{
  "best_global_step": 625,
  "best_metric": 1.0413788557052612,
  "best_model_checkpoint": ".../training_output/checkpoint-600",
  "epoch": 10.0,
  "eval_steps": 25,
  "global_step": 640,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15625,
      "grad_norm": 5.998435020446777,
      "learning_rate": 1.40625e-06,
      "loss": 4.4544,
      "step": 10
    },
    {
      "epoch": 0.3125,
      "grad_norm": 2.767270088195801,
      "learning_rate": 2.96875e-06,
      "loss": 4.2972,
      "step": 20
    },
    {
      "epoch": 0.390625,
      "eval_loss": 2.121493101119995,
      "eval_runtime": 4.6524,
      "eval_samples_per_second": 92.855,
      "eval_steps_per_second": 5.803,
      "step": 25
    },
    {
      "epoch": 0.46875,
      "grad_norm": 1.8981692790985107,
      "learning_rate": 4.53125e-06,
      "loss": 4.2058,
      "step": 30
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.126503348350525,
      "learning_rate": 6.093750000000001e-06,
      "loss": 4.1676,
      "step": 40
    },
    {
      "epoch": 0.78125,
      "grad_norm": 1.022411823272705,
      "learning_rate": 7.656250000000001e-06,
      "loss": 4.1589,
      "step": 50
    },
    {
      "epoch": 0.78125,
      "eval_loss": 2.0753986835479736,
      "eval_runtime": 4.6234,
      "eval_samples_per_second": 93.439,
      "eval_steps_per_second": 5.84,
      "step": 50
    },
    {
      "epoch": 0.9375,
      "grad_norm": 1.0090315341949463,
      "learning_rate": 9.21875e-06,
      "loss": 4.1537,
      "step": 60
    },
    {
      "epoch": 1.09375,
      "grad_norm": 1.3605090379714966,
      "learning_rate": 9.913194444444446e-06,
      "loss": 4.1414,
      "step": 70
    },
    {
      "epoch": 1.171875,
      "eval_loss": 2.052276372909546,
      "eval_runtime": 4.9117,
      "eval_samples_per_second": 87.953,
      "eval_steps_per_second": 5.497,
      "step": 75
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.2447760105133057,
      "learning_rate": 9.739583333333334e-06,
      "loss": 4.1134,
      "step": 80
    },
    {
      "epoch": 1.40625,
      "grad_norm": 3.2149338722229004,
      "learning_rate": 9.565972222222222e-06,
      "loss": 4.0985,
      "step": 90
    },
    {
      "epoch": 1.5625,
      "grad_norm": 4.0335187911987305,
      "learning_rate": 9.392361111111112e-06,
      "loss": 4.0703,
      "step": 100
    },
    {
      "epoch": 1.5625,
      "eval_loss": 2.0017964839935303,
      "eval_runtime": 4.8851,
      "eval_samples_per_second": 88.433,
      "eval_steps_per_second": 5.527,
      "step": 100
    },
    {
      "epoch": 1.71875,
      "grad_norm": 4.933378219604492,
      "learning_rate": 9.21875e-06,
      "loss": 4.0612,
      "step": 110
    },
    {
      "epoch": 1.875,
      "grad_norm": 5.589846134185791,
      "learning_rate": 9.045138888888889e-06,
      "loss": 4.0258,
      "step": 120
    },
    {
      "epoch": 1.953125,
      "eval_loss": 1.940491795539856,
      "eval_runtime": 4.949,
      "eval_samples_per_second": 87.29,
      "eval_steps_per_second": 5.456,
      "step": 125
    },
    {
      "epoch": 2.03125,
      "grad_norm": 6.472458839416504,
      "learning_rate": 8.871527777777779e-06,
      "loss": 3.9457,
      "step": 130
    },
    {
      "epoch": 2.1875,
      "grad_norm": 8.088825225830078,
      "learning_rate": 8.697916666666667e-06,
      "loss": 3.8651,
      "step": 140
    },
    {
      "epoch": 2.34375,
      "grad_norm": 10.9951171875,
      "learning_rate": 8.524305555555557e-06,
      "loss": 3.7917,
      "step": 150
    },
    {
      "epoch": 2.34375,
      "eval_loss": 1.874100685119629,
      "eval_runtime": 4.6019,
      "eval_samples_per_second": 93.874,
      "eval_steps_per_second": 5.867,
      "step": 150
    },
    {
      "epoch": 2.5,
      "grad_norm": 12.075238227844238,
      "learning_rate": 8.350694444444445e-06,
