{
  "best_metric": 0.6816287040710449,
  "best_model_checkpoint": "./checkpoints/llava-v1.6-vicuna-7b/checkpoint-225",
  "epoch": 8.0,
  "eval_steps": 1.0,
  "global_step": 256,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03125,
      "grad_norm": 1.3320099054231718,
      "learning_rate": 0.0,
      "loss": 1.3851,
      "step": 1
    },
    {
      "epoch": 0.03125,
      "eval_loss": 1.3910757303237915,
      "eval_runtime": 63.0135,
      "eval_samples_per_second": 3.174,
      "eval_steps_per_second": 0.397,
      "step": 1
    },
    {
      "epoch": 0.0625,
      "grad_norm": 1.0473401758450829,
      "learning_rate": 8.613531161467863e-06,
      "loss": 1.3255,
      "step": 2
    },
    {
      "epoch": 0.0625,
      "eval_loss": 1.3910757303237915,
      "eval_runtime": 56.9747,
      "eval_samples_per_second": 3.51,
      "eval_steps_per_second": 0.439,
      "step": 2
    },
    {
      "epoch": 0.09375,
      "grad_norm": 1.0429876090069883,
      "learning_rate": 1.3652123889719709e-05,
      "loss": 1.3737,
      "step": 3
    },
    {
      "epoch": 0.09375,
      "eval_loss": 1.3638323545455933,
      "eval_runtime": 56.6988,
      "eval_samples_per_second": 3.527,
      "eval_steps_per_second": 0.441,
      "step": 3
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.9193695742967616,
      "learning_rate": 1.7227062322935725e-05,
      "loss": 1.3309,
      "step": 4
    },
    {
      "epoch": 0.125,
      "eval_loss": 1.3227791786193848,
      "eval_runtime": 56.6188,
      "eval_samples_per_second": 3.532,
      "eval_steps_per_second": 0.442,
      "step": 4
    },
    {
      "epoch": 0.15625,
      "grad_norm": 1.0043594584185398,
      "learning_rate": 2e-05,
      "loss": 1.2984,
      "step": 5
    },
    {
      "epoch": 0.15625,
      "eval_loss": 1.2728056907653809,
      "eval_runtime": 58.8213,
      "eval_samples_per_second": 3.4,
      "eval_steps_per_second": 0.425,
      "step": 5
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.8222566364005439,
      "learning_rate": 2e-05,
      "loss": 1.2639,
      "step": 6
    },
    {
      "epoch": 0.1875,
      "eval_loss": 1.2296103239059448,
      "eval_runtime": 56.6504,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 6
    },
    {
      "epoch": 0.21875,
      "grad_norm": 0.6389176248800544,
      "learning_rate": 2e-05,
      "loss": 1.2314,
      "step": 7
    },
    {
      "epoch": 0.21875,
      "eval_loss": 1.1983529329299927,
      "eval_runtime": 56.5641,
      "eval_samples_per_second": 3.536,
      "eval_steps_per_second": 0.442,
      "step": 7
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.599291017991319,
      "learning_rate": 2e-05,
      "loss": 1.2037,
      "step": 8
    },
    {
      "epoch": 0.25,
      "eval_loss": 1.1734061241149902,
      "eval_runtime": 56.6005,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 0.442,
      "step": 8
    },
    {
      "epoch": 0.28125,
      "grad_norm": 0.4952974010296138,
      "learning_rate": 2e-05,
      "loss": 1.226,
      "step": 9
    },
    {
      "epoch": 0.28125,
      "eval_loss": 1.1502536535263062,
      "eval_runtime": 56.7524,
      "eval_samples_per_second": 3.524,
      "eval_steps_per_second": 0.441,
      "step": 9
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.4967350606769311,
      "learning_rate": 2e-05,
      "loss": 1.1613,
      "step": 10
    },
    {
      "epoch": 0.3125,
      "eval_loss": 1.127350091934204,
      "eval_runtime": 56.7569,
      "eval_samples_per_second": 3.524,
      "eval_steps_per_second": 0.44,
      "step": 10
    },
    {
      "epoch": 0.34375,
      "grad_norm": 0.43644425188108293,
      "learning_rate": 2e-05,
      "loss": 1.2077,
      "step": 11
    },
    {
      "epoch": 0.34375,
      "eval_loss": 1.104610562324524,
      "eval_runtime": 56.607,
      "eval_samples_per_second": 3.533,
      "eval_steps_per_second": 0.442,
      "step": 11
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.4763392566533296,
      "learning_rate": 2e-05,
      "loss": 1.1593,
      "step": 12
    },
    {
      "epoch": 0.375,
      "eval_loss": 1.0827140808105469,
      "eval_runtime": 56.6548,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 12
    },
    {
      "epoch": 0.40625,
      "grad_norm": 0.49138280391100253,
      "learning_rate": 2e-05,
      "loss": 1.1679,
      "step": 13
    },
    {
      "epoch": 0.40625,
      "eval_loss": 1.0621232986450195,
      "eval_runtime": 56.8147,
      "eval_samples_per_second": 3.52,
      "eval_steps_per_second": 0.44,
      "step": 13
    },
    {
      "epoch": 0.4375,
      "grad_norm": 0.4305508696222477,
      "learning_rate": 2e-05,
      "loss": 1.0008,
      "step": 14
    },
    {
      "epoch": 0.4375,
      "eval_loss": 1.0437134504318237,
      "eval_runtime": 56.7306,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 0.441,
      "step": 14
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.39438622708065774,
      "learning_rate": 2e-05,
      "loss": 1.1206,
      "step": 15
    },
    {
      "epoch": 0.46875,
      "eval_loss": 1.0277280807495117,
      "eval_runtime": 56.6499,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 15
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.40300919769454296,
      "learning_rate": 2e-05,
      "loss": 1.0501,
      "step": 16
    },
    {
      "epoch": 0.5,
      "eval_loss": 1.0134528875350952,
      "eval_runtime": 56.3333,
      "eval_samples_per_second": 3.55,
      "eval_steps_per_second": 0.444,
      "step": 16
    },
    {
      "epoch": 0.53125,
      "grad_norm": 0.35230570754831836,
      "learning_rate": 2e-05,
      "loss": 1.0593,
      "step": 17
    },
    {
      "epoch": 0.53125,
      "eval_loss": 1.0004419088363647,
      "eval_runtime": 56.6019,
      "eval_samples_per_second": 3.533,
      "eval_steps_per_second": 0.442,
      "step": 17
    },
    {
      "epoch": 0.5625,
      "grad_norm": 0.37606931260721715,
      "learning_rate": 2e-05,
      "loss": 1.0482,
      "step": 18
    },
    {
      "epoch": 0.5625,
      "eval_loss": 0.9879937767982483,
      "eval_runtime": 56.6945,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 0.441,
      "step": 18
    },
    {
      "epoch": 0.59375,
      "grad_norm": 0.2941404563021841,
      "learning_rate": 2e-05,
      "loss": 0.9707,
      "step": 19
    },
    {
      "epoch": 0.59375,
      "eval_loss": 0.976818859577179,
      "eval_runtime": 56.6805,
      "eval_samples_per_second": 3.529,
      "eval_steps_per_second": 0.441,
      "step": 19
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.2958263397509482,
      "learning_rate": 2e-05,
      "loss": 1.091,
      "step": 20
    },
    {
      "epoch": 0.625,
      "eval_loss": 0.9669834971427917,
      "eval_runtime": 57.6231,
      "eval_samples_per_second": 3.471,
      "eval_steps_per_second": 0.434,
      "step": 20
    },
    {
      "epoch": 0.65625,
      "grad_norm": 0.2485896802049987,
      "learning_rate": 2e-05,
      "loss": 1.0041,
      "step": 21
    },
    {
      "epoch": 0.65625,
      "eval_loss": 0.9583450555801392,
      "eval_runtime": 56.5142,
      "eval_samples_per_second": 3.539,
      "eval_steps_per_second": 0.442,
      "step": 21
    },
    {
      "epoch": 0.6875,
      "grad_norm": 0.296994298254859,
      "learning_rate": 2e-05,
      "loss": 1.055,
      "step": 22
    },
    {
      "epoch": 0.6875,
      "eval_loss": 0.9502925276756287,
      "eval_runtime": 56.6393,
      "eval_samples_per_second": 3.531,
      "eval_steps_per_second": 0.441,
      "step": 22
    },
    {
      "epoch": 0.71875,
      "grad_norm": 0.2499735192340966,
      "learning_rate": 2e-05,
      "loss": 1.04,
      "step": 23
    },
    {
      "epoch": 0.71875,
      "eval_loss": 0.9427899122238159,
      "eval_runtime": 56.5467,
      "eval_samples_per_second": 3.537,
      "eval_steps_per_second": 0.442,
      "step": 23
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.23614468035916372,
      "learning_rate": 2e-05,
      "loss": 1.0387,
      "step": 24
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.9359552264213562,
      "eval_runtime": 56.8371,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 24
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.2597610358499704,
      "learning_rate": 2e-05,
      "loss": 0.9821,
      "step": 25
    },
    {
      "epoch": 0.78125,
      "eval_loss": 0.929139256477356,
      "eval_runtime": 56.659,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 25
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.2483654904520099,
      "learning_rate": 2e-05,
      "loss": 1.0139,
      "step": 26
    },
    {
      "epoch": 0.8125,
      "eval_loss": 0.9226005673408508,
      "eval_runtime": 56.4669,
      "eval_samples_per_second": 3.542,
      "eval_steps_per_second": 0.443,
      "step": 26
    },
    {
      "epoch": 0.84375,
      "grad_norm": 0.2814780741041167,
      "learning_rate": 2e-05,
      "loss": 0.9374,
      "step": 27
    },
    {
      "epoch": 0.84375,
      "eval_loss": 0.9160022735595703,
      "eval_runtime": 56.6558,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 27
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.29993540247195477,
      "learning_rate": 2e-05,
      "loss": 0.948,
      "step": 28
    },
    {
      "epoch": 0.875,
      "eval_loss": 0.9092594981193542,
      "eval_runtime": 56.743,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 0.441,
      "step": 28
    },
    {
      "epoch": 0.90625,
      "grad_norm": 0.24302264777949295,
      "learning_rate": 2e-05,
      "loss": 0.9676,
      "step": 29
    },
    {
      "epoch": 0.90625,
      "eval_loss": 0.9028491377830505,
      "eval_runtime": 56.802,
      "eval_samples_per_second": 3.521,
      "eval_steps_per_second": 0.44,
      "step": 29
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.28001197555170687,
      "learning_rate": 2e-05,
      "loss": 1.0044,
      "step": 30
    },
    {
      "epoch": 0.9375,
      "eval_loss": 0.8969234228134155,
      "eval_runtime": 56.8402,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 30
    },
    {
      "epoch": 0.96875,
      "grad_norm": 0.26990828196944483,
      "learning_rate": 2e-05,
      "loss": 0.8417,
      "step": 31
    },
    {
      "epoch": 0.96875,
      "eval_loss": 0.890943169593811,
      "eval_runtime": 56.9987,
      "eval_samples_per_second": 3.509,
      "eval_steps_per_second": 0.439,
      "step": 31
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.25976007498641823,
      "learning_rate": 2e-05,
      "loss": 0.95,
      "step": 32
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8852173686027527,
      "eval_runtime": 56.722,
      "eval_samples_per_second": 3.526,
      "eval_steps_per_second": 0.441,
      "step": 32
    },
    {
      "epoch": 1.03125,
      "grad_norm": 0.29530149620990226,
      "learning_rate": 2e-05,
      "loss": 0.9931,
