{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 1,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": 181.0,
      "learning_rate": 2.5e-05,
      "loss": 1.1894,
      "step": 1
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.45,
      "eval_loss": 1.370854377746582,
      "eval_runtime": 0.9284,
      "eval_samples_per_second": 107.713,
      "eval_steps_per_second": 3.231,
      "step": 1
    },
    {
      "epoch": 0.4,
      "grad_norm": 214.0,
      "learning_rate": 5e-05,
      "loss": 1.3321,
      "step": 2
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.6,
      "eval_loss": 0.7611915469169617,
      "eval_runtime": 1.0247,
      "eval_samples_per_second": 97.591,
      "eval_steps_per_second": 2.928,
      "step": 2
    },
    {
      "epoch": 0.6,
      "grad_norm": 99.0,
      "learning_rate": 4.8958333333333335e-05,
      "loss": 0.864,
      "step": 3
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.63,
      "eval_loss": 0.8467197418212891,
      "eval_runtime": 1.0232,
      "eval_samples_per_second": 97.729,
      "eval_steps_per_second": 2.932,
      "step": 3
    },
    {
      "epoch": 0.8,
      "grad_norm": 207.0,
      "learning_rate": 4.791666666666667e-05,
      "loss": 0.9293,
      "step": 4
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.91,
      "eval_loss": 0.411680668592453,
      "eval_runtime": 1.0225,
      "eval_samples_per_second": 97.8,
      "eval_steps_per_second": 2.934,
      "step": 4
    },
    {
      "epoch": 1.0,
      "grad_norm": 76.5,
      "learning_rate": 4.6875e-05,
      "loss": 0.3438,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.92,
      "eval_loss": 0.2434273660182953,
      "eval_runtime": 0.9724,
      "eval_samples_per_second": 102.835,
      "eval_steps_per_second": 3.085,
      "step": 5
    },
    {
      "epoch": 1.2,
      "grad_norm": 12.3125,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.377,
      "step": 6
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.87,
      "eval_loss": 0.3455248177051544,
      "eval_runtime": 1.0215,
      "eval_samples_per_second": 97.895,
      "eval_steps_per_second": 2.937,
      "step": 6
    },
    {
      "epoch": 1.4,
      "grad_norm": 111.0,
      "learning_rate": 4.4791666666666673e-05,
      "loss": 0.6136,
      "step": 7
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.9,
      "eval_loss": 0.324820876121521,
      "eval_runtime": 1.0222,
      "eval_samples_per_second": 97.829,
      "eval_steps_per_second": 2.935,
      "step": 7
    },
    {
      "epoch": 1.6,
      "grad_norm": 37.25,
      "learning_rate": 4.375e-05,
      "loss": 0.135,
      "step": 8
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.92,
      "eval_loss": 0.24112339317798615,
      "eval_runtime": 0.9739,
      "eval_samples_per_second": 102.682,
      "eval_steps_per_second": 3.08,
      "step": 8
    },
    {
      "epoch": 1.8,
      "grad_norm": 8.3125,
      "learning_rate": 4.270833333333333e-05,
      "loss": 0.2028,
      "step": 9
    },
    {
      "epoch": 1.8,
      "eval_accuracy": 0.92,
      "eval_loss": 0.2563403844833374,
      "eval_runtime": 1.0204,
      "eval_samples_per_second": 97.997,
      "eval_steps_per_second": 2.94,
      "step": 9
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.5,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.0607,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.92,
      "eval_loss": 0.2764730155467987,
      "eval_runtime": 1.0225,
      "eval_samples_per_second": 97.804,
      "eval_steps_per_second": 2.934,
      "step": 10
    },
    {
      "epoch": 2.2,
      "grad_norm": 15.6875,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 0.0602,
      "step": 11
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.93,
      "eval_loss": 0.26724421977996826,
      "eval_runtime": 1.0214,
      "eval_samples_per_second": 97.907,
      "eval_steps_per_second": 2.937,
      "step": 11
    },
    {
      "epoch": 2.4,
      "grad_norm": 5.84375,
      "learning_rate": 3.958333333333333e-05,
      "loss": 0.0881,
      "step": 12
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.93,
      "eval_loss": 0.2725869417190552,
      "eval_runtime": 1.0234,
      "eval_samples_per_second": 97.712,
      "eval_steps_per_second": 2.931,
      "step": 12
    },
    {
      "epoch": 2.6,
      "grad_norm": 9.0625,
      "learning_rate": 3.854166666666667e-05,
      "loss": 0.0385,
      "step": 13
    },
    {
      "epoch": 2.6,
      "eval_accuracy": 0.93,
      "eval_loss": 0.2831783592700958,
      "eval_runtime": 1.023,
      "eval_samples_per_second": 97.753,
