771,
      "loss": 1.1784,
      "step": 164300
    },
    {
      "epoch": 33.06,
      "learning_rate": 0.0007780182924175748,
      "loss": 1.178,
      "step": 164400
    },
    {
      "epoch": 33.08,
      "learning_rate": 0.0007763938526795867,
      "loss": 1.177,
      "step": 164500
    },
    {
      "epoch": 33.1,
      "learning_rate": 0.0007747703463142225,
      "loss": 1.1826,
      "step": 164600
    },
    {
      "epoch": 33.12,
      "learning_rate": 0.0007731477765210531,
      "loss": 1.1774,
      "step": 164700
    },
    {
      "epoch": 33.14,
      "learning_rate": 0.0007715261464978038,
      "loss": 1.1816,
      "step": 164800
    },
    {
      "epoch": 33.16,
      "learning_rate": 0.0007699054594403476,
      "loss": 1.1792,
      "step": 164900
    },
    {
      "epoch": 33.18,
      "learning_rate": 0.0007682857185426989,
      "loss": 1.1801,
      "step": 165000
    },
    {
      "epoch": 33.2,
      "learning_rate": 0.0007666669269970071,
      "loss": 1.1801,
      "step": 165100
    },
    {
      "epoch": 33.22,
      "learning_rate": 0.0007650490879935517,
      "loss": 1.1833,
      "step": 165200
    },
    {
      "epoch": 33.24,
      "learning_rate": 0.0007634322047207346,
      "loss": 1.1818,
      "step": 165300
    },
    {
      "epoch": 33.26,
      "learning_rate": 0.0007618162803650734,
      "loss": 1.1805,
      "step": 165400
    },
    {
      "epoch": 33.28,
      "learning_rate": 0.0007602013181111966,
      "loss": 1.179,
      "step": 165500
    },
    {
      "epoch": 33.3,
      "learning_rate": 0.0007585873211418363,
      "loss": 1.1792,
      "step": 165600
    },
    {
      "epoch": 33.32,
      "learning_rate": 0.000756974292637823,
      "loss": 1.1775,
      "step": 165700
    },
    {
      "epoch": 33.34,
      "learning_rate": 0.0007553622357780774,
      "loss": 1.183,
      "step": 165800
    },
    {
      "epoch": 33.36,
      "learning_rate": 0.0007537511537396058,
      "loss": 1.1838,
      "step": 165900
    },
    {
      "epoch": 33.38,
      "learning_rate": 0.0007521410496974935,
      "loss": 1.182,
      "step": 166000
    },
    {
      "epoch": 33.4,
      "learning_rate": 0.0007505319268248988,
      "loss": 1.1809,
      "step": 166100
    },
    {
      "epoch": 33.42,
      "learning_rate": 0.0007489237882930453,
      "loss": 1.1785,
      "step": 166200
    },
    {
      "epoch": 33.44,
      "learning_rate": 0.0007473166372712171,
      "loss": 1.184,
      "step": 166300
    },
    {
      "epoch": 33.46,
      "learning_rate": 0.0007457104769267522,
      "loss": 1.1809,
      "step": 166400
    },
    {
      "epoch": 33.48,
      "learning_rate": 0.000744105310425037,
      "loss": 1.1809,
      "step": 166500
    },
    {
      "epoch": 33.5,
      "learning_rate": 0.0007425011409294979,
      "loss": 1.1788,
      "step": 166600
    },
    {
      "epoch": 33.52,
      "learning_rate": 0.0007408979716015968,
      "loss": 1.1821,
      "step": 166700
    },
    {
      "epoch": 33.54,
      "learning_rate": 0.000739295805600825,
      "loss": 1.1827,
      "step": 166800
    },
    {
      "epoch": 33.56,
      "learning_rate": 0.0007376946460846965,
      "loss": 1.1819,
      "step": 166900
    },
    {
      "epoch": 33.58,
      "learning_rate": 0.0007360944962087409,
      "loss": 1.1793,
      "step": 167000
    },
    {
      "epoch": 33.6,
      "learning_rate": 0.0007344953591264986,
      "loss": 1.1827,
      "step": 167100
    },
    {
      "epoch": 33.62,
      "learning_rate": 0.000732897237989514,
      "loss": 1.1813,
      "step": 167200
    },
    {
      "epoch": 33.64,
      "learning_rate": 0.0007313001359473295,
      "loss": 1.1835,
      "step": 167300
    },
    {
      "epoch": 33.66,
      "learning_rate": 0.0007297040561474782,
      "loss": 1.1772,
      "step": 167400
    },
    {
      "epoch": 33.68,
      "learning_rate": 0.0007281090017354799,
      "loss": 1.1828,
      "step": 167500
    },
    {
      "epoch": 33.7,
      "learning_rate": 0.0007265149758548325,
      "loss": 1.182,
      "step": 167600
    },
    {
      "epoch": 33.72,
      "learning_rate": 0.0007249219816470082,
      "loss": 1.1799,
      "step": 167700
    },
    {
      "epoch": 33.74,
      "learning_rate": 0.0007233300222514435,
      "loss": 1.1794,
      "step": 167800
    },
    {
      "epoch": 33.76,
      "learning_rate": 0.0007217391008055382,
      "loss": 1.1759,
      "step": 167900
    },
    {
      "epoch": 33.78,
      "learning_rate": 0.0007201492204446453,
      "loss": 1.1799,
      "step": 168000
    },
    {
      "epoch": 33.8,
      "learning_rate": 0.0007185603843020663,
      "loss": 1.1757,
      "step": 168100
    },
    {
      "epoch": 33.82,
      "learning_rate": 0.0007169725955090442,
      "loss": 1.1791,
      "step": 168200
    },
    {
      "epoch": 33.84,
      "learning_rate": 0.0007153858571947587,
      "loss": 1.1777,
      "step": 168300
    },
    {
      "epoch": 33.86,
      "learning_rate": 0.0007138160241077916,
      "loss": 1.1769,
      "step": 168400
    },
    {
      "epoch": 33.88,
      "learning_rate": 0.0007122313855474593,
      "loss": 1.1821,
      "step": 168500
    },
    {
      "epoch": 33.9,
      "learning_rate": 0.0007106478068097369,
      "loss": 1.1786,
      "step": 168600
    },
    {
      "epoch": 33.92,
      "learning_rate": 0.0007090652910155055,
      "loss": 1.1816,
      "step": 168700
    },
    {
      "epoch": 33.94,
      "learning_rate": 0.0007074838412835532,
      "loss": 1.1778,
      "step": 168800
    },
    {
      "epoch": 33.96,
      "learning_rate": 0.0007059034607305667,
      "loss": 1.1819,
      "step": 168900
    },
    {
      "epoch": 33.98,
      "learning_rate": 0.0007043241524711256,
      "loss": 1.1795,
      "step": 169000
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.4142251067155042,
      "eval_loss": 1.1453089714050293,
      "eval_runtime": 19.8806,
      "eval_samples_per_second": 4002.288,
      "eval_steps_per_second": 15.643,
      "step": 169090
    },
    {
      "epoch": 34.0,
      "learning_rate": 0.000702745919617695,
      "loss": 1.1752,
      "step": 169100
    },
    {
      "epoch": 34.02,
      "learning_rate": 0.0007011687652806225,
      "loss": 1.1677,
      "step": 169200
    },
    {
      "epoch": 34.04,
      "learning_rate": 0.000699592692568129,
      "loss": 1.17,
      "step": 169300
    },
    {
      "epoch": 34.06,
      "learning_rate": 0.0006980177045863047,
      "loss": 1.1725,
      "step": 169400
    },
    {
      "epoch": 34.08,
      "learning_rate": 0.0006964438044391006,
      "loss": 1.1729,
      "step": 169500
    },
    {
      "epoch": 34.1,
      "learning_rate": 0.0006948709952283247,
      "loss": 1.1716,
      "step": 169600
    },
    {
      "epoch": 34.12,
      "learning_rate": 0.0006932992800536353,
      "loss": 1.174,
      "step": 169700
    },
    {
      "epoch": 34.14,
      "learning_rate": 0.0006917286620125349,
      "loss": 1.1747,
      "step": 169800
    },
    {
      "epoch": 34.16,
      "learning_rate": 0.0006901591442003616,
      "loss": 1.1702,
      "step": 169900
    },
    {
      "epoch": 34.18,
      "learning_rate": 0.0006885907297102878,
      "loss": 1.1726,
      "step": 170000
    },
    {
      "epoch": 34.2,
      "learning_rate": 0.0006870234216333101,
      "loss": 1.1726,
      "step": 170100
    },
    {
      "epoch": 34.22,
      "learning_rate": 0.0006854572230582455,
      "loss": 1.1732,
      "step": 170200
    },
    {
      "epoch": 34.24,
      "learning_rate": 0.000683892137071723,
      "loss": 1.1736,
      "step": 170300
    },
    {
      "epoch": 34.26,
      "learning_rate": 0.00068232816675818,
      "loss": 1.1748,
      "step": 170400
    },
    {
      "epoch": 34.28,
      "learning_rate": 0.0006807653151998552,
      "loss": 1.1725,
      "step": 170500
    },
    {
      "epoch": 34.3,
      "learning_rate": 0.0006792035854767827,
      "loss": 1.1689,
      "step": 170600
    },
    {
      "epoch": 34.32,
      "learning_rate": 0.0006776429806667841,
      "loss": 1.1719,
      "step": 170700
    },
    {
      "epoch": 34.34,
      "learning_rate": 0.0006760835038454657,
      "loss": 1.1711,
      "step": 170800
    },
    {
      "epoch": 34.36,
      "learning_rate": 0.0006745407359349601,
      "loss": 1.1753,
      "step": 170900
    },
    {
      "epoch": 34.38,
      "learning_rate": 0.0006729835129523944,
      "loss": 1.1734,
      "step": 171000
    },
    {
      "epoch": 34.4,
      "learning_rate": 0.0006714274271412859,
      "loss": 1.174,
      "step": 171100
    },
    {
      "epoch": 34.42,
      "learning_rate": 0.0006698724815683352,
      "loss": 1.1727,
      "step": 171200
    },
    {
      "epoch": 34.44,
      "learning_rate": 0.0006683186792979937,
      "loss": 1.1714,
      "step": 171300
    },
    {
      "epoch": 34.46,
      "learning_rate": 0.0006667660233924612,
      "loss": 1.1702,
      "step": 171400
    },
    {
      "epoch": 34.48,
      "learning_rate": 0.0006652145169116783,
      "loss": 1.1687,
      "step": 171500
    },
    {
      "epoch": 34.5,
      "learning_rate": 0.0006636641629133204,
      "loss": 1.1715,
      "step": 171600
    },
    {
      "epoch": 34.52,
      "learning_rate": 0.0006621149644527902,
      "loss": 1.1716,
      "step": 171700
    },
    {
      "epoch": 34.54,
      "learning_rate": 0.0006605669245832145,
      "loss": 1.1719,
      "step": 171800
    },
    {
      "epoch": 34.56,
      "learning_rate": 0.0006590200463554366,
      "loss": 1.1726,
      "step": 171900
    },
    {
      "epoch": 34.59,
      "learning_rate": 0.0006574743328180105,
      "loss": 1.1751,
      "step": 172000
    },
    {
      "epoch": 34.61,
      "learning_rate": 0.0006559297870171938,
      "loss": 1.1714,
      "step": 172100
    },
    {
      "epoch": 34.63,
      "learning_rate": 0.0006543864119969438,
      "loss": 1.1752,
      "step": 172200
    },
    {
      "epoch": 34.65,
      "learning_rate": 0.0006528442107989105,
      "loss": 1.1756,
      "step": 172300
    },
    {
      "epoch": 34.67,
      "learning_rate": 0.0006513031864624303,
      "loss": 1.1723,
      "step": 172400
    },
    {
      "epoch": 34.69,
      "learning_rate": 0.0006497633420245197,
      "loss": 1.1723,
      "step": 172500
    },
    {
      "epoch": 34.71,
      "learning_rate": 0.0006482246805198708,
      "loss": 1.1735,
      "step": 172600
    },
    {
      "epoch": 34.73,
      "learning_rate": 0.0006466872049808438,
      "loss": 1.1732,
      "step": 172700
    },
    {
      "epoch": 34.75,
      "learning_rate": 0.0006451509184374624,
      "loss": 1.1732,
      "step": 172800
    },
    {
      "epoch": 34.77,
      "learning_rate": 0.0006436158239174055,
      "loss": 1.1751,
      "step": 172900
    },
    {
      "epoch": 34.79,
      "learning_rate": 0.0006420819244460042,
      "loss": 1.1721,
      "step": 173000
    },
    {
      "epoch": 34.81,
      "learning_rate": 0.0006405492230462343,
      "loss": 1.1709,
      "step": 173100
    },
    {
      "epoch": 34.83,
      "learning_rate": 0.0006390177227387101,
      "loss": 1.1696,
      "step": 173200
    },
    {
      "epoch": 34.85,
      "learning_rate": 0.0006374874265416783,
      "loss": 1.17,
      "step": 173300
    },
    {
      "epoch": 34.87,
      "learning_rate": 0.0006359583374710134,
      "loss": 1.1774,
      "step": 173400
    },
    {
      "epoch": 34.89,
      "learning_rate": 0.0006344304585402111,
      "loss": 1.1666,
      "step": 173500
    },
    {
      "epoch": 34.91,
      "learning_rate": 0.0006329037927603816,
      "loss": 1.1701,
      "step": 173600
    },
    {
      "epoch": 34.93,
      "learning_rate": 0.0006313783431402438,
      "loss": 1.1708,
      "step": 173700
    },
    {
      "epoch": 34.95,
      "learning_rate": 0.0006298541126861209,
      "loss": 1.1724,
      "step": 173800
    },
    {
      "epoch": 34.97,
      "learning_rate": 0.0006283311044019327,
      "loss": 1.1727,
      "step": 173900
    },
    {
      "epoch": 34.99,
      "learning_rate": 0.0006268093212891912,
      "loss": 1.1685,
      "step": 174000
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.41499333743989697,
      "eval_loss": 1.1372462511062622,
      "eval_runtime": 19.9974,
      "eval_samples_per_second": 3978.919,
      "eval_steps_per_second": 15.552,
      "step": 174063
    },
    {
      "epoch": 35.01,
      "learning_rate": 0.0006253039658071285,
      "loss": 1.1616,
      "step": 174100
    },
    {
      "epoch": 35.03,
      "learning_rate": 0.0006237846297056513,
      "loss": 1.1594,
      "step": 174200
    },
    {
      "epoch": 35.05,
      "learning_rate": 0.0006222665277357129,
      "loss": 1.1619,
      "step": 174300
    },
    {
      "epoch": 35.07,
      "learning_rate": 0.0006207496628891555,
      "loss": 1.1655,
      "step": 174400
    },
    {
      "epoch": 35.09,
      "learning_rate": 0.0006192340381553838,
      "loss": 1.1588,
      "step": 174500
    },
    {
      "epoch": 35.11,
      "learning_rate": 0.0006177196565213567,
      "loss": 1.1639,
      "step": 174600
    },
    {
      "epoch": 35.13,
      "learning_rate": 0.0006162065209715849,
      "loss": 1.1594,
      "step": 174700
    },
    {
      "epoch": 35.15,
      "learning_rate": 0.0006146946344881228,
      "loss": 1.164,
      "step": 174800
    },
    {
      "epoch": 35.17,
      "learning_rate": 0.0006131840000505637,
      "loss": 1.1602,
      "step": 174900
    },