      "step": 33
    },
    {
      "epoch": 1.03125,
      "eval_loss": 0.8795143961906433,
      "eval_runtime": 56.8541,
      "eval_samples_per_second": 3.518,
      "eval_steps_per_second": 0.44,
      "step": 33
    },
    {
      "epoch": 1.0625,
      "grad_norm": 0.2759239362577793,
      "learning_rate": 2e-05,
      "loss": 0.9978,
      "step": 34
    },
    {
      "epoch": 1.0625,
      "eval_loss": 0.8741766214370728,
      "eval_runtime": 56.7708,
      "eval_samples_per_second": 3.523,
      "eval_steps_per_second": 0.44,
      "step": 34
    },
    {
      "epoch": 1.09375,
      "grad_norm": 0.246531740102282,
      "learning_rate": 2e-05,
      "loss": 1.0163,
      "step": 35
    },
    {
      "epoch": 1.09375,
      "eval_loss": 0.8691757321357727,
      "eval_runtime": 56.8382,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 35
    },
    {
      "epoch": 1.125,
      "grad_norm": 0.2646078522027086,
      "learning_rate": 2e-05,
      "loss": 0.971,
      "step": 36
    },
    {
      "epoch": 1.125,
      "eval_loss": 0.8643682599067688,
      "eval_runtime": 56.689,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 0.441,
      "step": 36
    },
    {
      "epoch": 1.15625,
      "grad_norm": 0.2395171492146917,
      "learning_rate": 2e-05,
      "loss": 0.9227,
      "step": 37
    },
    {
      "epoch": 1.15625,
      "eval_loss": 0.8600785136222839,
      "eval_runtime": 56.72,
      "eval_samples_per_second": 3.526,
      "eval_steps_per_second": 0.441,
      "step": 37
    },
    {
      "epoch": 1.1875,
      "grad_norm": 0.28215229152733834,
      "learning_rate": 2e-05,
      "loss": 0.9308,
      "step": 38
    },
    {
      "epoch": 1.1875,
      "eval_loss": 0.8562959432601929,
      "eval_runtime": 56.8289,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 38
    },
    {
      "epoch": 1.21875,
      "grad_norm": 0.27116244597267625,
      "learning_rate": 2e-05,
      "loss": 0.9563,
      "step": 39
    },
    {
      "epoch": 1.21875,
      "eval_loss": 0.8526366949081421,
      "eval_runtime": 56.6829,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 0.441,
      "step": 39
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.2623711894386991,
      "learning_rate": 2e-05,
      "loss": 0.9535,
      "step": 40
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.8490655422210693,
      "eval_runtime": 56.6874,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 0.441,
      "step": 40
    },
    {
      "epoch": 1.28125,
      "grad_norm": 0.27251908150193377,
      "learning_rate": 2e-05,
      "loss": 0.9287,
      "step": 41
    },
    {
      "epoch": 1.28125,
      "eval_loss": 0.8451938629150391,
      "eval_runtime": 56.7117,
      "eval_samples_per_second": 3.527,
      "eval_steps_per_second": 0.441,
      "step": 41
    },
    {
      "epoch": 1.3125,
      "grad_norm": 0.2642817191103673,
      "learning_rate": 2e-05,
      "loss": 0.9186,
      "step": 42
    },
    {
      "epoch": 1.3125,
      "eval_loss": 0.8413894772529602,
      "eval_runtime": 56.9042,
      "eval_samples_per_second": 3.515,
      "eval_steps_per_second": 0.439,
      "step": 42
    },
    {
      "epoch": 1.34375,
      "grad_norm": 0.26857391288606197,
      "learning_rate": 2e-05,
      "loss": 0.8792,
      "step": 43
    },
    {
      "epoch": 1.34375,
      "eval_loss": 0.8373947739601135,
      "eval_runtime": 56.7211,
      "eval_samples_per_second": 3.526,
      "eval_steps_per_second": 0.441,
      "step": 43
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.2474531366673803,
      "learning_rate": 2e-05,
      "loss": 0.8965,
      "step": 44
    },
    {
      "epoch": 1.375,
      "eval_loss": 0.8339560031890869,
      "eval_runtime": 56.8277,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 44
    },
    {
      "epoch": 1.40625,
      "grad_norm": 0.26467660282496797,
      "learning_rate": 2e-05,
      "loss": 0.8762,
      "step": 45
    },
    {
      "epoch": 1.40625,
      "eval_loss": 0.8309465050697327,
      "eval_runtime": 56.7019,
      "eval_samples_per_second": 3.527,
      "eval_steps_per_second": 0.441,
      "step": 45
    },
    {
      "epoch": 1.4375,
      "grad_norm": 0.2652288034609541,
      "learning_rate": 2e-05,
      "loss": 0.9118,
      "step": 46
    },
    {
      "epoch": 1.4375,
      "eval_loss": 0.8279169201850891,
      "eval_runtime": 56.6271,
      "eval_samples_per_second": 3.532,
      "eval_steps_per_second": 0.441,
      "step": 46
    },
    {
      "epoch": 1.46875,
      "grad_norm": 0.27355995161173785,
      "learning_rate": 2e-05,
      "loss": 0.9249,
      "step": 47
    },
    {
      "epoch": 1.46875,
      "eval_loss": 0.8252391219139099,
      "eval_runtime": 56.6323,
      "eval_samples_per_second": 3.532,
      "eval_steps_per_second": 0.441,
      "step": 47
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.2588399009432225,
      "learning_rate": 2e-05,
      "loss": 0.8359,
      "step": 48
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.8225956559181213,
      "eval_runtime": 58.0142,
      "eval_samples_per_second": 3.447,
      "eval_steps_per_second": 0.431,
      "step": 48
    },
    {
      "epoch": 1.53125,
      "grad_norm": 0.28116473918910634,
      "learning_rate": 2e-05,
      "loss": 0.846,
      "step": 49
    },
    {
      "epoch": 1.53125,
      "eval_loss": 0.8198111057281494,
      "eval_runtime": 56.6785,
      "eval_samples_per_second": 3.529,
      "eval_steps_per_second": 0.441,
      "step": 49
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.30791508615928687,
      "learning_rate": 2e-05,
      "loss": 0.8364,
      "step": 50
    },
    {
      "epoch": 1.5625,
      "eval_loss": 0.816419243812561,
      "eval_runtime": 56.7867,
      "eval_samples_per_second": 3.522,
      "eval_steps_per_second": 0.44,
      "step": 50
    },
    {
      "epoch": 1.59375,
      "grad_norm": 0.2635774938006065,
      "learning_rate": 2e-05,
      "loss": 0.8565,
      "step": 51
    },
    {
      "epoch": 1.59375,
      "eval_loss": 0.8128839731216431,
      "eval_runtime": 56.5904,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 0.442,
      "step": 51
    },
    {
      "epoch": 1.625,
      "grad_norm": 0.25740594308086223,
      "learning_rate": 2e-05,
      "loss": 0.7573,
      "step": 52
    },
    {
      "epoch": 1.625,
      "eval_loss": 0.8096449971199036,
      "eval_runtime": 56.7381,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 0.441,
      "step": 52
    },
    {
      "epoch": 1.65625,
      "grad_norm": 0.25917235006885775,
      "learning_rate": 2e-05,
      "loss": 0.8982,
      "step": 53
    },
    {
      "epoch": 1.65625,
      "eval_loss": 0.8064478039741516,
      "eval_runtime": 57.4343,
      "eval_samples_per_second": 3.482,
      "eval_steps_per_second": 0.435,
      "step": 53
    },
    {
      "epoch": 1.6875,
      "grad_norm": 0.2831937064873763,
      "learning_rate": 2e-05,
      "loss": 0.8781,
      "step": 54
    },
    {
      "epoch": 1.6875,
      "eval_loss": 0.8034397959709167,
      "eval_runtime": 56.8346,
      "eval_samples_per_second": 3.519,
      "eval_steps_per_second": 0.44,
      "step": 54
    },
    {
      "epoch": 1.71875,
      "grad_norm": 0.2863024186152095,
      "learning_rate": 2e-05,
      "loss": 0.8861,
      "step": 55
    },
    {
      "epoch": 1.71875,
      "eval_loss": 0.800960898399353,
      "eval_runtime": 56.7424,
      "eval_samples_per_second": 3.525,
      "eval_steps_per_second": 0.441,
      "step": 55
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.28320211213029406,
      "learning_rate": 2e-05,
      "loss": 0.9514,
      "step": 56
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.7988448143005371,
      "eval_runtime": 57.0405,
      "eval_samples_per_second": 3.506,
      "eval_steps_per_second": 0.438,
      "step": 56
    },
    {
      "epoch": 1.78125,
      "grad_norm": 0.3204132014824286,
      "learning_rate": 2e-05,
      "loss": 0.8947,
      "step": 57
    },
    {
      "epoch": 1.78125,
      "eval_loss": 0.7971951365470886,
      "eval_runtime": 57.1716,
      "eval_samples_per_second": 3.498,
      "eval_steps_per_second": 0.437,
      "step": 57
    },
    {
      "epoch": 1.8125,
      "grad_norm": 0.29386668880511096,
      "learning_rate": 2e-05,
      "loss": 0.9125,
      "step": 58
    },
    {
      "epoch": 1.8125,
      "eval_loss": 0.7956165075302124,
      "eval_runtime": 57.3457,
      "eval_samples_per_second": 3.488,
      "eval_steps_per_second": 0.436,
      "step": 58
    },
    {
      "epoch": 1.84375,
      "grad_norm": 0.31091076146467406,
      "learning_rate": 2e-05,
      "loss": 0.8638,
      "step": 59
    },
    {
      "epoch": 1.84375,
      "eval_loss": 0.7935267090797424,
      "eval_runtime": 57.373,
      "eval_samples_per_second": 3.486,
      "eval_steps_per_second": 0.436,
      "step": 59
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.28779917523565474,
      "learning_rate": 2e-05,
      "loss": 0.9113,
      "step": 60
    },
    {
      "epoch": 1.875,
      "eval_loss": 0.7914787530899048,
      "eval_runtime": 57.2668,
      "eval_samples_per_second": 3.492,
      "eval_steps_per_second": 0.437,
      "step": 60
    },
    {
      "epoch": 1.90625,
      "grad_norm": 0.31820258275619673,
      "learning_rate": 2e-05,
      "loss": 0.8113,
      "step": 61
    },
    {
      "epoch": 1.90625,
      "eval_loss": 0.788929283618927,
      "eval_runtime": 57.2581,
      "eval_samples_per_second": 3.493,
      "eval_steps_per_second": 0.437,
      "step": 61
    },
    {
      "epoch": 1.9375,
      "grad_norm": 0.30186200117869055,
      "learning_rate": 2e-05,
      "loss": 0.8685,
      "step": 62
    },
    {
      "epoch": 1.9375,
      "eval_loss": 0.7862411737442017,
      "eval_runtime": 57.2688,
      "eval_samples_per_second": 3.492,
      "eval_steps_per_second": 0.437,
      "step": 62
    },
    {
      "epoch": 1.96875,
      "grad_norm": 0.27549296702686904,
      "learning_rate": 2e-05,
      "loss": 0.911,
      "step": 63
    },
    {
      "epoch": 1.96875,
      "eval_loss": 0.7838772535324097,
      "eval_runtime": 57.5102,
      "eval_samples_per_second": 3.478,
      "eval_steps_per_second": 0.435,
      "step": 63
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.29444542350221403,
      "learning_rate": 2e-05,
      "loss": 0.8877,
      "step": 64
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.7814672589302063,
      "eval_runtime": 57.3342,
      "eval_samples_per_second": 3.488,
      "eval_steps_per_second": 0.436,
      "step": 64
    },
    {
      "epoch": 2.03125,
      "grad_norm": 0.32976362380066954,
      "learning_rate": 2e-05,
      "loss": 0.836,
      "step": 65
    },
    {
      "epoch": 2.03125,
      "eval_loss": 0.7788661122322083,
      "eval_runtime": 57.6392,
      "eval_samples_per_second": 3.47,
      "eval_steps_per_second": 0.434,
