trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 87,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 26.318459725486353,
      "learning_rate": 5e-06,
      "loss": 1.1321,
      "step": 10
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 4.606141465604429,
      "learning_rate": 5e-06,
      "loss": 1.0305,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.966795027256012,
      "eval_runtime": 22.0564,
      "eval_samples_per_second": 35.454,
      "eval_steps_per_second": 0.589,
      "step": 29
    },
    {
      "epoch": 1.0344827586206897,
      "grad_norm": 3.982194680899152,
      "learning_rate": 5e-06,
      "loss": 0.992,
      "step": 30
    },
    {
      "epoch": 1.3793103448275863,
      "grad_norm": 1.544390231168174,
      "learning_rate": 5e-06,
      "loss": 0.9331,
      "step": 40
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 2.71003049663633,
      "learning_rate": 5e-06,
      "loss": 0.9029,
      "step": 50
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8883161544799805,
      "eval_runtime": 21.2114,
      "eval_samples_per_second": 36.867,
      "eval_steps_per_second": 0.613,
      "step": 58
    },
    {
      "epoch": 2.0689655172413794,
      "grad_norm": 1.4349517314666953,
      "learning_rate": 5e-06,
      "loss": 0.879,
      "step": 60
    },
    {
      "epoch": 2.413793103448276,
      "grad_norm": 1.3190035539077891,
      "learning_rate": 5e-06,
      "loss": 0.8358,
      "step": 70
    },
    {
      "epoch": 2.7586206896551726,
      "grad_norm": 1.2288764806105619,
      "learning_rate": 5e-06,
      "loss": 0.826,
      "step": 80
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.860611081123352,
      "eval_runtime": 19.6144,
      "eval_samples_per_second": 39.869,
      "eval_steps_per_second": 0.663,
      "step": 87
    },
    {
      "epoch": 3.0,
      "step": 87,
      "total_flos": 145518860697600.0,
      "train_loss": 0.9317608592153965,
      "train_runtime": 3194.413,
      "train_samples_per_second": 13.94,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 10,
  "max_steps": 87,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 145518860697600.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}