trainer_state.json

{
  "best_global_step": null,
  "best_metric": Infinity,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": NaN,
      "learning_rate": 4.804e-05,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.4,
      "grad_norm": NaN,
      "learning_rate": 4.604e-05,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 0.6,
      "grad_norm": NaN,
      "learning_rate": 4.4040000000000005e-05,
      "loss": 0.0,
      "step": 150
    },
    {
      "epoch": 0.8,
      "grad_norm": NaN,
      "learning_rate": 4.2040000000000004e-05,
      "loss": 0.0,
      "step": 200
    },
    {
      "epoch": 1.0,
      "grad_norm": NaN,
      "learning_rate": 4.004e-05,
      "loss": 0.0,
      "step": 250
    },
    {
      "epoch": 1.0,
      "eval_loss": NaN,
      "eval_runtime": 36.1377,
      "eval_samples_per_second": 27.672,
      "eval_steps_per_second": 3.459,
      "step": 250
    },
    {
      "epoch": 1.2,
      "grad_norm": NaN,
      "learning_rate": 3.804e-05,
      "loss": 0.0,
      "step": 300
    },
    {
      "epoch": 1.4,
      "grad_norm": NaN,
      "learning_rate": 3.604e-05,
      "loss": 0.0,
      "step": 350
    },
    {
      "epoch": 1.6,
      "grad_norm": NaN,
      "learning_rate": 3.404e-05,
      "loss": 0.0,
      "step": 400
    },
    {
      "epoch": 1.8,
      "grad_norm": NaN,
      "learning_rate": 3.2040000000000005e-05,
      "loss": 0.0,
      "step": 450
    },
    {
      "epoch": 2.0,
      "grad_norm": NaN,
      "learning_rate": 3.004e-05,
      "loss": 0.0,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": NaN,
      "eval_runtime": 36.1799,
      "eval_samples_per_second": 27.64,
      "eval_steps_per_second": 3.455,
      "step": 500
    },
    {
      "epoch": 2.2,
      "grad_norm": NaN,
      "learning_rate": 2.804e-05,
      "loss": 0.0,
      "step": 550
    },
    {
      "epoch": 2.4,
      "grad_norm": NaN,
      "learning_rate": 2.6040000000000005e-05,
      "loss": 0.0,
      "step": 600
    },
    {
      "epoch": 2.6,
      "grad_norm": NaN,
      "learning_rate": 2.404e-05,
      "loss": 0.0,
      "step": 650
    },
    {
      "epoch": 2.8,
      "grad_norm": NaN,
      "learning_rate": 2.2040000000000002e-05,
      "loss": 0.0,
      "step": 700
    },
    {
      "epoch": 3.0,
      "grad_norm": NaN,
      "learning_rate": 2.004e-05,
      "loss": 0.0,
      "step": 750
    },
    {
      "epoch": 3.0,
      "eval_loss": NaN,
      "eval_runtime": 36.1238,
      "eval_samples_per_second": 27.683,
      "eval_steps_per_second": 3.46,
      "step": 750
    },
    {
      "epoch": 3.2,
      "grad_norm": NaN,
      "learning_rate": 1.804e-05,
      "loss": 0.0,
      "step": 800
    },
    {
      "epoch": 3.4,
      "grad_norm": NaN,
      "learning_rate": 1.604e-05,
      "loss": 0.0,
      "step": 850
    },
    {
      "epoch": 3.6,
      "grad_norm": NaN,
      "learning_rate": 1.4040000000000001e-05,
      "loss": 0.0,
      "step": 900
    },
    {
      "epoch": 3.8,
      "grad_norm": NaN,
      "learning_rate": 1.204e-05,
      "loss": 0.0,
      "step": 950
    },
    {
      "epoch": 4.0,
      "grad_norm": NaN,
      "learning_rate": 1.004e-05,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_loss": NaN,
      "eval_runtime": 36.1443,
      "eval_samples_per_second": 27.667,
      "eval_steps_per_second": 3.458,
      "step": 1000
    },
    {
      "epoch": 4.2,
      "grad_norm": NaN,
      "learning_rate": 8.040000000000001e-06,
      "loss": 0.0,
      "step": 1050
    },
    {
      "epoch": 4.4,
      "grad_norm": NaN,
      "learning_rate": 6.040000000000001e-06,
      "loss": 0.0,
      "step": 1100
    },
    {
      "epoch": 4.6,
      "grad_norm": NaN,
      "learning_rate": 4.04e-06,
      "loss": 0.0,
      "step": 1150
    },
    {
      "epoch": 4.8,
      "grad_norm": NaN,
      "learning_rate": 2.0400000000000004e-06,
      "loss": 0.0,
      "step": 1200
    },
    {
      "epoch": 5.0,
      "grad_norm": NaN,
      "learning_rate": 4e-08,
      "loss": 0.0,
      "step": 1250
    },
    {
      "epoch": 5.0,
      "eval_loss": NaN,
      "eval_runtime": 36.1201,
      "eval_samples_per_second": 27.685,
      "eval_steps_per_second": 3.461,
      "step": 1250
    }
  ],
  "logging_steps": 50,
  "max_steps": 1250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.369514704896e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}