      "step": 65
    },
    {
      "epoch": 2.0625,
      "grad_norm": 0.3091109685624876,
      "learning_rate": 2e-05,
      "loss": 0.8565,
      "step": 66
    },
    {
      "epoch": 2.0625,
      "eval_loss": 0.7769085764884949,
      "eval_runtime": 57.2017,
      "eval_samples_per_second": 3.496,
      "eval_steps_per_second": 0.437,
      "step": 66
    },
    {
      "epoch": 2.09375,
      "grad_norm": 0.3011651623444141,
      "learning_rate": 2e-05,
      "loss": 0.8265,
      "step": 67
    },
    {
      "epoch": 2.09375,
      "eval_loss": 0.7751161456108093,
      "eval_runtime": 57.4125,
      "eval_samples_per_second": 3.484,
      "eval_steps_per_second": 0.435,
      "step": 67
    },
    {
      "epoch": 2.125,
      "grad_norm": 0.28278958612422994,
      "learning_rate": 2e-05,
      "loss": 0.8893,
      "step": 68
    },
    {
      "epoch": 2.125,
      "eval_loss": 0.7736042737960815,
      "eval_runtime": 57.2826,
      "eval_samples_per_second": 3.491,
      "eval_steps_per_second": 0.436,
      "step": 68
    },
    {
      "epoch": 2.15625,
      "grad_norm": 0.30212533045014006,
      "learning_rate": 2e-05,
      "loss": 0.8256,
      "step": 69
    },
    {
      "epoch": 2.15625,
      "eval_loss": 0.7718043327331543,
      "eval_runtime": 59.4842,
      "eval_samples_per_second": 3.362,
      "eval_steps_per_second": 0.42,
      "step": 69
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.32231592883907934,
      "learning_rate": 2e-05,
      "loss": 0.7754,
      "step": 70
    },
    {
      "epoch": 2.1875,
      "eval_loss": 0.7697712779045105,
      "eval_runtime": 57.2127,
      "eval_samples_per_second": 3.496,
      "eval_steps_per_second": 0.437,
      "step": 70
    },
    {
      "epoch": 2.21875,
      "grad_norm": 0.29880148326318595,
      "learning_rate": 2e-05,
      "loss": 0.864,
      "step": 71
    },
    {
      "epoch": 2.21875,
      "eval_loss": 0.7679712176322937,
      "eval_runtime": 57.1052,
      "eval_samples_per_second": 3.502,
      "eval_steps_per_second": 0.438,
      "step": 71
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.30389759178870646,
      "learning_rate": 2e-05,
      "loss": 0.7831,
      "step": 72
    },
    {
      "epoch": 2.25,
      "eval_loss": 0.7662644386291504,
      "eval_runtime": 57.37,
      "eval_samples_per_second": 3.486,
      "eval_steps_per_second": 0.436,
      "step": 72
    },
    {
      "epoch": 2.28125,
      "grad_norm": 0.3424258847516451,
      "learning_rate": 2e-05,
      "loss": 0.8311,
      "step": 73
    },
    {
      "epoch": 2.28125,
      "eval_loss": 0.7646127343177795,
      "eval_runtime": 57.1884,
      "eval_samples_per_second": 3.497,
      "eval_steps_per_second": 0.437,
      "step": 73
    },
    {
      "epoch": 2.3125,
      "grad_norm": 0.2831654885374943,
      "learning_rate": 2e-05,
      "loss": 0.8261,
      "step": 74
    },
    {
      "epoch": 2.3125,
      "eval_loss": 0.7631255388259888,
      "eval_runtime": 57.4573,
      "eval_samples_per_second": 3.481,
      "eval_steps_per_second": 0.435,
      "step": 74
    },
    {
      "epoch": 2.34375,
      "grad_norm": 0.29894569677081223,
      "learning_rate": 2e-05,
      "loss": 0.8801,
      "step": 75
    },
    {
      "epoch": 2.34375,
      "eval_loss": 0.7617875933647156,
      "eval_runtime": 57.1641,
      "eval_samples_per_second": 3.499,
      "eval_steps_per_second": 0.437,
      "step": 75
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.3030991848050202,
      "learning_rate": 2e-05,
      "loss": 0.7921,
      "step": 76
    },
    {
      "epoch": 2.375,
      "eval_loss": 0.7605040073394775,
      "eval_runtime": 57.0991,
      "eval_samples_per_second": 3.503,
      "eval_steps_per_second": 0.438,
      "step": 76
    },
    {
      "epoch": 2.40625,
      "grad_norm": 0.30216971620226146,
      "learning_rate": 2e-05,
      "loss": 0.8527,
      "step": 77
    },
    {
      "epoch": 2.40625,
      "eval_loss": 0.7591890096664429,
      "eval_runtime": 58.6087,
      "eval_samples_per_second": 3.412,
      "eval_steps_per_second": 0.427,
      "step": 77
    },
    {
      "epoch": 2.4375,
      "grad_norm": 0.34907486616204614,
      "learning_rate": 2e-05,
      "loss": 0.841,
      "step": 78
    },
    {
      "epoch": 2.4375,
      "eval_loss": 0.7577351331710815,
      "eval_runtime": 59.509,
      "eval_samples_per_second": 3.361,
      "eval_steps_per_second": 0.42,
      "step": 78
    },
    {
      "epoch": 2.46875,
      "grad_norm": 0.3356288667630128,
      "learning_rate": 2e-05,
      "loss": 0.8417,
      "step": 79
    },
    {
      "epoch": 2.46875,
      "eval_loss": 0.7571098208427429,
      "eval_runtime": 57.4972,
      "eval_samples_per_second": 3.478,
      "eval_steps_per_second": 0.435,
      "step": 79
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.3547770718977253,
      "learning_rate": 2e-05,
      "loss": 0.8865,
      "step": 80
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.7565757632255554,
      "eval_runtime": 57.4262,
      "eval_samples_per_second": 3.483,
      "eval_steps_per_second": 0.435,
      "step": 80
    },
    {
      "epoch": 2.53125,
      "grad_norm": 0.36400071548952273,
      "learning_rate": 2e-05,
      "loss": 0.8201,
      "step": 81
    },
    {
      "epoch": 2.53125,
      "eval_loss": 0.7553688287734985,
      "eval_runtime": 59.6772,
      "eval_samples_per_second": 3.351,
      "eval_steps_per_second": 0.419,
      "step": 81
    },
    {
      "epoch": 2.5625,
      "grad_norm": 0.32432854183732784,
      "learning_rate": 2e-05,
      "loss": 0.8705,
      "step": 82
    },
    {
      "epoch": 2.5625,
      "eval_loss": 0.7540337443351746,
      "eval_runtime": 58.1967,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 82
    },
    {
      "epoch": 2.59375,
      "grad_norm": 0.3367161155473714,
      "learning_rate": 2e-05,
      "loss": 0.8225,
      "step": 83
    },
    {
      "epoch": 2.59375,
      "eval_loss": 0.752601683139801,
      "eval_runtime": 59.728,
      "eval_samples_per_second": 3.349,
      "eval_steps_per_second": 0.419,
      "step": 83
    },
    {
      "epoch": 2.625,
      "grad_norm": 0.3542073894911913,
      "learning_rate": 2e-05,
      "loss": 0.7887,
      "step": 84
    },
    {
      "epoch": 2.625,
      "eval_loss": 0.750983715057373,
      "eval_runtime": 58.2468,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 84
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.3387577198880303,
      "learning_rate": 2e-05,
      "loss": 0.7594,
      "step": 85
    },
    {
      "epoch": 2.65625,
      "eval_loss": 0.7495383620262146,
      "eval_runtime": 58.3457,
      "eval_samples_per_second": 3.428,
      "eval_steps_per_second": 0.428,
      "step": 85
    },
    {
      "epoch": 2.6875,
      "grad_norm": 0.381221735797731,
      "learning_rate": 2e-05,
      "loss": 0.7911,
      "step": 86
    },
    {
      "epoch": 2.6875,
      "eval_loss": 0.7477438449859619,
      "eval_runtime": 58.0584,
      "eval_samples_per_second": 3.445,
      "eval_steps_per_second": 0.431,
      "step": 86
    },
    {
      "epoch": 2.71875,
      "grad_norm": 0.3782280426863171,
      "learning_rate": 2e-05,
      "loss": 0.8115,
      "step": 87
    },
    {
      "epoch": 2.71875,
      "eval_loss": 0.7464295029640198,
      "eval_runtime": 57.9835,
      "eval_samples_per_second": 3.449,
      "eval_steps_per_second": 0.431,
      "step": 87
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.3751127153118298,
      "learning_rate": 2e-05,
      "loss": 0.8896,
      "step": 88
    },
    {
      "epoch": 2.75,
      "eval_loss": 0.7451103329658508,
      "eval_runtime": 58.1947,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 88
    },
    {
      "epoch": 2.78125,
      "grad_norm": 0.3580034870691801,
      "learning_rate": 2e-05,
      "loss": 0.7964,
      "step": 89
    },
    {
      "epoch": 2.78125,
      "eval_loss": 0.744097113609314,
      "eval_runtime": 58.1644,
      "eval_samples_per_second": 3.439,
      "eval_steps_per_second": 0.43,
      "step": 89
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.3630926811819107,
      "learning_rate": 2e-05,
      "loss": 0.848,
      "step": 90
    },
    {
      "epoch": 2.8125,
      "eval_loss": 0.7432359457015991,
      "eval_runtime": 58.0811,
      "eval_samples_per_second": 3.443,
      "eval_steps_per_second": 0.43,
      "step": 90
    },
    {
      "epoch": 2.84375,
      "grad_norm": 0.3668484035124972,
      "learning_rate": 2e-05,
      "loss": 0.7444,
      "step": 91
    },
    {
      "epoch": 2.84375,
      "eval_loss": 0.7424789667129517,
      "eval_runtime": 59.6811,
      "eval_samples_per_second": 3.351,
      "eval_steps_per_second": 0.419,
      "step": 91
    },
    {
      "epoch": 2.875,
      "grad_norm": 0.37526030248163283,
      "learning_rate": 2e-05,
      "loss": 0.8381,
      "step": 92
    },
    {
      "epoch": 2.875,
      "eval_loss": 0.7417113780975342,
      "eval_runtime": 58.1209,
      "eval_samples_per_second": 3.441,
      "eval_steps_per_second": 0.43,
      "step": 92
    },
    {
      "epoch": 2.90625,
      "grad_norm": 0.36285898832422037,
      "learning_rate": 2e-05,
      "loss": 0.7797,
      "step": 93
    },
    {
      "epoch": 2.90625,
      "eval_loss": 0.7411203980445862,
      "eval_runtime": 58.3212,
      "eval_samples_per_second": 3.429,
      "eval_steps_per_second": 0.429,
      "step": 93
    },
    {
      "epoch": 2.9375,
      "grad_norm": 0.39983168875602654,
      "learning_rate": 2e-05,
      "loss": 0.8571,
      "step": 94
    },
    {
      "epoch": 2.9375,
      "eval_loss": 0.7402496933937073,
      "eval_runtime": 58.0746,
      "eval_samples_per_second": 3.444,
      "eval_steps_per_second": 0.43,
      "step": 94
    },
    {
      "epoch": 2.96875,
      "grad_norm": 0.3697896026052261,
      "learning_rate": 2e-05,
      "loss": 0.7917,
      "step": 95
    },
    {
      "epoch": 2.96875,
      "eval_loss": 0.7398749589920044,
      "eval_runtime": 59.8008,
      "eval_samples_per_second": 3.344,
      "eval_steps_per_second": 0.418,
      "step": 95
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.39419135002625816,
      "learning_rate": 2e-05,
      "loss": 0.7987,
      "step": 96
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7384353876113892,
      "eval_runtime": 58.3389,
      "eval_samples_per_second": 3.428,
      "eval_steps_per_second": 0.429,
      "step": 96
    },
    {
      "epoch": 3.03125,
      "grad_norm": 0.40732207424611727,
      "learning_rate": 2e-05,
      "loss": 0.7205,
      "step": 97
    },
    {
      "epoch": 3.03125,
      "eval_loss": 0.73604416847229,
      "eval_runtime": 58.2114,
      "eval_samples_per_second": 3.436,
      "eval_steps_per_second": 0.429,
      "step": 97
    },
    {
      "epoch": 3.0625,
      "grad_norm": 0.3641635271623762,
