trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.8656716417910446,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23880597014925373,
      "grad_norm": 0.585221529006958,
      "learning_rate": 9.829629131445342e-05,
      "loss": 1.4111,
      "step": 2
    },
    {
      "epoch": 0.47761194029850745,
      "grad_norm": 0.6329778432846069,
      "learning_rate": 9.330127018922194e-05,
      "loss": 1.4368,
      "step": 4
    },
    {
      "epoch": 0.7164179104477612,
      "grad_norm": 0.6056985259056091,
      "learning_rate": 8.535533905932738e-05,
      "loss": 1.2286,
      "step": 6
    },
    {
      "epoch": 0.9552238805970149,
      "grad_norm": 0.5265995860099792,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.1773,
      "step": 8
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 0.3901197910308838,
      "learning_rate": 6.294095225512603e-05,
      "loss": 0.9521,
      "step": 10
    },
    {
      "epoch": 1.4328358208955223,
      "grad_norm": 0.342200368642807,
      "learning_rate": 5e-05,
      "loss": 0.9986,
      "step": 12
    },
    {
      "epoch": 1.671641791044776,
      "grad_norm": 0.29529017210006714,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.9659,
      "step": 14
    },
    {
      "epoch": 1.9104477611940298,
      "grad_norm": 0.27741575241088867,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.9646,
      "step": 16
    },
    {
      "epoch": 2.1492537313432836,
      "grad_norm": 0.308491051197052,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.9083,
      "step": 18
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.30398857593536377,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.9453,
      "step": 20
    },
    {
      "epoch": 2.626865671641791,
      "grad_norm": 0.27445322275161743,
      "learning_rate": 1.70370868554659e-06,
      "loss": 0.9429,
      "step": 22
    },
    {
      "epoch": 2.8656716417910446,
      "grad_norm": 0.2563874423503876,
      "learning_rate": 0.0,
      "loss": 0.8827,
      "step": 24
    },
    {
      "epoch": 2.8656716417910446,
      "step": 24,
      "total_flos": 1.2058224578297856e+16,
      "train_loss": 1.0678408990303676,
      "train_runtime": 102.2691,
      "train_samples_per_second": 11.675,
      "train_steps_per_second": 0.235
    }
  ],
  "logging_steps": 2,
  "max_steps": 24,
  "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": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2058224578297856e+16,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}