      "loss": 3.6884,
      "step": 160
    },
    {
      "epoch": 2.65625,
      "grad_norm": 13.287556648254395,
      "learning_rate": 8.177083333333335e-06,
      "loss": 3.6679,
      "step": 170
    },
    {
      "epoch": 2.734375,
      "eval_loss": 1.779481291770935,
      "eval_runtime": 5.5138,
      "eval_samples_per_second": 78.348,
      "eval_steps_per_second": 4.897,
      "step": 175
    },
    {
      "epoch": 2.8125,
      "grad_norm": 13.939103126525879,
      "learning_rate": 8.003472222222223e-06,
      "loss": 3.6507,
      "step": 180
    },
    {
      "epoch": 2.96875,
      "grad_norm": 14.256585121154785,
      "learning_rate": 7.829861111111112e-06,
      "loss": 3.5815,
      "step": 190
    },
    {
      "epoch": 3.125,
      "grad_norm": 16.98706817626953,
      "learning_rate": 7.656250000000001e-06,
      "loss": 3.3862,
      "step": 200
    },
    {
      "epoch": 3.125,
      "eval_loss": 1.6903235912322998,
      "eval_runtime": 4.8218,
      "eval_samples_per_second": 89.592,
      "eval_steps_per_second": 5.6,
      "step": 200
    },
    {
      "epoch": 3.28125,
      "grad_norm": 19.495948791503906,
      "learning_rate": 7.482638888888889e-06,
      "loss": 3.3398,
      "step": 210
    },
    {
      "epoch": 3.4375,
      "grad_norm": 18.759845733642578,
      "learning_rate": 7.309027777777779e-06,
      "loss": 3.23,
      "step": 220
    },
    {
      "epoch": 3.515625,
      "eval_loss": 1.664551854133606,
      "eval_runtime": 5.0578,
      "eval_samples_per_second": 85.413,
      "eval_steps_per_second": 5.338,
      "step": 225
    },
    {
      "epoch": 3.59375,
      "grad_norm": 19.621652603149414,
      "learning_rate": 7.135416666666667e-06,
      "loss": 3.2616,
      "step": 230
    },
    {
      "epoch": 3.75,
      "grad_norm": 20.33112335205078,
      "learning_rate": 6.961805555555556e-06,
      "loss": 3.1892,
      "step": 240
    },
    {
      "epoch": 3.90625,
      "grad_norm": 21.586963653564453,
      "learning_rate": 6.788194444444444e-06,
      "loss": 3.1402,
      "step": 250
    },
    {
      "epoch": 3.90625,
      "eval_loss": 1.548005223274231,
      "eval_runtime": 4.7767,
      "eval_samples_per_second": 90.438,
      "eval_steps_per_second": 5.652,
      "step": 250
    },
    {
      "epoch": 4.0625,
      "grad_norm": 20.794466018676758,
      "learning_rate": 6.614583333333334e-06,
      "loss": 3.0188,
      "step": 260
    },
    {
      "epoch": 4.21875,
      "grad_norm": 21.908910751342773,
      "learning_rate": 6.4409722222222226e-06,
      "loss": 2.8433,
      "step": 270
    },
    {
      "epoch": 4.296875,
      "eval_loss": 1.49290931224823,
      "eval_runtime": 4.6928,
      "eval_samples_per_second": 92.055,
      "eval_steps_per_second": 5.753,
      "step": 275
    },
    {
      "epoch": 4.375,
      "grad_norm": 24.68117332458496,
      "learning_rate": 6.2673611111111125e-06,
      "loss": 2.8547,
      "step": 280
    },
    {
      "epoch": 4.53125,
      "grad_norm": 24.355791091918945,
      "learning_rate": 6.093750000000001e-06,
      "loss": 2.8257,
      "step": 290
    },
    {
      "epoch": 4.6875,
      "grad_norm": 23.48269271850586,
      "learning_rate": 5.920138888888889e-06,
      "loss": 2.7926,
      "step": 300
    },
    {
      "epoch": 4.6875,
      "eval_loss": 1.4208338260650635,
      "eval_runtime": 4.6392,
      "eval_samples_per_second": 93.12,
      "eval_steps_per_second": 5.82,
      "step": 300
    },
    {
      "epoch": 4.84375,
      "grad_norm": 25.30280303955078,
      "learning_rate": 5.746527777777778e-06,
      "loss": 2.8348,
      "step": 310
    },
    {
      "epoch": 5.0,
      "grad_norm": 26.121788024902344,
      "learning_rate": 5.572916666666667e-06,
      "loss": 2.7752,
      "step": 320
    },
    {
      "epoch": 5.078125,
      "eval_loss": 1.3631044626235962,