      "learning_rate": 2e-05,
      "loss": 0.8062,
      "step": 98
    },
    {
      "epoch": 3.0625,
      "eval_loss": 0.7333144545555115,
      "eval_runtime": 59.7484,
      "eval_samples_per_second": 3.347,
      "eval_steps_per_second": 0.418,
      "step": 98
    },
    {
      "epoch": 3.09375,
      "grad_norm": 0.3556866449584765,
      "learning_rate": 2e-05,
      "loss": 0.7681,
      "step": 99
    },
    {
      "epoch": 3.09375,
      "eval_loss": 0.7306910157203674,
      "eval_runtime": 58.141,
      "eval_samples_per_second": 3.44,
      "eval_steps_per_second": 0.43,
      "step": 99
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.3826129743685834,
      "learning_rate": 2e-05,
      "loss": 0.7961,
      "step": 100
    },
    {
      "epoch": 3.125,
      "eval_loss": 0.7283279895782471,
      "eval_runtime": 58.1482,
      "eval_samples_per_second": 3.439,
      "eval_steps_per_second": 0.43,
      "step": 100
    },
    {
      "epoch": 3.15625,
      "grad_norm": 0.35166540759020914,
      "learning_rate": 2e-05,
      "loss": 0.7382,
      "step": 101
    },
    {
      "epoch": 3.15625,
      "eval_loss": 0.7267993688583374,
      "eval_runtime": 57.8007,
      "eval_samples_per_second": 3.46,
      "eval_steps_per_second": 0.433,
      "step": 101
    },
    {
      "epoch": 3.1875,
      "grad_norm": 0.38414476136018477,
      "learning_rate": 2e-05,
      "loss": 0.7999,
      "step": 102
    },
    {
      "epoch": 3.1875,
      "eval_loss": 0.7261015176773071,
      "eval_runtime": 57.9723,
      "eval_samples_per_second": 3.45,
      "eval_steps_per_second": 0.431,
      "step": 102
    },
    {
      "epoch": 3.21875,
      "grad_norm": 0.40218377868187477,
      "learning_rate": 2e-05,
      "loss": 0.8115,
      "step": 103
    },
    {
      "epoch": 3.21875,
      "eval_loss": 0.7257917523384094,
      "eval_runtime": 58.0394,
      "eval_samples_per_second": 3.446,
      "eval_steps_per_second": 0.431,
      "step": 103
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.41934721904445194,
      "learning_rate": 2e-05,
      "loss": 0.7228,
      "step": 104
    },
    {
      "epoch": 3.25,
      "eval_loss": 0.7251278758049011,
      "eval_runtime": 59.2828,
      "eval_samples_per_second": 3.374,
      "eval_steps_per_second": 0.422,
      "step": 104
    },
    {
      "epoch": 3.28125,
      "grad_norm": 0.3882012129329853,
      "learning_rate": 2e-05,
      "loss": 0.7658,
      "step": 105
    },
    {
      "epoch": 3.28125,
      "eval_loss": 0.724635899066925,
      "eval_runtime": 59.0543,
      "eval_samples_per_second": 3.387,
      "eval_steps_per_second": 0.423,
      "step": 105
    },
    {
      "epoch": 3.3125,
      "grad_norm": 0.4068559748805906,
      "learning_rate": 2e-05,
      "loss": 0.7977,
      "step": 106
    },
    {
      "epoch": 3.3125,
      "eval_loss": 0.7242235541343689,
      "eval_runtime": 58.5527,
      "eval_samples_per_second": 3.416,
      "eval_steps_per_second": 0.427,
      "step": 106
    },
    {
      "epoch": 3.34375,
      "grad_norm": 0.4620335365938039,
      "learning_rate": 2e-05,
      "loss": 0.7015,
      "step": 107
    },
    {
      "epoch": 3.34375,
      "eval_loss": 0.7226566076278687,
      "eval_runtime": 58.8135,
      "eval_samples_per_second": 3.401,
      "eval_steps_per_second": 0.425,
      "step": 107
    },
    {
      "epoch": 3.375,
      "grad_norm": 0.4009314815042761,
      "learning_rate": 2e-05,
      "loss": 0.7488,
      "step": 108
    },
    {
      "epoch": 3.375,
      "eval_loss": 0.7213454246520996,
      "eval_runtime": 58.735,
      "eval_samples_per_second": 3.405,
      "eval_steps_per_second": 0.426,
      "step": 108
    },
    {
      "epoch": 3.40625,
      "grad_norm": 0.456822567760836,
      "learning_rate": 2e-05,
      "loss": 0.7307,
      "step": 109
    },
    {
      "epoch": 3.40625,
      "eval_loss": 0.719496488571167,
      "eval_runtime": 58.9211,
      "eval_samples_per_second": 3.394,
      "eval_steps_per_second": 0.424,
      "step": 109
    },
    {
      "epoch": 3.4375,
      "grad_norm": 0.45520197938839,
      "learning_rate": 2e-05,
      "loss": 0.7348,
      "step": 110
    },
    {
      "epoch": 3.4375,
      "eval_loss": 0.7171263098716736,
      "eval_runtime": 58.9274,
      "eval_samples_per_second": 3.394,
      "eval_steps_per_second": 0.424,
      "step": 110
    },
    {
      "epoch": 3.46875,
      "grad_norm": 0.4421606621837213,
      "learning_rate": 2e-05,
      "loss": 0.8011,
      "step": 111
    },
    {
      "epoch": 3.46875,
      "eval_loss": 0.7155402898788452,
      "eval_runtime": 58.4009,
      "eval_samples_per_second": 3.425,
      "eval_steps_per_second": 0.428,
      "step": 111
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.4111011701354251,
      "learning_rate": 2e-05,
      "loss": 0.7829,
      "step": 112
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.714958667755127,
      "eval_runtime": 58.3143,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.429,
      "step": 112
    },
    {
      "epoch": 3.53125,
      "grad_norm": 0.40366265866888357,
      "learning_rate": 2e-05,
      "loss": 0.8596,
      "step": 113
    },
    {
      "epoch": 3.53125,
      "eval_loss": 0.7153159976005554,
      "eval_runtime": 58.5749,
      "eval_samples_per_second": 3.414,
      "eval_steps_per_second": 0.427,
      "step": 113
    },
    {
      "epoch": 3.5625,
      "grad_norm": 0.44914251592864773,
      "learning_rate": 2e-05,
      "loss": 0.7268,
      "step": 114
    },
    {
      "epoch": 3.5625,
      "eval_loss": 0.7159590721130371,
      "eval_runtime": 58.6872,
      "eval_samples_per_second": 3.408,
      "eval_steps_per_second": 0.426,
      "step": 114
    },
    {
      "epoch": 3.59375,
      "grad_norm": 0.4062399312752312,
      "learning_rate": 2e-05,
      "loss": 0.7875,
      "step": 115
    },
    {
      "epoch": 3.59375,
      "eval_loss": 0.7165355086326599,
      "eval_runtime": 58.4703,
      "eval_samples_per_second": 3.421,
      "eval_steps_per_second": 0.428,
      "step": 115
    },
    {
      "epoch": 3.625,
      "grad_norm": 0.44817350106485787,
      "learning_rate": 2e-05,
      "loss": 0.7623,
      "step": 116
    },
    {
      "epoch": 3.625,
      "eval_loss": 0.716560423374176,
      "eval_runtime": 58.5904,
      "eval_samples_per_second": 3.414,
      "eval_steps_per_second": 0.427,
      "step": 116
    },
    {
      "epoch": 3.65625,
      "grad_norm": 0.4309671248224914,
      "learning_rate": 2e-05,
      "loss": 0.7604,
      "step": 117
    },
    {
      "epoch": 3.65625,
      "eval_loss": 0.7165713310241699,
      "eval_runtime": 58.5214,
      "eval_samples_per_second": 3.418,
      "eval_steps_per_second": 0.427,
      "step": 117
    },
    {
      "epoch": 3.6875,
      "grad_norm": 0.44823929530189277,
      "learning_rate": 2e-05,
      "loss": 0.7751,
      "step": 118
    },
    {
      "epoch": 3.6875,
      "eval_loss": 0.7170334458351135,
      "eval_runtime": 58.7428,
      "eval_samples_per_second": 3.405,
      "eval_steps_per_second": 0.426,
      "step": 118
    },
    {
      "epoch": 3.71875,
      "grad_norm": 0.4369363559974751,
      "learning_rate": 2e-05,
      "loss": 0.8321,
      "step": 119
    },
    {
      "epoch": 3.71875,
      "eval_loss": 0.7169127464294434,
      "eval_runtime": 58.6794,
      "eval_samples_per_second": 3.408,
      "eval_steps_per_second": 0.426,
      "step": 119
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.43105130939689645,
      "learning_rate": 2e-05,
      "loss": 0.7722,
      "step": 120
    },
    {
      "epoch": 3.75,
      "eval_loss": 0.7162806987762451,
      "eval_runtime": 58.7674,
      "eval_samples_per_second": 3.403,
      "eval_steps_per_second": 0.425,
      "step": 120
    },
    {
      "epoch": 3.78125,
      "grad_norm": 0.43789804607163635,
      "learning_rate": 2e-05,
      "loss": 0.7548,
      "step": 121
    },
    {
      "epoch": 3.78125,
      "eval_loss": 0.7144981622695923,
      "eval_runtime": 58.3815,
      "eval_samples_per_second": 3.426,
      "eval_steps_per_second": 0.428,
      "step": 121
    },
    {
      "epoch": 3.8125,
      "grad_norm": 0.46941128815266536,
      "learning_rate": 2e-05,
      "loss": 0.8189,
      "step": 122
    },
    {
      "epoch": 3.8125,
      "eval_loss": 0.712846040725708,
      "eval_runtime": 58.5034,
      "eval_samples_per_second": 3.419,
      "eval_steps_per_second": 0.427,
      "step": 122
    },
    {
      "epoch": 3.84375,
      "grad_norm": 0.4415453126320104,
      "learning_rate": 2e-05,
      "loss": 0.7484,
      "step": 123
    },
    {
      "epoch": 3.84375,
      "eval_loss": 0.7111316919326782,
      "eval_runtime": 58.566,
      "eval_samples_per_second": 3.415,
      "eval_steps_per_second": 0.427,
      "step": 123
    },
    {
      "epoch": 3.875,
      "grad_norm": 0.4237981688992312,
      "learning_rate": 2e-05,
      "loss": 0.77,
      "step": 124
    },
    {
      "epoch": 3.875,
      "eval_loss": 0.7098332047462463,
      "eval_runtime": 58.5232,
      "eval_samples_per_second": 3.417,
      "eval_steps_per_second": 0.427,
      "step": 124
    },
    {
      "epoch": 3.90625,
      "grad_norm": 0.49069037639672286,
      "learning_rate": 2e-05,
      "loss": 0.8059,
      "step": 125
    },
    {
      "epoch": 3.90625,
      "eval_loss": 0.7086107730865479,
      "eval_runtime": 59.8651,
      "eval_samples_per_second": 3.341,
      "eval_steps_per_second": 0.418,
      "step": 125
    },
    {
      "epoch": 3.9375,
      "grad_norm": 0.48569295378281013,
      "learning_rate": 2e-05,
      "loss": 0.7799,
      "step": 126
    },
    {
      "epoch": 3.9375,
      "eval_loss": 0.7077484726905823,
      "eval_runtime": 58.4449,
      "eval_samples_per_second": 3.422,
      "eval_steps_per_second": 0.428,
      "step": 126
    },
    {
      "epoch": 3.96875,
      "grad_norm": 0.47224685972430797,
      "learning_rate": 2e-05,
      "loss": 0.7381,
      "step": 127
    },
    {
      "epoch": 3.96875,
      "eval_loss": 0.7073386907577515,
      "eval_runtime": 58.5961,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 0.427,
      "step": 127
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.48833051814427636,
      "learning_rate": 2e-05,
      "loss": 0.678,
      "step": 128
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.706765353679657,
      "eval_runtime": 60.6877,
      "eval_samples_per_second": 3.296,
      "eval_steps_per_second": 0.412,
      "step": 128
    },
    {
      "epoch": 4.03125,
      "grad_norm": 0.4116173650136014,
      "learning_rate": 2e-05,
      "loss": 0.7582,
      "step": 129
    },
    {
      "epoch": 4.03125,
      "eval_loss": 0.7067686319351196,
      "eval_runtime": 58.4349,
      "eval_samples_per_second": 3.423,
      "eval_steps_per_second": 0.428,
      "step": 129
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.46176556383782513,
