{
  "best_global_step": 820,
  "best_metric": 0.6319929957389832,
  "best_model_checkpoint": "bert_base_rand_10_v2_qnli/checkpoint-820",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2870,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.7229080200195312,
      "learning_rate": 4.9e-05,
      "loss": 0.6642,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6214534138751602,
      "eval_loss": 0.6441905498504639,
      "eval_runtime": 3.5397,
      "eval_samples_per_second": 1543.346,
      "eval_steps_per_second": 6.215,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7834525108337402,
      "learning_rate": 4.8e-05,
      "loss": 0.6249,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6401244737323815,
      "eval_loss": 0.6319929957389832,
      "eval_runtime": 3.4456,
      "eval_samples_per_second": 1585.497,
      "eval_steps_per_second": 6.385,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.34190559387207,
      "learning_rate": 4.7e-05,
      "loss": 0.5583,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6447007138934652,
      "eval_loss": 0.63871830701828,
      "eval_runtime": 3.4579,
      "eval_samples_per_second": 1579.864,
      "eval_steps_per_second": 6.362,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.13766622543335,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4548,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6490939044481054,
      "eval_loss": 0.723526656627655,
      "eval_runtime": 3.4688,
      "eval_samples_per_second": 1574.918,
      "eval_steps_per_second": 6.342,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.328271865844727,
      "learning_rate": 4.5e-05,
      "loss": 0.341,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.651473549331869,
      "eval_loss": 0.8323887586593628,
      "eval_runtime": 3.4751,
      "eval_samples_per_second": 1572.032,
      "eval_steps_per_second": 6.331,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.726317405700684,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.242,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6414058209774849,
      "eval_loss": 1.1305311918258667,
      "eval_runtime": 3.4879,
      "eval_samples_per_second": 1566.253,
      "eval_steps_per_second": 6.307,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 11.080260276794434,
      "learning_rate": 4.3e-05,
      "loss": 0.175,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6403075233388248,
      "eval_loss": 1.2281441688537598,
      "eval_runtime": 3.4473,
      "eval_samples_per_second": 1584.732,
      "eval_steps_per_second": 6.382,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "step": 2870,
      "total_flos": 9.645664445050368e+16,
      "train_loss": 0.43717852070771857,
      "train_runtime": 1226.3758,
      "train_samples_per_second": 4270.428,
      "train_steps_per_second": 16.716
    }
  ],
  "logging_steps": 1,
  "max_steps": 20500,
  "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": 9.645664445050368e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}