      "eval_runtime": 4.6263,
      "eval_samples_per_second": 93.379,
      "eval_steps_per_second": 5.836,
      "step": 325
    },
    {
      "epoch": 5.15625,
      "grad_norm": 23.74436378479004,
      "learning_rate": 5.399305555555556e-06,
      "loss": 2.5492,
      "step": 330
    },
    {
      "epoch": 5.3125,
      "grad_norm": 25.19687843322754,
      "learning_rate": 5.2256944444444445e-06,
      "loss": 2.5594,
      "step": 340
    },
    {
      "epoch": 5.46875,
      "grad_norm": 24.556848526000977,
      "learning_rate": 5.0520833333333344e-06,
      "loss": 2.5388,
      "step": 350
    },
    {
      "epoch": 5.46875,
      "eval_loss": 1.3000701665878296,
      "eval_runtime": 4.6145,
      "eval_samples_per_second": 93.618,
      "eval_steps_per_second": 5.851,
      "step": 350
    },
    {
      "epoch": 5.625,
      "grad_norm": 25.34222984313965,
      "learning_rate": 4.878472222222223e-06,
      "loss": 2.4945,
      "step": 360
    },
    {
      "epoch": 5.78125,
      "grad_norm": 26.028766632080078,
      "learning_rate": 4.704861111111112e-06,
      "loss": 2.4622,
      "step": 370
    },
    {
      "epoch": 5.859375,
      "eval_loss": 1.2737797498703003,
      "eval_runtime": 4.6455,
      "eval_samples_per_second": 92.994,
      "eval_steps_per_second": 5.812,
      "step": 375
    },
    {
      "epoch": 5.9375,
      "grad_norm": 25.235429763793945,
      "learning_rate": 4.53125e-06,
      "loss": 2.4739,
      "step": 380
    },
    {
      "epoch": 6.09375,
      "grad_norm": 24.631134033203125,
      "learning_rate": 4.357638888888889e-06,
      "loss": 2.3813,
      "step": 390
    },
    {
      "epoch": 6.25,
      "grad_norm": 27.127145767211914,
      "learning_rate": 4.184027777777778e-06,
      "loss": 2.25,
      "step": 400
    },
    {
      "epoch": 6.25,
      "eval_loss": 1.220424771308899,
      "eval_runtime": 4.705,
      "eval_samples_per_second": 91.817,
      "eval_steps_per_second": 5.739,
      "step": 400
    },
    {
      "epoch": 6.40625,
      "grad_norm": 24.72281837463379,
      "learning_rate": 4.010416666666667e-06,
      "loss": 2.3371,
      "step": 410
    },
    {
      "epoch": 6.5625,
      "grad_norm": 25.333667755126953,
      "learning_rate": 3.836805555555556e-06,
      "loss": 2.2495,
      "step": 420
    },
    {
      "epoch": 6.640625,
      "eval_loss": 1.1890788078308105,
      "eval_runtime": 4.9032,
      "eval_samples_per_second": 88.105,
      "eval_steps_per_second": 5.507,
      "step": 425
    },
    {
      "epoch": 6.71875,
      "grad_norm": 25.572025299072266,
      "learning_rate": 3.6631944444444446e-06,
      "loss": 2.2528,
      "step": 430
    },
    {
      "epoch": 6.875,
      "grad_norm": 26.69953155517578,
      "learning_rate": 3.4895833333333333e-06,
      "loss": 2.2632,
      "step": 440
    },
    {
      "epoch": 7.03125,
      "grad_norm": 25.635225296020508,
      "learning_rate": 3.3159722222222224e-06,
      "loss": 2.1912,
      "step": 450
    },
    {
      "epoch": 7.03125,
      "eval_loss": 1.143013596534729,
      "eval_runtime": 4.7149,
      "eval_samples_per_second": 91.624,
      "eval_steps_per_second": 5.727,
      "step": 450
    },
    {
      "epoch": 7.1875,
      "grad_norm": 26.367721557617188,
      "learning_rate": 3.1423611111111115e-06,
      "loss": 2.0829,
      "step": 460
    },
    {
      "epoch": 7.34375,
      "grad_norm": 26.74110984802246,
      "learning_rate": 2.96875e-06,
      "loss": 2.0772,
      "step": 470
    },
    {
      "epoch": 7.421875,
      "eval_loss": 1.130937933921814,
      "eval_runtime": 4.8054,
      "eval_samples_per_second": 89.9,
      "eval_steps_per_second": 5.619,
      "step": 475
    },
    {
      "epoch": 7.5,
      "grad_norm": 25.550865173339844,
      "learning_rate": 2.7951388888888893e-06,
      "loss": 2.0683,
      "step": 480
