{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 307,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0163265306122449,
      "grad_norm": 64.88935089111328,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 14.3746,
      "step": 5
    },
    {
      "epoch": 0.0326530612244898,
      "grad_norm": 45.59659957885742,
      "learning_rate": 3.2142857142857147e-06,
      "loss": 13.7736,
      "step": 10
    },
    {
      "epoch": 0.04897959183673469,
      "grad_norm": 34.90564727783203,
      "learning_rate": 5e-06,
      "loss": 11.5266,
      "step": 15
    },
    {
      "epoch": 0.0653061224489796,
      "grad_norm": 26.288707733154297,
      "learning_rate": 6.785714285714287e-06,
      "loss": 8.9275,
      "step": 20
    },
    {
      "epoch": 0.08163265306122448,
      "grad_norm": 20.026254653930664,
      "learning_rate": 8.571428571428571e-06,
      "loss": 7.6264,
      "step": 25
    },
    {
      "epoch": 0.09795918367346938,
      "grad_norm": 16.850696563720703,
      "learning_rate": 9.99996905885448e-06,
      "loss": 6.2223,
      "step": 30
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 13.888376235961914,
      "learning_rate": 9.9988861589696e-06,
      "loss": 5.7476,
      "step": 35
    },
    {
      "epoch": 0.1306122448979592,
      "grad_norm": 14.347899436950684,
      "learning_rate": 9.996256584729157e-06,
      "loss": 5.0648,
      "step": 40
    },
    {
      "epoch": 0.1469387755102041,
      "grad_norm": 12.692639350891113,
      "learning_rate": 9.992081149733404e-06,
      "loss": 4.9457,
      "step": 45
    },
    {
      "epoch": 0.16326530612244897,
      "grad_norm": 13.268131256103516,
      "learning_rate": 9.986361145877783e-06,
      "loss": 4.7859,
      "step": 50
    },
    {
      "epoch": 0.17959183673469387,
      "grad_norm": 12.126714706420898,
      "learning_rate": 9.979098342953198e-06,
      "loss": 4.3966,
      "step": 55
    },
    {
      "epoch": 0.19591836734693877,
      "grad_norm": 11.363078117370605,
      "learning_rate": 9.970294988098452e-06,
      "loss": 4.5139,
      "step": 60
    },
    {
      "epoch": 0.21224489795918366,
      "grad_norm": 12.150185585021973,
      "learning_rate": 9.959953805104953e-06,
      "loss": 4.1404,
      "step": 65
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 12.580989837646484,
      "learning_rate": 9.948077993573983e-06,
      "loss": 4.2476,
      "step": 70
    },
    {
      "epoch": 0.24489795918367346,
      "grad_norm": 13.421459197998047,
      "learning_rate": 9.934671227926714e-06,
      "loss": 4.3114,
      "step": 75
    },
    {
      "epoch": 0.2612244897959184,
      "grad_norm": 12.11119270324707,
      "learning_rate": 9.919737656267335e-06,
      "loss": 4.1571,
      "step": 80
    },
    {
      "epoch": 0.27755102040816326,
      "grad_norm": 11.114103317260742,
      "learning_rate": 9.90328189909961e-06,
      "loss": 4.3387,
      "step": 85
    },
    {
      "epoch": 0.2938775510204082,
      "grad_norm": 10.965707778930664,
      "learning_rate": 9.885309047897285e-06,
      "loss": 4.129,
      "step": 90
    },
    {
      "epoch": 0.31020408163265306,
      "grad_norm": 10.521084785461426,
      "learning_rate": 9.86582466352875e-06,
      "loss": 3.7176,
      "step": 95
    },
    {
      "epoch": 0.32653061224489793,
      "grad_norm": 11.604248046875,
      "learning_rate": 9.844834774536503e-06,
      "loss": 4.0923,
      "step": 100
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 10.684212684631348,
