trainer_state.json

{
  "best_global_step": 3222,
  "best_metric": 0.9287502085766728,
  "best_model_checkpoint": "./results/checkpoint-3222",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 4296,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4655493482309125,
      "grad_norm": 27.341506958007812,
      "learning_rate": 9.535381750465549e-07,
      "loss": 0.5434,
      "step": 500
    },
    {
      "epoch": 0.931098696461825,
      "grad_norm": 20.316381454467773,
      "learning_rate": 9.069832402234636e-07,
      "loss": 0.32,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.900884365092608,
      "eval_loss": 0.270067423582077,
      "eval_runtime": 119.0728,
      "eval_samples_per_second": 50.331,
      "eval_steps_per_second": 3.149,
      "step": 1074
    },
    {
      "epoch": 1.3966480446927374,
      "grad_norm": 59.5791015625,
      "learning_rate": 8.604283054003724e-07,
      "loss": 0.2702,
      "step": 1500
    },
    {
      "epoch": 1.86219739292365,
      "grad_norm": 35.37834548950195,
      "learning_rate": 8.138733705772811e-07,
      "loss": 0.2547,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9229100617386952,
      "eval_loss": 0.21657520532608032,
      "eval_runtime": 118.9079,
      "eval_samples_per_second": 50.4,
      "eval_steps_per_second": 3.154,
      "step": 2148
    },
    {
      "epoch": 2.3277467411545625,
      "grad_norm": 51.73606491088867,
      "learning_rate": 7.673184357541899e-07,
      "loss": 0.2446,
      "step": 2500
    },
    {
      "epoch": 2.793296089385475,
      "grad_norm": 39.03384017944336,
      "learning_rate": 7.207635009310986e-07,
      "loss": 0.2238,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9287502085766728,
      "eval_loss": 0.21104617416858673,
      "eval_runtime": 118.9557,
      "eval_samples_per_second": 50.38,
      "eval_steps_per_second": 3.152,
      "step": 3222
    },
    {
      "epoch": 3.2588454376163876,
      "grad_norm": 18.250083923339844,
      "learning_rate": 6.742085661080075e-07,
      "loss": 0.226,
      "step": 3500
    },
    {
      "epoch": 3.7243947858473,
      "grad_norm": 14.896623611450195,
      "learning_rate": 6.276536312849162e-07,
      "loss": 0.2153,
      "step": 4000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9237443684298348,
      "eval_loss": 0.23883432149887085,
      "eval_runtime": 118.9658,
      "eval_samples_per_second": 50.376,
      "eval_steps_per_second": 3.152,
      "step": 4296
    }
  ],
  "logging_steps": 500,
  "max_steps": 10740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 6.401995693518029e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}