    {
      "epoch": 35.19,
      "learning_rate": 0.0006116746206360317,
      "loss": 1.1629,
      "step": 175000
    },
    {
      "epoch": 35.21,
      "learning_rate": 0.0006101664992191795,
      "loss": 1.1656,
      "step": 175100
    },
    {
      "epoch": 35.23,
      "learning_rate": 0.0006086596387721796,
      "loss": 1.166,
      "step": 175200
    },
    {
      "epoch": 35.25,
      "learning_rate": 0.0006071540422647201,
      "loss": 1.1632,
      "step": 175300
    },
    {
      "epoch": 35.27,
      "learning_rate": 0.0006056497126639966,
      "loss": 1.1649,
      "step": 175400
    },
    {
      "epoch": 35.29,
      "learning_rate": 0.0006041466529347094,
      "loss": 1.1621,
      "step": 175500
    },
    {
      "epoch": 35.31,
      "learning_rate": 0.0006026448660390557,
      "loss": 1.1657,
      "step": 175600
    },
    {
      "epoch": 35.33,
      "learning_rate": 0.0006011443549367248,
      "loss": 1.1663,
      "step": 175700
    },
    {
      "epoch": 35.35,
      "learning_rate": 0.0005996451225848903,
      "loss": 1.1682,
      "step": 175800
    },
    {
      "epoch": 35.37,
      "learning_rate": 0.0005981471719382066,
      "loss": 1.1618,
      "step": 175900
    },
    {
      "epoch": 35.39,
      "learning_rate": 0.0005966505059488022,
      "loss": 1.1687,
      "step": 176000
    },
    {
      "epoch": 35.41,
      "learning_rate": 0.0005951551275662743,
      "loss": 1.162,
      "step": 176100
    },
    {
      "epoch": 35.43,
      "learning_rate": 0.0005936610397376806,
      "loss": 1.1656,
      "step": 176200
    },
    {
      "epoch": 35.45,
      "learning_rate": 0.0005921682454075374,
      "loss": 1.1663,
      "step": 176300
    },
    {
      "epoch": 35.47,
      "learning_rate": 0.0005906767475178108,
      "loss": 1.163,
      "step": 176400
    },
    {
      "epoch": 35.49,
      "learning_rate": 0.0005891865490079131,
      "loss": 1.1663,
      "step": 176500
    },
    {
      "epoch": 35.51,
      "learning_rate": 0.0005876976528146937,
      "loss": 1.1646,
      "step": 176600
    },
    {
      "epoch": 35.53,
      "learning_rate": 0.000586210061872437,
      "loss": 1.1609,
      "step": 176700
    },
    {
      "epoch": 35.55,
      "learning_rate": 0.0005847237791128547,
      "loss": 1.1629,
      "step": 176800
    },
    {
      "epoch": 35.57,
      "learning_rate": 0.0005832388074650808,
      "loss": 1.1615,
      "step": 176900
    },
    {
      "epoch": 35.59,
      "learning_rate": 0.0005817551498556642,
      "loss": 1.1688,
      "step": 177000
    },
    {
      "epoch": 35.61,
      "learning_rate": 0.0005802728092085649,
      "loss": 1.1621,
      "step": 177100
    },
    {
      "epoch": 35.63,
      "learning_rate": 0.0005787917884451475,
      "loss": 1.1638,
      "step": 177200
    },
    {
      "epoch": 35.65,
      "learning_rate": 0.000577312090484176,
      "loss": 1.1656,
      "step": 177300
    },
    {
      "epoch": 35.67,
      "learning_rate": 0.0005758337182418055,
      "loss": 1.1625,
      "step": 177400
    },
    {
      "epoch": 35.69,
      "learning_rate": 0.0005743566746315804,
      "loss": 1.1613,
      "step": 177500
    },
    {
      "epoch": 35.71,
      "learning_rate": 0.0005728809625644257,
      "loss": 1.167,
      "step": 177600
    },
    {
      "epoch": 35.73,
      "learning_rate": 0.0005714065849486429,
      "loss": 1.1632,
      "step": 177700
    },
    {
      "epoch": 35.75,
      "learning_rate": 0.0005699335446899022,
      "loss": 1.1662,
      "step": 177800
    },
    {
      "epoch": 35.77,
      "learning_rate": 0.0005684618446912396,
      "loss": 1.1628,
      "step": 177900
    },
    {
      "epoch": 35.79,
      "learning_rate": 0.0005669914878530493,
      "loss": 1.1689,
      "step": 178000
    },
    {
      "epoch": 35.81,
      "learning_rate": 0.0005655224770730786,
      "loss": 1.1652,
      "step": 178100
    },
    {
      "epoch": 35.83,
      "learning_rate": 0.0005640694851778694,
      "loss": 1.1631,
      "step": 178200
    },
    {
      "epoch": 35.85,
      "learning_rate": 0.0005626031616641948,
      "loss": 1.1611,
      "step": 178300
    },
    {
      "epoch": 35.87,
      "learning_rate": 0.0005611381928571567,
      "loss": 1.1598,
      "step": 178400
    },
    {
      "epoch": 35.89,
      "learning_rate": 0.0005596745816438834,
      "loss": 1.1639,
      "step": 178500
    },
    {
      "epoch": 35.91,
      "learning_rate": 0.0005582123309088284,
      "loss": 1.1606,
      "step": 178600
    },
    {
      "epoch": 35.93,
      "learning_rate": 0.0005567514435337624,
      "loss": 1.1622,
      "step": 178700
    },
    {
      "epoch": 35.95,
      "learning_rate": 0.000555291922397771,
      "loss": 1.162,
      "step": 178800
    },
    {
      "epoch": 35.97,
      "learning_rate": 0.000553833770377246,
      "loss": 1.164,
      "step": 178900
    },
    {
      "epoch": 35.99,
      "learning_rate": 0.0005523769903458824,
      "loss": 1.1658,
      "step": 179000
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.4156597525371182,
      "eval_loss": 1.1301237344741821,
      "eval_runtime": 19.5705,
      "eval_samples_per_second": 4065.719,
      "eval_steps_per_second": 15.891,
      "step": 179037
    },
    {
      "epoch": 36.01,
      "learning_rate": 0.0005509215851746693,
      "loss": 1.1566,
      "step": 179100
    },
    {
      "epoch": 36.03,
      "learning_rate": 0.0005494675577318875,
      "loss": 1.1495,
      "step": 179200
    },
    {
      "epoch": 36.05,
      "learning_rate": 0.0005480149108831029,
      "loss": 1.1467,
      "step": 179300
    },
    {
      "epoch": 36.07,
      "learning_rate": 0.0005465636474911603,
      "loss": 1.1519,
      "step": 179400
    },
    {
      "epoch": 36.09,
      "learning_rate": 0.000545113770416177,
      "loss": 1.1524,
      "step": 179500
    },
    {
      "epoch": 36.11,
      "learning_rate": 0.0005436652825155394,
      "loss": 1.1548,
      "step": 179600
    },
    {
      "epoch": 36.13,
      "learning_rate": 0.0005422181866438958,
      "loss": 1.1576,
      "step": 179700
    },
    {
      "epoch": 36.15,
      "learning_rate": 0.0005407724856531514,
      "loss": 1.1558,
      "step": 179800
    },
    {
      "epoch": 36.17,
      "learning_rate": 0.0005393281823924612,
      "loss": 1.1539,
      "step": 179900
    },
    {
      "epoch": 36.19,
      "learning_rate": 0.000537885279708227,
      "loss": 1.1562,
      "step": 180000
    },
    {
      "epoch": 36.21,
      "learning_rate": 0.0005364437804440896,
      "loss": 1.1551,
      "step": 180100
    },
    {
      "epoch": 36.23,
      "learning_rate": 0.0005350036874409248,
      "loss": 1.1554,
      "step": 180200
    },
    {
      "epoch": 36.25,
      "learning_rate": 0.0005335650035368354,
      "loss": 1.1547,
      "step": 180300
    },
    {
      "epoch": 36.27,
      "learning_rate": 0.0005321277315671485,
      "loss": 1.159,
      "step": 180400
    },
    {
      "epoch": 36.29,
      "learning_rate": 0.0005306918743644085,
      "loss": 1.1551,
      "step": 180500
    },
    {
      "epoch": 36.31,
      "learning_rate": 0.0005292574347583714,
      "loss": 1.1557,
      "step": 180600
    },
    {
      "epoch": 36.33,
      "learning_rate": 0.0005278244155759988,
      "loss": 1.1572,
      "step": 180700
    },
    {
      "epoch": 36.35,
      "learning_rate": 0.0005263928196414538,
      "loss": 1.1536,
      "step": 180800
    },
    {
      "epoch": 36.37,
      "learning_rate": 0.0005249626497760943,
      "loss": 1.1544,
      "step": 180900
    },
    {
      "epoch": 36.39,
      "learning_rate": 0.0005235339087984682,
      "loss": 1.1571,
      "step": 181000
    },
    {
      "epoch": 36.41,
      "learning_rate": 0.0005221065995243063,
      "loss": 1.1575,
      "step": 181100
    },
    {
      "epoch": 36.43,
      "learning_rate": 0.0005206807247665185,
      "loss": 1.1507,
      "step": 181200
    },
    {
      "epoch": 36.46,
      "learning_rate": 0.0005192562873351877,
      "loss": 1.1566,
      "step": 181300
    },
    {
      "epoch": 36.48,
      "learning_rate": 0.0005178332900375647,
      "loss": 1.1561,
      "step": 181400
    },
    {
      "epoch": 36.5,
      "learning_rate": 0.0005164117356780603,
      "loss": 1.1568,
      "step": 181500
    },
    {
      "epoch": 36.52,
      "learning_rate": 0.000514991627058243,
      "loss": 1.1599,
      "step": 181600
    },
    {
      "epoch": 36.54,
      "learning_rate": 0.0005135729669768321,
      "loss": 1.1601,
      "step": 181700
    },
    {
      "epoch": 36.56,
      "learning_rate": 0.0005121557582296915,
      "loss": 1.1561,
      "step": 181800
    },
    {
      "epoch": 36.58,
      "learning_rate": 0.0005107400036098259,
      "loss": 1.1579,
      "step": 181900
    },
    {
      "epoch": 36.6,
      "learning_rate": 0.0005093257059073723,
      "loss": 1.1532,
      "step": 182000
    },
    {
      "epoch": 36.62,
      "learning_rate": 0.0005079128679095983,
      "loss": 1.159,
      "step": 182100
    },
    {
      "epoch": 36.64,
      "learning_rate": 0.0005065014924008942,
      "loss": 1.1549,
      "step": 182200
    },
    {
      "epoch": 36.66,
      "learning_rate": 0.0005051056740029337,
      "loss": 1.1537,
      "step": 182300
    },
    {
      "epoch": 36.68,
      "learning_rate": 0.000503697217119769,
      "loss": 1.1526,
      "step": 182400
    },
    {
      "epoch": 36.7,
      "learning_rate": 0.0005022902310337858,
      "loss": 1.1616,
      "step": 182500
    },
    {
      "epoch": 36.72,
      "learning_rate": 0.000500884718517842,
      "loss": 1.1551,
      "step": 182600
    },
    {
      "epoch": 36.74,
      "learning_rate": 0.0004994806823418908,
      "loss": 1.1537,
      "step": 182700
    },
    {
      "epoch": 36.76,
      "learning_rate": 0.0004980781252729766,
      "loss": 1.1562,
      "step": 182800
    },
    {
      "epoch": 36.78,
      "learning_rate": 0.000496677050075227,
      "loss": 1.1569,
      "step": 182900
    },
    {
      "epoch": 36.8,
      "learning_rate": 0.000495277459509851,
      "loss": 1.1549,
      "step": 183000
    },
    {
      "epoch": 36.82,
      "learning_rate": 0.0004938793563351308,
      "loss": 1.1525,
      "step": 183100
    },
    {
      "epoch": 36.84,
      "learning_rate": 0.0004924827433064183,
      "loss": 1.1588,
      "step": 183200
    },
    {
      "epoch": 36.86,
      "learning_rate": 0.0004910876231761266,
      "loss": 1.1492,
      "step": 183300
    },
    {
      "epoch": 36.88,
      "learning_rate": 0.0004896939986937287,
      "loss": 1.1494,
      "step": 183400
    },
    {
      "epoch": 36.9,
      "learning_rate": 0.000488301872605749,
      "loss": 1.1526,
      "step": 183500
    },
    {
      "epoch": 36.92,
      "learning_rate": 0.00048691124765575953,
      "loss": 1.1514,
      "step": 183600
    },
    {
      "epoch": 36.94,
      "learning_rate": 0.0004855221265843726,
      "loss": 1.1562,
      "step": 183700
    },
    {
      "epoch": 36.96,
      "learning_rate": 0.00048413451212923826,
      "loss": 1.1559,
      "step": 183800
    },
    {
      "epoch": 36.98,
      "learning_rate": 0.0004827622605958196,
      "loss": 1.1564,
      "step": 183900
    },
    {
      "epoch": 37.0,
      "learning_rate": 0.0004813776524399174,
      "loss": 1.1529,
      "step": 184000
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.416447232778523,
      "eval_loss": 1.121274709701538,
      "eval_runtime": 19.5197,
      "eval_samples_per_second": 4076.295,
      "eval_steps_per_second": 15.933,
      "step": 184010
    },
    {
      "epoch": 37.02,
      "learning_rate": 0.0004799945590681072,
      "loss": 1.1429,
      "step": 184100
    },
    {
      "epoch": 37.04,
      "learning_rate": 0.0004786129832061578,
      "loss": 1.143,
      "step": 184200
    },
    {
      "epoch": 37.06,
      "learning_rate": 0.00047723292757684944,
      "loss": 1.1453,
      "step": 184300
    },
    {
      "epoch": 37.08,
      "learning_rate": 0.00047585439489996554,
      "loss": 1.1486,
      "step": 184400
    },
    {
      "epoch": 37.1,
      "learning_rate": 0.0004744773878922883,
      "loss": 1.1424,
      "step": 184500
    },
    {
      "epoch": 37.12,
      "learning_rate": 0.0004731019092675921,
      "loss": 1.1469,
      "step": 184600
    },
    {
      "epoch": 37.14,
      "learning_rate": 0.00047172796173664076,
      "loss": 1.1446,
      "step": 184700
    },
    {
      "epoch": 37.16,
      "learning_rate": 0.0004703555480071799,
      "loss": 1.1482,
      "step": 184800
    },
    {
      "epoch": 37.18,
      "learning_rate": 0.00046898467078393294,
      "loss": 1.1476,
      "step": 184900
    },
    {
      "epoch": 37.2,
      "learning_rate": 0.00046761533276859366,
      "loss": 1.1463,
      "step": 185000
    },
    {
      "epoch": 37.22,
      "learning_rate": 0.0004662475366598239,
      "loss": 1.1455,
      "step": 185100
    },
    {
      "epoch": 37.24,
      "learning_rate": 0.00046488128515324634,
      "loss": 1.1469,
      "step": 185200
    },
    {
      "epoch": 37.26,
      "learning_rate": 0.00046351658094144005,
      "loss": 1.1439,
      "step": 185300
    },
    {
      "epoch": 37.28,
      "learning_rate": 0.0004621534267139332,
      "loss": 1.1458,
      "step": 185400
    },
    {
      "epoch": 37.3,
      "learning_rate": 0.00046079182515720076,
      "loss": 1.1454,
      "step": 185500
    },
    {
      "epoch": 37.32,
