last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0046475113832626785,
  "eval_steps": 5,
  "global_step": 37,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00012560841576385618,
      "eval_loss": NaN,
      "eval_runtime": 896.5663,
      "eval_samples_per_second": 3.74,
      "eval_steps_per_second": 1.87,
      "step": 1
    },
    {
      "epoch": 0.00037682524729156853,
      "grad_norm": NaN,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0,
      "step": 3
    },
    {
      "epoch": 0.0006280420788192808,
      "eval_loss": NaN,
      "eval_runtime": 896.2034,
      "eval_samples_per_second": 3.741,
      "eval_steps_per_second": 1.871,
      "step": 5
    },
    {
      "epoch": 0.0007536504945831371,
      "grad_norm": NaN,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0,
      "step": 6
    },
    {
      "epoch": 0.0011304757418747057,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 9
    },
    {
      "epoch": 0.0012560841576385617,
      "eval_loss": NaN,
      "eval_runtime": 894.8458,
      "eval_samples_per_second": 3.747,
      "eval_steps_per_second": 1.874,
      "step": 10
    },
    {
      "epoch": 0.0015073009891662741,
      "grad_norm": NaN,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.0,
      "step": 12
    },
    {
      "epoch": 0.0018841262364578427,
      "grad_norm": NaN,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.0018841262364578427,
      "eval_loss": NaN,
      "eval_runtime": 894.9467,
      "eval_samples_per_second": 3.747,
      "eval_steps_per_second": 1.874,
      "step": 15
    },
    {
      "epoch": 0.0022609514837494114,
      "grad_norm": NaN,
      "learning_rate": 0.0002,
      "loss": 0.0,
      "step": 18
    },
    {
      "epoch": 0.0025121683152771234,
      "eval_loss": NaN,
      "eval_runtime": 895.7932,
      "eval_samples_per_second": 3.743,
      "eval_steps_per_second": 1.872,
      "step": 20
    },
    {
      "epoch": 0.0026377767310409796,
      "grad_norm": NaN,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.0,
      "step": 21
    },
    {
      "epoch": 0.0030146019783325482,
      "grad_norm": NaN,
      "learning_rate": 0.00015469481581224272,
      "loss": 0.0,
      "step": 24
    },
    {
      "epoch": 0.0031402103940964044,
      "eval_loss": NaN,
      "eval_runtime": 895.6403,
      "eval_samples_per_second": 3.744,
      "eval_steps_per_second": 1.872,
      "step": 25
    },
    {
      "epoch": 0.003391427225624117,
      "grad_norm": NaN,
      "learning_rate": 0.00010825793454723325,
      "loss": 0.0,
      "step": 27
    },
    {
      "epoch": 0.0037682524729156855,
      "grad_norm": NaN,
      "learning_rate": 5.983045753470308e-05,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.0037682524729156855,
      "eval_loss": NaN,
      "eval_runtime": 895.6125,
      "eval_samples_per_second": 3.744,
      "eval_steps_per_second": 1.872,
      "step": 30
    },
    {
      "epoch": 0.004145077720207254,
      "grad_norm": NaN,
      "learning_rate": 2.1085949060360654e-05,
      "loss": 0.0,
      "step": 33
    },
    {
      "epoch": 0.004396294551734966,
      "eval_loss": NaN,
      "eval_runtime": 895.5724,
      "eval_samples_per_second": 3.744,
      "eval_steps_per_second": 1.873,
      "step": 35
    },
    {
      "epoch": 0.004521902967498823,
      "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": 2.111466030445363e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}