    },
    {
      "epoch": 7.65625,
      "grad_norm": 26.189870834350586,
      "learning_rate": 2.621527777777778e-06,
      "loss": 2.1433,
      "step": 490
    },
    {
      "epoch": 7.8125,
      "grad_norm": 25.142404556274414,
      "learning_rate": 2.4479166666666666e-06,
      "loss": 2.1272,
      "step": 500
    },
    {
      "epoch": 7.8125,
      "eval_loss": 1.1044453382492065,
      "eval_runtime": 4.7489,
      "eval_samples_per_second": 90.969,
      "eval_steps_per_second": 5.686,
      "step": 500
    },
    {
      "epoch": 7.96875,
      "grad_norm": 26.091402053833008,
      "learning_rate": 2.2743055555555557e-06,
      "loss": 2.0564,
      "step": 510
    },
    {
      "epoch": 8.125,
      "grad_norm": 25.8810977935791,
      "learning_rate": 2.1006944444444448e-06,
      "loss": 1.9743,
      "step": 520
    },
    {
      "epoch": 8.203125,
      "eval_loss": 1.0846548080444336,
      "eval_runtime": 4.9525,
      "eval_samples_per_second": 87.229,
      "eval_steps_per_second": 5.452,
      "step": 525
    },
    {
      "epoch": 8.28125,
      "grad_norm": 25.896873474121094,
      "learning_rate": 1.9270833333333334e-06,
      "loss": 1.9884,
      "step": 530
    },
    {
      "epoch": 8.4375,
      "grad_norm": 27.100982666015625,
      "learning_rate": 1.7534722222222223e-06,
      "loss": 1.9915,
      "step": 540
    },
    {
      "epoch": 8.59375,
      "grad_norm": 27.147930145263672,
      "learning_rate": 1.5798611111111112e-06,
      "loss": 2.007,
      "step": 550
    },
    {
      "epoch": 8.59375,
      "eval_loss": 1.0685715675354004,
      "eval_runtime": 4.6413,
      "eval_samples_per_second": 93.076,
      "eval_steps_per_second": 5.817,
      "step": 550
    },
    {
      "epoch": 8.75,
      "grad_norm": 26.25347137451172,
      "learning_rate": 1.40625e-06,
      "loss": 1.9727,
      "step": 560
    },
    {
      "epoch": 8.90625,
      "grad_norm": 26.688894271850586,
      "learning_rate": 1.232638888888889e-06,
      "loss": 1.9395,
      "step": 570
    },
    {
      "epoch": 8.984375,
      "eval_loss": 1.061125636100769,
      "eval_runtime": 4.6582,
      "eval_samples_per_second": 92.74,
      "eval_steps_per_second": 5.796,
      "step": 575
    },
    {
      "epoch": 9.0625,
      "grad_norm": 27.173229217529297,
      "learning_rate": 1.0590277777777778e-06,
      "loss": 1.9538,
      "step": 580
    },
    {
      "epoch": 9.21875,
      "grad_norm": 25.862943649291992,
      "learning_rate": 8.854166666666668e-07,
      "loss": 1.9069,
      "step": 590
    },
    {
      "epoch": 9.375,
      "grad_norm": 25.96297264099121,
      "learning_rate": 7.118055555555556e-07,
      "loss": 1.8768,
      "step": 600
    },
    {
      "epoch": 9.375,
      "eval_loss": 1.0467159748077393,
      "eval_runtime": 4.6503,
      "eval_samples_per_second": 92.896,
      "eval_steps_per_second": 5.806,
      "step": 600
    },
    {
      "epoch": 9.53125,
      "grad_norm": 26.35511589050293,
      "learning_rate": 5.381944444444445e-07,
      "loss": 1.8875,
      "step": 610
    },
    {
      "epoch": 9.6875,
      "grad_norm": 26.81864356994629,
      "learning_rate": 3.6458333333333337e-07,
      "loss": 1.8867,
      "step": 620
    },
    {
      "epoch": 9.765625,
      "eval_loss": 1.0413788557052612,
      "eval_runtime": 4.6336,
      "eval_samples_per_second": 93.232,
      "eval_steps_per_second": 5.827,
      "step": 625
    },
    {
      "epoch": 9.84375,
      "grad_norm": 24.75679588317871,
      "learning_rate": 1.9097222222222225e-07,
      "loss": 1.846,
      "step": 630
    },
    {
      "epoch": 10.0,
      "grad_norm": 24.3472843170166,
      "learning_rate": 1.736111111111111e-08,
      "loss": 1.9162,
      "step": 640
    }
  ],
  "logging_steps": 10,
  "max_steps": 640,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}