      "learning_rate": 0.00045943177895465734,
      "loss": 1.1487,
      "step": 185600
    },
    {
      "epoch": 37.34,
      "learning_rate": 0.0004580732907866525,
      "loss": 1.1434,
      "step": 185700
    },
    {
      "epoch": 37.36,
      "learning_rate": 0.00045671636333046426,
      "loss": 1.152,
      "step": 185800
    },
    {
      "epoch": 37.38,
      "learning_rate": 0.00045536099926029585,
      "loss": 1.1452,
      "step": 185900
    },
    {
      "epoch": 37.4,
      "learning_rate": 0.00045400720124726915,
      "loss": 1.1455,
      "step": 186000
    },
    {
      "epoch": 37.42,
      "learning_rate": 0.00045265497195942,
      "loss": 1.146,
      "step": 186100
    },
    {
      "epoch": 37.44,
      "learning_rate": 0.00045130431406169156,
      "loss": 1.1436,
      "step": 186200
    },
    {
      "epoch": 37.46,
      "learning_rate": 0.00044995523021593126,
      "loss": 1.1471,
      "step": 186300
    },
    {
      "epoch": 37.48,
      "learning_rate": 0.0004486077230808838,
      "loss": 1.1461,
      "step": 186400
    },
    {
      "epoch": 37.5,
      "learning_rate": 0.00044726179531218727,
      "loss": 1.1437,
      "step": 186500
    },
    {
      "epoch": 37.52,
      "learning_rate": 0.00044591744956236595,
      "loss": 1.1465,
      "step": 186600
    },
    {
      "epoch": 37.54,
      "learning_rate": 0.00044457468848082713,
      "loss": 1.1502,
      "step": 186700
    },
    {
      "epoch": 37.56,
      "learning_rate": 0.00044323351471385527,
      "loss": 1.1474,
      "step": 186800
    },
    {
      "epoch": 37.58,
      "learning_rate": 0.0004418939309046065,
      "loss": 1.1465,
      "step": 186900
    },
    {
      "epoch": 37.6,
      "learning_rate": 0.00044056931171319787,
      "loss": 1.1458,
      "step": 187000
    },
    {
      "epoch": 37.62,
      "learning_rate": 0.0004392328997709341,
      "loss": 1.1453,
      "step": 187100
    },
    {
      "epoch": 37.64,
      "learning_rate": 0.00043789808567071725,
      "loss": 1.1459,
      "step": 187200
    },
    {
      "epoch": 37.66,
      "learning_rate": 0.0004365648720431699,
      "loss": 1.144,
      "step": 187300
    },
    {
      "epoch": 37.68,
      "learning_rate": 0.0004352332615157606,
      "loss": 1.1478,
      "step": 187400
    },
    {
      "epoch": 37.7,
      "learning_rate": 0.00043390325671279736,
      "loss": 1.1437,
      "step": 187500
    },
    {
      "epoch": 37.72,
      "learning_rate": 0.00043257486025542497,
      "loss": 1.1467,
      "step": 187600
    },
    {
      "epoch": 37.74,
      "learning_rate": 0.0004312480747616181,
      "loss": 1.1486,
      "step": 187700
    },
    {
      "epoch": 37.76,
      "learning_rate": 0.00042992290284617695,
      "loss": 1.1459,
      "step": 187800
    },
    {
      "epoch": 37.78,
      "learning_rate": 0.00042859934712072045,
      "loss": 1.1466,
      "step": 187900
    },
    {
      "epoch": 37.8,
      "learning_rate": 0.00042727741019368354,
      "loss": 1.147,
      "step": 188000
    },
    {
      "epoch": 37.82,
      "learning_rate": 0.0004259570946703109,
      "loss": 1.1458,
      "step": 188100
    },
    {
      "epoch": 37.84,
      "learning_rate": 0.00042463840315265153,
      "loss": 1.1443,
      "step": 188200
    },
    {
      "epoch": 37.86,
      "learning_rate": 0.00042332133823955317,
      "loss": 1.1469,
      "step": 188300
    },
    {
      "epoch": 37.88,
      "learning_rate": 0.00042200590252665886,
      "loss": 1.1453,
      "step": 188400
    },
    {
      "epoch": 37.9,
      "learning_rate": 0.0004206920986064004,
      "loss": 1.1479,
      "step": 188500
    },
    {
      "epoch": 37.92,
      "learning_rate": 0.00041937992906799436,
      "loss": 1.1438,
      "step": 188600
    },
    {
      "epoch": 37.94,
      "learning_rate": 0.00041806939649743484,
      "loss": 1.1443,
      "step": 188700
    },
    {
      "epoch": 37.96,
      "learning_rate": 0.00041676050347749116,
      "loss": 1.1442,
      "step": 188800
    },
    {
      "epoch": 37.98,
      "learning_rate": 0.00041545325258770095,
      "loss": 1.1463,
      "step": 188900
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.41712748843756287,
      "eval_loss": 1.113813042640686,
      "eval_runtime": 19.8226,
      "eval_samples_per_second": 4013.998,
      "eval_steps_per_second": 15.689,
      "step": 188983
    },
    {
      "epoch": 38.0,
      "learning_rate": 0.00041414764640436614,
      "loss": 1.1435,
      "step": 189000
    },
    {
      "epoch": 38.02,
      "learning_rate": 0.0004128436875005459,
      "loss": 1.1341,
      "step": 189100
    },
    {
      "epoch": 38.04,
      "learning_rate": 0.0004115413784460545,
      "loss": 1.1316,
      "step": 189200
    },
    {
      "epoch": 38.06,
      "learning_rate": 0.00041024072180745373,
      "loss": 1.1341,
      "step": 189300
    },
    {
      "epoch": 38.08,
      "learning_rate": 0.00040894172014805,
      "loss": 1.1296,
      "step": 189400
    },
    {
      "epoch": 38.1,
      "learning_rate": 0.00040764437602788627,
      "loss": 1.1366,
      "step": 189500
    },
    {
      "epoch": 38.12,
      "learning_rate": 0.00040634869200374056,
      "loss": 1.1403,
      "step": 189600
    },
    {
      "epoch": 38.14,
      "learning_rate": 0.00040505467062911833,
      "loss": 1.1389,
      "step": 189700
    },
    {
      "epoch": 38.16,
      "learning_rate": 0.00040376231445424904,
      "loss": 1.1316,
      "step": 189800
    },
    {
      "epoch": 38.18,
      "learning_rate": 0.00040247162602607923,
      "loss": 1.1382,
      "step": 189900
    },
    {
      "epoch": 38.2,
      "learning_rate": 0.00040118260788826965,
      "loss": 1.1355,
      "step": 190000
    },
    {
      "epoch": 38.22,
      "learning_rate": 0.00039989526258118915,
      "loss": 1.1365,
      "step": 190100
    },
    {
      "epoch": 38.24,
      "learning_rate": 0.00039860959264191007,
      "loss": 1.136,
      "step": 190200
    },
    {
      "epoch": 38.26,
      "learning_rate": 0.0003973256006042017,
      "loss": 1.1405,
      "step": 190300
    },
    {
      "epoch": 38.28,
      "learning_rate": 0.00039604328899852786,
      "loss": 1.1402,
      "step": 190400
    },
    {
      "epoch": 38.3,
      "learning_rate": 0.00039476266035204003,
      "loss": 1.1375,
      "step": 190500
    },
    {
      "epoch": 38.33,
      "learning_rate": 0.00039348371718857386,
      "loss": 1.1352,
      "step": 190600
    },
    {
      "epoch": 38.35,
      "learning_rate": 0.0003922064620286414,
      "loss": 1.1363,
      "step": 190700
    },
    {
      "epoch": 38.37,
      "learning_rate": 0.00039093089738942975,
      "loss": 1.1377,
      "step": 190800
    },
    {
      "epoch": 38.39,
      "learning_rate": 0.0003896570257847937,
      "loss": 1.1385,
      "step": 190900
    },
    {
      "epoch": 38.41,
      "learning_rate": 0.00038839756308466366,
      "loss": 1.1418,
      "step": 191000
    },
    {
      "epoch": 38.43,
      "learning_rate": 0.0003871270680844698,
      "loss": 1.1369,
      "step": 191100
    },
    {
      "epoch": 38.45,
      "learning_rate": 0.0003858582736153553,
      "loss": 1.1408,
      "step": 191200
    },
    {
      "epoch": 38.47,
      "learning_rate": 0.00038459118217783184,
      "loss": 1.1376,
      "step": 191300
    },
    {
      "epoch": 38.49,
      "learning_rate": 0.000383325796269056,
      "loss": 1.1401,
      "step": 191400
    },
    {
      "epoch": 38.51,
      "learning_rate": 0.0003820621183828224,
      "loss": 1.1371,
      "step": 191500
    },
    {
      "epoch": 38.53,
      "learning_rate": 0.0003808001510095603,
      "loss": 1.1346,
      "step": 191600
    },
    {
      "epoch": 38.55,
      "learning_rate": 0.0003795398966363266,
      "loss": 1.1439,
      "step": 191700
    },
    {
      "epoch": 38.57,
      "learning_rate": 0.0003782813577468032,
      "loss": 1.1426,
      "step": 191800
    },
    {
      "epoch": 38.59,
      "learning_rate": 0.00037702453682129114,
      "loss": 1.1383,
      "step": 191900
    },
    {
      "epoch": 38.61,
      "learning_rate": 0.00037576943633670605,
      "loss": 1.1362,
      "step": 192000
    },
    {
      "epoch": 38.63,
      "learning_rate": 0.0003745160587665715,
      "loss": 1.1356,
      "step": 192100
    },
    {
      "epoch": 38.65,
      "learning_rate": 0.000373264406581017,
      "loss": 1.1333,
      "step": 192200
    },
    {
      "epoch": 38.67,
      "learning_rate": 0.00037201448224677153,
      "loss": 1.14,
      "step": 192300
    },
    {
      "epoch": 38.69,
      "learning_rate": 0.0003707662882271585,
      "loss": 1.1349,
      "step": 192400
    },
    {
      "epoch": 38.71,
      "learning_rate": 0.00036951982698209094,
      "loss": 1.1371,
      "step": 192500
    },
    {
      "epoch": 38.73,
      "learning_rate": 0.0003682751009680678,
      "loss": 1.1391,
      "step": 192600
    },
    {
      "epoch": 38.75,
      "learning_rate": 0.0003670321126381676,
      "loss": 1.134,
      "step": 192700
    },
    {
      "epoch": 38.77,
      "learning_rate": 0.000365790864442045,
      "loss": 1.1358,
      "step": 192800
    },
    {
      "epoch": 38.79,
      "learning_rate": 0.0003645513588259242,
      "loss": 1.1331,
      "step": 192900
    },
    {
      "epoch": 38.81,
      "learning_rate": 0.0003633135982325958,
      "loss": 1.1376,
      "step": 193000
    },
    {
      "epoch": 38.83,
      "learning_rate": 0.0003620775851014114,
      "loss": 1.1353,
      "step": 193100
    },
    {
      "epoch": 38.85,
      "learning_rate": 0.00036084332186827895,
      "loss": 1.1376,
      "step": 193200
    },
    {
      "epoch": 38.87,
      "learning_rate": 0.0003596231273926687,
      "loss": 1.1346,
      "step": 193300
    },
    {
      "epoch": 38.89,
      "learning_rate": 0.0003583923536899551,
      "loss": 1.1375,
      "step": 193400
    },
    {
      "epoch": 38.91,
      "learning_rate": 0.0003571633371480666,
      "loss": 1.1387,
      "step": 193500
    },
    {
      "epoch": 38.93,
      "learning_rate": 0.0003559360801891219,
      "loss": 1.1375,
      "step": 193600
    },
    {
      "epoch": 38.95,
      "learning_rate": 0.00035471058523177265,
      "loss": 1.1357,
      "step": 193700
    },
    {
      "epoch": 38.97,
      "learning_rate": 0.00035348685469119664,
      "loss": 1.136,
      "step": 193800
    },
    {
      "epoch": 38.99,
      "learning_rate": 0.0003522648909790957,
      "loss": 1.1352,
      "step": 193900
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.41790908246302183,
      "eval_loss": 1.1047524213790894,
      "eval_runtime": 19.8895,
      "eval_samples_per_second": 4000.506,
      "eval_steps_per_second": 15.636,
      "step": 193956
    },
    {
      "epoch": 39.01,
      "learning_rate": 0.000351044696503689,
      "loss": 1.1315,
      "step": 194000
    },
    {
      "epoch": 39.03,
      "learning_rate": 0.0003498262736697096,
      "loss": 1.1213,
      "step": 194100
    },
    {
      "epoch": 39.05,
      "learning_rate": 0.00034860962487839767,
      "loss": 1.126,
      "step": 194200
    },
    {
      "epoch": 39.07,
      "learning_rate": 0.00034739475252749854,
      "loss": 1.125,
      "step": 194300
    },
    {
      "epoch": 39.09,
      "learning_rate": 0.0003461816590112558,
      "loss": 1.1259,
      "step": 194400
    },
    {
      "epoch": 39.11,
      "learning_rate": 0.00034497034672040824,
      "loss": 1.1266,
      "step": 194500
    },
    {
      "epoch": 39.13,
      "learning_rate": 0.00034376081804218263,
      "loss": 1.1286,
      "step": 194600
    },
    {
      "epoch": 39.15,
      "learning_rate": 0.00034255307536029206,
      "loss": 1.1286,
      "step": 194700
    },
    {
      "epoch": 39.17,
      "learning_rate": 0.00034134712105492946,
      "loss": 1.1294,
      "step": 194800
    },
    {
      "epoch": 39.19,
      "learning_rate": 0.00034014295750276326,
      "loss": 1.1256,
      "step": 194900
    },
    {
      "epoch": 39.21,
      "learning_rate": 0.0003389405870769323,
      "loss": 1.1315,
      "step": 195000
    },
    {
      "epoch": 39.23,
      "learning_rate": 0.00033774001214704184,
      "loss": 1.1328,
      "step": 195100
    },
    {
      "epoch": 39.25,
      "learning_rate": 0.00033654123507915887,
      "loss": 1.1278,
      "step": 195200
    },
    {
      "epoch": 39.27,
      "learning_rate": 0.0003353562190853805,
      "loss": 1.1284,
      "step": 195300
    },
    {
      "epoch": 39.29,
      "learning_rate": 0.00033416102678803547,
      "loss": 1.1292,
      "step": 195400
    },
    {
      "epoch": 39.31,
      "learning_rate": 0.00033296763940608335,
      "loss": 1.1335,
      "step": 195500
    },
    {
      "epoch": 39.33,
      "learning_rate": 0.0003317760592914265,
      "loss": 1.1264,
      "step": 195600
    },
    {
      "epoch": 39.35,
      "learning_rate": 0.00033058628879240477,
      "loss": 1.1293,
      "step": 195700
    },
    {
      "epoch": 39.37,
      "learning_rate": 0.0003293983302537923,
      "loss": 1.1275,
      "step": 195800
    },
    {
      "epoch": 39.39,
      "learning_rate": 0.00032821218601679066,
      "loss": 1.127,
      "step": 195900
    },
    {
      "epoch": 39.41,
      "learning_rate": 0.00032702785841902783,
      "loss": 1.1287,
      "step": 196000
    },
    {
      "epoch": 39.43,
      "learning_rate": 0.0003258453497945503,
      "loss": 1.1278,
      "step": 196100
    },
    {
      "epoch": 39.45,
