run-2/checkpoint-1263/trainer_state.json

{
  "best_metric": 0.8887614678899083,
  "best_model_checkpoint": "bert-base-uncased-finetuned-sst2/run-2/checkpoint-1263",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1263,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8761467889908257,
      "eval_loss": 0.3037863075733185,
      "eval_runtime": 2.2668,
      "eval_samples_per_second": 384.685,
      "eval_steps_per_second": 24.263,
      "step": 421
    },
    {
      "epoch": 1.19,
      "grad_norm": 24.56251335144043,
      "learning_rate": 5.135971766809031e-05,
      "loss": 0.3191,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8830275229357798,
      "eval_loss": 0.4693216383457184,
      "eval_runtime": 2.3381,
      "eval_samples_per_second": 372.946,
      "eval_steps_per_second": 23.523,
      "step": 842
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.9397337436676025,
      "learning_rate": 3.5359805621956256e-05,
      "loss": 0.1393,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8887614678899083,
      "eval_loss": 0.533748984336853,
      "eval_runtime": 2.3644,
      "eval_samples_per_second": 368.807,
      "eval_steps_per_second": 23.262,
      "step": 1263
    }
  ],
  "logging_steps": 500,
  "max_steps": 2105,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 292344132499080.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 6.735962971422436e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 16,
    "seed": 2
  }
}