trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 99,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 8.059077118463017,
      "learning_rate": 5e-06,
      "loss": 0.732,
      "step": 10
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.4614798499656616,
      "learning_rate": 5e-06,
      "loss": 0.6327,
      "step": 20
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 2.08216893424911,
      "learning_rate": 5e-06,
      "loss": 0.5881,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5583714842796326,
      "eval_runtime": 22.4257,
      "eval_samples_per_second": 39.642,
      "eval_steps_per_second": 0.624,
      "step": 33
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 1.5369606421179118,
      "learning_rate": 5e-06,
      "loss": 0.5367,
      "step": 40
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 1.208061629541289,
      "learning_rate": 5e-06,
      "loss": 0.5128,
      "step": 50
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 1.1652875666383131,
      "learning_rate": 5e-06,
      "loss": 0.4975,
      "step": 60
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5151528120040894,
      "eval_runtime": 22.6124,
      "eval_samples_per_second": 39.315,
      "eval_steps_per_second": 0.619,
      "step": 66
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 1.405112627235947,
      "learning_rate": 5e-06,
      "loss": 0.4743,
      "step": 70
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 1.0655283861506202,
      "learning_rate": 5e-06,
      "loss": 0.4348,
      "step": 80
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 1.1549525346676852,
      "learning_rate": 5e-06,
      "loss": 0.4332,
      "step": 90
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5040238499641418,
      "eval_runtime": 21.7197,
      "eval_samples_per_second": 40.931,
      "eval_steps_per_second": 0.645,
      "step": 99
    },
    {
      "epoch": 3.0,
      "step": 99,
      "total_flos": 165619307642880.0,
      "train_loss": 0.5283013642436326,
      "train_runtime": 3529.0044,
      "train_samples_per_second": 14.355,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 99,
  "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": 165619307642880.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}