last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.05366726296958855,
  "eval_steps": 8,
  "global_step": 30,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017889087656529517,
      "eval_loss": 8.153331756591797,
      "eval_runtime": 33.2147,
      "eval_samples_per_second": 7.105,
      "eval_steps_per_second": 3.553,
      "step": 1
    },
    {
      "epoch": 0.005366726296958855,
      "grad_norm": 19.788150787353516,
      "learning_rate": 6e-05,
      "loss": 32.8462,
      "step": 3
    },
    {
      "epoch": 0.01073345259391771,
      "grad_norm": 33.15983581542969,
      "learning_rate": 0.00012,
      "loss": 29.6124,
      "step": 6
    },
    {
      "epoch": 0.014311270125223614,
      "eval_loss": 6.523018836975098,
      "eval_runtime": 33.7776,
      "eval_samples_per_second": 6.987,
      "eval_steps_per_second": 3.493,
      "step": 8
    },
    {
      "epoch": 0.016100178890876567,
      "grad_norm": 47.39445877075195,
      "learning_rate": 0.00018,
      "loss": 24.4004,
      "step": 9
    },
    {
      "epoch": 0.02146690518783542,
      "grad_norm": 42.159324645996094,
      "learning_rate": 0.00019510565162951537,
      "loss": 18.1195,
      "step": 12
    },
    {
      "epoch": 0.026833631484794274,
      "grad_norm": 13.451399803161621,
      "learning_rate": 0.00017071067811865476,
      "loss": 9.1384,
      "step": 15
    },
    {
      "epoch": 0.028622540250447227,
      "eval_loss": 0.9095311760902405,
      "eval_runtime": 33.888,
      "eval_samples_per_second": 6.964,
      "eval_steps_per_second": 3.482,
      "step": 16
    },
    {
      "epoch": 0.03220035778175313,
      "grad_norm": 14.610176086425781,
      "learning_rate": 0.00013090169943749476,
      "loss": 3.727,
      "step": 18
    },
    {
      "epoch": 0.03756708407871199,
      "grad_norm": 16.35560417175293,
      "learning_rate": 8.435655349597689e-05,
      "loss": 2.9105,
      "step": 21
    },
    {
      "epoch": 0.04293381037567084,
      "grad_norm": 11.3445405960083,
      "learning_rate": 4.12214747707527e-05,
      "loss": 1.9107,
      "step": 24
    },
    {
      "epoch": 0.04293381037567084,
      "eval_loss": 0.3514160215854645,
      "eval_runtime": 33.9061,
      "eval_samples_per_second": 6.96,
      "eval_steps_per_second": 3.48,
      "step": 24
    },
    {
      "epoch": 0.04830053667262969,
      "grad_norm": 9.279090881347656,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 1.3966,
      "step": 27
    },
    {
      "epoch": 0.05366726296958855,
      "grad_norm": 8.892674446105957,
      "learning_rate": 0.0,
      "loss": 0.9105,
      "step": 30
    }
  ],
  "logging_steps": 3,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 70,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9772495452241920.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}