trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 5,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.0,
      "eval_loss": 3.9932637214660645,
      "eval_runtime": 0.7232,
      "eval_samples_per_second": 6.913,
      "eval_steps_per_second": 1.383,
      "step": 5
    },
    {
      "epoch": 10.0,
      "grad_norm": 447.8272848066355,
      "learning_rate": 9.698463103929542e-06,
      "loss": 3.3273,
      "step": 10
    },
    {
      "epoch": 10.0,
      "eval_loss": 2.1877388954162598,
      "eval_runtime": 1.187,
      "eval_samples_per_second": 4.212,
      "eval_steps_per_second": 0.842,
      "step": 10
    },
    {
      "epoch": 15.0,
      "eval_loss": 2.1877388954162598,
      "eval_runtime": 0.7203,
      "eval_samples_per_second": 6.941,
      "eval_steps_per_second": 1.388,
      "step": 15
    },
    {
      "epoch": 20.0,
      "grad_norm": 196.9862804871096,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.7156,
      "step": 20
    },
    {
      "epoch": 20.0,
      "eval_loss": 1.5716664791107178,
      "eval_runtime": 0.8611,
      "eval_samples_per_second": 5.806,
      "eval_steps_per_second": 1.161,
      "step": 20
    },
    {
      "epoch": 25.0,
      "eval_loss": 1.5716664791107178,
      "eval_runtime": 0.7273,
      "eval_samples_per_second": 6.875,
      "eval_steps_per_second": 1.375,
      "step": 25
    },
    {
      "epoch": 30.0,
      "grad_norm": 195.98568962834594,
      "learning_rate": 4.131759111665349e-06,
      "loss": 1.3707,
      "step": 30
    },
    {
      "epoch": 30.0,
      "eval_loss": 1.355377435684204,
      "eval_runtime": 1.3581,
      "eval_samples_per_second": 3.682,
      "eval_steps_per_second": 0.736,
      "step": 30
    },
    {
      "epoch": 35.0,
      "eval_loss": 1.355377435684204,
      "eval_runtime": 0.7213,
      "eval_samples_per_second": 6.932,
      "eval_steps_per_second": 1.386,
      "step": 35
    },
    {
      "epoch": 40.0,
      "grad_norm": 131.74931566909007,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 1.1402,
      "step": 40
    },
    {
      "epoch": 40.0,
      "eval_loss": 1.2805026769638062,
      "eval_runtime": 0.9702,
      "eval_samples_per_second": 5.153,
      "eval_steps_per_second": 1.031,
      "step": 40
    },
    {
      "epoch": 45.0,
      "eval_loss": 1.2805026769638062,
      "eval_runtime": 0.715,
      "eval_samples_per_second": 6.993,
      "eval_steps_per_second": 1.399,
      "step": 45
    },
    {
      "epoch": 50.0,
      "grad_norm": 107.57180038242872,
      "learning_rate": 0.0,
      "loss": 1.0501,
      "step": 50
    },
    {
      "epoch": 50.0,
      "eval_loss": 1.2817479372024536,
      "eval_runtime": 1.1532,
      "eval_samples_per_second": 4.336,
      "eval_steps_per_second": 0.867,
      "step": 50
    },
    {
      "epoch": 50.0,
      "step": 50,
      "total_flos": 2010511278080.0,
      "train_loss": 1.720783920288086,
      "train_runtime": 395.0445,
      "train_samples_per_second": 5.696,
      "train_steps_per_second": 0.127
    }
  ],
  "logging_steps": 10,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2010511278080.0,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}