trainer_state.json

{
  "best_global_step": 17,
  "best_metric": 0.4835864,
  "best_model_checkpoint": "/mnt/disks/sdb/dayoon_ko/output/nqpara-all-wnl/lr-3e-5-batch-4/checkpoint-17",
  "epoch": 1.0,
  "eval_steps": 2,
  "global_step": 17,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.061895551257253385,
      "grad_norm": 18.75,
      "learning_rate": 3e-05,
      "loss": 1.1349077224731445,
      "memory(GiB)": 96.58,
      "step": 1,
      "token_acc": 0.7732092589081256,
      "train_speed(iter/s)": 0.007516
    },
    {
      "epoch": 0.12379110251450677,
      "eval_loss": 0.7309088110923767,
      "eval_runtime": 19.3212,
      "eval_samples_per_second": 2.226,
      "eval_steps_per_second": 0.569,
      "step": 2
    },
    {
      "epoch": 0.24758220502901354,
      "eval_loss": 0.5805338025093079,
      "eval_runtime": 19.4233,
      "eval_samples_per_second": 2.214,
      "eval_steps_per_second": 0.566,
      "step": 4
    },
    {
      "epoch": 0.3713733075435203,
      "eval_loss": 0.5194844603538513,
      "eval_runtime": 18.8319,
      "eval_samples_per_second": 2.283,
      "eval_steps_per_second": 0.584,
      "step": 6
    },
    {
      "epoch": 0.4951644100580271,
      "eval_loss": 0.4991206228733063,
      "eval_runtime": 19.2405,
      "eval_samples_per_second": 2.235,
      "eval_steps_per_second": 0.572,
      "step": 8
    },
    {
      "epoch": 0.6189555125725339,
      "eval_loss": 0.4898950457572937,
      "eval_runtime": 19.974,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.551,
      "step": 10
    },
    {
      "epoch": 0.7427466150870407,
      "eval_loss": 0.4852791726589203,
      "eval_runtime": 19.2596,
      "eval_samples_per_second": 2.233,
      "eval_steps_per_second": 0.571,
      "step": 12
    },
    {
      "epoch": 0.8665377176015474,
      "eval_loss": 0.48366209864616394,
      "eval_runtime": 18.8268,
      "eval_samples_per_second": 2.284,
      "eval_steps_per_second": 0.584,
      "step": 14
    },
    {
      "epoch": 0.9903288201160542,
      "eval_loss": 0.48381921648979187,
      "eval_runtime": 18.2985,
      "eval_samples_per_second": 2.35,
      "eval_steps_per_second": 0.601,
      "step": 16
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.4835864007472992,
      "eval_runtime": 18.4797,
      "eval_samples_per_second": 2.327,
      "eval_steps_per_second": 0.595,
      "step": 17
    }
  ],
  "logging_steps": 64,
  "max_steps": 17,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7653160524694323e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}