      "learning_rate": 0.0003246646624738207,
      "loss": 1.1296,
      "step": 196200
    },
    {
      "epoch": 39.47,
      "learning_rate": 0.00032348579878371114,
      "loss": 1.1309,
      "step": 196300
    },
    {
      "epoch": 39.49,
      "learning_rate": 0.00032230876104750046,
      "loss": 1.1274,
      "step": 196400
    },
    {
      "epoch": 39.51,
      "learning_rate": 0.000321133551584869,
      "loss": 1.1295,
      "step": 196500
    },
    {
      "epoch": 39.53,
      "learning_rate": 0.00031996017271189437,
      "loss": 1.1271,
      "step": 196600
    },
    {
      "epoch": 39.55,
      "learning_rate": 0.0003187886267410448,
      "loss": 1.1254,
      "step": 196700
    },
    {
      "epoch": 39.57,
      "learning_rate": 0.0003176189159811782,
      "loss": 1.1303,
      "step": 196800
    },
    {
      "epoch": 39.59,
      "learning_rate": 0.00031645104273753474,
      "loss": 1.1273,
      "step": 196900
    },
    {
      "epoch": 39.61,
      "learning_rate": 0.0003152850093117338,
      "loss": 1.13,
      "step": 197000
    },
    {
      "epoch": 39.63,
      "learning_rate": 0.0003141208180017678,
      "loss": 1.1253,
      "step": 197100
    },
    {
      "epoch": 39.65,
      "learning_rate": 0.00031295847110199976,
      "loss": 1.1286,
      "step": 197200
    },
    {
      "epoch": 39.67,
      "learning_rate": 0.0003117979709031578,
      "loss": 1.1295,
      "step": 197300
    },
    {
      "epoch": 39.69,
      "learning_rate": 0.0003106393196923304,
      "loss": 1.1279,
      "step": 197400
    },
    {
      "epoch": 39.71,
      "learning_rate": 0.00030949407858107237,
      "loss": 1.1279,
      "step": 197500
    },
    {
      "epoch": 39.73,
      "learning_rate": 0.00030833911364617274,
      "loss": 1.1268,
      "step": 197600
    },
    {
      "epoch": 39.75,
      "learning_rate": 0.0003071860045159272,
      "loss": 1.1229,
      "step": 197700
    },
    {
      "epoch": 39.77,
      "learning_rate": 0.00030603475346285824,
      "loss": 1.1263,
      "step": 197800
    },
    {
      "epoch": 39.79,
      "learning_rate": 0.0003048853627558264,
      "loss": 1.1254,
      "step": 197900
    },
    {
      "epoch": 39.81,
      "learning_rate": 0.0003037378346600259,
      "loss": 1.1294,
      "step": 198000
    },
    {
      "epoch": 39.83,
      "learning_rate": 0.0003025921714369792,
      "loss": 1.1278,
      "step": 198100
    },
    {
      "epoch": 39.85,
      "learning_rate": 0.0003014483753445349,
      "loss": 1.1253,
      "step": 198200
    },
    {
      "epoch": 39.87,
      "learning_rate": 0.000300306448636861,
      "loss": 1.1272,
      "step": 198300
    },
    {
      "epoch": 39.89,
      "learning_rate": 0.0002991663935644423,
      "loss": 1.1225,
      "step": 198400
    },
    {
      "epoch": 39.91,
      "learning_rate": 0.0002980282123740735,
      "loss": 1.1233,
      "step": 198500
    },
    {
      "epoch": 39.93,
      "learning_rate": 0.00029689190730885747,
      "loss": 1.1229,
      "step": 198600
    },
    {
      "epoch": 39.95,
      "learning_rate": 0.00029575748060819946,
      "loss": 1.1273,
      "step": 198700
    },
    {
      "epoch": 39.97,
      "learning_rate": 0.00029462493450780307,
      "loss": 1.1271,
      "step": 198800
    },
    {
      "epoch": 39.99,
      "learning_rate": 0.0002934942712396647,
      "loss": 1.1259,
      "step": 198900
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.41850995375025135,
      "eval_loss": 1.096311330795288,
      "eval_runtime": 19.5694,
      "eval_samples_per_second": 4065.946,
      "eval_steps_per_second": 15.892,
      "step": 198930
    },
    {
      "epoch": 40.01,
      "learning_rate": 0.0002923654930320711,
      "loss": 1.116,
      "step": 199000
    },
    {
      "epoch": 40.03,
      "learning_rate": 0.0002912386021095936,
      "loss": 1.1166,
      "step": 199100
    },
    {
      "epoch": 40.05,
      "learning_rate": 0.00029011360069308446,
      "loss": 1.1156,
      "step": 199200
    },
    {
      "epoch": 40.07,
      "learning_rate": 0.0002889904909996713,
      "loss": 1.1193,
      "step": 199300
    },
    {
      "epoch": 40.09,
      "learning_rate": 0.0002878692752427543,
      "loss": 1.1196,
      "step": 199400
    },
    {
      "epoch": 40.11,
      "learning_rate": 0.0002867611394349388,
      "loss": 1.123,
      "step": 199500
    },
    {
      "epoch": 40.13,
      "learning_rate": 0.0002856436991818517,
      "loss": 1.117,
      "step": 199600
    },
    {
      "epoch": 40.15,
      "learning_rate": 0.00028452815946104537,
      "loss": 1.1189,
      "step": 199700
    },
    {
      "epoch": 40.17,
      "learning_rate": 0.00028341452247099996,
      "loss": 1.1162,
      "step": 199800
    },
    {
      "epoch": 40.2,
      "learning_rate": 0.0002823027904064475,
      "loss": 1.1149,
      "step": 199900
    },
    {
      "epoch": 40.22,
      "learning_rate": 0.00028119296545836515,
      "loss": 1.1151,
      "step": 200000
    },
    {
      "epoch": 40.24,
      "learning_rate": 0.000280085049813972,
      "loss": 1.1168,
      "step": 200100
    },
    {
      "epoch": 40.26,
      "learning_rate": 0.00027897904565672314,
      "loss": 1.1182,
      "step": 200200
    },
    {
      "epoch": 40.28,
      "learning_rate": 0.00027787495516630776,
      "loss": 1.1149,
      "step": 200300
    },
    {
      "epoch": 40.3,
      "learning_rate": 0.0002767727805186432,
      "loss": 1.1155,
      "step": 200400
    },
    {
      "epoch": 40.32,
      "learning_rate": 0.0002756725238858715,
      "loss": 1.1182,
      "step": 200500
    },
    {
      "epoch": 40.34,
      "learning_rate": 0.00027457418743635374,
      "loss": 1.1201,
      "step": 200600
    },
    {
      "epoch": 40.36,
      "learning_rate": 0.00027347777333466746,
      "loss": 1.1188,
      "step": 200700
    },
    {
      "epoch": 40.38,
      "learning_rate": 0.0002723832837416017,
      "loss": 1.1222,
      "step": 200800
    },
    {
      "epoch": 40.4,
      "learning_rate": 0.0002712907208141528,
      "loss": 1.1186,
      "step": 200900
    },
    {
      "epoch": 40.42,
      "learning_rate": 0.00027020008670551935,
      "loss": 1.121,
      "step": 201000
    },
    {
      "epoch": 40.44,
      "learning_rate": 0.0002691113835650995,
      "loss": 1.1172,
      "step": 201100
    },
    {
      "epoch": 40.46,
      "learning_rate": 0.0002680246135384853,
      "loss": 1.1223,
      "step": 201200
    },
    {
      "epoch": 40.48,
      "learning_rate": 0.00026693977876746,
      "loss": 1.118,
      "step": 201300
    },
    {
      "epoch": 40.5,
      "learning_rate": 0.0002658568813899914,
      "loss": 1.1155,
      "step": 201400
    },
    {
      "epoch": 40.52,
      "learning_rate": 0.00026478672351106777,
      "loss": 1.1191,
      "step": 201500
    },
    {
      "epoch": 40.54,
      "learning_rate": 0.00026370768789222914,
      "loss": 1.1167,
      "step": 201600
    },
    {
      "epoch": 40.56,
      "learning_rate": 0.0002626305960366815,
      "loss": 1.1194,
      "step": 201700
    },
    {
      "epoch": 40.58,
      "learning_rate": 0.00026155545006713424,
      "loss": 1.1171,
      "step": 201800
    },
    {
      "epoch": 40.6,
      "learning_rate": 0.00026048225210246063,
      "loss": 1.1199,
      "step": 201900
    },
    {
      "epoch": 40.62,
      "learning_rate": 0.0002594110042576962,
      "loss": 1.1161,
      "step": 202000
    },
    {
      "epoch": 40.64,
      "learning_rate": 0.00025834170864403287,
      "loss": 1.1146,
      "step": 202100
    },
    {
      "epoch": 40.66,
      "learning_rate": 0.00025727436736881505,
      "loss": 1.117,
      "step": 202200
    },
    {
      "epoch": 40.68,
      "learning_rate": 0.00025620898253553515,
      "loss": 1.1194,
      "step": 202300
    },
    {
      "epoch": 40.7,
      "learning_rate": 0.00025514555624383053,
      "loss": 1.1188,
      "step": 202400
    },
    {
      "epoch": 40.72,
      "learning_rate": 0.0002540840905894784,
      "loss": 1.1166,
      "step": 202500
    },
    {
      "epoch": 40.74,
      "learning_rate": 0.0002530245876643923,
      "loss": 1.1188,
      "step": 202600
    },
    {
      "epoch": 40.76,
      "learning_rate": 0.0002519670495566169,
      "loss": 1.1191,
      "step": 202700
    },
    {
      "epoch": 40.78,
      "learning_rate": 0.00025091147835032526,
      "loss": 1.1151,
      "step": 202800
    },
    {
      "epoch": 40.8,
      "learning_rate": 0.00024985787612581423,
      "loss": 1.1194,
      "step": 202900
    },
    {
      "epoch": 40.82,
      "learning_rate": 0.00024880624495950024,
      "loss": 1.1167,
      "step": 203000
    },
    {
      "epoch": 40.84,
      "learning_rate": 0.00024775658692391416,
      "loss": 1.1208,
      "step": 203100
    },
    {
      "epoch": 40.86,
      "learning_rate": 0.0002467089040876995,
      "loss": 1.1195,
      "step": 203200
    },
    {
      "epoch": 40.88,
      "learning_rate": 0.00024566319851560675,
      "loss": 1.1142,
      "step": 203300
    },
    {
      "epoch": 40.9,
      "learning_rate": 0.00024461947226848984,
      "loss": 1.1159,
      "step": 203400
    },
    {
      "epoch": 40.92,
      "learning_rate": 0.00024357772740330076,
      "loss": 1.1104,
      "step": 203500
    },
    {
      "epoch": 40.94,
      "learning_rate": 0.00024254835376265563,
      "loss": 1.1169,
      "step": 203600
    },
    {
      "epoch": 40.96,
      "learning_rate": 0.00024151055795158666,
      "loss": 1.1174,
      "step": 203700
    },
    {
      "epoch": 40.98,
      "learning_rate": 0.00024047474964942626,
      "loss": 1.1196,
      "step": 203800
    },
    {
      "epoch": 41.0,
      "learning_rate": 0.00023944093089752302,
      "loss": 1.1194,
      "step": 203900
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.41941595374516055,
      "eval_loss": 1.0883480310440063,
      "eval_runtime": 19.8865,
      "eval_samples_per_second": 4001.112,
      "eval_steps_per_second": 15.639,
      "step": 203903
    },
    {
      "epoch": 41.02,
      "learning_rate": 0.00023840910373330374,
      "loss": 1.1042,
      "step": 204000
    },
    {
      "epoch": 41.04,
      "learning_rate": 0.00023737927019027105,
      "loss": 1.1048,
      "step": 204100
    },
    {
      "epoch": 41.06,
      "learning_rate": 0.00023635143229799844,
      "loss": 1.1014,
      "step": 204200
    },
    {
      "epoch": 41.08,
      "learning_rate": 0.0002353255920821265,
      "loss": 1.1069,
      "step": 204300
    },
    {
      "epoch": 41.1,
      "learning_rate": 0.0002343017515643582,
      "loss": 1.107,
      "step": 204400
    },
    {
      "epoch": 41.12,
      "learning_rate": 0.00023327991276245617,
      "loss": 1.1049,
      "step": 204500
    },
    {
      "epoch": 41.14,
      "learning_rate": 0.00023226007769023795,
      "loss": 1.106,
      "step": 204600
    },
    {
      "epoch": 41.16,
      "learning_rate": 0.00023124224835757254,
      "loss": 1.109,
      "step": 204700
    },
    {
      "epoch": 41.18,
      "learning_rate": 0.00023022642677037505,
      "loss": 1.1107,
      "step": 204800
    },
    {
      "epoch": 41.2,
      "learning_rate": 0.0002292126149306048,
      "loss": 1.1064,
      "step": 204900
    },
    {
      "epoch": 41.22,
      "learning_rate": 0.00022820081483625993,
      "loss": 1.1077,
      "step": 205000
    },
    {
      "epoch": 41.24,
      "learning_rate": 0.00022719102848137426,
      "loss": 1.1066,
      "step": 205100
    },
    {
      "epoch": 41.26,
      "learning_rate": 0.0002261832578560119,
      "loss": 1.1104,
      "step": 205200
    },
    {
      "epoch": 41.28,
      "learning_rate": 0.00022517750494626537,
      "loss": 1.1109,
      "step": 205300
    },
    {
      "epoch": 41.3,
      "learning_rate": 0.00022417377173425068,
      "loss": 1.1095,
      "step": 205400
    },
    {
      "epoch": 41.32,
      "learning_rate": 0.00022317206019810355,
      "loss": 1.1118,
      "step": 205500
    },
    {
      "epoch": 41.34,
      "learning_rate": 0.00022218235916729556,
      "loss": 1.1085,
      "step": 205600
    },
    {
      "epoch": 41.36,
      "learning_rate": 0.00022118467663540713,
      "loss": 1.1091,
      "step": 205700
    },
    {
      "epoch": 41.38,
      "learning_rate": 0.00022018902167022874,
      "loss": 1.1113,
      "step": 205800
    },
    {
      "epoch": 41.4,
      "learning_rate": 0.00021919539623397572,
      "loss": 1.1041,
      "step": 205900
    },
    {
      "epoch": 41.42,
      "learning_rate": 0.00021820380228486328,
      "loss": 1.1076,
      "step": 206000
    },
    {
      "epoch": 41.44,
      "learning_rate": 0.00021721424177710335,
      "loss": 1.1081,
      "step": 206100
    },
    {
      "epoch": 41.46,
      "learning_rate": 0.00021622671666089957,
      "loss": 1.1087,
      "step": 206200
    },
    {
      "epoch": 41.48,
      "learning_rate": 0.00021524122888244514,
      "loss": 1.1065,
      "step": 206300
    },
    {
      "epoch": 41.5,
      "learning_rate": 0.00021425778038391768,
      "loss": 1.1096,
      "step": 206400
    },
    {
      "epoch": 41.52,
      "learning_rate": 0.00021327637310347648,
      "loss": 1.1105,
      "step": 206500
    },
    {
      "epoch": 41.54,
      "learning_rate": 0.0002122970089752567,
      "loss": 1.109,
      "step": 206600
    },
    {
      "epoch": 41.56,
      "learning_rate": 0.00021131968992936804,
      "loss": 1.1114,
      "step": 206700
    },
    {
