{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 1000,
  "global_step": 10896,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.27540622418066646,
      "grad_norm": 13.115499496459961,
      "learning_rate": 1.5259938837920492e-05,
      "loss": 1.0427,
      "step": 500
    },
    {
      "epoch": 0.5508124483613329,
      "grad_norm": 10.038908958435059,
      "learning_rate": 1.932630345635618e-05,
      "loss": 0.3874,
      "step": 1000
    },
    {
      "epoch": 0.5508124483613329,
      "eval_accuracy": 0.9134971070852699,
      "eval_f1": 0.875406283856988,
      "eval_loss": 0.36551177501678467,
      "eval_precision": 0.8548972188633616,
      "eval_recall": 0.8969235648588646,
      "eval_runtime": 5.0106,
      "eval_samples_per_second": 544.045,
      "eval_steps_per_second": 34.128,
      "step": 1000
    },
    {
      "epoch": 0.8262186725419994,
      "grad_norm": 20.2281436920166,
      "learning_rate": 1.8349931653973834e-05,
      "loss": 0.3175,
      "step": 1500
    },
    {
      "epoch": 1.1013494904984853,
      "grad_norm": 21.382644653320312,
      "learning_rate": 1.737355985159149e-05,
      "loss": 0.2691,
      "step": 2000
    },
    {
      "epoch": 1.1013494904984853,
      "eval_accuracy": 0.9356919282936546,
      "eval_f1": 0.9239401496259352,
      "eval_loss": 0.2406017929315567,
      "eval_precision": 0.9083665338645418,
      "eval_recall": 0.9400570884871551,
      "eval_runtime": 4.8938,
      "eval_samples_per_second": 557.03,
      "eval_steps_per_second": 34.942,
      "step": 2000
    },
    {
      "epoch": 1.3767557146791518,
      "grad_norm": 0.7748643159866333,
      "learning_rate": 1.639718804920914e-05,
      "loss": 0.2279,
      "step": 2500
    },
    {
      "epoch": 1.6521619388598183,
      "grad_norm": 4.197641372680664,
      "learning_rate": 1.5420816246826794e-05,
      "loss": 0.2276,
      "step": 3000
    },
    {
      "epoch": 1.6521619388598183,
      "eval_accuracy": 0.938347718865598,
      "eval_f1": 0.9255401834291933,
      "eval_loss": 0.2395576536655426,
      "eval_precision": 0.9076219512195122,
      "eval_recall": 0.9441801458928005,
      "eval_runtime": 4.9765,
      "eval_samples_per_second": 547.778,
      "eval_steps_per_second": 34.362,
      "step": 3000
    },
    {
      "epoch": 1.9275681630404846,
      "grad_norm": 25.588197708129883,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 0.2075,
      "step": 3500
    },
    {
      "epoch": 2.2026989809969706,
      "grad_norm": 22.494050979614258,
      "learning_rate": 1.3468072642062098e-05,
      "loss": 0.1709,
      "step": 4000
    },
    {
      "epoch": 2.2026989809969706,
      "eval_accuracy": 0.9443232476524709,
      "eval_f1": 0.9338362407669338,
      "eval_loss": 0.23579855263233185,
      "eval_precision": 0.9255451713395638,
      "eval_recall": 0.9422771963209642,
      "eval_runtime": 4.9528,
      "eval_samples_per_second": 550.395,
      "eval_steps_per_second": 34.526,
      "step": 4000
    },
    {
      "epoch": 2.478105205177637,
      "grad_norm": 10.178144454956055,
      "learning_rate": 1.249170083967975e-05,
      "loss": 0.1547,
      "step": 4500
    },
    {
      "epoch": 2.7535114293583036,
      "grad_norm": 50.71146774291992,
      "learning_rate": 1.1515329037297404e-05,
      "loss": 0.1613,
      "step": 5000
    },
    {
      "epoch": 2.7535114293583036,
      "eval_accuracy": 0.9422365550602295,
      "eval_f1": 0.9349874371859297,
      "eval_loss": 0.283699095249176,
      "eval_precision": 0.9259720062208399,
      "eval_recall": 0.9441801458928005,
      "eval_runtime": 5.001,
      "eval_samples_per_second": 545.09,
      "eval_steps_per_second": 34.193,
      "step": 5000
    },
    {
      "epoch": 3.028642247314789,
      "grad_norm": 2.2809739112854004,
      "learning_rate": 1.0538957234915056e-05,
