last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.015240448975388735,
  "eval_steps": 5,
  "global_step": 37,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004119040263618577,
      "eval_loss": NaN,
      "eval_runtime": 318.8369,
      "eval_samples_per_second": 3.209,
      "eval_steps_per_second": 1.606,
      "step": 1
    },
    {
      "epoch": 0.001235712079085573,
      "grad_norm": NaN,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0,
      "step": 3
    },
    {
      "epoch": 0.0020595201318092883,
      "eval_loss": NaN,
      "eval_runtime": 317.9954,
      "eval_samples_per_second": 3.217,
      "eval_steps_per_second": 1.61,
      "step": 5
    },
    {
      "epoch": 0.002471424158171146,
      "grad_norm": NaN,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0,
      "step": 6
    },
    {
      "epoch": 0.0037071362372567192,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 9
    },
    {
      "epoch": 0.004119040263618577,
      "eval_loss": NaN,
      "eval_runtime": 317.8623,
      "eval_samples_per_second": 3.218,
      "eval_steps_per_second": 1.611,
      "step": 10
    },
    {
      "epoch": 0.004942848316342292,
      "grad_norm": NaN,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.0,
      "step": 12
    },
    {
      "epoch": 0.006178560395427865,
      "grad_norm": NaN,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.006178560395427865,
      "eval_loss": NaN,
      "eval_runtime": 317.7047,
      "eval_samples_per_second": 3.22,
      "eval_steps_per_second": 1.612,
      "step": 15
    },
    {
      "epoch": 0.0074142724745134385,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 18
    },
    {
      "epoch": 0.008238080527237153,
      "eval_loss": NaN,
      "eval_runtime": 317.6442,
      "eval_samples_per_second": 3.221,
      "eval_steps_per_second": 1.612,
      "step": 20
    },
    {
      "epoch": 0.008649984553599012,
      "grad_norm": NaN,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.0,
      "step": 21
    },
    {
      "epoch": 0.009885696632684585,
      "grad_norm": NaN,
      "learning_rate": 0.00015469481581224272,
      "loss": 0.0,
      "step": 24
    },
    {
      "epoch": 0.010297600659046443,
      "eval_loss": NaN,
      "eval_runtime": 318.0146,
      "eval_samples_per_second": 3.217,
      "eval_steps_per_second": 1.61,
      "step": 25
    },
    {
      "epoch": 0.011121408711770158,
      "grad_norm": NaN,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.0,
      "step": 27
    },
    {
      "epoch": 0.01235712079085573,
      "grad_norm": NaN,
      "learning_rate": 5.983045753470308e-05,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.01235712079085573,
      "eval_loss": NaN,
      "eval_runtime": 317.4705,
      "eval_samples_per_second": 3.222,
      "eval_steps_per_second": 1.613,
      "step": 30
    },
    {
      "epoch": 0.013592832869941304,
      "grad_norm": NaN,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.0,
      "step": 33
    },
    {
      "epoch": 0.014416640922665019,
      "eval_loss": NaN,
      "eval_runtime": 317.6364,
      "eval_samples_per_second": 3.221,
      "eval_steps_per_second": 1.612,
      "step": 35
    },
    {
      "epoch": 0.014828544949026877,
      "grad_norm": NaN,
      "learning_rate": 1.3638696597277679e-06,
      "loss": 0.0,
      "step": 36
    }
  ],
  "logging_steps": 3,
  "max_steps": 37,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 18,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6197187817897984e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}