run-18/checkpoint-1375/trainer_state.json

{
  "best_metric": 0.8612244897959184,
  "best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-18\\checkpoint-825",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1375,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7510204081632653,
      "eval_loss": 1.3183083534240723,
      "eval_runtime": 1.358,
      "eval_samples_per_second": 180.413,
      "eval_steps_per_second": 11.782,
      "step": 275
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.5764411106222026e-05,
      "loss": 1.5279,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8448979591836735,
      "eval_loss": 0.8408748507499695,
      "eval_runtime": 1.338,
      "eval_samples_per_second": 183.109,
      "eval_steps_per_second": 11.958,
      "step": 550
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8612244897959184,
      "eval_loss": 0.6363237500190735,
      "eval_runtime": 1.364,
      "eval_samples_per_second": 179.619,
      "eval_steps_per_second": 11.73,
      "step": 825
    },
    {
      "epoch": 3.64,
      "learning_rate": 6.756176188380868e-06,
      "loss": 0.7297,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8612244897959184,
      "eval_loss": 0.5622856020927429,
      "eval_runtime": 1.317,
      "eval_samples_per_second": 186.029,
      "eval_steps_per_second": 12.149,
      "step": 1100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8612244897959184,
      "eval_loss": 0.5402354001998901,
      "eval_runtime": 1.193,
      "eval_samples_per_second": 205.364,
      "eval_steps_per_second": 13.412,
      "step": 1375
    }
  ],
  "logging_steps": 500,
  "max_steps": 1375,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 20324049665850.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 2.4772646024063183e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 8,
    "seed": 40
  }
}