      "epoch": 41.58,
      "learning_rate": 0.00021034441789188953,
      "loss": 1.1096,
      "step": 206800
    },
    {
      "epoch": 41.6,
      "learning_rate": 0.00020937119478486616,
      "loss": 1.1161,
      "step": 206900
    },
    {
      "epoch": 41.62,
      "learning_rate": 0.00020840002252630428,
      "loss": 1.1108,
      "step": 207000
    },
    {
      "epoch": 41.64,
      "learning_rate": 0.00020743090303016923,
      "loss": 1.1092,
      "step": 207100
    },
    {
      "epoch": 41.66,
      "learning_rate": 0.00020646383820638022,
      "loss": 1.1069,
      "step": 207200
    },
    {
      "epoch": 41.68,
      "learning_rate": 0.00020549882996080793,
      "loss": 1.108,
      "step": 207300
    },
    {
      "epoch": 41.7,
      "learning_rate": 0.0002045358801952689,
      "loss": 1.1072,
      "step": 207400
    },
    {
      "epoch": 41.72,
      "learning_rate": 0.00020357499080752366,
      "loss": 1.108,
      "step": 207500
    },
    {
      "epoch": 41.74,
      "learning_rate": 0.00020261616369127183,
      "loss": 1.1074,
      "step": 207600
    },
    {
      "epoch": 41.76,
      "learning_rate": 0.0002016689581419083,
      "loss": 1.1075,
      "step": 207700
    },
    {
      "epoch": 41.78,
      "learning_rate": 0.00020071424056369538,
      "loss": 1.1085,
      "step": 207800
    },
    {
      "epoch": 41.8,
      "learning_rate": 0.00019976159089487967,
      "loss": 1.1082,
      "step": 207900
    },
    {
      "epoch": 41.82,
      "learning_rate": 0.0001988110110129225,
      "loss": 1.1102,
      "step": 208000
    },
    {
      "epoch": 41.84,
      "learning_rate": 0.00019786250279120606,
      "loss": 1.1108,
      "step": 208100
    },
    {
      "epoch": 41.86,
      "learning_rate": 0.00019691606809902964,
      "loss": 1.108,
      "step": 208200
    },
    {
      "epoch": 41.88,
      "learning_rate": 0.00019597170880160568,
      "loss": 1.1053,
      "step": 208300
    },
    {
      "epoch": 41.9,
      "learning_rate": 0.00019502942676005705,
      "loss": 1.1093,
      "step": 208400
    },
    {
      "epoch": 41.92,
      "learning_rate": 0.00019408922383141235,
      "loss": 1.109,
      "step": 208500
    },
    {
      "epoch": 41.94,
      "learning_rate": 0.0001931511018686033,
      "loss": 1.1064,
      "step": 208600
    },
    {
      "epoch": 41.96,
      "learning_rate": 0.0001922150627204597,
      "loss": 1.1058,
      "step": 208700
    },
    {
      "epoch": 41.98,
      "learning_rate": 0.00019128110823170717,
      "loss": 1.1035,
      "step": 208800
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.42008968684058473,
      "eval_loss": 1.080112099647522,
      "eval_runtime": 19.7792,
      "eval_samples_per_second": 4022.819,
      "eval_steps_per_second": 15.724,
      "step": 208876
    },
    {
      "epoch": 42.0,
      "learning_rate": 0.00019034924024296282,
      "loss": 1.1021,
      "step": 208900
    },
    {
      "epoch": 42.02,
      "learning_rate": 0.00018941946059073192,
      "loss": 1.097,
      "step": 209000
    },
    {
      "epoch": 42.04,
      "learning_rate": 0.0001884917711074033,
      "loss": 1.0949,
      "step": 209100
    },
    {
      "epoch": 42.07,
      "learning_rate": 0.00018756617362124722,
      "loss": 1.094,
      "step": 209200
    },
    {
      "epoch": 42.09,
      "learning_rate": 0.00018664266995641125,
      "loss": 1.0974,
      "step": 209300
    },
    {
      "epoch": 42.11,
      "learning_rate": 0.00018572126193291613,
      "loss": 1.1003,
      "step": 209400
    },
    {
      "epoch": 42.13,
      "learning_rate": 0.00018480195136665227,
      "loss": 1.0985,
      "step": 209500
    },
    {
      "epoch": 42.15,
      "learning_rate": 0.00018388474006937703,
      "loss": 1.0984,
      "step": 209600
    },
    {
      "epoch": 42.17,
      "learning_rate": 0.0001829787705446627,
      "loss": 1.0988,
      "step": 209700
    },
    {
      "epoch": 42.19,
      "learning_rate": 0.0001820657421663692,
      "loss": 1.0996,
      "step": 209800
    },
    {
      "epoch": 42.21,
      "learning_rate": 0.00018115481844952584,
      "loss": 1.099,
      "step": 209900
    },
    {
      "epoch": 42.23,
      "learning_rate": 0.00018024600118936058,
      "loss": 1.0987,
      "step": 210000
    },
    {
      "epoch": 42.25,
      "learning_rate": 0.0001793392921769506,
      "loss": 1.1014,
      "step": 210100
    },
    {
      "epoch": 42.27,
      "learning_rate": 0.0001784346931992184,
      "loss": 1.1018,
      "step": 210200
    },
    {
      "epoch": 42.29,
      "learning_rate": 0.0001775322060389281,
      "loss": 1.0979,
      "step": 210300
    },
    {
      "epoch": 42.31,
      "learning_rate": 0.000176631832474681,
      "loss": 1.1027,
      "step": 210400
    },
    {
      "epoch": 42.33,
      "learning_rate": 0.00017573357428091394,
      "loss": 1.0946,
      "step": 210500
    },
    {
      "epoch": 42.35,
      "learning_rate": 0.0001748374332278946,
      "loss": 1.0991,
      "step": 210600
    },
    {
      "epoch": 42.37,
      "learning_rate": 0.00017394341108171812,
      "loss": 1.1001,
      "step": 210700
    },
    {
      "epoch": 42.39,
      "learning_rate": 0.00017305150960430338,
      "loss": 1.0999,
      "step": 210800
    },
    {
      "epoch": 42.41,
      "learning_rate": 0.00017216173055339067,
      "loss": 1.1005,
      "step": 210900
    },
    {
      "epoch": 42.43,
      "learning_rate": 0.0001712740756825369,
      "loss": 1.0971,
      "step": 211000
    },
    {
      "epoch": 42.45,
      "learning_rate": 0.0001703885467411133,
      "loss": 1.1002,
      "step": 211100
    },
    {
      "epoch": 42.47,
      "learning_rate": 0.00016950514547430003,
      "loss": 1.1011,
      "step": 211200
    },
    {
      "epoch": 42.49,
      "learning_rate": 0.00016862387362308565,
      "loss": 1.1001,
      "step": 211300
    },
    {
      "epoch": 42.51,
      "learning_rate": 0.00016774473292426124,
      "loss": 1.0965,
      "step": 211400
    },
    {
      "epoch": 42.53,
      "learning_rate": 0.00016686772511041823,
      "loss": 1.0978,
      "step": 211500
    },
    {
      "epoch": 42.55,
      "learning_rate": 0.00016599285190994393,
      "loss": 1.098,
      "step": 211600
    },
    {
      "epoch": 42.57,
      "learning_rate": 0.00016512011504701954,
      "loss": 1.0967,
      "step": 211700
    },
    {
      "epoch": 42.59,
      "learning_rate": 0.00016425821164064812,
      "loss": 1.0975,
      "step": 211800
    },
    {
      "epoch": 42.61,
      "learning_rate": 0.00016338973120231013,
      "loss": 1.1008,
      "step": 211900
    },
    {
      "epoch": 42.63,
      "learning_rate": 0.00016252339223169542,
      "loss": 1.1004,
      "step": 212000
    },
    {
      "epoch": 42.65,
      "learning_rate": 0.0001616591964361662,
      "loss": 1.0986,
      "step": 212100
    },
    {
      "epoch": 42.67,
      "learning_rate": 0.00016079714551885991,
      "loss": 1.0978,
      "step": 212200
    },
    {
      "epoch": 42.69,
      "learning_rate": 0.00015993724117868786,
      "loss": 1.1004,
      "step": 212300
    },
    {
      "epoch": 42.71,
      "learning_rate": 0.00015907948511033082,
      "loss": 1.0987,
      "step": 212400
    },
    {
      "epoch": 42.73,
      "learning_rate": 0.00015822387900423591,
      "loss": 1.1042,
      "step": 212500
    },
    {
      "epoch": 42.75,
      "learning_rate": 0.00015737042454661232,
      "loss": 1.0976,
      "step": 212600
    },
    {
      "epoch": 42.77,
      "learning_rate": 0.00015652762576620364,
      "loss": 1.1006,
      "step": 212700
    },
    {
      "epoch": 42.79,
      "learning_rate": 0.00015567845808881514,
      "loss": 1.1024,
      "step": 212800
    },
    {
      "epoch": 42.81,
      "learning_rate": 0.00015483144707635803,
      "loss": 1.0976,
      "step": 212900
    },
    {
      "epoch": 42.83,
      "learning_rate": 0.0001539865943981035,
      "loss": 1.0966,
      "step": 213000
    },
    {
      "epoch": 42.85,
      "learning_rate": 0.0001531439017190679,
      "loss": 1.1029,
      "step": 213100
    },
    {
      "epoch": 42.87,
      "learning_rate": 0.0001523033707000121,
      "loss": 1.0952,
      "step": 213200
    },
    {
      "epoch": 42.89,
      "learning_rate": 0.00015146500299743584,
      "loss": 1.0991,
      "step": 213300
    },
    {
      "epoch": 42.91,
      "learning_rate": 0.0001506288002635764,
      "loss": 1.0977,
      "step": 213400
    },
    {
      "epoch": 42.93,
      "learning_rate": 0.00014979476414640313,
      "loss": 1.095,
      "step": 213500
    },
    {
      "epoch": 42.95,
      "learning_rate": 0.00014896289628961654,
      "loss": 1.0932,
      "step": 213600
    },
    {
      "epoch": 42.97,
      "learning_rate": 0.00014813319833264339,
      "loss": 1.0971,
      "step": 213700
    },
    {
      "epoch": 42.99,
      "learning_rate": 0.00014730567191063472,
      "loss": 1.0962,
      "step": 213800
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.4207685107173674,
      "eval_loss": 1.071601152420044,
      "eval_runtime": 19.6492,
      "eval_samples_per_second": 4049.421,
      "eval_steps_per_second": 15.828,
      "step": 213849
    },
    {
      "epoch": 43.01,
      "learning_rate": 0.00014648031865446083,
      "loss": 1.0912,
      "step": 213900
    },
    {
      "epoch": 43.03,
      "learning_rate": 0.00014565714019071016,
      "loss": 1.0872,
      "step": 214000
    },
    {
      "epoch": 43.05,
      "learning_rate": 0.00014483613814168474,
      "loss": 1.0874,
      "step": 214100
    },
    {
      "epoch": 43.07,
      "learning_rate": 0.0001440173141253978,
      "loss": 1.0936,
      "step": 214200
    },
    {
      "epoch": 43.09,
      "learning_rate": 0.00014320066975556919,
      "loss": 1.0912,
      "step": 214300
    },
    {
      "epoch": 43.11,
      "learning_rate": 0.000142386206641624,
      "loss": 1.0885,
      "step": 214400
    },
    {
      "epoch": 43.13,
      "learning_rate": 0.00014157392638868823,
      "loss": 1.0879,
      "step": 214500
    },
    {
      "epoch": 43.15,
      "learning_rate": 0.00014076383059758642,
      "loss": 1.0882,
      "step": 214600
    },
    {
      "epoch": 43.17,
      "learning_rate": 0.0001399559208648371,
      "loss": 1.0894,
      "step": 214700
    },
    {
      "epoch": 43.19,
      "learning_rate": 0.00013915019878265114,
      "loss": 1.0883,
      "step": 214800
    },
    {
      "epoch": 43.21,
      "learning_rate": 0.0001383466659389282,
      "loss": 1.0865,
      "step": 214900
    },
    {
      "epoch": 43.23,
      "learning_rate": 0.00013754532391725333,
      "loss": 1.0899,
      "step": 215000
    },
    {
      "epoch": 43.25,
      "learning_rate": 0.0001367461742968934,
      "loss": 1.09,
      "step": 215100
    },
    {
      "epoch": 43.27,
      "learning_rate": 0.0001359492186527951,
      "loss": 1.0892,
      "step": 215200
    },
    {
      "epoch": 43.29,
      "learning_rate": 0.0001351544585555814,
      "loss": 1.0893,
      "step": 215300
    },
    {
      "epoch": 43.31,
      "learning_rate": 0.00013436189557154823,
      "loss": 1.0895,
      "step": 215400
    },
    {
      "epoch": 43.33,
      "learning_rate": 0.0001335715312626608,
      "loss": 1.0932,
      "step": 215500
    },
    {
      "epoch": 43.35,
      "learning_rate": 0.00013278336718655206,
      "loss": 1.0897,
      "step": 215600
    },
    {
      "epoch": 43.37,
      "learning_rate": 0.00013199740489651862,
      "loss": 1.0899,
      "step": 215700
    },
    {
      "epoch": 43.39,
      "learning_rate": 0.0001312136459415178,
      "loss": 1.0911,
      "step": 215800
    },
    {
      "epoch": 43.41,
      "learning_rate": 0.00013043209186616432,
      "loss": 1.0859,
      "step": 215900
    },
    {
      "epoch": 43.43,
      "learning_rate": 0.000129652744210728,
      "loss": 1.0889,
      "step": 216000
    },
    {
      "epoch": 43.45,
      "learning_rate": 0.00012887560451113,
      "loss": 1.0887,
      "step": 216100
    },
    {
      "epoch": 43.47,
      "learning_rate": 0.0001281006742989406,
      "loss": 1.0937,
      "step": 216200
    },
    {
      "epoch": 43.49,
      "learning_rate": 0.0001273279551013748,
      "loss": 1.0875,
      "step": 216300
    },
    {
      "epoch": 43.51,
      "learning_rate": 0.00012655744844129082,
      "loss": 1.0894,
      "step": 216400
    },
    {
      "epoch": 43.53,
      "learning_rate": 0.0001257891558371864,
      "loss": 1.088,
      "step": 216500
    },
    {
      "epoch": 43.55,
      "learning_rate": 0.00012502307880319613,
      "loss": 1.0907,
      "step": 216600
    },
    {
      "epoch": 43.57,
      "learning_rate": 0.00012426684646913629,
      "loss": 1.0884,
      "step": 216700
    },
    {
      "epoch": 43.59,
      "learning_rate": 0.0001235051829070172,
      "loss": 1.0873,
      "step": 216800
    },
    {
      "epoch": 43.61,
      "learning_rate": 0.00012274573941621584,
      "loss": 1.0884,
      "step": 216900
    },
    {
      "epoch": 43.63,
      "learning_rate": 0.00012198851749342654,
      "loss": 1.0917,
      "step": 217000
    },
    {
      "epoch": 43.65,
      "learning_rate": 0.00012123351863096627,
      "loss": 1.0934,
      "step": 217100
    },
    {
      "epoch": 43.67,
      "learning_rate": 0.0001204807443167695,
      "loss": 1.0904,
      "step": 217200
    },
    {
      "epoch": 43.69,
      "learning_rate": 0.00011973019603438773,
      "loss": 1.0906,
      "step": 217300