      "learning_rate": 9.822345875271884e-06,
      "loss": 3.8132,
      "step": 105
    },
    {
      "epoch": 0.35918367346938773,
      "grad_norm": 10.826234817504883,
      "learning_rate": 9.798364923885703e-06,
      "loss": 3.743,
      "step": 110
    },
    {
      "epoch": 0.37551020408163266,
      "grad_norm": 10.982087135314941,
      "learning_rate": 9.772899340175362e-06,
      "loss": 3.7181,
      "step": 115
    },
    {
      "epoch": 0.39183673469387753,
      "grad_norm": 11.453109741210938,
      "learning_rate": 9.745957003289138e-06,
      "loss": 3.6383,
      "step": 120
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 10.43824577331543,
      "learning_rate": 9.71754624928835e-06,
      "loss": 3.7914,
      "step": 125
    },
    {
      "epoch": 0.42448979591836733,
      "grad_norm": 10.69090461730957,
      "learning_rate": 9.687675868568145e-06,
      "loss": 3.6724,
      "step": 130
    },
    {
      "epoch": 0.44081632653061226,
      "grad_norm": 11.608959197998047,
      "learning_rate": 9.65635510313772e-06,
      "loss": 3.6466,
      "step": 135
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 11.002665519714355,
      "learning_rate": 9.623593643760805e-06,
      "loss": 3.6917,
      "step": 140
    },
    {
      "epoch": 0.47346938775510206,
      "grad_norm": 11.66286849975586,
      "learning_rate": 9.589401626957309e-06,
      "loss": 3.6477,
      "step": 145
    },
    {
      "epoch": 0.4897959183673469,
      "grad_norm": 10.273452758789062,
      "learning_rate": 9.553789631867039e-06,
      "loss": 3.8276,
      "step": 150
    },
    {
      "epoch": 0.5061224489795918,
      "grad_norm": 9.553406715393066,
      "learning_rate": 9.516768676976476e-06,
      "loss": 3.5288,
      "step": 155
    },
    {
      "epoch": 0.5224489795918368,
      "grad_norm": 10.327857971191406,
      "learning_rate": 9.478350216709609e-06,
      "loss": 3.4926,
      "step": 160
    },
    {
      "epoch": 0.5387755102040817,
      "grad_norm": 9.799028396606445,
      "learning_rate": 9.438546137883898e-06,
      "loss": 3.8356,
      "step": 165
    },
    {
      "epoch": 0.5551020408163265,
      "grad_norm": 10.007837295532227,
      "learning_rate": 9.397368756032445e-06,
      "loss": 3.5522,
      "step": 170
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 11.845935821533203,
      "learning_rate": 9.354830811593527e-06,
      "loss": 3.5413,
      "step": 175
    },
    {
      "epoch": 0.5877551020408164,
      "grad_norm": 9.952783584594727,
      "learning_rate": 9.31094546596864e-06,
      "loss": 3.6146,
      "step": 180
    },
    {
      "epoch": 0.6040816326530613,
      "grad_norm": 11.10181999206543,
      "learning_rate": 9.265726297450332e-06,
      "loss": 3.4828,
      "step": 185
    },
    {
      "epoch": 0.6204081632653061,
      "grad_norm": 10.211321830749512,
      "learning_rate": 9.219187297021015e-06,
      "loss": 3.6493,
      "step": 190
    },
    {
      "epoch": 0.636734693877551,
      "grad_norm": 9.923507690429688,
      "learning_rate": 9.171342864024103e-06,
      "loss": 3.2479,
      "step": 195
    },
    {
      "epoch": 0.6530612244897959,
      "grad_norm": 10.2088623046875,
      "learning_rate": 9.122207801708802e-06,
      "loss": 3.4984,
      "step": 200
    },
    {
      "epoch": 0.6693877551020408,
      "grad_norm": 10.336106300354004,
      "learning_rate": 9.071797312649934e-06,
      "loss": 3.5449,
      "step": 205