      "learning_rate": 2e-05,
      "loss": 0.7749,
      "step": 130
    },
    {
      "epoch": 4.0625,
      "eval_loss": 0.7066690325737,
      "eval_runtime": 58.7029,
      "eval_samples_per_second": 3.407,
      "eval_steps_per_second": 0.426,
      "step": 130
    },
    {
      "epoch": 4.09375,
      "grad_norm": 0.4454696779432102,
      "learning_rate": 2e-05,
      "loss": 0.773,
      "step": 131
    },
    {
      "epoch": 4.09375,
      "eval_loss": 0.7064326405525208,
      "eval_runtime": 61.252,
      "eval_samples_per_second": 3.265,
      "eval_steps_per_second": 0.408,
      "step": 131
    },
    {
      "epoch": 4.125,
      "grad_norm": 0.5015422163334902,
      "learning_rate": 2e-05,
      "loss": 0.7369,
      "step": 132
    },
    {
      "epoch": 4.125,
      "eval_loss": 0.7057382464408875,
      "eval_runtime": 59.411,
      "eval_samples_per_second": 3.366,
      "eval_steps_per_second": 0.421,
      "step": 132
    },
    {
      "epoch": 4.15625,
      "grad_norm": 0.472373878055723,
      "learning_rate": 2e-05,
      "loss": 0.8262,
      "step": 133
    },
    {
      "epoch": 4.15625,
      "eval_loss": 0.7050846815109253,
      "eval_runtime": 59.2996,
      "eval_samples_per_second": 3.373,
      "eval_steps_per_second": 0.422,
      "step": 133
    },
    {
      "epoch": 4.1875,
      "grad_norm": 0.5384950553698907,
      "learning_rate": 2e-05,
      "loss": 0.74,
      "step": 134
    },
    {
      "epoch": 4.1875,
      "eval_loss": 0.7045766711235046,
      "eval_runtime": 59.2928,
      "eval_samples_per_second": 3.373,
      "eval_steps_per_second": 0.422,
      "step": 134
    },
    {
      "epoch": 4.21875,
      "grad_norm": 0.4692662892631433,
      "learning_rate": 2e-05,
      "loss": 0.7443,
      "step": 135
    },
    {
      "epoch": 4.21875,
      "eval_loss": 0.7045109272003174,
      "eval_runtime": 59.525,
      "eval_samples_per_second": 3.36,
      "eval_steps_per_second": 0.42,
      "step": 135
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.49707639799158876,
      "learning_rate": 2e-05,
      "loss": 0.733,
      "step": 136
    },
    {
      "epoch": 4.25,
      "eval_loss": 0.7047656178474426,
      "eval_runtime": 60.1718,
      "eval_samples_per_second": 3.324,
      "eval_steps_per_second": 0.415,
      "step": 136
    },
    {
      "epoch": 4.28125,
      "grad_norm": 0.5042999858449994,
      "learning_rate": 2e-05,
      "loss": 0.7303,
      "step": 137
    },
    {
      "epoch": 4.28125,
      "eval_loss": 0.7046284675598145,
      "eval_runtime": 60.01,
      "eval_samples_per_second": 3.333,
      "eval_steps_per_second": 0.417,
      "step": 137
    },
    {
      "epoch": 4.3125,
      "grad_norm": 0.5236583357740581,
      "learning_rate": 2e-05,
      "loss": 0.7254,
      "step": 138
    },
    {
      "epoch": 4.3125,
      "eval_loss": 0.7038366794586182,
      "eval_runtime": 60.3496,
      "eval_samples_per_second": 3.314,
      "eval_steps_per_second": 0.414,
      "step": 138
    },
    {
      "epoch": 4.34375,
      "grad_norm": 0.5197559530441114,
      "learning_rate": 2e-05,
      "loss": 0.6956,
      "step": 139
    },
    {
      "epoch": 4.34375,
      "eval_loss": 0.7023048400878906,
      "eval_runtime": 60.3808,
      "eval_samples_per_second": 3.312,
      "eval_steps_per_second": 0.414,
      "step": 139
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.5214546280852583,
      "learning_rate": 2e-05,
      "loss": 0.7243,
      "step": 140
    },
    {
      "epoch": 4.375,
      "eval_loss": 0.7011681199073792,
      "eval_runtime": 60.1368,
      "eval_samples_per_second": 3.326,
      "eval_steps_per_second": 0.416,
      "step": 140
    },
    {
      "epoch": 4.40625,
      "grad_norm": 0.47638616269940814,
      "learning_rate": 2e-05,
      "loss": 0.7442,
      "step": 141
    },
    {
      "epoch": 4.40625,
      "eval_loss": 0.7005561590194702,
      "eval_runtime": 61.003,
      "eval_samples_per_second": 3.279,
      "eval_steps_per_second": 0.41,
      "step": 141
    },
    {
      "epoch": 4.4375,
      "grad_norm": 0.5067672241908349,
      "learning_rate": 2e-05,
      "loss": 0.693,
      "step": 142
    },
    {
      "epoch": 4.4375,
      "eval_loss": 0.7004985809326172,
      "eval_runtime": 60.1646,
      "eval_samples_per_second": 3.324,
      "eval_steps_per_second": 0.416,
      "step": 142
    },
    {
      "epoch": 4.46875,
      "grad_norm": 0.5323088696033406,
      "learning_rate": 2e-05,
      "loss": 0.7019,
      "step": 143
    },
    {
      "epoch": 4.46875,
      "eval_loss": 0.7001196146011353,
      "eval_runtime": 59.9527,
      "eval_samples_per_second": 3.336,
      "eval_steps_per_second": 0.417,
      "step": 143
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.4994538125400832,
      "learning_rate": 2e-05,
      "loss": 0.684,
      "step": 144
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.6989223957061768,
      "eval_runtime": 59.7753,
      "eval_samples_per_second": 3.346,
      "eval_steps_per_second": 0.418,
      "step": 144
    },
    {
      "epoch": 4.53125,
      "grad_norm": 0.5328972466603664,
      "learning_rate": 2e-05,
      "loss": 0.7581,
      "step": 145
    },
    {
      "epoch": 4.53125,
      "eval_loss": 0.697172999382019,
      "eval_runtime": 59.678,
      "eval_samples_per_second": 3.351,
      "eval_steps_per_second": 0.419,
      "step": 145
    },
    {
      "epoch": 4.5625,
      "grad_norm": 0.557725244530984,
      "learning_rate": 2e-05,
      "loss": 0.6562,
      "step": 146
    },
    {
      "epoch": 4.5625,
      "eval_loss": 0.6954514980316162,
      "eval_runtime": 59.6753,
      "eval_samples_per_second": 3.351,
      "eval_steps_per_second": 0.419,
      "step": 146
    },
    {
      "epoch": 4.59375,
      "grad_norm": 0.520999668899182,
      "learning_rate": 2e-05,
      "loss": 0.7108,
      "step": 147
    },
    {
      "epoch": 4.59375,
      "eval_loss": 0.6949453949928284,
      "eval_runtime": 59.7891,
      "eval_samples_per_second": 3.345,
      "eval_steps_per_second": 0.418,
      "step": 147
    },
    {
      "epoch": 4.625,
      "grad_norm": 0.513677589761833,
      "learning_rate": 2e-05,
      "loss": 0.6697,
      "step": 148
    },
    {
      "epoch": 4.625,
      "eval_loss": 0.6953239440917969,
      "eval_runtime": 59.7415,
      "eval_samples_per_second": 3.348,
      "eval_steps_per_second": 0.418,
      "step": 148
    },
    {
      "epoch": 4.65625,
      "grad_norm": 0.5054488117701784,
      "learning_rate": 2e-05,
      "loss": 0.7793,
      "step": 149
    },
    {
      "epoch": 4.65625,
      "eval_loss": 0.6959659457206726,
      "eval_runtime": 59.9711,
      "eval_samples_per_second": 3.335,
      "eval_steps_per_second": 0.417,
      "step": 149
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.5962123257952582,
      "learning_rate": 2e-05,
      "loss": 0.7068,
      "step": 150
    },
    {
      "epoch": 4.6875,
      "eval_loss": 0.6952192783355713,
      "eval_runtime": 59.6824,
      "eval_samples_per_second": 3.351,
      "eval_steps_per_second": 0.419,
      "step": 150
    },
    {
      "epoch": 4.71875,
      "grad_norm": 0.6009619303481951,
      "learning_rate": 2e-05,
      "loss": 0.7261,
      "step": 151
    },
    {
      "epoch": 4.71875,
      "eval_loss": 0.6935360431671143,
      "eval_runtime": 59.5352,
      "eval_samples_per_second": 3.359,
      "eval_steps_per_second": 0.42,
      "step": 151
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5670117266130251,
      "learning_rate": 2e-05,
      "loss": 0.744,
      "step": 152
    },
    {
      "epoch": 4.75,
      "eval_loss": 0.6924968957901001,
      "eval_runtime": 61.2965,
      "eval_samples_per_second": 3.263,
      "eval_steps_per_second": 0.408,
      "step": 152
    },
    {
      "epoch": 4.78125,
      "grad_norm": 0.5564998515626721,
      "learning_rate": 2e-05,
      "loss": 0.6982,
      "step": 153
    },
    {
      "epoch": 4.78125,
      "eval_loss": 0.6924961805343628,
      "eval_runtime": 61.2731,
      "eval_samples_per_second": 3.264,
      "eval_steps_per_second": 0.408,
      "step": 153
    },
    {
      "epoch": 4.8125,
      "grad_norm": 0.528752035989291,
      "learning_rate": 2e-05,
      "loss": 0.7109,
      "step": 154
    },
    {
      "epoch": 4.8125,
      "eval_loss": 0.6933311223983765,
      "eval_runtime": 59.8859,
      "eval_samples_per_second": 3.34,
      "eval_steps_per_second": 0.417,
      "step": 154
    },
    {
      "epoch": 4.84375,
      "grad_norm": 0.5868388300709311,
      "learning_rate": 2e-05,
      "loss": 0.6592,
      "step": 155
    },
    {
      "epoch": 4.84375,
      "eval_loss": 0.6933980584144592,
      "eval_runtime": 59.9915,
      "eval_samples_per_second": 3.334,
      "eval_steps_per_second": 0.417,
      "step": 155
    },
    {
      "epoch": 4.875,
      "grad_norm": 0.5602090329210427,
      "learning_rate": 2e-05,
      "loss": 0.7682,
      "step": 156
    },
    {
      "epoch": 4.875,
      "eval_loss": 0.6923888921737671,
      "eval_runtime": 61.499,
      "eval_samples_per_second": 3.252,
      "eval_steps_per_second": 0.407,
      "step": 156
    },
    {
      "epoch": 4.90625,
      "grad_norm": 0.5051330890531748,
      "learning_rate": 2e-05,
      "loss": 0.7491,
      "step": 157
    },
    {
      "epoch": 4.90625,
      "eval_loss": 0.69191575050354,
      "eval_runtime": 59.6969,
      "eval_samples_per_second": 3.35,
      "eval_steps_per_second": 0.419,
      "step": 157
    },
    {
      "epoch": 4.9375,
      "grad_norm": 0.5377224007409029,
      "learning_rate": 2e-05,
      "loss": 0.7501,
      "step": 158
    },
    {
      "epoch": 4.9375,
      "eval_loss": 0.69122314453125,
      "eval_runtime": 60.1345,
      "eval_samples_per_second": 3.326,
      "eval_steps_per_second": 0.416,
      "step": 158
    },
    {
      "epoch": 4.96875,
      "grad_norm": 0.544576473903093,
      "learning_rate": 2e-05,
      "loss": 0.714,
      "step": 159
    },
    {
      "epoch": 4.96875,
      "eval_loss": 0.6905286908149719,
      "eval_runtime": 59.9667,
      "eval_samples_per_second": 3.335,
      "eval_steps_per_second": 0.417,
      "step": 159
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5027197538560159,
      "learning_rate": 2e-05,
      "loss": 0.7181,
      "step": 160
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.6906802654266357,
      "eval_runtime": 60.0766,
      "eval_samples_per_second": 3.329,
      "eval_steps_per_second": 0.416,
      "step": 160
    },
    {
      "epoch": 5.03125,
      "grad_norm": 0.5041535532115543,
      "learning_rate": 2e-05,
      "loss": 0.6636,
      "step": 161
    },
    {
      "epoch": 5.03125,
      "eval_loss": 0.6912646293640137,
      "eval_runtime": 63.5855,
      "eval_samples_per_second": 3.145,
      "eval_steps_per_second": 0.393,
      "step": 161