    },
    {
      "epoch": 43.71,
      "learning_rate": 0.00011898187526298495,
      "loss": 1.0889,
      "step": 217400
    },
    {
      "epoch": 43.73,
      "learning_rate": 0.00011823578347733579,
      "loss": 1.09,
      "step": 217500
    },
    {
      "epoch": 43.75,
      "learning_rate": 0.00011749192214782101,
      "loss": 1.0912,
      "step": 217600
    },
    {
      "epoch": 43.77,
      "learning_rate": 0.00011675029274042654,
      "loss": 1.0905,
      "step": 217700
    },
    {
      "epoch": 43.79,
      "learning_rate": 0.0001160108967167392,
      "loss": 1.0874,
      "step": 217800
    },
    {
      "epoch": 43.81,
      "learning_rate": 0.0001152737355339449,
      "loss": 1.0902,
      "step": 217900
    },
    {
      "epoch": 43.83,
      "learning_rate": 0.00011453881064482418,
      "loss": 1.088,
      "step": 218000
    },
    {
      "epoch": 43.85,
      "learning_rate": 0.0001138061234977511,
      "loss": 1.0878,
      "step": 218100
    },
    {
      "epoch": 43.87,
      "learning_rate": 0.00011307567553668963,
      "loss": 1.0927,
      "step": 218200
    },
    {
      "epoch": 43.89,
      "learning_rate": 0.00011234746820119087,
      "loss": 1.0882,
      "step": 218300
    },
    {
      "epoch": 43.91,
      "learning_rate": 0.00011162150292638924,
      "loss": 1.0927,
      "step": 218400
    },
    {
      "epoch": 43.94,
      "learning_rate": 0.00011089778114300134,
      "loss": 1.0885,
      "step": 218500
    },
    {
      "epoch": 43.96,
      "learning_rate": 0.00011017630427732253,
      "loss": 1.0884,
      "step": 218600
    },
    {
      "epoch": 43.98,
      "learning_rate": 0.00010945707375122376,
      "loss": 1.0878,
      "step": 218700
    },
    {
      "epoch": 44.0,
      "learning_rate": 0.0001087472496787982,
      "loss": 1.0855,
      "step": 218800
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.4214934698002123,
      "eval_loss": 1.0645169019699097,
      "eval_runtime": 19.7931,
      "eval_samples_per_second": 4019.996,
      "eval_steps_per_second": 15.713,
      "step": 218823
    },
    {
      "epoch": 44.02,
      "learning_rate": 0.00010803249358108017,
      "loss": 1.0767,
      "step": 218900
    },
    {
      "epoch": 44.04,
      "learning_rate": 0.00010731998804791671,
      "loss": 1.0805,
      "step": 219000
    },
    {
      "epoch": 44.06,
      "learning_rate": 0.00010660973448349847,
      "loss": 1.0767,
      "step": 219100
    },
    {
      "epoch": 44.08,
      "learning_rate": 0.00010590173428757774,
      "loss": 1.0796,
      "step": 219200
    },
    {
      "epoch": 44.1,
      "learning_rate": 0.00010519598885546585,
      "loss": 1.0782,
      "step": 219300
    },
    {
      "epoch": 44.12,
      "learning_rate": 0.00010449249957803011,
      "loss": 1.0765,
      "step": 219400
    },
    {
      "epoch": 44.14,
      "learning_rate": 0.00010379126784169191,
      "loss": 1.0828,
      "step": 219500
    },
    {
      "epoch": 44.16,
      "learning_rate": 0.00010309229502842355,
      "loss": 1.077,
      "step": 219600
    },
    {
      "epoch": 44.18,
      "learning_rate": 0.00010239558251574535,
      "loss": 1.0754,
      "step": 219700
    },
    {
      "epoch": 44.2,
      "learning_rate": 0.00010170113167672274,
      "loss": 1.0823,
      "step": 219800
    },
    {
      "epoch": 44.22,
      "learning_rate": 0.00010100894387996454,
      "loss": 1.082,
      "step": 219900
    },
    {
      "epoch": 44.24,
      "learning_rate": 0.00010031902048961913,
      "loss": 1.0795,
      "step": 220000
    },
    {
      "epoch": 44.26,
      "learning_rate": 9.963136286537278e-05,
      "loss": 1.0816,
      "step": 220100
    },
    {
      "epoch": 44.28,
      "learning_rate": 9.894597236244558e-05,
      "loss": 1.079,
      "step": 220200
    },
    {
      "epoch": 44.3,
      "learning_rate": 9.826285033159035e-05,
      "loss": 1.0816,
      "step": 220300
    },
    {
      "epoch": 44.32,
      "learning_rate": 9.758199811908924e-05,
      "loss": 1.0845,
      "step": 220400
    },
    {
      "epoch": 44.34,
      "learning_rate": 9.690341706675043e-05,
      "loss": 1.0838,
      "step": 220500
    },
    {
      "epoch": 44.36,
      "learning_rate": 9.622710851190694e-05,
      "loss": 1.0781,
      "step": 220600
    },
    {
      "epoch": 44.38,
      "learning_rate": 9.555307378741259e-05,
      "loss": 1.0841,
      "step": 220700
    },
    {
      "epoch": 44.4,
      "learning_rate": 9.488802055091186e-05,
      "loss": 1.0813,
      "step": 220800
    },
    {
      "epoch": 44.42,
      "learning_rate": 9.421851469638642e-05,
      "loss": 1.0834,
      "step": 220900
    },
    {
      "epoch": 44.44,
      "learning_rate": 9.355128663070311e-05,
      "loss": 1.0805,
      "step": 221000
    },
    {
      "epoch": 44.46,
      "learning_rate": 9.288633766882021e-05,
      "loss": 1.0837,
      "step": 221100
    },
    {
      "epoch": 44.48,
      "learning_rate": 9.222366912120445e-05,
      "loss": 1.079,
      "step": 221200
    },
    {
      "epoch": 44.5,
      "learning_rate": 9.156328229382879e-05,
      "loss": 1.082,
      "step": 221300
    },
    {
      "epoch": 44.52,
      "learning_rate": 9.090517848816912e-05,
      "loss": 1.0806,
      "step": 221400
    },
    {
      "epoch": 44.54,
      "learning_rate": 9.024935900120185e-05,
      "loss": 1.082,
      "step": 221500
    },
    {
      "epoch": 44.56,
      "learning_rate": 8.95958251254017e-05,
      "loss": 1.0818,
      "step": 221600
    },
    {
      "epoch": 44.58,
      "learning_rate": 8.894457814873885e-05,
      "loss": 1.0804,
      "step": 221700
    },
    {
      "epoch": 44.6,
      "learning_rate": 8.829561935467689e-05,
      "loss": 1.0799,
      "step": 221800
    },
    {
      "epoch": 44.62,
      "learning_rate": 8.76489500221693e-05,
      "loss": 1.0803,
      "step": 221900
    },
    {
      "epoch": 44.64,
      "learning_rate": 8.700457142565774e-05,
      "loss": 1.0811,
      "step": 222000
    },
    {
      "epoch": 44.66,
      "learning_rate": 8.63624848350697e-05,
      "loss": 1.0835,
      "step": 222100
    },
    {
      "epoch": 44.68,
      "learning_rate": 8.57226915158156e-05,
      "loss": 1.0782,
      "step": 222200
    },
    {
      "epoch": 44.7,
      "learning_rate": 8.508519272878545e-05,
      "loss": 1.0845,
      "step": 222300
    },
    {
      "epoch": 44.72,
      "learning_rate": 8.444998973034829e-05,
      "loss": 1.0814,
      "step": 222400
    },
    {
      "epoch": 44.74,
      "learning_rate": 8.381708377234828e-05,
      "loss": 1.0828,
      "step": 222500
    },
    {
      "epoch": 44.76,
      "learning_rate": 8.318647610210284e-05,
      "loss": 1.0845,
      "step": 222600
    },
    {
      "epoch": 44.78,
      "learning_rate": 8.25581679623992e-05,
      "loss": 1.0805,
      "step": 222700
    },
    {
      "epoch": 44.8,
      "learning_rate": 8.193216059149355e-05,
      "loss": 1.0768,
      "step": 222800
    },
    {
      "epoch": 44.82,
      "learning_rate": 8.131468087783922e-05,
      "loss": 1.0814,
      "step": 222900
    },
    {
      "epoch": 44.84,
      "learning_rate": 8.06932557027712e-05,
      "loss": 1.0778,
      "step": 223000
    },
    {
      "epoch": 44.86,
      "learning_rate": 8.007413497182939e-05,
      "loss": 1.083,
      "step": 223100
    },
    {
      "epoch": 44.88,
      "learning_rate": 7.94573199051636e-05,
      "loss": 1.075,
      "step": 223200
    },
    {
      "epoch": 44.9,
      "learning_rate": 7.884281171837912e-05,
      "loss": 1.0767,
      "step": 223300
    },
    {
      "epoch": 44.92,
      "learning_rate": 7.82306116225355e-05,
      "loss": 1.0833,
      "step": 223400
    },
    {
      "epoch": 44.94,
      "learning_rate": 7.762072082414307e-05,
      "loss": 1.0793,
      "step": 223500
    },
    {
      "epoch": 44.96,
      "learning_rate": 7.701314052516146e-05,
      "loss": 1.0785,
      "step": 223600
    },
    {
      "epoch": 44.98,
      "learning_rate": 7.640787192299645e-05,
      "loss": 1.0801,
      "step": 223700
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.4221078634754255,
      "eval_loss": 1.0577867031097412,
      "eval_runtime": 19.8376,
      "eval_samples_per_second": 4010.968,
      "eval_steps_per_second": 15.677,
      "step": 223796
    },
    {
      "epoch": 45.0,
      "learning_rate": 7.580491621049817e-05,
      "loss": 1.0812,
      "step": 223800
    },
    {
      "epoch": 45.02,
      "learning_rate": 7.52042745759586e-05,
      "loss": 1.0668,
      "step": 223900
    },
    {
      "epoch": 45.04,
      "learning_rate": 7.460594820310931e-05,
      "loss": 1.0678,
      "step": 224000
    },
    {
      "epoch": 45.06,
      "learning_rate": 7.400993827111833e-05,
      "loss": 1.0686,
      "step": 224100
    },
    {
      "epoch": 45.08,
      "learning_rate": 7.341624595458923e-05,
      "loss": 1.0737,
      "step": 224200
    },
    {
      "epoch": 45.1,
      "learning_rate": 7.282487242355806e-05,
      "loss": 1.0732,
      "step": 224300
    },
    {
      "epoch": 45.12,
      "learning_rate": 7.223581884349059e-05,
      "loss": 1.0704,
      "step": 224400
    },
    {
      "epoch": 45.14,
      "learning_rate": 7.16490863752807e-05,
      "loss": 1.0695,
      "step": 224500
    },
    {
      "epoch": 45.16,
      "learning_rate": 7.106467617524787e-05,
      "loss": 1.0738,
      "step": 224600
    },
    {
      "epoch": 45.18,
      "learning_rate": 7.048258939513497e-05,
      "loss": 1.07,
      "step": 224700
    },
    {
      "epoch": 45.2,
      "learning_rate": 6.990282718210616e-05,
      "loss": 1.0724,
      "step": 224800
    },
    {
      "epoch": 45.22,
      "learning_rate": 6.93311535277738e-05,
      "loss": 1.071,
      "step": 224900
    },
    {
      "epoch": 45.24,
      "learning_rate": 6.875602059798258e-05,
      "loss": 1.0762,
      "step": 225000
    },
    {
      "epoch": 45.26,
      "learning_rate": 6.818321563795887e-05,
      "loss": 1.074,
      "step": 225100
    },
    {
      "epoch": 45.28,
      "learning_rate": 6.761273977657439e-05,
      "loss": 1.069,
      "step": 225200
    },
    {
      "epoch": 45.3,
      "learning_rate": 6.704459413811065e-05,
      "loss": 1.0751,
      "step": 225300
    },
    {
      "epoch": 45.32,
      "learning_rate": 6.647877984225656e-05,
      "loss": 1.0755,
      "step": 225400
    },
    {
      "epoch": 45.34,
      "learning_rate": 6.59152980041064e-05,
      "loss": 1.0733,
      "step": 225500
    },
    {
      "epoch": 45.36,
      "learning_rate": 6.535414973415796e-05,
      "loss": 1.0741,
      "step": 225600
    },
    {
      "epoch": 45.38,
      "learning_rate": 6.479533613830982e-05,
      "loss": 1.0741,
      "step": 225700
    },
    {
      "epoch": 45.4,
      "learning_rate": 6.423885831785992e-05,
      "loss": 1.076,
      "step": 225800
    },
    {
      "epoch": 45.42,
      "learning_rate": 6.368471736950224e-05,
      "loss": 1.073,
      "step": 225900
    },
    {
      "epoch": 45.44,
      "learning_rate": 6.313291438532556e-05,
      "loss": 1.075,
      "step": 226000
    },
    {
      "epoch": 45.46,
      "learning_rate": 6.258345045281138e-05,
      "loss": 1.0719,
      "step": 226100
    },
    {
      "epoch": 45.48,
      "learning_rate": 6.203632665483128e-05,
      "loss": 1.0774,
      "step": 226200
    },
    {
      "epoch": 45.5,
      "learning_rate": 6.149154406964438e-05,
      "loss": 1.0707,
      "step": 226300
    },
    {
      "epoch": 45.52,
      "learning_rate": 6.09491037708966e-05,
      "loss": 1.0762,
      "step": 226400
    },
    {
      "epoch": 45.54,
      "learning_rate": 6.040900682761738e-05,
      "loss": 1.07,
      "step": 226500
    },
    {
      "epoch": 45.56,
      "learning_rate": 5.987125430421819e-05,
      "loss": 1.0754,
      "step": 226600
    },
    {
      "epoch": 45.58,
      "learning_rate": 5.9335847260489354e-05,
      "loss": 1.0736,
      "step": 226700
    },
    {
      "epoch": 45.6,
      "learning_rate": 5.8802786751599724e-05,
      "loss": 1.0737,
      "step": 226800
    },
    {
      "epoch": 45.62,
      "learning_rate": 5.827207382809313e-05,
      "loss": 1.0699,
      "step": 226900
    },
    {
      "epoch": 45.64,
      "learning_rate": 5.774370953588723e-05,
      "loss": 1.0726,
      "step": 227000
    },
    {
      "epoch": 45.66,
      "learning_rate": 5.72176949162706e-05,
      "loss": 1.0727,
      "step": 227100
    },
    {
      "epoch": 45.68,
      "learning_rate": 5.669403100590123e-05,
      "loss": 1.0772,
      "step": 227200
    },
    {
      "epoch": 45.7,
      "learning_rate": 5.617792031399907e-05,
      "loss": 1.0748,
      "step": 227300
    },
    {
      "epoch": 45.72,
      "learning_rate": 5.565893738080891e-05,
      "loss": 1.0718,
      "step": 227400
    },
    {
      "epoch": 45.74,
      "learning_rate": 5.5142308228831315e-05,
      "loss": 1.0739,
      "step": 227500
    },
    {
      "epoch": 45.76,
      "learning_rate": 5.4628033876227974e-05,
      "loss": 1.075,
      "step": 227600
    },
    {
      "epoch": 45.78,
      "learning_rate": 5.411611533651911e-05,
      "loss": 1.0738,
      "step": 227700
    },
    {
      "epoch": 45.81,
      "learning_rate": 5.3606553618582714e-05,
      "loss": 1.0752,
      "step": 227800
    },
    {
      "epoch": 45.83,
      "learning_rate": 5.309934972665201e-05,
      "loss": 1.0727,
      "step": 227900
    },
    {
      "epoch": 45.85,