      "eval_steps_per_second": 2.933,
      "step": 13
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.203125,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0081,
      "step": 14
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 0.93,
      "eval_loss": 0.27984124422073364,
      "eval_runtime": 1.0215,
      "eval_samples_per_second": 97.894,
      "eval_steps_per_second": 2.937,
      "step": 14
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.46875,
      "learning_rate": 3.6458333333333336e-05,
      "loss": 0.0815,
      "step": 15
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.93,
      "eval_loss": 0.2797885239124298,
      "eval_runtime": 1.0235,
      "eval_samples_per_second": 97.705,
      "eval_steps_per_second": 2.931,
      "step": 15
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.625,
      "learning_rate": 3.541666666666667e-05,
      "loss": 0.0029,
      "step": 16
    },
    {
      "epoch": 3.2,
      "eval_accuracy": 0.93,
      "eval_loss": 0.29290881752967834,
      "eval_runtime": 1.0228,
      "eval_samples_per_second": 97.77,
      "eval_steps_per_second": 2.933,
      "step": 16
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.5390625,
      "learning_rate": 3.4375e-05,
      "loss": 0.0027,
      "step": 17
    },
    {
      "epoch": 3.4,
      "eval_accuracy": 0.93,
      "eval_loss": 0.3011457622051239,
      "eval_runtime": 1.0225,
      "eval_samples_per_second": 97.796,
      "eval_steps_per_second": 2.934,
      "step": 17
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.42578125,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0013,
      "step": 18
    },
    {
      "epoch": 3.6,
      "eval_accuracy": 0.93,
      "eval_loss": 0.31652283668518066,
      "eval_runtime": 1.023,
      "eval_samples_per_second": 97.748,
      "eval_steps_per_second": 2.932,
      "step": 18
    },
    {
      "epoch": 3.8,
      "grad_norm": 5.8125,
      "learning_rate": 3.229166666666667e-05,
      "loss": 0.0151,
      "step": 19
    },
    {
      "epoch": 3.8,
      "eval_accuracy": 0.95,
      "eval_loss": 0.31321465969085693,
      "eval_runtime": 1.022,
      "eval_samples_per_second": 97.844,
      "eval_steps_per_second": 2.935,
      "step": 19
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2216796875,
      "learning_rate": 3.125e-05,
      "loss": 0.0005,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.95,
      "eval_loss": 0.3194425702095032,
      "eval_runtime": 1.0227,
      "eval_samples_per_second": 97.78,
      "eval_steps_per_second": 2.933,
      "step": 20
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.52734375,
      "learning_rate": 3.0208333333333334e-05,
      "loss": 0.0008,
      "step": 21
    },
    {
      "epoch": 4.2,
      "eval_accuracy": 0.95,
      "eval_loss": 0.33766308426856995,
      "eval_runtime": 1.0213,
      "eval_samples_per_second": 97.913,
      "eval_steps_per_second": 2.937,
      "step": 21
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.052734375,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.0001,
      "step": 22
    },
    {
      "epoch": 4.4,
      "eval_accuracy": 0.95,
      "eval_loss": 0.34586212038993835,
      "eval_runtime": 1.0224,
      "eval_samples_per_second": 97.805,
      "eval_steps_per_second": 2.934,
      "step": 22
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.043212890625,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 0.0001,
      "step": 23
    },
    {
      "epoch": 4.6,
      "eval_accuracy": 0.95,
      "eval_loss": 0.3585689067840576,
      "eval_runtime": 0.9713,
      "eval_samples_per_second": 102.955,
      "eval_steps_per_second": 3.089,
      "step": 23
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.00946044921875,
      "learning_rate": 2.7083333333333332e-05,
      "loss": 0.0,
      "step": 24
    },
    {
      "epoch": 4.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.3643890619277954,
      "eval_runtime": 1.0221,
      "eval_samples_per_second": 97.84,
      "eval_steps_per_second": 2.935,
      "step": 24
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0556640625,
      "learning_rate": 2.604166666666667e-05,
      "loss": 0.0001,
      "step": 25
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.3764663636684418,
      "eval_runtime": 1.0239,
      "eval_samples_per_second": 97.666,
      "eval_steps_per_second": 2.93,
      "step": 25
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.0042724609375,
      "learning_rate": 2.5e-05,
      "loss": 0.0,
      "step": 26
    },
    {
      "epoch": 5.2,
