{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 380,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2631578947368421,
      "grad_norm": Infinity,
      "learning_rate": 4.9473684210526315e-05,
      "loss": 2.4394,
      "step": 5
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 6.091742515563965,
      "learning_rate": 4.881578947368421e-05,
      "loss": 1.776,
      "step": 10
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 6.011572360992432,
      "learning_rate": 4.8157894736842105e-05,
      "loss": 1.6479,
      "step": 15
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 5.232929706573486,
      "learning_rate": 4.75e-05,
      "loss": 1.3513,
      "step": 20
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 3.361309289932251,
      "learning_rate": 4.68421052631579e-05,
      "loss": 1.2262,
      "step": 25
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 3.8729989528656006,
      "learning_rate": 4.618421052631579e-05,
      "loss": 1.1481,
      "step": 30
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 8.15366268157959,
      "learning_rate": 4.552631578947369e-05,
      "loss": 0.9278,
      "step": 35
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 5.728829383850098,
      "learning_rate": 4.486842105263158e-05,
      "loss": 0.8307,
      "step": 40
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 4.391445159912109,
      "learning_rate": 4.421052631578947e-05,
      "loss": 0.8832,
      "step": 45
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 5.178050518035889,
      "learning_rate": 4.355263157894737e-05,
      "loss": 0.626,
      "step": 50
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 4.203180313110352,
      "learning_rate": 4.289473684210527e-05,
      "loss": 0.6467,
      "step": 55
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 2.902172327041626,
      "learning_rate": 4.223684210526316e-05,
      "loss": 0.3657,
      "step": 60
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 1.7815818786621094,
      "learning_rate": 4.157894736842106e-05,
      "loss": 0.3913,
      "step": 65
    },
    {
      "epoch": 3.6842105263157894,
      "grad_norm": 5.220995903015137,
      "learning_rate": 4.092105263157895e-05,
      "loss": 0.6332,
      "step": 70
    },
    {
      "epoch": 3.9473684210526314,
      "grad_norm": 2.012242555618286,
      "learning_rate": 4.026315789473684e-05,
      "loss": 0.4408,
      "step": 75
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 1.9606434106826782,
      "learning_rate": 3.960526315789474e-05,
      "loss": 0.5089,
      "step": 80
    },
    {
      "epoch": 4.473684210526316,
      "grad_norm": 1.584269404411316,
      "learning_rate": 3.894736842105263e-05,
      "loss": 0.297,
      "step": 85
    },
    {
      "epoch": 4.7368421052631575,
      "grad_norm": 2.2993006706237793,
      "learning_rate": 3.828947368421053e-05,
      "loss": 0.2525,
      "step": 90
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7839508056640625,
      "learning_rate": 3.7631578947368425e-05,
      "loss": 0.3642,
      "step": 95
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 2.162219285964966,
      "learning_rate": 3.6973684210526316e-05,
      "loss": 0.2416,
      "step": 100
    },
    {
      "epoch": 5.526315789473684,
      "grad_norm": 2.49100399017334,
      "learning_rate": 3.6315789473684214e-05,
      "loss": 0.2607,
      "step": 105
    },
    {
      "epoch": 5.7894736842105265,
      "grad_norm": 6.302850723266602,
      "learning_rate": 3.5657894736842106e-05,
      "loss": 0.3316,
      "step": 110
    },
    {
      "epoch": 6.052631578947368,
      "grad_norm": 1.1700443029403687,
      "learning_rate": 3.5e-05,
      "loss": 0.2009,
      "step": 115
    },
    {
      "epoch": 6.315789473684211,
      "grad_norm": 1.686787724494934,
      "learning_rate": 3.4342105263157895e-05,
      "loss": 0.2794,
      "step": 120
    },
    {
      "epoch": 6.578947368421053,
      "grad_norm": 6.972183704376221,
      "learning_rate": 3.368421052631579e-05,
      "loss": 0.3693,
      "step": 125
    },
    {
      "epoch": 6.842105263157895,
      "grad_norm": 3.670428991317749,
      "learning_rate": 3.302631578947369e-05,
      "loss": 0.2268,
      "step": 130
    },
    {
      "epoch": 7.105263157894737,
