last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0005162622612287042,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.162622612287042e-05,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 17.3727,
      "step": 1
    },
    {
      "epoch": 5.162622612287042e-05,
      "eval_loss": 4.458253383636475,
      "eval_runtime": 1160.1948,
      "eval_samples_per_second": 3.515,
      "eval_steps_per_second": 3.515,
      "step": 1
    },
    {
      "epoch": 0.00010325245224574084,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 19.5813,
      "step": 2
    },
    {
      "epoch": 0.00015487867836861126,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 16.4565,
      "step": 3
    },
    {
      "epoch": 0.00015487867836861126,
      "eval_loss": 4.458253383636475,
      "eval_runtime": 1161.4102,
      "eval_samples_per_second": 3.511,
      "eval_steps_per_second": 3.511,
      "step": 3
    },
    {
      "epoch": 0.00020650490449148167,
      "grad_norm": 33.112857818603516,
      "learning_rate": 2e-05,
      "loss": 15.6053,
      "step": 4
    },
    {
      "epoch": 0.0002581311306143521,
      "grad_norm": NaN,
      "learning_rate": 2e-05,
      "loss": 16.6773,
      "step": 5
    },
    {
      "epoch": 0.0003097573567372225,
      "grad_norm": 36.85125732421875,
      "learning_rate": 4e-05,
      "loss": 17.8976,
      "step": 6
    },
    {
      "epoch": 0.0003097573567372225,
      "eval_loss": 4.365196228027344,
      "eval_runtime": 1155.9375,
      "eval_samples_per_second": 3.528,
      "eval_steps_per_second": 3.528,
      "step": 6
    },
    {
      "epoch": 0.00036138358286009293,
      "grad_norm": 51.97016525268555,
      "learning_rate": 6e-05,
      "loss": 20.0777,
      "step": 7
    },
    {
      "epoch": 0.00041300980898296334,
      "grad_norm": NaN,
      "learning_rate": 6e-05,
      "loss": 17.7553,
      "step": 8
    },
    {
      "epoch": 0.00046463603510583376,
      "grad_norm": 45.40068817138672,
      "learning_rate": 8e-05,
      "loss": 18.4153,
      "step": 9
    },
    {
      "epoch": 0.00046463603510583376,
      "eval_loss": 3.409105062484741,
      "eval_runtime": 1155.3327,
      "eval_samples_per_second": 3.53,
      "eval_steps_per_second": 3.53,
      "step": 9
    },
    {
      "epoch": 0.0005162622612287042,
      "grad_norm": 98.54031372070312,
      "learning_rate": 0.0001,
      "loss": 11.2608,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.14364023570432e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}