      "learning_rate": 5.259450466031324e-05,
      "loss": 1.071,
      "step": 228000
    },
    {
      "epoch": 45.87,
      "learning_rate": 5.2092019414504e-05,
      "loss": 1.074,
      "step": 228100
    },
    {
      "epoch": 45.89,
      "learning_rate": 5.1591894979511055e-05,
      "loss": 1.0699,
      "step": 228200
    },
    {
      "epoch": 45.91,
      "learning_rate": 5.109413234096888e-05,
      "loss": 1.0727,
      "step": 228300
    },
    {
      "epoch": 45.93,
      "learning_rate": 5.059873247985722e-05,
      "loss": 1.0721,
      "step": 228400
    },
    {
      "epoch": 45.95,
      "learning_rate": 5.010569637249912e-05,
      "loss": 1.0708,
      "step": 228500
    },
    {
      "epoch": 45.97,
      "learning_rate": 4.961502499055928e-05,
      "loss": 1.0701,
      "step": 228600
    },
    {
      "epoch": 45.99,
      "learning_rate": 4.912671930104237e-05,
      "loss": 1.072,
      "step": 228700
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.42260501009247403,
      "eval_loss": 1.052234172821045,
      "eval_runtime": 19.8604,
      "eval_samples_per_second": 4006.368,
      "eval_steps_per_second": 15.659,
      "step": 228769
    },
    {
      "epoch": 46.01,
      "learning_rate": 4.864078026629054e-05,
      "loss": 1.0673,
      "step": 228800
    },
    {
      "epoch": 46.03,
      "learning_rate": 4.8157208843981476e-05,
      "loss": 1.0645,
      "step": 228900
    },
    {
      "epoch": 46.05,
      "learning_rate": 4.767600598712743e-05,
      "loss": 1.0635,
      "step": 229000
    },
    {
      "epoch": 46.07,
      "learning_rate": 4.719717264407245e-05,
      "loss": 1.0651,
      "step": 229100
    },
    {
      "epoch": 46.09,
      "learning_rate": 4.672070975849069e-05,
      "loss": 1.0623,
      "step": 229200
    },
    {
      "epoch": 46.11,
      "learning_rate": 4.625134744279142e-05,
      "loss": 1.0643,
      "step": 229300
    },
    {
      "epoch": 46.13,
      "learning_rate": 4.5779604556573094e-05,
      "loss": 1.0661,
      "step": 229400
    },
    {
      "epoch": 46.15,
      "learning_rate": 4.5310234921539935e-05,
      "loss": 1.0668,
      "step": 229500
    },
    {
      "epoch": 46.17,
      "learning_rate": 4.4843239462715455e-05,
      "loss": 1.0682,
      "step": 229600
    },
    {
      "epoch": 46.19,
      "learning_rate": 4.437861910044444e-05,
      "loss": 1.0681,
      "step": 229700
    },
    {
      "epoch": 46.21,
      "learning_rate": 4.3916374750390256e-05,
      "loss": 1.0666,
      "step": 229800
    },
    {
      "epoch": 46.23,
      "learning_rate": 4.345650732353393e-05,
      "loss": 1.0656,
      "step": 229900
    },
    {
      "epoch": 46.25,
      "learning_rate": 4.299901772617215e-05,
      "loss": 1.0653,
      "step": 230000
    },
    {
      "epoch": 46.27,
      "learning_rate": 4.2543906859915384e-05,
      "loss": 1.0678,
      "step": 230100
    },
    {
      "epoch": 46.29,
      "learning_rate": 4.209117562168643e-05,
      "loss": 1.064,
      "step": 230200
    },
    {
      "epoch": 46.31,
      "learning_rate": 4.1640824903717566e-05,
      "loss": 1.071,
      "step": 230300
    },
    {
      "epoch": 46.33,
      "learning_rate": 4.119285559355049e-05,
      "loss": 1.0661,
      "step": 230400
    },
    {
      "epoch": 46.35,
      "learning_rate": 4.0747268574033294e-05,
      "loss": 1.0667,
      "step": 230500
    },
    {
      "epoch": 46.37,
      "learning_rate": 4.0304064723319104e-05,
      "loss": 1.0685,
      "step": 230600
    },
    {
      "epoch": 46.39,
      "learning_rate": 3.986324491486421e-05,
      "loss": 1.0662,
      "step": 230700
    },
    {
      "epoch": 46.41,
      "learning_rate": 3.942481001742673e-05,
      "loss": 1.0661,
      "step": 230800
    },
    {
      "epoch": 46.43,
      "learning_rate": 3.8988760895064675e-05,
      "loss": 1.0645,
      "step": 230900
    },
    {
      "epoch": 46.45,
      "learning_rate": 3.8555098407134085e-05,
      "loss": 1.0628,
      "step": 231000
    },
    {
      "epoch": 46.47,
      "learning_rate": 3.8123823408287294e-05,
      "loss": 1.0712,
      "step": 231100
    },
    {
      "epoch": 46.49,
      "learning_rate": 3.7694936748471633e-05,
      "loss": 1.0657,
      "step": 231200
    },
    {
      "epoch": 46.51,
      "learning_rate": 3.726843927292778e-05,
      "loss": 1.0666,
      "step": 231300
    },
    {
      "epoch": 46.53,
      "learning_rate": 3.6848561063323876e-05,
      "loss": 1.0681,
      "step": 231400
    },
    {
      "epoch": 46.55,
      "learning_rate": 3.6426820560480634e-05,
      "loss": 1.0666,
      "step": 231500
    },
    {
      "epoch": 46.57,
      "learning_rate": 3.600747174108493e-05,
      "loss": 1.0647,
      "step": 231600
    },
    {
      "epoch": 46.59,
      "learning_rate": 3.5590515431579846e-05,
      "loss": 1.0647,
      "step": 231700
    },
    {
      "epoch": 46.61,
      "learning_rate": 3.51759524536939e-05,
      "loss": 1.0643,
      "step": 231800
    },
    {
      "epoch": 46.63,
      "learning_rate": 3.476378362443869e-05,
      "loss": 1.0623,
      "step": 231900
    },
    {
      "epoch": 46.65,
      "learning_rate": 3.435400975610778e-05,
      "loss": 1.0654,
      "step": 232000
    },
    {
      "epoch": 46.67,
      "learning_rate": 3.394663165627407e-05,
      "loss": 1.0645,
      "step": 232100
    },
    {
      "epoch": 46.69,
      "learning_rate": 3.3541650127789646e-05,
      "loss": 1.0678,
      "step": 232200
    },
    {
      "epoch": 46.71,
      "learning_rate": 3.31390659687833e-05,
      "loss": 1.0657,
      "step": 232300
    },
    {
      "epoch": 46.73,
      "learning_rate": 3.2738879972659116e-05,
      "loss": 1.068,
      "step": 232400
    },
    {
      "epoch": 46.75,
      "learning_rate": 3.234109292809456e-05,
      "loss": 1.0686,
      "step": 232500
    },
    {
      "epoch": 46.77,
      "learning_rate": 3.194570561904003e-05,
      "loss": 1.0659,
      "step": 232600
    },
    {
      "epoch": 46.79,
      "learning_rate": 3.1552718824715834e-05,
      "loss": 1.0674,
      "step": 232700
    },
    {
      "epoch": 46.81,
      "learning_rate": 3.116213331961215e-05,
      "loss": 1.0668,
      "step": 232800
    },
    {
      "epoch": 46.83,
      "learning_rate": 3.077394987348589e-05,
      "loss": 1.0674,
      "step": 232900
    },
    {
      "epoch": 46.85,
      "learning_rate": 3.0388169251360788e-05,
      "loss": 1.0628,
      "step": 233000
    },
    {
      "epoch": 46.87,
      "learning_rate": 3.0004792213524645e-05,
      "loss": 1.0639,
      "step": 233100
    },
    {
      "epoch": 46.89,
      "learning_rate": 2.9623819515528783e-05,
      "loss": 1.0649,
      "step": 233200
    },
    {
      "epoch": 46.91,
      "learning_rate": 2.9245251908185526e-05,
      "loss": 1.0651,
      "step": 233300
    },
    {
      "epoch": 46.93,
      "learning_rate": 2.8872839843944315e-05,
      "loss": 1.067,
      "step": 233400
    },
    {
      "epoch": 46.95,
      "learning_rate": 2.8499060581948928e-05,
      "loss": 1.0659,
      "step": 233500
    },
    {
      "epoch": 46.97,
      "learning_rate": 2.812768862725676e-05,
      "loss": 1.0669,
      "step": 233600
    },
    {
      "epoch": 46.99,
      "learning_rate": 2.7758724711759303e-05,
      "loss": 1.0625,
      "step": 233700
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.42298172791300875,
      "eval_loss": 1.0480923652648926,
      "eval_runtime": 19.887,
      "eval_samples_per_second": 4001.005,
      "eval_steps_per_second": 15.638,
      "step": 233742
    },
    {
      "epoch": 47.01,
      "learning_rate": 2.7392169562602682e-05,
      "loss": 1.0617,
      "step": 233800
    },
    {
      "epoch": 47.03,
      "learning_rate": 2.7028023902185562e-05,
      "loss": 1.0595,
      "step": 233900
    },
    {
      "epoch": 47.05,
      "learning_rate": 2.6666288448158464e-05,
      "loss": 1.0587,
      "step": 234000
    },
    {
      "epoch": 47.07,
      "learning_rate": 2.6306963913421827e-05,
      "loss": 1.0602,
      "step": 234100
    },
    {
      "epoch": 47.09,
      "learning_rate": 2.5950051006124048e-05,
      "loss": 1.0601,
      "step": 234200
    },
    {
      "epoch": 47.11,
      "learning_rate": 2.5595550429661775e-05,
      "loss": 1.0611,
      "step": 234300
    },
    {
      "epoch": 47.13,
      "learning_rate": 2.5243462882676703e-05,
      "loss": 1.0591,
      "step": 234400
    },
    {
      "epoch": 47.15,
      "learning_rate": 2.4893789059055454e-05,
      "loss": 1.0635,
      "step": 234500
    },
    {
      "epoch": 47.17,
      "learning_rate": 2.4546529647927335e-05,
      "loss": 1.059,
      "step": 234600
    },
    {
      "epoch": 47.19,
      "learning_rate": 2.4201685333663654e-05,
      "loss": 1.0587,
      "step": 234700
    },
    {
      "epoch": 47.21,
      "learning_rate": 2.3859256795876057e-05,
      "loss": 1.0579,
      "step": 234800
    },
    {
      "epoch": 47.23,
      "learning_rate": 2.351924470941541e-05,
      "loss": 1.0601,
      "step": 234900
    },
    {
      "epoch": 47.25,
      "learning_rate": 2.318164974436987e-05,
      "loss": 1.06,
      "step": 235000
    },
    {
      "epoch": 47.27,
      "learning_rate": 2.2846472566064037e-05,
      "loss": 1.0604,
      "step": 235100
    },
    {
      "epoch": 47.29,
      "learning_rate": 2.2513713835058124e-05,
      "loss": 1.058,
      "step": 235200
    },
    {
      "epoch": 47.31,
      "learning_rate": 2.2183374207145472e-05,
      "loss": 1.0605,
      "step": 235300
    },
    {
      "epoch": 47.33,
      "learning_rate": 2.1858721552181353e-05,
      "loss": 1.0602,
      "step": 235400
    },
    {
      "epoch": 47.35,
      "learning_rate": 2.153319787157798e-05,
      "loss": 1.0587,
      "step": 235500
    },
    {
      "epoch": 47.37,
      "learning_rate": 2.1210095226447403e-05,
      "loss": 1.0582,
      "step": 235600
    },
    {
      "epoch": 47.39,
      "learning_rate": 2.0889414253553036e-05,
      "loss": 1.0639,
      "step": 235700
    },
    {
      "epoch": 47.41,
      "learning_rate": 2.057115558488601e-05,
      "loss": 1.0615,
      "step": 235800
    },
    {
      "epoch": 47.43,
      "learning_rate": 2.0255319847663906e-05,
      "loss": 1.0598,
      "step": 235900
    },
    {
      "epoch": 47.45,
      "learning_rate": 1.9941907664328407e-05,
      "loss": 1.0658,
      "step": 236000
    },
    {
      "epoch": 47.47,
      "learning_rate": 1.9630919652545432e-05,
      "loss": 1.0611,
      "step": 236100
    },
    {
      "epoch": 47.49,
      "learning_rate": 1.9322356425203607e-05,
      "loss": 1.0624,
      "step": 236200
    },
    {
      "epoch": 47.51,
      "learning_rate": 1.901621859041247e-05,
      "loss": 1.0621,
      "step": 236300
    },
    {
      "epoch": 47.53,
      "learning_rate": 1.8712506751501767e-05,
      "loss": 1.0599,
      "step": 236400
    },
    {
      "epoch": 47.55,
      "learning_rate": 1.841122150702007e-05,
      "loss": 1.0564,
      "step": 236500
    },
    {
      "epoch": 47.57,
      "learning_rate": 1.8112363450733667e-05,
      "loss": 1.0621,
      "step": 236600
    },
    {
      "epoch": 47.59,
      "learning_rate": 1.781593317162572e-05,
      "loss": 1.0629,
      "step": 236700
    },
    {
      "epoch": 47.61,
      "learning_rate": 1.7521931253894342e-05,
      "loss": 1.0607,
      "step": 236800
    },
    {
      "epoch": 47.63,
      "learning_rate": 1.7230358276952156e-05,
      "loss": 1.0589,
      "step": 236900
    },
    {
      "epoch": 47.65,
      "learning_rate": 1.6941214815424793e-05,
      "loss": 1.0609,
      "step": 237000
    },
    {
      "epoch": 47.68,
      "learning_rate": 1.66545014391499e-05,
      "loss": 1.0602,
      "step": 237100
    },
    {
      "epoch": 47.7,
      "learning_rate": 1.6370218713175762e-05,
      "loss": 1.066,
      "step": 237200
    },
    {
      "epoch": 47.72,
      "learning_rate": 1.6088367197760607e-05,
      "loss": 1.0584,
      "step": 237300
    },
    {
      "epoch": 47.74,
      "learning_rate": 1.5808947448371218e-05,
      "loss": 1.0585,
      "step": 237400
    },
    {
      "epoch": 47.76,
      "learning_rate": 1.553471784824498e-05,
      "loss": 1.0647,
      "step": 237500
    },
    {
      "epoch": 47.78,
      "learning_rate": 1.5260138946824596e-05,
      "loss": 1.0597,
      "step": 237600
    },
    {
      "epoch": 47.8,
      "learning_rate": 1.4987993443684039e-05,
      "loss": 1.0618,
      "step": 237700
    },
    {
      "epoch": 47.82,
      "learning_rate": 1.4718281875161916e-05,
      "loss": 1.0587,
      "step": 237800
    },
    {
      "epoch": 47.84,
      "learning_rate": 1.4451004772799565e-05,
      "loss": 1.0562,
      "step": 237900
    },
    {
      "epoch": 47.86,
      "learning_rate": 1.418616266334133e-05,
      "loss": 1.0588,
      "step": 238000
    },
    {
      "epoch": 47.88,
      "learning_rate": 1.392375606873178e-05,
      "loss": 1.0633,
      "step": 238100
    },
    {
      "epoch": 47.9,
      "learning_rate": 1.3663785506116133e-05,
      "loss": 1.0609,
      "step": 238200
    },
    {
      "epoch": 47.92,
      "learning_rate": 1.3406251487839e-05,
      "loss": 1.0614,
      "step": 238300
    },
    {
      "epoch": 47.94,
      "learning_rate": 1.3151154521442582e-05,
      "loss": 1.0631,
      "step": 238400
    },
    {
      "epoch": 47.96,
      "learning_rate": 1.2898495109666397e-05,
      "loss": 1.06,
      "step": 238500
    },
    {