      "grad_norm": 0.7313272953033447,
      "learning_rate": 3.236842105263158e-05,
      "loss": 0.1025,
      "step": 135
    },
    {
      "epoch": 7.368421052631579,
      "grad_norm": 2.2111823558807373,
      "learning_rate": 3.1710526315789473e-05,
      "loss": 0.2386,
      "step": 140
    },
    {
      "epoch": 7.631578947368421,
      "grad_norm": 0.6066373586654663,
      "learning_rate": 3.105263157894737e-05,
      "loss": 0.2176,
      "step": 145
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 1.489353060722351,
      "learning_rate": 3.0394736842105266e-05,
      "loss": 0.1689,
      "step": 150
    },
    {
      "epoch": 8.157894736842104,
      "grad_norm": 0.5530461668968201,
      "learning_rate": 2.9736842105263157e-05,
      "loss": 0.1457,
      "step": 155
    },
    {
      "epoch": 8.421052631578947,
      "grad_norm": 2.413187026977539,
      "learning_rate": 2.9078947368421055e-05,
      "loss": 0.2149,
      "step": 160
    },
    {
      "epoch": 8.68421052631579,
      "grad_norm": 0.7150534987449646,
      "learning_rate": 2.842105263157895e-05,
      "loss": 0.1202,
      "step": 165
    },
    {
      "epoch": 8.947368421052632,
      "grad_norm": 5.491703510284424,
      "learning_rate": 2.776315789473684e-05,
      "loss": 0.2105,
      "step": 170
    },
    {
      "epoch": 9.210526315789474,
      "grad_norm": 0.81364506483078,
      "learning_rate": 2.710526315789474e-05,
      "loss": 0.0898,
      "step": 175
    },
    {
      "epoch": 9.473684210526315,
      "grad_norm": 0.8343147039413452,
      "learning_rate": 2.644736842105263e-05,
      "loss": 0.1286,
      "step": 180
    },
    {
      "epoch": 9.736842105263158,
      "grad_norm": 0.5138881206512451,
      "learning_rate": 2.578947368421053e-05,
      "loss": 0.1681,
      "step": 185
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.5581791400909424,
      "learning_rate": 2.5131578947368423e-05,
      "loss": 0.1773,
      "step": 190
    },
    {
      "epoch": 10.263157894736842,
      "grad_norm": 0.6555180549621582,
      "learning_rate": 2.4473684210526318e-05,
      "loss": 0.0637,
      "step": 195
    },
    {
      "epoch": 10.526315789473685,
      "grad_norm": 1.0848801136016846,
      "learning_rate": 2.3815789473684212e-05,
      "loss": 0.1363,
      "step": 200
    },
    {
      "epoch": 10.789473684210526,
      "grad_norm": 0.3970225751399994,
      "learning_rate": 2.3157894736842107e-05,
      "loss": 0.0784,
      "step": 205
    },
    {
      "epoch": 11.052631578947368,
      "grad_norm": 2.0135111808776855,
      "learning_rate": 2.25e-05,
      "loss": 0.3391,
      "step": 210
    },
    {
      "epoch": 11.31578947368421,
      "grad_norm": 2.3919425010681152,
      "learning_rate": 2.1842105263157896e-05,
      "loss": 0.0865,
      "step": 215
    },
    {
      "epoch": 11.578947368421053,
      "grad_norm": 1.5701793432235718,
      "learning_rate": 2.118421052631579e-05,
      "loss": 0.2226,
      "step": 220
    },
    {
      "epoch": 11.842105263157894,
      "grad_norm": 0.6800135374069214,
      "learning_rate": 2.0526315789473685e-05,
      "loss": 0.0702,
      "step": 225
    },
    {
      "epoch": 12.105263157894736,
      "grad_norm": 6.435481071472168,
      "learning_rate": 1.986842105263158e-05,
      "loss": 0.1457,
      "step": 230
    },
    {
      "epoch": 12.368421052631579,
      "grad_norm": 0.4558697044849396,
      "learning_rate": 1.9210526315789474e-05,
      "loss": 0.07,
      "step": 235
    },
    {
      "epoch": 12.631578947368421,
      "grad_norm": 0.35098525881767273,
      "learning_rate": 1.855263157894737e-05,
      "loss": 0.0722,
      "step": 240
    },
    {
      "epoch": 12.894736842105264,
      "grad_norm": 13.637933731079102,
      "learning_rate": 1.7894736842105264e-05,
      "loss": 0.1701,
      "step": 245
    },
    {
      "epoch": 13.157894736842104,
      "grad_norm": 30.400209426879883,
      "learning_rate": 1.723684210526316e-05,
      "loss": 0.0991,
      "step": 250
    },
    {
      "epoch": 13.421052631578947,
      "grad_norm": 0.5243188738822937,
      "learning_rate": 1.6578947368421053e-05,
      "loss": 0.3336,
      "step": 255
    },
    {
      "epoch": 13.68421052631579,
      "grad_norm": 0.6692495346069336,
