trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 105,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 3.9646038631444687,
      "learning_rate": 5e-06,
      "loss": 0.996,
      "step": 10
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 4.918642638371639,
      "learning_rate": 5e-06,
      "loss": 0.8944,
      "step": 20
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.3108211729045733,
      "learning_rate": 5e-06,
      "loss": 0.8572,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8312200307846069,
      "eval_runtime": 110.2707,
      "eval_samples_per_second": 8.561,
      "eval_steps_per_second": 0.272,
      "step": 35
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 1.3985649962465836,
      "learning_rate": 5e-06,
      "loss": 0.8205,
      "step": 40
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.990368274403578,
      "learning_rate": 5e-06,
      "loss": 0.792,
      "step": 50
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.9663231906613171,
      "learning_rate": 5e-06,
      "loss": 0.7814,
      "step": 60
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.6499452853977579,
      "learning_rate": 5e-06,
      "loss": 0.7768,
      "step": 70
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8017017245292664,
      "eval_runtime": 110.3066,
      "eval_samples_per_second": 8.558,
      "eval_steps_per_second": 0.272,
      "step": 70
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 1.1076482815278739,
      "learning_rate": 5e-06,
      "loss": 0.723,
      "step": 80
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.6444807327009624,
      "learning_rate": 5e-06,
      "loss": 0.7163,
      "step": 90
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6105018774007208,
      "learning_rate": 5e-06,
      "loss": 0.7145,
      "step": 100
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7955389618873596,
      "eval_runtime": 120.7051,
      "eval_samples_per_second": 7.821,
      "eval_steps_per_second": 0.249,
      "step": 105
    },
    {
      "epoch": 3.0,
      "step": 105,
      "total_flos": 175669531115520.0,
      "train_loss": 0.23945580210004533,
      "train_runtime": 5809.6465,
      "train_samples_per_second": 9.253,
      "train_steps_per_second": 0.018
    }
  ],
  "logging_steps": 10,
  "max_steps": 105,
  "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": 175669531115520.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}