{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 10560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.3427704572677612,
      "learning_rate": 4.75e-05,
      "loss": 0.7018,
      "step": 528
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9115075027327835,
      "eval_f1": 0.5138413685847589,
      "eval_loss": 0.33528366684913635,
      "eval_precision": 0.5528781793842035,
      "eval_recall": 0.4799535153980244,
      "eval_runtime": 4.4519,
      "eval_samples_per_second": 210.021,
      "eval_steps_per_second": 3.369,
      "step": 528
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.260859727859497,
      "learning_rate": 4.5e-05,
      "loss": 0.2639,
      "step": 1056
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9412203120341847,
      "eval_f1": 0.7251732101616628,
      "eval_loss": 0.1912250965833664,
      "eval_precision": 0.6493566176470589,
      "eval_recall": 0.821034282393957,
      "eval_runtime": 4.5644,
      "eval_samples_per_second": 204.846,
      "eval_steps_per_second": 3.286,
      "step": 1056
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1796255111694336,
      "learning_rate": 4.25e-05,
      "loss": 0.1862,
      "step": 1584
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9465865050183842,
      "eval_f1": 0.7531402204562931,
      "eval_loss": 0.1671685427427292,
      "eval_precision": 0.6738532110091743,
      "eval_recall": 0.8535735037768739,
      "eval_runtime": 4.5056,
      "eval_samples_per_second": 207.52,
      "eval_steps_per_second": 3.329,
      "step": 1584
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.1285958290100098,
      "learning_rate": 4e-05,
      "loss": 0.1612,
      "step": 2112
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9539401768856206,
      "eval_f1": 0.7823765020026703,
      "eval_loss": 0.14461010694503784,
      "eval_precision": 0.7238142292490118,
      "eval_recall": 0.8512492736780941,
      "eval_runtime": 4.5298,
      "eval_samples_per_second": 206.41,
      "eval_steps_per_second": 3.311,
      "step": 2112
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1812392473220825,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.1439,
      "step": 2640
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9545364205505317,
      "eval_f1": 0.7862656374767102,
      "eval_loss": 0.13903872668743134,
      "eval_precision": 0.7254420432220039,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.6125,
      "eval_samples_per_second": 202.712,
      "eval_steps_per_second": 3.252,
      "step": 2640
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.9920209646224976,
      "learning_rate": 3.5e-05,
      "loss": 0.1358,
      "step": 3168
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9551326642154427,
      "eval_f1": 0.7892923403127484,
      "eval_loss": 0.1392282098531723,
      "eval_precision": 0.7256335282651072,
      "eval_recall": 0.8651946542707728,
      "eval_runtime": 4.8436,
      "eval_samples_per_second": 193.037,
      "eval_steps_per_second": 3.097,
      "step": 3168
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.6399483680725098,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.129,
      "step": 3696
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9561264036569611,
      "eval_f1": 0.7918540068764879,
      "eval_loss": 0.13837336003780365,
      "eval_precision": 0.7266990291262136,
      "eval_recall": 0.8698431144683324,
      "eval_runtime": 4.5425,
      "eval_samples_per_second": 205.834,
      "eval_steps_per_second": 3.302,
      "step": 3696
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.901687741279602,
      "learning_rate": 3e-05,
      "loss": 0.1228,
      "step": 4224
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.957517638875087,
      "eval_f1": 0.7969124301304232,
      "eval_loss": 0.13390584290027618,
      "eval_precision": 0.7352652259332023,
      "eval_recall": 0.8698431144683324,
      "eval_runtime": 4.5586,
      "eval_samples_per_second": 205.105,
      "eval_steps_per_second": 3.29,
      "step": 4224
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.3437026739120483,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.1168,
      "step": 4752
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9577163867633907,
      "eval_f1": 0.7960010807889759,
      "eval_loss": 0.13209262490272522,
      "eval_precision": 0.7439393939393939,
      "eval_recall": 0.8558977338756537,
      "eval_runtime": 4.5415,
      "eval_samples_per_second": 205.88,
      "eval_steps_per_second": 3.303,
      "step": 4752
    },
    {
      "epoch": 10.0,
      "grad_norm": 3.851469039916992,
      "learning_rate": 2.5e-05,
      "loss": 0.1146,
      "step": 5280
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.958113882539998,
      "eval_f1": 0.7973009446693656,
      "eval_loss": 0.12995323538780212,
      "eval_precision": 0.7444556451612904,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.5491,
      "eval_samples_per_second": 205.534,
      "eval_steps_per_second": 3.297,
      "step": 5280
    },
    {
      "epoch": 11.0,
      "grad_norm": 2.413081645965576,
      "learning_rate": 2.25e-05,
      "loss": 0.1105,
      "step": 5808