      "eval_accuracy": 0.96,
      "eval_loss": 0.3840982913970947,
      "eval_runtime": 1.0218,
      "eval_samples_per_second": 97.864,
      "eval_steps_per_second": 2.936,
      "step": 26
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.004913330078125,
      "learning_rate": 2.3958333333333334e-05,
      "loss": 0.0,
      "step": 27
    },
    {
      "epoch": 5.4,
      "eval_accuracy": 0.96,
      "eval_loss": 0.38938531279563904,
      "eval_runtime": 1.0225,
      "eval_samples_per_second": 97.795,
      "eval_steps_per_second": 2.934,
      "step": 27
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.05419921875,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.0001,
      "step": 28
    },
    {
      "epoch": 5.6,
      "eval_accuracy": 0.96,
      "eval_loss": 0.39918196201324463,
      "eval_runtime": 1.0221,
      "eval_samples_per_second": 97.835,
      "eval_steps_per_second": 2.935,
      "step": 28
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.0015106201171875,
      "learning_rate": 2.1875e-05,
      "loss": 0.0,
      "step": 29
    },
    {
      "epoch": 5.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4027714431285858,
      "eval_runtime": 0.9742,
      "eval_samples_per_second": 102.648,
      "eval_steps_per_second": 3.079,
      "step": 29
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.00124359130859375,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4107368588447571,
      "eval_runtime": 1.0223,
      "eval_samples_per_second": 97.82,
      "eval_steps_per_second": 2.935,
      "step": 30
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.00616455078125,
      "learning_rate": 1.9791666666666665e-05,
      "loss": 0.0,
      "step": 31
    },
    {
      "epoch": 6.2,
      "eval_accuracy": 0.96,
      "eval_loss": 0.417328417301178,
      "eval_runtime": 1.0242,
      "eval_samples_per_second": 97.634,
      "eval_steps_per_second": 2.929,
      "step": 31
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.0005950927734375,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0,
      "step": 32
    },
    {
      "epoch": 6.4,
      "eval_accuracy": 0.96,
      "eval_loss": 0.41879531741142273,
      "eval_runtime": 1.0235,
      "eval_samples_per_second": 97.705,
      "eval_steps_per_second": 2.931,
      "step": 32
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.000667572021484375,
      "learning_rate": 1.7708333333333335e-05,
      "loss": 0.0,
      "step": 33
    },
    {
      "epoch": 6.6,
      "eval_accuracy": 0.96,
      "eval_loss": 0.42711684107780457,
      "eval_runtime": 1.0243,
      "eval_samples_per_second": 97.627,
      "eval_steps_per_second": 2.929,
      "step": 33
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.0023651123046875,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0,
      "step": 34
    },
    {
      "epoch": 6.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.42084625363349915,
      "eval_runtime": 1.021,
      "eval_samples_per_second": 97.941,
      "eval_steps_per_second": 2.938,
      "step": 34
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.000560760498046875,
      "learning_rate": 1.5625e-05,
      "loss": 0.0,
      "step": 35
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.42686185240745544,
      "eval_runtime": 1.0218,
      "eval_samples_per_second": 97.868,
      "eval_steps_per_second": 2.936,
      "step": 35
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.0002498626708984375,
      "learning_rate": 1.4583333333333335e-05,
      "loss": 0.0,
      "step": 36
    },
    {
      "epoch": 7.2,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4289308190345764,
      "eval_runtime": 1.0218,
      "eval_samples_per_second": 97.865,
      "eval_steps_per_second": 2.936,
      "step": 36
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.00040435791015625,
      "learning_rate": 1.3541666666666666e-05,
      "loss": 0.0,
      "step": 37
    },
    {
      "epoch": 7.4,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43295109272003174,
      "eval_runtime": 1.0228,
      "eval_samples_per_second": 97.775,
      "eval_steps_per_second": 2.933,
      "step": 37
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.00019741058349609375,
      "learning_rate": 1.25e-05,
      "loss": 0.0,
      "step": 38
    },
    {
      "epoch": 7.6,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43248245120048523,
      "eval_runtime": 1.0208,
      "eval_samples_per_second": 97.959,
      "eval_steps_per_second": 2.939,
      "step": 38
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.0015869140625,