    },
    {
      "epoch": 5.0625,
      "grad_norm": 0.5286650599348627,
      "learning_rate": 2e-05,
      "loss": 0.8107,
      "step": 162
    },
    {
      "epoch": 5.0625,
      "eval_loss": 0.6922540068626404,
      "eval_runtime": 56.5364,
      "eval_samples_per_second": 3.538,
      "eval_steps_per_second": 0.442,
      "step": 162
    },
    {
      "epoch": 5.09375,
      "grad_norm": 0.588785168960039,
      "learning_rate": 2e-05,
      "loss": 0.6169,
      "step": 163
    },
    {
      "epoch": 5.09375,
      "eval_loss": 0.692643404006958,
      "eval_runtime": 56.5005,
      "eval_samples_per_second": 3.54,
      "eval_steps_per_second": 0.442,
      "step": 163
    },
    {
      "epoch": 5.125,
      "grad_norm": 0.5752677936578872,
      "learning_rate": 2e-05,
      "loss": 0.7473,
      "step": 164
    },
    {
      "epoch": 5.125,
      "eval_loss": 0.6927568912506104,
      "eval_runtime": 58.5386,
      "eval_samples_per_second": 3.417,
      "eval_steps_per_second": 0.427,
      "step": 164
    },
    {
      "epoch": 5.15625,
      "grad_norm": 0.6487162117437294,
      "learning_rate": 2e-05,
      "loss": 0.588,
      "step": 165
    },
    {
      "epoch": 5.15625,
      "eval_loss": 0.692574143409729,
      "eval_runtime": 56.4611,
      "eval_samples_per_second": 3.542,
      "eval_steps_per_second": 0.443,
      "step": 165
    },
    {
      "epoch": 5.1875,
      "grad_norm": 0.6353608377871973,
      "learning_rate": 2e-05,
      "loss": 0.6933,
      "step": 166
    },
    {
      "epoch": 5.1875,
      "eval_loss": 0.6932590007781982,
      "eval_runtime": 56.5989,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 0.442,
      "step": 166
    },
    {
      "epoch": 5.21875,
      "grad_norm": 0.5450036592535661,
      "learning_rate": 2e-05,
      "loss": 0.7175,
      "step": 167
    },
    {
      "epoch": 5.21875,
      "eval_loss": 0.6944625973701477,
      "eval_runtime": 56.5362,
      "eval_samples_per_second": 3.538,
      "eval_steps_per_second": 0.442,
      "step": 167
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.6095734786538398,
      "learning_rate": 2e-05,
      "loss": 0.7478,
      "step": 168
    },
    {
      "epoch": 5.25,
      "eval_loss": 0.695120632648468,
      "eval_runtime": 56.465,
      "eval_samples_per_second": 3.542,
      "eval_steps_per_second": 0.443,
      "step": 168
    },
    {
      "epoch": 5.28125,
      "grad_norm": 0.5879704367364821,
      "learning_rate": 2e-05,
      "loss": 0.674,
      "step": 169
    },
    {
      "epoch": 5.28125,
      "eval_loss": 0.6956540942192078,
      "eval_runtime": 56.6007,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 0.442,
      "step": 169
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.6595426789183463,
      "learning_rate": 2e-05,
      "loss": 0.6536,
      "step": 170
    },
    {
      "epoch": 5.3125,
      "eval_loss": 0.6957553029060364,
      "eval_runtime": 56.4722,
      "eval_samples_per_second": 3.542,
      "eval_steps_per_second": 0.443,
      "step": 170
    },
    {
      "epoch": 5.34375,
      "grad_norm": 0.7708120772721636,
      "learning_rate": 2e-05,
      "loss": 0.666,
      "step": 171
    },
    {
      "epoch": 5.34375,
      "eval_loss": 0.693030834197998,
      "eval_runtime": 56.3518,
      "eval_samples_per_second": 3.549,
      "eval_steps_per_second": 0.444,
      "step": 171
    },
    {
      "epoch": 5.375,
      "grad_norm": 0.666091377671071,
      "learning_rate": 2e-05,
      "loss": 0.7422,
      "step": 172
    },
    {
      "epoch": 5.375,
      "eval_loss": 0.6900334358215332,
      "eval_runtime": 56.5395,
      "eval_samples_per_second": 3.537,
      "eval_steps_per_second": 0.442,
      "step": 172
    },
    {
      "epoch": 5.40625,
      "grad_norm": 0.6203365868953359,
      "learning_rate": 2e-05,
      "loss": 0.7069,
      "step": 173
    },
    {
      "epoch": 5.40625,
      "eval_loss": 0.6880744099617004,
      "eval_runtime": 56.4675,
      "eval_samples_per_second": 3.542,
      "eval_steps_per_second": 0.443,
      "step": 173
    },
    {
      "epoch": 5.4375,
      "grad_norm": 0.6299525495855296,
      "learning_rate": 2e-05,
      "loss": 0.7422,
      "step": 174
    },
    {
      "epoch": 5.4375,
      "eval_loss": 0.686725378036499,
      "eval_runtime": 56.671,
      "eval_samples_per_second": 3.529,
      "eval_steps_per_second": 0.441,
      "step": 174
    },
    {
      "epoch": 5.46875,
      "grad_norm": 0.6415660970283229,
      "learning_rate": 2e-05,
      "loss": 0.7347,
      "step": 175
    },
    {
      "epoch": 5.46875,
      "eval_loss": 0.6870352029800415,
      "eval_runtime": 56.5976,
      "eval_samples_per_second": 3.534,
      "eval_steps_per_second": 0.442,
      "step": 175
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.6569935128967318,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 176
    },
    {
      "epoch": 5.5,
      "eval_loss": 0.6870338320732117,
      "eval_runtime": 57.2325,
      "eval_samples_per_second": 3.495,
      "eval_steps_per_second": 0.437,
      "step": 176
    },
    {
      "epoch": 5.53125,
      "grad_norm": 0.6895239904364278,
      "learning_rate": 2e-05,
      "loss": 0.7106,
      "step": 177
    },
    {
      "epoch": 5.53125,
      "eval_loss": 0.6859387755393982,
      "eval_runtime": 57.3075,
      "eval_samples_per_second": 3.49,
      "eval_steps_per_second": 0.436,
      "step": 177
    },
    {
      "epoch": 5.5625,
      "grad_norm": 0.5855839234707383,
      "learning_rate": 2e-05,
      "loss": 0.7361,
      "step": 178
    },
    {
      "epoch": 5.5625,
      "eval_loss": 0.6856819987297058,
      "eval_runtime": 57.5973,
      "eval_samples_per_second": 3.472,
      "eval_steps_per_second": 0.434,
      "step": 178
    },
    {
      "epoch": 5.59375,
      "grad_norm": 0.6198072484940144,
      "learning_rate": 2e-05,
      "loss": 0.6386,
      "step": 179
    },
    {
      "epoch": 5.59375,
      "eval_loss": 0.6865841746330261,
      "eval_runtime": 57.4429,
      "eval_samples_per_second": 3.482,
      "eval_steps_per_second": 0.435,
      "step": 179
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.6169444945747248,
      "learning_rate": 2e-05,
      "loss": 0.6455,
      "step": 180
    },
    {
      "epoch": 5.625,
      "eval_loss": 0.6871997714042664,
      "eval_runtime": 57.3975,
      "eval_samples_per_second": 3.484,
      "eval_steps_per_second": 0.436,
      "step": 180
    },
    {
      "epoch": 5.65625,
      "grad_norm": 0.6524804251939137,
      "learning_rate": 2e-05,
      "loss": 0.6588,
      "step": 181
    },
    {
      "epoch": 5.65625,
      "eval_loss": 0.6873356103897095,
      "eval_runtime": 57.4579,
      "eval_samples_per_second": 3.481,
      "eval_steps_per_second": 0.435,
      "step": 181
    },
    {
      "epoch": 5.6875,
      "grad_norm": 0.6578787618504525,
      "learning_rate": 2e-05,
      "loss": 0.6274,
      "step": 182
    },
    {
      "epoch": 5.6875,
      "eval_loss": 0.6880214214324951,
      "eval_runtime": 57.5735,
      "eval_samples_per_second": 3.474,
      "eval_steps_per_second": 0.434,
      "step": 182
    },
    {
      "epoch": 5.71875,
      "grad_norm": 0.732160801451622,
      "learning_rate": 2e-05,
      "loss": 0.6623,
      "step": 183
    },
    {
      "epoch": 5.71875,
      "eval_loss": 0.6879817247390747,
      "eval_runtime": 57.5801,
      "eval_samples_per_second": 3.473,
      "eval_steps_per_second": 0.434,
      "step": 183
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.7294753965107613,
      "learning_rate": 2e-05,
      "loss": 0.6562,
      "step": 184
    },
    {
      "epoch": 5.75,
      "eval_loss": 0.6870495676994324,
      "eval_runtime": 57.6659,
      "eval_samples_per_second": 3.468,
      "eval_steps_per_second": 0.434,
      "step": 184
    },
    {
      "epoch": 5.78125,
      "grad_norm": 0.6947870304881401,
      "learning_rate": 2e-05,
      "loss": 0.695,
      "step": 185
    },
    {
      "epoch": 5.78125,
      "eval_loss": 0.6856162548065186,
      "eval_runtime": 57.4452,
      "eval_samples_per_second": 3.482,
      "eval_steps_per_second": 0.435,
      "step": 185
    },
    {
      "epoch": 5.8125,
      "grad_norm": 0.7085011414361884,
      "learning_rate": 2e-05,
      "loss": 0.6634,
      "step": 186
    },
    {
      "epoch": 5.8125,
      "eval_loss": 0.6839439272880554,
      "eval_runtime": 57.3621,
      "eval_samples_per_second": 3.487,
      "eval_steps_per_second": 0.436,
      "step": 186
    },
    {
      "epoch": 5.84375,
      "grad_norm": 0.6548606152047736,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 187
    },
    {
      "epoch": 5.84375,
      "eval_loss": 0.6837204098701477,
      "eval_runtime": 57.3849,
      "eval_samples_per_second": 3.485,
      "eval_steps_per_second": 0.436,
      "step": 187
    },
    {
      "epoch": 5.875,
      "grad_norm": 0.6662179186613736,
      "learning_rate": 2e-05,
      "loss": 0.6528,
      "step": 188
    },
    {
      "epoch": 5.875,
      "eval_loss": 0.6844826340675354,
      "eval_runtime": 57.3173,
      "eval_samples_per_second": 3.489,
      "eval_steps_per_second": 0.436,
      "step": 188
    },
    {
      "epoch": 5.90625,
      "grad_norm": 0.6638311768585444,
      "learning_rate": 2e-05,
      "loss": 0.6582,
      "step": 189
    },
    {
      "epoch": 5.90625,
      "eval_loss": 0.6846724152565002,
      "eval_runtime": 57.5354,
      "eval_samples_per_second": 3.476,
      "eval_steps_per_second": 0.435,
      "step": 189
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.7007259768118588,
      "learning_rate": 2e-05,
      "loss": 0.6742,
      "step": 190
    },
    {
      "epoch": 5.9375,
      "eval_loss": 0.6834731101989746,
      "eval_runtime": 57.4134,
      "eval_samples_per_second": 3.484,
      "eval_steps_per_second": 0.435,
      "step": 190
    },
    {
      "epoch": 5.96875,
      "grad_norm": 0.6563132346432226,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 191
    },
    {
      "epoch": 5.96875,
      "eval_loss": 0.6817070245742798,
      "eval_runtime": 56.6649,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 0.441,
      "step": 191
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6349703649303867,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 192
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.6804311871528625,
      "eval_runtime": 56.4378,
      "eval_samples_per_second": 3.544,
      "eval_steps_per_second": 0.443,
      "step": 192
    },
    {
      "epoch": 6.03125,
      "grad_norm": 0.6716039243820887,
      "learning_rate": 2e-05,
      "loss": 0.7145,
      "step": 193
    },
    {
      "epoch": 6.03125,
      "eval_loss": 0.6804825067520142,
      "eval_runtime": 56.6403,
      "eval_samples_per_second": 3.531,
      "eval_steps_per_second": 0.441,
      "step": 193
    },
    {
      "epoch": 6.0625,