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9571201430984796,
      "eval_f1": 0.7947830715996806,
      "eval_loss": 0.13270916044712067,
      "eval_precision": 0.7333005893909627,
      "eval_recall": 0.8675188843695526,
      "eval_runtime": 4.536,
      "eval_samples_per_second": 206.127,
      "eval_steps_per_second": 3.307,
      "step": 5808
    },
    {
      "epoch": 12.0,
      "grad_norm": 1.1331512928009033,
      "learning_rate": 2e-05,
      "loss": 0.1083,
      "step": 6336
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9569213952101759,
      "eval_f1": 0.794345158708989,
      "eval_loss": 0.1333465278148651,
      "eval_precision": 0.7342209072978304,
      "eval_recall": 0.8651946542707728,
      "eval_runtime": 4.5053,
      "eval_samples_per_second": 207.533,
      "eval_steps_per_second": 3.329,
      "step": 6336
    },
    {
      "epoch": 13.0,
      "grad_norm": 1.8731575012207031,
      "learning_rate": 1.75e-05,
      "loss": 0.106,
      "step": 6864
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9591076219815164,
      "eval_f1": 0.7998916869753587,
      "eval_loss": 0.12651574611663818,
      "eval_precision": 0.7489858012170385,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.4726,
      "eval_samples_per_second": 209.052,
      "eval_steps_per_second": 3.354,
      "step": 6864
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.8700233697891235,
      "learning_rate": 1.5e-05,
      "loss": 0.1032,
      "step": 7392
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9589088740932128,
      "eval_f1": 0.7973009446693656,
      "eval_loss": 0.12690864503383636,
      "eval_precision": 0.7444556451612904,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.5513,
      "eval_samples_per_second": 205.436,
      "eval_steps_per_second": 3.296,
      "step": 7392
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.2827842235565186,
      "learning_rate": 1.25e-05,
      "loss": 0.1023,
      "step": 7920
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.9585113783166054,
      "eval_f1": 0.7998922704012928,
      "eval_loss": 0.12912563979625702,
      "eval_precision": 0.7454819277108434,
      "eval_recall": 0.862870424171993,
      "eval_runtime": 4.5324,
      "eval_samples_per_second": 206.293,
      "eval_steps_per_second": 3.31,
      "step": 7920
    },
    {
      "epoch": 16.0,
      "grad_norm": 1.694359302520752,
      "learning_rate": 1e-05,
      "loss": 0.1014,
      "step": 8448
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.957517638875087,
      "eval_f1": 0.7947269303201507,
      "eval_loss": 0.12707427144050598,
      "eval_precision": 0.7399799599198397,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.562,
      "eval_samples_per_second": 204.955,
      "eval_steps_per_second": 3.288,
      "step": 8448
    },
    {
      "epoch": 17.0,
      "grad_norm": 1.139172911643982,
      "learning_rate": 7.5e-06,
      "loss": 0.1002,
      "step": 8976
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9589088740932128,
      "eval_f1": 0.8041789445486203,
      "eval_loss": 0.12810933589935303,
      "eval_precision": 0.7460238568588469,
      "eval_recall": 0.8721673445671121,
      "eval_runtime": 4.5254,
      "eval_samples_per_second": 206.614,
      "eval_steps_per_second": 3.315,
      "step": 8976
    },
    {
      "epoch": 18.0,
      "grad_norm": 1.6876777410507202,
      "learning_rate": 5e-06,
      "loss": 0.0986,
      "step": 9504
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.9573188909867832,
      "eval_f1": 0.8016021361815754,
      "eval_loss": 0.13038571178913116,
      "eval_precision": 0.741600790513834,
      "eval_recall": 0.8721673445671121,
      "eval_runtime": 4.5245,
      "eval_samples_per_second": 206.653,
      "eval_steps_per_second": 3.315,
      "step": 9504
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.5088372230529785,
      "learning_rate": 2.5e-06,
      "loss": 0.0978,
      "step": 10032
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9589088740932128,
      "eval_f1": 0.8046473925965956,
      "eval_loss": 0.1270894557237625,
      "eval_precision": 0.752020202020202,
      "eval_recall": 0.8651946542707728,
      "eval_runtime": 4.5248,
      "eval_samples_per_second": 206.64,
      "eval_steps_per_second": 3.315,
      "step": 10032
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.988950490951538,
      "learning_rate": 0.0,
      "loss": 0.0984,
      "step": 10560
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.9579151346516943,
      "eval_f1": 0.8007549204637368,
      "eval_loss": 0.12812790274620056,
      "eval_precision": 0.7469818913480886,
      "eval_recall": 0.862870424171993,
      "eval_runtime": 4.5277,
      "eval_samples_per_second": 206.507,
      "eval_steps_per_second": 3.313,
      "step": 10560
    },
    {
      "epoch": 20.0,
      "step": 10560,
      "total_flos": 4552961808488766.0,
      "train_loss": 0.1551312410470211,
      "train_runtime": 1231.1625,
      "train_samples_per_second": 137.057,
      "train_steps_per_second": 8.577
    }
  ],
  "logging_steps": 500,
  "max_steps": 10560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 4552961808488766.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}