trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.979683972911964,
  "eval_steps": 500,
  "global_step": 165,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18058690744920994,
      "grad_norm": 4.922645091392012,
      "learning_rate": 5e-06,
      "loss": 0.7687,
      "step": 10
    },
    {
      "epoch": 0.3611738148984199,
      "grad_norm": 0.9395652150073383,
      "learning_rate": 5e-06,
      "loss": 0.671,
      "step": 20
    },
    {
      "epoch": 0.5417607223476298,
      "grad_norm": 1.046038701193381,
      "learning_rate": 5e-06,
      "loss": 0.6367,
      "step": 30
    },
    {
      "epoch": 0.7223476297968398,
      "grad_norm": 0.594532259113549,
      "learning_rate": 5e-06,
      "loss": 0.6256,
      "step": 40
    },
    {
      "epoch": 0.9029345372460497,
      "grad_norm": 0.5730093717758814,
      "learning_rate": 5e-06,
      "loss": 0.6154,
      "step": 50
    },
    {
      "epoch": 0.9932279909706546,
      "eval_loss": 0.5993303656578064,
      "eval_runtime": 39.1462,
      "eval_samples_per_second": 38.062,
      "eval_steps_per_second": 0.613,
      "step": 55
    },
    {
      "epoch": 1.0835214446952597,
      "grad_norm": 0.7543972397193625,
      "learning_rate": 5e-06,
      "loss": 0.6293,
      "step": 60
    },
    {
      "epoch": 1.2641083521444696,
      "grad_norm": 0.6738473289304389,
      "learning_rate": 5e-06,
      "loss": 0.5656,
      "step": 70
    },
    {
      "epoch": 1.4446952595936795,
      "grad_norm": 0.7714560381826885,
      "learning_rate": 5e-06,
      "loss": 0.5588,
      "step": 80
    },
    {
      "epoch": 1.6252821670428894,
      "grad_norm": 0.5914735318576527,
      "learning_rate": 5e-06,
      "loss": 0.561,
      "step": 90
    },
    {
      "epoch": 1.8058690744920993,
      "grad_norm": 0.6021441110951579,
      "learning_rate": 5e-06,
      "loss": 0.5526,
      "step": 100
    },
    {
      "epoch": 1.9864559819413092,
      "grad_norm": 0.6858905012136822,
      "learning_rate": 5e-06,
      "loss": 0.5525,
      "step": 110
    },
    {
      "epoch": 1.9864559819413092,
      "eval_loss": 0.5813493132591248,
      "eval_runtime": 39.0789,
      "eval_samples_per_second": 38.128,
      "eval_steps_per_second": 0.614,
      "step": 110
    },
    {
      "epoch": 2.1670428893905194,
      "grad_norm": 0.8968088455298097,
      "learning_rate": 5e-06,
      "loss": 0.549,
      "step": 120
    },
    {
      "epoch": 2.3476297968397293,
      "grad_norm": 1.110817374207432,
      "learning_rate": 5e-06,
      "loss": 0.5085,
      "step": 130
    },
    {
      "epoch": 2.528216704288939,
      "grad_norm": 0.6217391415622974,
      "learning_rate": 5e-06,
      "loss": 0.5079,
      "step": 140
    },
    {
      "epoch": 2.708803611738149,
      "grad_norm": 0.7055441296859828,
      "learning_rate": 5e-06,
      "loss": 0.5073,
      "step": 150
    },
    {
      "epoch": 2.889390519187359,
      "grad_norm": 0.7396133060004412,
      "learning_rate": 5e-06,
      "loss": 0.5138,
      "step": 160
    },
    {
      "epoch": 2.979683972911964,
      "eval_loss": 0.5825901627540588,
      "eval_runtime": 38.2401,
      "eval_samples_per_second": 38.964,
      "eval_steps_per_second": 0.628,
      "step": 165
    },
    {
      "epoch": 2.979683972911964,
      "step": 165,
      "total_flos": 276171765841920.0,
      "train_loss": 0.5809369983095112,
      "train_runtime": 5706.7095,
      "train_samples_per_second": 14.877,
      "train_steps_per_second": 0.029
    }
  ],
  "logging_steps": 10,
  "max_steps": 165,
  "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": 276171765841920.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}