trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9991469396459802,
  "eval_steps": 500,
  "global_step": 3516,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.28435345134001566,
      "grad_norm": 0.656255841255188,
      "learning_rate": 9.294251565167901e-05,
      "loss": 1.2378,
      "step": 250
    },
    {
      "epoch": 0.5687069026800313,
      "grad_norm": 0.543454647064209,
      "learning_rate": 8.582811610700058e-05,
      "loss": 0.6375,
      "step": 500
    },
    {
      "epoch": 0.8530603540200469,
      "grad_norm": 0.43407657742500305,
      "learning_rate": 7.871371656232215e-05,
      "loss": 0.5158,
      "step": 750
    },
    {
      "epoch": 1.1374138053600626,
      "grad_norm": 0.4036562740802765,
      "learning_rate": 7.159931701764372e-05,
      "loss": 0.4885,
      "step": 1000
    },
    {
      "epoch": 1.421767256700078,
      "grad_norm": 0.40085434913635254,
      "learning_rate": 6.448491747296529e-05,
      "loss": 0.4759,
      "step": 1250
    },
    {
      "epoch": 1.706120708040094,
      "grad_norm": 0.4041031002998352,
      "learning_rate": 5.737051792828686e-05,
      "loss": 0.4629,
      "step": 1500
    },
    {
      "epoch": 1.9904741593801094,
      "grad_norm": 0.4215051233768463,
      "learning_rate": 5.025611838360843e-05,
      "loss": 0.4555,
      "step": 1750
    },
    {
      "epoch": 2.2748276107201253,
      "grad_norm": 0.4411364495754242,
      "learning_rate": 4.3141718838929996e-05,
      "loss": 0.4451,
      "step": 2000
    },
    {
      "epoch": 2.5591810620601407,
      "grad_norm": 0.4689200818538666,
      "learning_rate": 3.602731929425157e-05,
      "loss": 0.433,
      "step": 2250
    },
    {
      "epoch": 2.843534513400156,
      "grad_norm": 0.4492100775241852,
      "learning_rate": 2.8912919749573137e-05,
      "loss": 0.4217,
      "step": 2500
    },
    {
      "epoch": 3.127887964740172,
      "grad_norm": 0.4875541627407074,
      "learning_rate": 2.1798520204894708e-05,
      "loss": 0.4116,
      "step": 2750
    },
    {
      "epoch": 3.4122414160801875,
      "grad_norm": 0.48542124032974243,
      "learning_rate": 1.4684120660216277e-05,
      "loss": 0.4012,
      "step": 3000
    },
    {
      "epoch": 3.6965948674202034,
      "grad_norm": 0.5120503306388855,
      "learning_rate": 7.569721115537849e-06,
      "loss": 0.3939,
      "step": 3250
    },
    {
      "epoch": 3.980948318760219,
      "grad_norm": 0.510686993598938,
      "learning_rate": 4.553215708594195e-07,
      "loss": 0.3892,
      "step": 3500
    },
    {
      "epoch": 3.9991469396459802,
      "step": 3516,
      "total_flos": 9.473878732962447e+17,
      "train_loss": 0.5115500152856959,
      "train_runtime": 22624.5633,
      "train_samples_per_second": 39.791,
      "train_steps_per_second": 0.155
    }
  ],
  "logging_steps": 250,
  "max_steps": 3516,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.473878732962447e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}