      "grad_norm": 0.5950395984856348,
      "learning_rate": 2e-05,
      "loss": 0.6768,
      "step": 194
    },
    {
      "epoch": 6.0625,
      "eval_loss": 0.6823931932449341,
      "eval_runtime": 56.5459,
      "eval_samples_per_second": 3.537,
      "eval_steps_per_second": 0.442,
      "step": 194
    },
    {
      "epoch": 6.09375,
      "grad_norm": 0.6787703014730869,
      "learning_rate": 2e-05,
      "loss": 0.6158,
      "step": 195
    },
    {
      "epoch": 6.09375,
      "eval_loss": 0.6854414939880371,
      "eval_runtime": 56.5293,
      "eval_samples_per_second": 3.538,
      "eval_steps_per_second": 0.442,
      "step": 195
    },
    {
      "epoch": 6.125,
      "grad_norm": 0.6526684210082853,
      "learning_rate": 2e-05,
      "loss": 0.6479,
      "step": 196
    },
    {
      "epoch": 6.125,
      "eval_loss": 0.6892845034599304,
      "eval_runtime": 56.5099,
      "eval_samples_per_second": 3.539,
      "eval_steps_per_second": 0.442,
      "step": 196
    },
    {
      "epoch": 6.15625,
      "grad_norm": 0.6997704487164051,
      "learning_rate": 2e-05,
      "loss": 0.6706,
      "step": 197
    },
    {
      "epoch": 6.15625,
      "eval_loss": 0.6941932439804077,
      "eval_runtime": 58.514,
      "eval_samples_per_second": 3.418,
      "eval_steps_per_second": 0.427,
      "step": 197
    },
    {
      "epoch": 6.1875,
      "grad_norm": 0.7511370305129338,
      "learning_rate": 2e-05,
      "loss": 0.7418,
      "step": 198
    },
    {
      "epoch": 6.1875,
      "eval_loss": 0.6964046955108643,
      "eval_runtime": 58.4428,
      "eval_samples_per_second": 3.422,
      "eval_steps_per_second": 0.428,
      "step": 198
    },
    {
      "epoch": 6.21875,
      "grad_norm": 0.8468482037911412,
      "learning_rate": 2e-05,
      "loss": 0.618,
      "step": 199
    },
    {
      "epoch": 6.21875,
      "eval_loss": 0.6947888731956482,
      "eval_runtime": 56.6921,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 0.441,
      "step": 199
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.80366391754735,
      "learning_rate": 2e-05,
      "loss": 0.6712,
      "step": 200
    },
    {
      "epoch": 6.25,
      "eval_loss": 0.691255509853363,
      "eval_runtime": 56.7536,
      "eval_samples_per_second": 3.524,
      "eval_steps_per_second": 0.441,
      "step": 200
    },
    {
      "epoch": 6.28125,
      "grad_norm": 0.7123001788838409,
      "learning_rate": 2e-05,
      "loss": 0.6886,
      "step": 201
    },
    {
      "epoch": 6.28125,
      "eval_loss": 0.6888566613197327,
      "eval_runtime": 57.4537,
      "eval_samples_per_second": 3.481,
      "eval_steps_per_second": 0.435,
      "step": 201
    },
    {
      "epoch": 6.3125,
      "grad_norm": 0.7785807978964993,
      "learning_rate": 2e-05,
      "loss": 0.6096,
      "step": 202
    },
    {
      "epoch": 6.3125,
      "eval_loss": 0.6869829297065735,
      "eval_runtime": 57.3967,
      "eval_samples_per_second": 3.485,
      "eval_steps_per_second": 0.436,
      "step": 202
    },
    {
      "epoch": 6.34375,
      "grad_norm": 0.6771659776183533,
      "learning_rate": 2e-05,
      "loss": 0.7328,
      "step": 203
    },
    {
      "epoch": 6.34375,
      "eval_loss": 0.6867367625236511,
      "eval_runtime": 57.5277,
      "eval_samples_per_second": 3.477,
      "eval_steps_per_second": 0.435,
      "step": 203
    },
    {
      "epoch": 6.375,
      "grad_norm": 0.8106446356590065,
      "learning_rate": 2e-05,
      "loss": 0.5931,
      "step": 204
    },
    {
      "epoch": 6.375,
      "eval_loss": 0.6862130165100098,
      "eval_runtime": 57.4868,
      "eval_samples_per_second": 3.479,
      "eval_steps_per_second": 0.435,
      "step": 204
    },
    {
      "epoch": 6.40625,
      "grad_norm": 0.6600674902481064,
      "learning_rate": 2e-05,
      "loss": 0.5789,
      "step": 205
    },
    {
      "epoch": 6.40625,
      "eval_loss": 0.6866827607154846,
      "eval_runtime": 57.4287,
      "eval_samples_per_second": 3.483,
      "eval_steps_per_second": 0.435,
      "step": 205
    },
    {
      "epoch": 6.4375,
      "grad_norm": 0.8177118767015663,
      "learning_rate": 2e-05,
      "loss": 0.6395,
      "step": 206
    },
    {
      "epoch": 6.4375,
      "eval_loss": 0.6866394281387329,
      "eval_runtime": 57.0918,
      "eval_samples_per_second": 3.503,
      "eval_steps_per_second": 0.438,
      "step": 206
    },
    {
      "epoch": 6.46875,
      "grad_norm": 0.7284237801181533,
      "learning_rate": 2e-05,
      "loss": 0.6835,
      "step": 207
    },
    {
      "epoch": 6.46875,
      "eval_loss": 0.6864017248153687,
      "eval_runtime": 57.1565,
      "eval_samples_per_second": 3.499,
      "eval_steps_per_second": 0.437,
      "step": 207
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.7603002790103086,
      "learning_rate": 2e-05,
      "loss": 0.6347,
      "step": 208
    },
    {
      "epoch": 6.5,
      "eval_loss": 0.6871703267097473,
      "eval_runtime": 57.4181,
      "eval_samples_per_second": 3.483,
      "eval_steps_per_second": 0.435,
      "step": 208
    },
    {
      "epoch": 6.53125,
      "grad_norm": 0.8359766442946917,
      "learning_rate": 2e-05,
      "loss": 0.6088,
      "step": 209
    },
    {
      "epoch": 6.53125,
      "eval_loss": 0.6878347992897034,
      "eval_runtime": 57.4837,
      "eval_samples_per_second": 3.479,
      "eval_steps_per_second": 0.435,
      "step": 209
    },
    {
      "epoch": 6.5625,
      "grad_norm": 0.7778968951616311,
      "learning_rate": 2e-05,
      "loss": 0.5912,
      "step": 210
    },
    {
      "epoch": 6.5625,
      "eval_loss": 0.6893374919891357,
      "eval_runtime": 57.6159,
      "eval_samples_per_second": 3.471,
      "eval_steps_per_second": 0.434,
      "step": 210
    },
    {
      "epoch": 6.59375,
      "grad_norm": 0.8300437291816744,
      "learning_rate": 2e-05,
      "loss": 0.6299,
      "step": 211
    },
    {
      "epoch": 6.59375,
      "eval_loss": 0.6899804472923279,
      "eval_runtime": 57.1491,
      "eval_samples_per_second": 3.5,
      "eval_steps_per_second": 0.437,
      "step": 211
    },
    {
      "epoch": 6.625,
      "grad_norm": 0.7994430152763061,
      "learning_rate": 2e-05,
      "loss": 0.6073,
      "step": 212
    },
    {
      "epoch": 6.625,
      "eval_loss": 0.6889459490776062,
      "eval_runtime": 57.3773,
      "eval_samples_per_second": 3.486,
      "eval_steps_per_second": 0.436,
      "step": 212
    },
    {
      "epoch": 6.65625,
      "grad_norm": 0.7475670453371858,
      "learning_rate": 2e-05,
      "loss": 0.6774,
      "step": 213
    },
    {
      "epoch": 6.65625,
      "eval_loss": 0.6873544454574585,
      "eval_runtime": 57.4114,
      "eval_samples_per_second": 3.484,
      "eval_steps_per_second": 0.435,
      "step": 213
    },
    {
      "epoch": 6.6875,
      "grad_norm": 0.7281375343651885,
      "learning_rate": 2e-05,
      "loss": 0.6404,
      "step": 214
    },
    {
      "epoch": 6.6875,
      "eval_loss": 0.6867469549179077,
      "eval_runtime": 57.2899,
      "eval_samples_per_second": 3.491,
      "eval_steps_per_second": 0.436,
      "step": 214
    },
    {
      "epoch": 6.71875,
      "grad_norm": 0.7684115091080507,
      "learning_rate": 2e-05,
      "loss": 0.6382,
      "step": 215
    },
    {
      "epoch": 6.71875,
      "eval_loss": 0.6860084533691406,
      "eval_runtime": 57.38,
      "eval_samples_per_second": 3.486,
      "eval_steps_per_second": 0.436,
      "step": 215
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.7962356695445627,
      "learning_rate": 2e-05,
      "loss": 0.6398,
      "step": 216
    },
    {
      "epoch": 6.75,
      "eval_loss": 0.6856002807617188,
      "eval_runtime": 57.2399,
      "eval_samples_per_second": 3.494,
      "eval_steps_per_second": 0.437,
      "step": 216
    },
    {
      "epoch": 6.78125,
      "grad_norm": 0.7893826807634562,
      "learning_rate": 2e-05,
      "loss": 0.59,
      "step": 217
    },
    {
      "epoch": 6.78125,
      "eval_loss": 0.6870043873786926,
      "eval_runtime": 57.1671,
      "eval_samples_per_second": 3.499,
      "eval_steps_per_second": 0.437,
      "step": 217
    },
    {
      "epoch": 6.8125,
      "grad_norm": 0.8329644141570051,
      "learning_rate": 2e-05,
      "loss": 0.5932,
      "step": 218
    },
    {
      "epoch": 6.8125,
      "eval_loss": 0.6870229840278625,
      "eval_runtime": 57.3642,
      "eval_samples_per_second": 3.486,
      "eval_steps_per_second": 0.436,
      "step": 218
    },
    {
      "epoch": 6.84375,
      "grad_norm": 0.9075127715796286,
      "learning_rate": 2e-05,
      "loss": 0.669,
      "step": 219
    },
    {
      "epoch": 6.84375,
      "eval_loss": 0.6856889128684998,
      "eval_runtime": 57.4226,
      "eval_samples_per_second": 3.483,
      "eval_steps_per_second": 0.435,
      "step": 219
    },
    {
      "epoch": 6.875,
      "grad_norm": 0.8464505810718659,
      "learning_rate": 2e-05,
      "loss": 0.686,
      "step": 220
    },
    {
      "epoch": 6.875,
      "eval_loss": 0.6835823059082031,
      "eval_runtime": 57.2105,
      "eval_samples_per_second": 3.496,
      "eval_steps_per_second": 0.437,
      "step": 220
    },
    {
      "epoch": 6.90625,
      "grad_norm": 0.7799140952562077,
      "learning_rate": 2e-05,
      "loss": 0.6503,
      "step": 221
    },
    {
      "epoch": 6.90625,
      "eval_loss": 0.6825523376464844,
      "eval_runtime": 57.0985,
      "eval_samples_per_second": 3.503,
      "eval_steps_per_second": 0.438,
      "step": 221
    },
    {
      "epoch": 6.9375,
      "grad_norm": 0.8495343756184095,
      "learning_rate": 2e-05,
      "loss": 0.6533,
      "step": 222
    },
    {
      "epoch": 6.9375,
      "eval_loss": 0.6813305616378784,
      "eval_runtime": 57.1896,
      "eval_samples_per_second": 3.497,
      "eval_steps_per_second": 0.437,
      "step": 222
    },
    {
      "epoch": 6.96875,
      "grad_norm": 0.8191950862245413,
      "learning_rate": 2e-05,
      "loss": 0.6627,
      "step": 223
    },
    {
      "epoch": 6.96875,
      "eval_loss": 0.6800451874732971,
      "eval_runtime": 57.3904,
      "eval_samples_per_second": 3.485,
      "eval_steps_per_second": 0.436,
      "step": 223
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.8196747980504347,
      "learning_rate": 2e-05,
      "loss": 0.7337,
      "step": 224
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.6801488399505615,
      "eval_runtime": 59.0121,
      "eval_samples_per_second": 3.389,
      "eval_steps_per_second": 0.424,
      "step": 224
    },
    {
      "epoch": 7.03125,
      "grad_norm": 0.7095908101379159,
      "learning_rate": 2e-05,
      "loss": 0.6203,
      "step": 225
    },
    {
      "epoch": 7.03125,
      "eval_loss": 0.6816287040710449,
      "eval_runtime": 57.1754,
      "eval_samples_per_second": 3.498,
      "eval_steps_per_second": 0.437,