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 10.09914493560791,
      "learning_rate": 9.020126994044194e-06,
      "loss": 3.369,
      "step": 210
    },
    {
      "epoch": 0.7020408163265306,
      "grad_norm": 9.624906539916992,
      "learning_rate": 8.96721283288432e-06,
      "loss": 3.2544,
      "step": 215
    },
    {
      "epoch": 0.7183673469387755,
      "grad_norm": 10.576995849609375,
      "learning_rate": 8.91307120101266e-06,
      "loss": 3.4857,
      "step": 220
    },
    {
      "epoch": 0.7346938775510204,
      "grad_norm": 9.93189811706543,
      "learning_rate": 8.857718850055663e-06,
      "loss": 3.4717,
      "step": 225
    },
    {
      "epoch": 0.7510204081632653,
      "grad_norm": 11.030794143676758,
      "learning_rate": 8.801172906240857e-06,
      "loss": 3.3982,
      "step": 230
    },
    {
      "epoch": 0.7673469387755102,
      "grad_norm": 9.748960494995117,
      "learning_rate": 8.743450865097929e-06,
      "loss": 3.4867,
      "step": 235
    },
    {
      "epoch": 0.7836734693877551,
      "grad_norm": 9.28258228302002,
      "learning_rate": 8.68457058604555e-06,
      "loss": 3.4942,
      "step": 240
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.684868812561035,
      "learning_rate": 8.624550286865592e-06,
      "loss": 3.5008,
      "step": 245
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 9.635165214538574,
      "learning_rate": 8.563408538066486e-06,
      "loss": 3.5459,
      "step": 250
    },
    {
      "epoch": 0.8326530612244898,
      "grad_norm": 9.419096946716309,
      "learning_rate": 8.501164257137431e-06,
      "loss": 3.4455,
      "step": 255
    },
    {
      "epoch": 0.8489795918367347,
      "grad_norm": 9.14213752746582,
      "learning_rate": 8.437836702695253e-06,
      "loss": 3.2428,
      "step": 260
    },
    {
      "epoch": 0.8653061224489796,
      "grad_norm": 9.956143379211426,
      "learning_rate": 8.373445468525719e-06,
      "loss": 3.4352,
      "step": 265
    },
    {
      "epoch": 0.8816326530612245,
      "grad_norm": 9.325640678405762,
      "learning_rate": 8.30801047752113e-06,
      "loss": 3.4512,
      "step": 270
    },
    {
      "epoch": 0.8979591836734694,
      "grad_norm": 10.310145378112793,
      "learning_rate": 8.241551975516133e-06,
      "loss": 3.654,
      "step": 275
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 10.872224807739258,
      "learning_rate": 8.17409052502355e-06,
      "loss": 3.4794,
      "step": 280
    },
    {
      "epoch": 0.9306122448979591,
      "grad_norm": 9.897497177124023,
      "learning_rate": 8.105646998872275e-06,
      "loss": 3.4442,
      "step": 285
    },
    {
      "epoch": 0.9469387755102041,
      "grad_norm": 9.942341804504395,
      "learning_rate": 8.036242573749142e-06,
      "loss": 3.2163,
      "step": 290
    },
    {
      "epoch": 0.963265306122449,
      "grad_norm": 10.519649505615234,
      "learning_rate": 7.965898723646777e-06,
      "loss": 3.6575,
      "step": 295
    },
    {
      "epoch": 0.9795918367346939,
      "grad_norm": 9.339863777160645,
      "learning_rate": 7.894637213219454e-06,
      "loss": 3.1936,
      "step": 300
    },
    {
      "epoch": 0.9959183673469387,
      "grad_norm": 10.674720764160156,
      "learning_rate": 7.82248009104904e-06,
      "loss": 3.5039,
      "step": 305
    }
  ],
  "logging_steps": 5,
  "max_steps": 921,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.68790515866665e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}