trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.4,
  "eval_steps": 500,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4,
      "grad_norm": 1.2139655351638794,
      "learning_rate": 0.0002,
      "loss": 1.8904,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8287366032600403,
      "eval_runtime": 0.886,
      "eval_samples_per_second": 3.386,
      "eval_steps_per_second": 1.129,
      "step": 3
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.7967442274093628,
      "learning_rate": 0.00011736481776669306,
      "loss": 1.4592,
      "step": 5
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.7421586513519287,
      "eval_runtime": 0.8866,
      "eval_samples_per_second": 3.384,
      "eval_steps_per_second": 1.128,
      "step": 6
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7162426114082336,
      "eval_runtime": 0.8865,
      "eval_samples_per_second": 3.384,
      "eval_steps_per_second": 1.128,
      "step": 9
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.5492648482322693,
      "learning_rate": 0.0,
      "loss": 1.0477,
      "step": 10
    },
    {
      "epoch": 3.4,
      "eval_loss": 0.7149809002876282,
      "eval_runtime": 0.9232,
      "eval_samples_per_second": 3.25,
      "eval_steps_per_second": 1.083,
      "step": 10
    },
    {
      "epoch": 3.4,
      "step": 10,
      "total_flos": 5976600671682560.0,
      "train_loss": 1.2965626955032348,
      "train_runtime": 74.9653,
      "train_samples_per_second": 1.334,
      "train_steps_per_second": 0.133
    }
  ],
  "logging_steps": 5,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 5976600671682560.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}