      "step": 225
    },
    {
      "epoch": 7.0625,
      "grad_norm": 0.7916901149958031,
      "learning_rate": 2e-05,
      "loss": 0.5489,
      "step": 226
    },
    {
      "epoch": 7.0625,
      "eval_loss": 0.6857742071151733,
      "eval_runtime": 58.0461,
      "eval_samples_per_second": 3.446,
      "eval_steps_per_second": 0.431,
      "step": 226
    },
    {
      "epoch": 7.09375,
      "grad_norm": 0.8190252103616696,
      "learning_rate": 2e-05,
      "loss": 0.613,
      "step": 227
    },
    {
      "epoch": 7.09375,
      "eval_loss": 0.6924745440483093,
      "eval_runtime": 58.351,
      "eval_samples_per_second": 3.428,
      "eval_steps_per_second": 0.428,
      "step": 227
    },
    {
      "epoch": 7.125,
      "grad_norm": 0.9385023798254423,
      "learning_rate": 2e-05,
      "loss": 0.5647,
      "step": 228
    },
    {
      "epoch": 7.125,
      "eval_loss": 0.7020445466041565,
      "eval_runtime": 58.1868,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 228
    },
    {
      "epoch": 7.15625,
      "grad_norm": 1.178887354836488,
      "learning_rate": 2e-05,
      "loss": 0.5957,
      "step": 229
    },
    {
      "epoch": 7.15625,
      "eval_loss": 0.7064430117607117,
      "eval_runtime": 58.3297,
      "eval_samples_per_second": 3.429,
      "eval_steps_per_second": 0.429,
      "step": 229
    },
    {
      "epoch": 7.1875,
      "grad_norm": 1.0054198258359948,
      "learning_rate": 2e-05,
      "loss": 0.5667,
      "step": 230
    },
    {
      "epoch": 7.1875,
      "eval_loss": 0.7060463428497314,
      "eval_runtime": 58.3212,
      "eval_samples_per_second": 3.429,
      "eval_steps_per_second": 0.429,
      "step": 230
    },
    {
      "epoch": 7.21875,
      "grad_norm": 1.005055760217432,
      "learning_rate": 2e-05,
      "loss": 0.6546,
      "step": 231
    },
    {
      "epoch": 7.21875,
      "eval_loss": 0.7029504179954529,
      "eval_runtime": 58.0188,
      "eval_samples_per_second": 3.447,
      "eval_steps_per_second": 0.431,
      "step": 231
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.9458472260674603,
      "learning_rate": 2e-05,
      "loss": 0.6503,
      "step": 232
    },
    {
      "epoch": 7.25,
      "eval_loss": 0.6988745927810669,
      "eval_runtime": 58.3149,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.429,
      "step": 232
    },
    {
      "epoch": 7.28125,
      "grad_norm": 1.022594832986886,
      "learning_rate": 2e-05,
      "loss": 0.611,
      "step": 233
    },
    {
      "epoch": 7.28125,
      "eval_loss": 0.6943955421447754,
      "eval_runtime": 58.3693,
      "eval_samples_per_second": 3.426,
      "eval_steps_per_second": 0.428,
      "step": 233
    },
    {
      "epoch": 7.3125,
      "grad_norm": 0.8953283498269817,
      "learning_rate": 2e-05,
      "loss": 0.6438,
      "step": 234
    },
    {
      "epoch": 7.3125,
      "eval_loss": 0.6924715638160706,
      "eval_runtime": 58.214,
      "eval_samples_per_second": 3.436,
      "eval_steps_per_second": 0.429,
      "step": 234
    },
    {
      "epoch": 7.34375,
      "grad_norm": 0.9094812403228425,
      "learning_rate": 2e-05,
      "loss": 0.6123,
      "step": 235
    },
    {
      "epoch": 7.34375,
      "eval_loss": 0.690609335899353,
      "eval_runtime": 58.6042,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 0.427,
      "step": 235
    },
    {
      "epoch": 7.375,
      "grad_norm": 0.9433427892139121,
      "learning_rate": 2e-05,
      "loss": 0.5772,
      "step": 236
    },
    {
      "epoch": 7.375,
      "eval_loss": 0.6895288825035095,
      "eval_runtime": 58.0083,
      "eval_samples_per_second": 3.448,
      "eval_steps_per_second": 0.431,
      "step": 236
    },
    {
      "epoch": 7.40625,
      "grad_norm": 0.9654218046347709,
      "learning_rate": 2e-05,
      "loss": 0.62,
      "step": 237
    },
    {
      "epoch": 7.40625,
      "eval_loss": 0.6887797713279724,
      "eval_runtime": 58.1374,
      "eval_samples_per_second": 3.44,
      "eval_steps_per_second": 0.43,
      "step": 237
    },
    {
      "epoch": 7.4375,
      "grad_norm": 1.033591761626784,
      "learning_rate": 2e-05,
      "loss": 0.6163,
      "step": 238
    },
    {
      "epoch": 7.4375,
      "eval_loss": 0.6888651847839355,
      "eval_runtime": 58.2539,
      "eval_samples_per_second": 3.433,
      "eval_steps_per_second": 0.429,
      "step": 238
    },
    {
      "epoch": 7.46875,
      "grad_norm": 0.9059638854254064,
      "learning_rate": 2e-05,
      "loss": 0.6364,
      "step": 239
    },
    {
      "epoch": 7.46875,
      "eval_loss": 0.6905540227890015,
      "eval_runtime": 58.0992,
      "eval_samples_per_second": 3.442,
      "eval_steps_per_second": 0.43,
      "step": 239
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.9193726862314907,
      "learning_rate": 2e-05,
      "loss": 0.5845,
      "step": 240
    },
    {
      "epoch": 7.5,
      "eval_loss": 0.693742036819458,
      "eval_runtime": 58.2336,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.429,
      "step": 240
    },
    {
      "epoch": 7.53125,
      "grad_norm": 0.8539139714986941,
      "learning_rate": 2e-05,
      "loss": 0.6344,
      "step": 241
    },
    {
      "epoch": 7.53125,
      "eval_loss": 0.696897566318512,
      "eval_runtime": 59.3124,
      "eval_samples_per_second": 3.372,
      "eval_steps_per_second": 0.421,
      "step": 241
    },
    {
      "epoch": 7.5625,
      "grad_norm": 0.9552275495908527,
      "learning_rate": 2e-05,
      "loss": 0.6159,
      "step": 242
    },
    {
      "epoch": 7.5625,
      "eval_loss": 0.6991227865219116,
      "eval_runtime": 58.1037,
      "eval_samples_per_second": 3.442,
      "eval_steps_per_second": 0.43,
      "step": 242
    },
    {
      "epoch": 7.59375,
      "grad_norm": 0.8953175982318474,
      "learning_rate": 2e-05,
      "loss": 0.5934,
      "step": 243
    },
    {
      "epoch": 7.59375,
      "eval_loss": 0.7009669542312622,
      "eval_runtime": 59.9178,
      "eval_samples_per_second": 3.338,
      "eval_steps_per_second": 0.417,
      "step": 243
    },
    {
      "epoch": 7.625,
      "grad_norm": 1.1254017430464345,
      "learning_rate": 2e-05,
      "loss": 0.6721,
      "step": 244
    },
    {
      "epoch": 7.625,
      "eval_loss": 0.7003803253173828,
      "eval_runtime": 59.9278,
      "eval_samples_per_second": 3.337,
      "eval_steps_per_second": 0.417,
      "step": 244
    },
    {
      "epoch": 7.65625,
      "grad_norm": 0.9666525684896161,
      "learning_rate": 2e-05,
      "loss": 0.5793,
      "step": 245
    },
    {
      "epoch": 7.65625,
      "eval_loss": 0.6997054815292358,
      "eval_runtime": 58.3355,
      "eval_samples_per_second": 3.428,
      "eval_steps_per_second": 0.429,
      "step": 245
    },
    {
      "epoch": 7.6875,
      "grad_norm": 1.0500213825228455,
      "learning_rate": 2e-05,
      "loss": 0.6262,
      "step": 246
    },
    {
      "epoch": 7.6875,
      "eval_loss": 0.6956760883331299,
      "eval_runtime": 57.9053,
      "eval_samples_per_second": 3.454,
      "eval_steps_per_second": 0.432,
      "step": 246
    },
    {
      "epoch": 7.71875,
      "grad_norm": 1.0445166827193935,
      "learning_rate": 2e-05,
      "loss": 0.6111,
      "step": 247
    },
    {
      "epoch": 7.71875,
      "eval_loss": 0.6909776329994202,
      "eval_runtime": 58.1856,
      "eval_samples_per_second": 3.437,
      "eval_steps_per_second": 0.43,
      "step": 247
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.8935484171996528,
      "learning_rate": 2e-05,
      "loss": 0.6036,
      "step": 248
    },
    {
      "epoch": 7.75,
      "eval_loss": 0.6887417435646057,
      "eval_runtime": 58.1651,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.43,
      "step": 248
    },
    {
      "epoch": 7.78125,
      "grad_norm": 0.9329951454150782,
      "learning_rate": 2e-05,
      "loss": 0.6434,
      "step": 249
    },
    {
      "epoch": 7.78125,
      "eval_loss": 0.6893429756164551,
      "eval_runtime": 58.4106,
      "eval_samples_per_second": 3.424,
      "eval_steps_per_second": 0.428,
      "step": 249
    },
    {
      "epoch": 7.8125,
      "grad_norm": 0.8799352767832798,
      "learning_rate": 2e-05,
      "loss": 0.6519,
      "step": 250
    },
    {
      "epoch": 7.8125,
      "eval_loss": 0.6929408311843872,
      "eval_runtime": 58.3105,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.429,
      "step": 250
    },
    {
      "epoch": 7.84375,
      "grad_norm": 0.9322996227983372,
      "learning_rate": 2e-05,
      "loss": 0.5684,
      "step": 251
    },
    {
      "epoch": 7.84375,
      "eval_loss": 0.6954038739204407,
      "eval_runtime": 57.8998,
      "eval_samples_per_second": 3.454,
      "eval_steps_per_second": 0.432,
      "step": 251
    },
    {
      "epoch": 7.875,
      "grad_norm": 1.0904651907324217,
      "learning_rate": 2e-05,
      "loss": 0.5851,
      "step": 252
    },
    {
      "epoch": 7.875,
      "eval_loss": 0.6938650012016296,
      "eval_runtime": 58.4905,
      "eval_samples_per_second": 3.419,
      "eval_steps_per_second": 0.427,
      "step": 252
    },
    {
      "epoch": 7.90625,
      "grad_norm": 1.0103592741616823,
      "learning_rate": 2e-05,
      "loss": 0.6655,
      "step": 253
    },
    {
      "epoch": 7.90625,
      "eval_loss": 0.6909225583076477,
      "eval_runtime": 58.1801,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.43,
      "step": 253
    },
    {
      "epoch": 7.9375,
      "grad_norm": 0.9208541649120607,
      "learning_rate": 2e-05,
      "loss": 0.6051,
      "step": 254
    },
    {
      "epoch": 7.9375,
      "eval_loss": 0.6913868188858032,
      "eval_runtime": 58.4224,
      "eval_samples_per_second": 3.423,
      "eval_steps_per_second": 0.428,
      "step": 254
    },
    {
      "epoch": 7.96875,
      "grad_norm": 0.9567638724372727,
      "learning_rate": 2e-05,
      "loss": 0.5529,
      "step": 255
    },
    {
      "epoch": 7.96875,
      "eval_loss": 0.6918243169784546,
      "eval_runtime": 58.1569,
      "eval_samples_per_second": 3.439,
      "eval_steps_per_second": 0.43,
      "step": 255
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.8913592607849594,
      "learning_rate": 2e-05,
      "loss": 0.6076,
      "step": 256
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.6921086311340332,
      "eval_runtime": 58.8193,
      "eval_samples_per_second": 3.4,
      "eval_steps_per_second": 0.425,
      "step": 256
    },
    {
      "epoch": 8.0,
      "step": 256,
      "total_flos": 77213396434944.0,
      "train_loss": 0.24315254529938102,
      "train_runtime": 7656.702,
      "train_samples_per_second": 1.045,
      "train_steps_per_second": 0.033
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 256,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 77213396434944.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}