      "epoch": 47.98,
      "learning_rate": 1.2648273750446026e-05,
      "loss": 1.0606,
      "step": 238600
    },
    {
      "epoch": 48.0,
      "learning_rate": 1.2400490936912284e-05,
      "loss": 1.0639,
      "step": 238700
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.4231766093868917,
      "eval_loss": 1.045819640159607,
      "eval_runtime": 19.861,
      "eval_samples_per_second": 4006.252,
      "eval_steps_per_second": 15.659,
      "step": 238716
    },
    {
      "epoch": 48.02,
      "learning_rate": 1.2155147157390245e-05,
      "loss": 1.0555,
      "step": 238800
    },
    {
      "epoch": 48.04,
      "learning_rate": 1.1912242895397857e-05,
      "loss": 1.057,
      "step": 238900
    },
    {
      "epoch": 48.06,
      "learning_rate": 1.1671778629645525e-05,
      "loss": 1.0555,
      "step": 239000
    },
    {
      "epoch": 48.08,
      "learning_rate": 1.1433754834035137e-05,
      "loss": 1.0566,
      "step": 239100
    },
    {
      "epoch": 48.1,
      "learning_rate": 1.1198171977658822e-05,
      "loss": 1.0548,
      "step": 239200
    },
    {
      "epoch": 48.12,
      "learning_rate": 1.096503052479783e-05,
      "loss": 1.0572,
      "step": 239300
    },
    {
      "epoch": 48.14,
      "learning_rate": 1.0734330934922404e-05,
      "loss": 1.0535,
      "step": 239400
    },
    {
      "epoch": 48.16,
      "learning_rate": 1.0508344144459226e-05,
      "loss": 1.0572,
      "step": 239500
    },
    {
      "epoch": 48.18,
      "learning_rate": 1.0282505209828901e-05,
      "loss": 1.0567,
      "step": 239600
    },
    {
      "epoch": 48.2,
      "learning_rate": 1.0059109483290113e-05,
      "loss": 1.061,
      "step": 239700
    },
    {
      "epoch": 48.22,
      "learning_rate": 9.838157405106102e-06,
      "loss": 1.0535,
      "step": 239800
    },
    {
      "epoch": 48.24,
      "learning_rate": 9.619649410724658e-06,
      "loss": 1.0575,
      "step": 239900
    },
    {
      "epoch": 48.26,
      "learning_rate": 9.40358593077631e-06,
      "loss": 1.0578,
      "step": 240000
    },
    {
      "epoch": 48.28,
      "learning_rate": 9.189967391074332e-06,
      "loss": 1.0559,
      "step": 240100
    },
    {
      "epoch": 48.3,
      "learning_rate": 8.978794212613355e-06,
      "loss": 1.0572,
      "step": 240200
    },
    {
      "epoch": 48.32,
      "learning_rate": 8.770066811569083e-06,
      "loss": 1.0544,
      "step": 240300
    },
    {
      "epoch": 48.34,
      "learning_rate": 8.563785599296769e-06,
      "loss": 1.0586,
      "step": 240400
    },
    {
      "epoch": 48.36,
      "learning_rate": 8.359950982330805e-06,
      "loss": 1.0592,
      "step": 240500
    },
    {
      "epoch": 48.38,
      "learning_rate": 8.158563362384158e-06,
      "loss": 1.0586,
      "step": 240600
    },
    {
      "epoch": 48.4,
      "learning_rate": 7.959623136347128e-06,
      "loss": 1.0547,
      "step": 240700
    },
    {
      "epoch": 48.42,
      "learning_rate": 7.763130696286647e-06,
      "loss": 1.0525,
      "step": 240800
    },
    {
      "epoch": 48.44,
      "learning_rate": 7.56908642944587e-06,
      "loss": 1.0569,
      "step": 240900
    },
    {
      "epoch": 48.46,
      "learning_rate": 7.3774907182427856e-06,
      "loss": 1.055,
      "step": 241000
    },
    {
      "epoch": 48.48,
      "learning_rate": 7.18834394027007e-06,
      "loss": 1.0579,
      "step": 241100
    },
    {
      "epoch": 48.5,
      "learning_rate": 7.001646468294265e-06,
      "loss": 1.0554,
      "step": 241200
    },
    {
      "epoch": 48.52,
      "learning_rate": 6.817398670254382e-06,
      "loss": 1.0551,
      "step": 241300
    },
    {
      "epoch": 48.54,
      "learning_rate": 6.635600909262185e-06,
      "loss": 1.0599,
      "step": 241400
    },
    {
      "epoch": 48.56,
      "learning_rate": 6.456253543600521e-06,
      "loss": 1.0554,
      "step": 241500
    },
    {
      "epoch": 48.58,
      "learning_rate": 6.281113760537027e-06,
      "loss": 1.0548,
      "step": 241600
    },
    {
      "epoch": 48.6,
      "learning_rate": 6.106643728384375e-06,
      "loss": 1.0615,
      "step": 241700
    },
    {
      "epoch": 48.62,
      "learning_rate": 5.934625134019766e-06,
      "loss": 1.0538,
      "step": 241800
    },
    {
      "epoch": 48.64,
      "learning_rate": 5.76505831645327e-06,
      "loss": 1.0599,
      "step": 241900
    },
    {
      "epoch": 48.66,
      "learning_rate": 5.597943609863821e-06,
      "loss": 1.0539,
      "step": 242000
    },
    {
      "epoch": 48.68,
      "learning_rate": 5.433281343597135e-06,
      "loss": 1.0565,
      "step": 242100
    },
    {
      "epoch": 48.7,
      "learning_rate": 5.2710718421662696e-06,
      "loss": 1.0596,
      "step": 242200
    },
    {
      "epoch": 48.72,
      "learning_rate": 5.111315425249952e-06,
      "loss": 1.0548,
      "step": 242300
    },
    {
      "epoch": 48.74,
      "learning_rate": 4.954012407692721e-06,
      "loss": 1.0547,
      "step": 242400
    },
    {
      "epoch": 48.76,
      "learning_rate": 4.799163099503956e-06,
      "loss": 1.0514,
      "step": 242500
    },
    {
      "epoch": 48.78,
      "learning_rate": 4.646767805857183e-06,
      "loss": 1.0587,
      "step": 242600
    },
    {
      "epoch": 48.8,
      "learning_rate": 4.496826827089795e-06,
      "loss": 1.0602,
      "step": 242700
    },
    {
      "epoch": 48.82,
      "learning_rate": 4.349340458702084e-06,
      "loss": 1.0576,
      "step": 242800
    },
    {
      "epoch": 48.84,
      "learning_rate": 4.204308991357098e-06,
      "loss": 1.0576,
      "step": 242900
    },
    {
      "epoch": 48.86,
      "learning_rate": 4.061732710879673e-06,
      "loss": 1.0592,
      "step": 243000
    },
    {
      "epoch": 48.88,
      "learning_rate": 3.921611898255878e-06,
      "loss": 1.0601,
      "step": 243100
    },
    {
      "epoch": 48.9,
      "learning_rate": 3.783946829632734e-06,
      "loss": 1.0588,
      "step": 243200
    },
    {
      "epoch": 48.92,
      "learning_rate": 3.648737776317801e-06,
      "loss": 1.0558,
      "step": 243300
    },
    {
      "epoch": 48.94,
      "learning_rate": 3.5159850047777885e-06,
      "loss": 1.057,
      "step": 243400
    },
    {
      "epoch": 48.96,
      "learning_rate": 3.3856887766392507e-06,
      "loss": 1.0593,
      "step": 243500
    },
    {
      "epoch": 48.98,
      "learning_rate": 3.259115580974137e-06,
      "loss": 1.0585,
      "step": 243600
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.4232847884907615,
      "eval_loss": 1.044702172279358,
      "eval_runtime": 19.8772,
      "eval_samples_per_second": 4002.975,
      "eval_steps_per_second": 15.646,
      "step": 243689
    },
    {
      "epoch": 49.0,
      "learning_rate": 3.1337086333987908e-06,
      "loss": 1.0534,
      "step": 243700
    },
    {
      "epoch": 49.02,
      "learning_rate": 3.0107589826065816e-06,
      "loss": 1.0535,
      "step": 243800
    },
    {
      "epoch": 49.04,
      "learning_rate": 2.8902668709041013e-06,
      "loss": 1.0535,
      "step": 243900
    },
    {
      "epoch": 49.06,
      "learning_rate": 2.772232535754593e-06,
      "loss": 1.0536,
      "step": 244000
    },
    {
      "epoch": 49.08,
      "learning_rate": 2.6566562097773973e-06,
      "loss": 1.0529,
      "step": 244100
    },
    {
      "epoch": 49.1,
      "learning_rate": 2.5435381207479514e-06,
      "loss": 1.057,
      "step": 244200
    },
    {
      "epoch": 49.12,
      "learning_rate": 2.432878491596957e-06,
      "loss": 1.0544,
      "step": 244300
    },
    {
      "epoch": 49.14,
      "learning_rate": 2.3246775404098252e-06,
      "loss": 1.0568,
      "step": 244400
    },
    {
      "epoch": 49.16,
      "learning_rate": 2.218935480426676e-06,
      "loss": 1.0521,
      "step": 244500
    },
    {
      "epoch": 49.18,
      "learning_rate": 2.1156525200416444e-06,
      "loss": 1.0557,
      "step": 244600
    },
    {
      "epoch": 49.2,
      "learning_rate": 2.0148288628026036e-06,
      "loss": 1.0553,
      "step": 244700
    },
    {
      "epoch": 49.22,
      "learning_rate": 1.9164647074104702e-06,
      "loss": 1.0539,
      "step": 244800
    },
    {
      "epoch": 49.24,
      "learning_rate": 1.8205602477193439e-06,
      "loss": 1.0517,
      "step": 244900
    },
    {
      "epoch": 49.26,
      "learning_rate": 1.7271156727355353e-06,
      "loss": 1.0535,
      "step": 245000
    },
    {
      "epoch": 49.28,
      "learning_rate": 1.6361311666174272e-06,
      "loss": 1.0565,
      "step": 245100
    },
    {
      "epoch": 49.3,
      "learning_rate": 1.5476069086756139e-06,
      "loss": 1.0513,
      "step": 245200
    },
    {
      "epoch": 49.32,
      "learning_rate": 1.4615430733713742e-06,
      "loss": 1.0539,
      "step": 245300
    },
    {
      "epoch": 49.34,
      "learning_rate": 1.3779398303177814e-06,
      "loss": 1.0565,
      "step": 245400
    },
    {
      "epoch": 49.36,
      "learning_rate": 1.296797344278039e-06,
      "loss": 1.0572,
      "step": 245500
    },
    {
      "epoch": 49.38,
      "learning_rate": 1.218115775166173e-06,
      "loss": 1.0538,
      "step": 245600
    },
    {
      "epoch": 49.4,
      "learning_rate": 1.1426453002136294e-06,
      "loss": 1.0577,
      "step": 245700
    },
    {
      "epoch": 49.42,
      "learning_rate": 1.0688614123491314e-06,
      "loss": 1.0547,
      "step": 245800
    },
    {
      "epoch": 49.44,
      "learning_rate": 9.97538890623456e-07,
      "loss": 1.0539,
      "step": 245900
    },
    {
      "epoch": 49.46,
      "learning_rate": 9.286778755976388e-07,
      "loss": 1.0524,
      "step": 246000
    },
    {
      "epoch": 49.48,
      "learning_rate": 8.622785029814562e-07,
      "loss": 1.0569,
      "step": 246100
    },
    {
      "epoch": 49.5,
      "learning_rate": 7.983409036331491e-07,
      "loss": 1.0525,
      "step": 246200
    },
    {
      "epoch": 49.52,
      "learning_rate": 7.368652035597001e-07,
      "loss": 1.0557,
      "step": 246300
    },
    {
      "epoch": 49.55,
      "learning_rate": 6.778515239161398e-07,
      "loss": 1.0568,
      "step": 246400
    },
    {
      "epoch": 49.57,
      "learning_rate": 6.212999810051301e-07,
      "loss": 1.058,
      "step": 246500
    },
    {
      "epoch": 49.59,
      "learning_rate": 5.672106862772419e-07,
      "loss": 1.0585,
      "step": 246600
    },
    {
      "epoch": 49.61,
      "learning_rate": 5.155837463306778e-07,
      "loss": 1.0561,
      "step": 246700
    },
    {
      "epoch": 49.63,
      "learning_rate": 4.664192629104391e-07,
      "loss": 1.057,
      "step": 246800
    },
    {
      "epoch": 49.65,
      "learning_rate": 4.201721622633381e-07,
      "loss": 1.0523,
      "step": 246900
    },
    {
      "epoch": 49.67,
      "learning_rate": 3.7590825082547965e-07,
      "loss": 1.0533,
      "step": 247000
    },
    {
      "epoch": 49.69,
      "learning_rate": 3.3410707118347595e-07,
      "loss": 1.0538,
      "step": 247100
    },
    {
      "epoch": 49.71,
      "learning_rate": 2.951498981840217e-07,
      "loss": 1.0602,
      "step": 247200
    },
    {
      "epoch": 49.73,
      "learning_rate": 2.5824979513769386e-07,
      "loss": 1.0534,
      "step": 247300
    },
    {
      "epoch": 49.75,
      "learning_rate": 2.238126557659037e-07,
      "loss": 1.0558,
      "step": 247400
    },
    {
      "epoch": 49.77,
      "learning_rate": 1.9183854793672352e-07,
      "loss": 1.0555,
      "step": 247500
    },
    {
      "epoch": 49.79,
      "learning_rate": 1.6232753466377536e-07,
      "loss": 1.0578,
      "step": 247600
    },
    {
      "epoch": 49.81,
      "learning_rate": 1.352796741069251e-07,
      "loss": 1.0576,
      "step": 247700
    },
    {
      "epoch": 49.83,
      "learning_rate": 1.1069501957144956e-07,
      "loss": 1.0551,
      "step": 247800
    },
    {
      "epoch": 49.85,
      "learning_rate": 8.857361950831422e-08,
      "loss": 1.0557,
      "step": 247900
    },
    {
      "epoch": 49.87,
      "learning_rate": 6.89155175137568e-08,
      "loss": 1.0538,
      "step": 248000
    },
    {
      "epoch": 49.89,
      "learning_rate": 5.172075232956486e-08,
      "loss": 1.0543,
      "step": 248100
    },
    {
      "epoch": 49.91,
      "learning_rate": 3.698935784279822e-08,
      "loss": 1.056,
      "step": 248200
    },
    {
      "epoch": 49.93,
      "learning_rate": 2.472136308592776e-08,
      "loss": 1.055,
      "step": 248300
    },
    {
      "epoch": 49.95,
      "learning_rate": 1.4916792236141507e-08,
      "loss": 1.0544,
      "step": 248400
    },
    {
      "epoch": 49.97,
      "learning_rate": 7.575664616454869e-09,
      "loss": 1.0545,
      "step": 248500
    },
    {
      "epoch": 49.99,
      "learning_rate": 2.6979946943228584e-09,
      "loss": 1.0521,
      "step": 248600
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.42330467435544344,
      "eval_loss": 1.0445035696029663,
      "eval_runtime": 19.4961,
      "eval_samples_per_second": 4081.225,
      "eval_steps_per_second": 15.952,
      "step": 248650
    },
    {
      "epoch": 50.0,
      "step": 248650,
      "total_flos": 6.912086038408397e+16,
      "train_loss": 1.2392261905236766,
      "train_runtime": 25350.8633,
      "train_samples_per_second": 2510.968,
      "train_steps_per_second": 9.808
    }
  ],
  "max_steps": 248650,
  "num_train_epochs": 50,
  "total_flos": 6.912086038408397e+16,
  "trial_name": null,
  "trial_params": null
}