      "learning_rate": 1.1458333333333333e-05,
      "loss": 0.0,
      "step": 39
    },
    {
      "epoch": 7.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4331529140472412,
      "eval_runtime": 0.9726,
      "eval_samples_per_second": 102.82,
      "eval_steps_per_second": 3.085,
      "step": 39
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.000469207763671875,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.0,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4309934973716736,
      "eval_runtime": 1.0232,
      "eval_samples_per_second": 97.73,
      "eval_steps_per_second": 2.932,
      "step": 40
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.00037384033203125,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.0,
      "step": 41
    },
    {
      "epoch": 8.2,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4338625967502594,
      "eval_runtime": 1.0229,
      "eval_samples_per_second": 97.763,
      "eval_steps_per_second": 2.933,
      "step": 41
    },
    {
      "epoch": 8.4,
      "grad_norm": 0.00142669677734375,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0,
      "step": 42
    },
    {
      "epoch": 8.4,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43012726306915283,
      "eval_runtime": 1.058,
      "eval_samples_per_second": 94.52,
      "eval_steps_per_second": 2.836,
      "step": 42
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.00075531005859375,
      "learning_rate": 7.2916666666666674e-06,
      "loss": 0.0,
      "step": 43
    },
    {
      "epoch": 8.6,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43235713243484497,
      "eval_runtime": 1.0267,
      "eval_samples_per_second": 97.401,
      "eval_steps_per_second": 2.922,
      "step": 43
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.0002593994140625,
      "learning_rate": 6.25e-06,
      "loss": 0.0,
      "step": 44
    },
    {
      "epoch": 8.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43538880348205566,
      "eval_runtime": 0.9747,
      "eval_samples_per_second": 102.591,
      "eval_steps_per_second": 3.078,
      "step": 44
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.0004329681396484375,
      "learning_rate": 5.208333333333334e-06,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4357055723667145,
      "eval_runtime": 1.0226,
      "eval_samples_per_second": 97.791,
      "eval_steps_per_second": 2.934,
      "step": 45
    },
    {
      "epoch": 9.2,
      "grad_norm": 0.000652313232421875,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0,
      "step": 46
    },
    {
      "epoch": 9.2,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43566077947616577,
      "eval_runtime": 1.0229,
      "eval_samples_per_second": 97.757,
      "eval_steps_per_second": 2.933,
      "step": 46
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.0002994537353515625,
      "learning_rate": 3.125e-06,
      "loss": 0.0,
      "step": 47
    },
    {
      "epoch": 9.4,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4365979731082916,
      "eval_runtime": 1.0208,
      "eval_samples_per_second": 97.958,
      "eval_steps_per_second": 2.939,
      "step": 47
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.0004177093505859375,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.0,
      "step": 48
    },
    {
      "epoch": 9.6,
      "eval_accuracy": 0.96,
      "eval_loss": 0.43618661165237427,
      "eval_runtime": 0.9713,
      "eval_samples_per_second": 102.953,
      "eval_steps_per_second": 3.089,
      "step": 48
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.000568389892578125,
      "learning_rate": 1.0416666666666667e-06,
      "loss": 0.0,
      "step": 49
    },
    {
      "epoch": 9.8,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4351811110973358,
      "eval_runtime": 1.0214,
      "eval_samples_per_second": 97.903,
      "eval_steps_per_second": 2.937,
      "step": 49
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.00014495849609375,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.4397030174732208,
      "eval_runtime": 1.0204,
      "eval_samples_per_second": 98.0,
      "eval_steps_per_second": 2.94,
      "step": 50
    },
    {
      "epoch": 10.0,
      "step": 50,
      "total_flos": 9439761349476352.0,
      "train_loss": 0.126956181311009,
      "train_runtime": 137.6145,
      "train_samples_per_second": 29.067,
      "train_steps_per_second": 0.363
    }
  ],
  "logging_steps": 1,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9439761349476352.0,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}