      "learning_rate": 1.5921052631578948e-05,
      "loss": 0.0951,
      "step": 260
    },
    {
      "epoch": 13.947368421052632,
      "grad_norm": 0.45694392919540405,
      "learning_rate": 1.5263157894736842e-05,
      "loss": 0.0598,
      "step": 265
    },
    {
      "epoch": 14.210526315789474,
      "grad_norm": 0.9331479668617249,
      "learning_rate": 1.4605263157894739e-05,
      "loss": 0.265,
      "step": 270
    },
    {
      "epoch": 14.473684210526315,
      "grad_norm": 0.40160542726516724,
      "learning_rate": 1.3947368421052631e-05,
      "loss": 0.1644,
      "step": 275
    },
    {
      "epoch": 14.736842105263158,
      "grad_norm": 0.38934916257858276,
      "learning_rate": 1.3289473684210526e-05,
      "loss": 0.0577,
      "step": 280
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.3913317322731018,
      "learning_rate": 1.2631578947368422e-05,
      "loss": 0.0656,
      "step": 285
    },
    {
      "epoch": 15.263157894736842,
      "grad_norm": 0.317749947309494,
      "learning_rate": 1.1973684210526315e-05,
      "loss": 0.0441,
      "step": 290
    },
    {
      "epoch": 15.526315789473685,
      "grad_norm": 2.1456081867218018,
      "learning_rate": 1.1315789473684212e-05,
      "loss": 0.0586,
      "step": 295
    },
    {
      "epoch": 15.789473684210526,
      "grad_norm": 0.7734329700469971,
      "learning_rate": 1.0657894736842106e-05,
      "loss": 0.2439,
      "step": 300
    },
    {
      "epoch": 16.05263157894737,
      "grad_norm": 0.4043017029762268,
      "learning_rate": 1e-05,
      "loss": 0.1657,
      "step": 305
    },
    {
      "epoch": 16.31578947368421,
      "grad_norm": 0.3189968466758728,
      "learning_rate": 9.342105263157895e-06,
      "loss": 0.0942,
      "step": 310
    },
    {
      "epoch": 16.57894736842105,
      "grad_norm": 2.8793745040893555,
      "learning_rate": 8.68421052631579e-06,
      "loss": 0.1211,
      "step": 315
    },
    {
      "epoch": 16.842105263157894,
      "grad_norm": 3.2684664726257324,
      "learning_rate": 8.026315789473685e-06,
      "loss": 0.0432,
      "step": 320
    },
    {
      "epoch": 17.105263157894736,
      "grad_norm": 0.29015296697616577,
      "learning_rate": 7.3684210526315784e-06,
      "loss": 0.1026,
      "step": 325
    },
    {
      "epoch": 17.36842105263158,
      "grad_norm": 9.337630271911621,
      "learning_rate": 6.710526315789474e-06,
      "loss": 0.0841,
      "step": 330
    },
    {
      "epoch": 17.63157894736842,
      "grad_norm": 0.33837392926216125,
      "learning_rate": 6.0526315789473685e-06,
      "loss": 0.08,
      "step": 335
    },
    {
      "epoch": 17.894736842105264,
      "grad_norm": 8.001954078674316,
      "learning_rate": 5.394736842105263e-06,
      "loss": 0.1835,
      "step": 340
    },
    {
      "epoch": 18.157894736842106,
      "grad_norm": 5.349979400634766,
      "learning_rate": 4.736842105263159e-06,
      "loss": 0.1063,
      "step": 345
    },
    {
      "epoch": 18.42105263157895,
      "grad_norm": 6.384610652923584,
      "learning_rate": 4.078947368421053e-06,
      "loss": 0.1241,
      "step": 350
    },
    {
      "epoch": 18.68421052631579,
      "grad_norm": 0.4049034118652344,
      "learning_rate": 3.421052631578948e-06,
      "loss": 0.0441,
      "step": 355
    },
    {
      "epoch": 18.94736842105263,
      "grad_norm": 0.38923588395118713,
      "learning_rate": 2.763157894736842e-06,
      "loss": 0.1969,
      "step": 360
    },
    {
      "epoch": 19.210526315789473,
      "grad_norm": 1.6115776300430298,
      "learning_rate": 2.105263157894737e-06,
      "loss": 0.2516,
      "step": 365
    },
    {
      "epoch": 19.473684210526315,
      "grad_norm": 1.166119933128357,
      "learning_rate": 1.4473684210526317e-06,
      "loss": 0.0834,
      "step": 370
    },
    {
      "epoch": 19.736842105263158,
      "grad_norm": 0.624567449092865,
      "learning_rate": 7.894736842105264e-07,
      "loss": 0.0486,
      "step": 375
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.30260634422302246,
      "learning_rate": 1.3157894736842107e-07,
      "loss": 0.038,
      "step": 380
    }
  ],
  "logging_steps": 5,
  "max_steps": 380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5389832079360.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}