trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.8,
  "eval_steps": 20,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "grad_norm": 0.026561671867966652,
      "learning_rate": 0.001894736842105263,
      "loss": 2.4537,
      "step": 2
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.1137104481458664,
      "learning_rate": 0.0016842105263157893,
      "loss": 2.4379,
      "step": 4
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.24436840415000916,
      "learning_rate": 0.0014736842105263158,
      "loss": 2.3347,
      "step": 6
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.30128273367881775,
      "learning_rate": 0.0012631578947368421,
      "loss": 2.1954,
      "step": 8
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.4249122440814972,
      "learning_rate": 0.0010526315789473684,
      "loss": 2.0775,
      "step": 10
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4739112854003906,
      "learning_rate": 0.0008421052631578947,
      "loss": 1.9817,
      "step": 12
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.44454509019851685,
      "learning_rate": 0.0006315789473684211,
      "loss": 1.9018,
      "step": 14
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.3762497007846832,
      "learning_rate": 0.00042105263157894734,
      "loss": 1.8925,
      "step": 16
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.34596943855285645,
      "learning_rate": 0.00021052631578947367,
      "loss": 1.8335,
      "step": 18
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.3068355619907379,
      "learning_rate": 0.0,
      "loss": 1.8221,
      "step": 20
    },
    {
      "epoch": 6.8,
      "eval_loss": 1.905713677406311,
      "eval_runtime": 1.8411,
      "eval_samples_per_second": 43.452,
      "eval_steps_per_second": 5.432,
      "step": 20
    },
    {
      "epoch": 6.8,
      "step": 20,
      "total_flos": 2.448849371136e+16,
      "train_loss": 2.093080127239227,
      "train_runtime": 118.9098,
      "train_samples_per_second": 53.822,
      "train_steps_per_second": 0.168
    },
    {
      "epoch": 6.8,
      "eval_loss": 1.905713677406311,
      "eval_runtime": 1.8743,
      "eval_samples_per_second": 42.682,
      "eval_steps_per_second": 5.335,
      "step": 20
    },
    {
      "epoch": 6.8,
      "eval_loss": 1.8781611919403076,
      "eval_runtime": 1.8586,
      "eval_samples_per_second": 43.043,
      "eval_steps_per_second": 5.38,
      "step": 20
    }
  ],
  "logging_steps": 2,
  "max_steps": 20,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.448849371136e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}