      "loss": 0.1459,
      "step": 5500
    },
    {
      "epoch": 3.304048471495456,
      "grad_norm": 0.0738934576511383,
      "learning_rate": 9.56258543253271e-06,
      "loss": 0.1086,
      "step": 6000
    },
    {
      "epoch": 3.304048471495456,
      "eval_accuracy": 0.9451768946220241,
      "eval_f1": 0.935539253941002,
      "eval_loss": 0.26784300804138184,
      "eval_precision": 0.921020282728949,
      "eval_recall": 0.950523311132255,
      "eval_runtime": 4.848,
      "eval_samples_per_second": 562.288,
      "eval_steps_per_second": 35.272,
      "step": 6000
    },
    {
      "epoch": 3.579454695676122,
      "grad_norm": 5.788504123687744,
      "learning_rate": 8.586213630150362e-06,
      "loss": 0.1089,
      "step": 6500
    },
    {
      "epoch": 3.854860919856789,
      "grad_norm": 15.219040870666504,
      "learning_rate": 7.609841827768014e-06,
      "loss": 0.1072,
      "step": 7000
    },
    {
      "epoch": 3.854860919856789,
      "eval_accuracy": 0.9431850516930665,
      "eval_f1": 0.9340901923683381,
      "eval_loss": 0.3070509731769562,
      "eval_precision": 0.9288178112260896,
      "eval_recall": 0.9394227719632097,
      "eval_runtime": 4.9423,
      "eval_samples_per_second": 551.561,
      "eval_steps_per_second": 34.599,
      "step": 7000
    },
    {
      "epoch": 4.129991737813275,
      "grad_norm": 0.6474624276161194,
      "learning_rate": 6.633470025385668e-06,
      "loss": 0.0953,
      "step": 7500
    },
    {
      "epoch": 4.405397961993941,
      "grad_norm": 1.6168689727783203,
      "learning_rate": 5.65709822300332e-06,
      "loss": 0.072,
      "step": 8000
    },
    {
      "epoch": 4.405397961993941,
      "eval_accuracy": 0.9465996395712796,
      "eval_f1": 0.9409547738693468,
      "eval_loss": 0.31703799962997437,
      "eval_precision": 0.9318818040435459,
      "eval_recall": 0.9502061528702823,
      "eval_runtime": 4.9657,
      "eval_samples_per_second": 548.971,
      "eval_steps_per_second": 34.437,
      "step": 8000
    },
    {
      "epoch": 4.6808041861746075,
      "grad_norm": 0.7237133979797363,
      "learning_rate": 4.680726420620973e-06,
      "loss": 0.0689,
      "step": 8500
    },
    {
      "epoch": 4.956210410355274,
      "grad_norm": 0.029142044484615326,
      "learning_rate": 3.7043546182386252e-06,
      "loss": 0.0846,
      "step": 9000
    },
    {
      "epoch": 4.956210410355274,
      "eval_accuracy": 0.9466944892345632,
      "eval_f1": 0.9408805031446541,
      "eval_loss": 0.3410860598087311,
      "eval_precision": 0.9329591518553165,
      "eval_recall": 0.9489375198223914,
      "eval_runtime": 5.1283,
      "eval_samples_per_second": 531.558,
      "eval_steps_per_second": 33.344,
      "step": 9000
    },
    {
      "epoch": 5.23134122831176,
      "grad_norm": 12.02783489227295,
      "learning_rate": 2.7279828158562783e-06,
      "loss": 0.0496,
      "step": 9500
    },
    {
      "epoch": 5.5067474524924265,
      "grad_norm": 0.016711721196770668,
      "learning_rate": 1.7516110134739312e-06,
      "loss": 0.0575,
      "step": 10000
    },
    {
      "epoch": 5.5067474524924265,
      "eval_accuracy": 0.9502039267760599,
      "eval_f1": 0.9438166980539863,
      "eval_loss": 0.33306461572647095,
      "eval_precision": 0.9341410375893134,
      "eval_recall": 0.9536948937519822,
      "eval_runtime": 5.6128,
      "eval_samples_per_second": 485.677,
      "eval_steps_per_second": 30.466,
      "step": 10000
    },
    {
      "epoch": 5.782153676673093,
      "grad_norm": 0.039173971861600876,
      "learning_rate": 7.752392110915838e-07,
      "loss": 0.0528,
      "step": 10500
    }
  ],
  "logging_steps": 500,
  "max_steps": 10896,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2832424021242060.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}