trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4252040684223175,
  "best_model_checkpoint": "bert_base_rand_5_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 9.30614948272705,
      "learning_rate": 4.9e-05,
      "loss": 0.386,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.4252040684223175,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 1423.35,
      "eval_steps_per_second": 6.529,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.963630676269531,
      "learning_rate": 4.8e-05,
      "loss": 0.2155,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 0.5008359551429749,
      "eval_runtime": 0.565,
      "eval_samples_per_second": 1543.328,
      "eval_steps_per_second": 7.079,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.330244541168213,
      "learning_rate": 4.7e-05,
      "loss": 0.1605,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.5880488753318787,
      "eval_runtime": 0.5799,
      "eval_samples_per_second": 1503.716,
      "eval_steps_per_second": 6.898,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.19688606262207,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1218,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.5611222982406616,
      "eval_runtime": 0.5666,
      "eval_samples_per_second": 1538.918,
      "eval_steps_per_second": 7.059,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.650278091430664,
      "learning_rate": 4.5e-05,
      "loss": 0.0951,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.578220009803772,
      "eval_runtime": 0.5756,
      "eval_samples_per_second": 1514.993,
      "eval_steps_per_second": 6.95,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 9.958868980407715,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.077,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.843362033367157,
      "eval_runtime": 0.5847,
      "eval_samples_per_second": 1491.441,
      "eval_steps_per_second": 6.841,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.17597288675982542,
      "train_runtime": 668.1932,
      "train_samples_per_second": 5039.635,
      "train_steps_per_second": 19.755
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}