trainer_state.json

{
  "best_metric": 0.7736,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0942/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 119.6313705444336,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 3.8459,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3616,
      "eval_loss": 3.712421417236328,
      "eval_runtime": 15.3776,
      "eval_samples_per_second": 243.862,
      "eval_steps_per_second": 3.837,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 293.9961242675781,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 3.26,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5522666666666667,
      "eval_loss": 2.5889902114868164,
      "eval_runtime": 15.3384,
      "eval_samples_per_second": 244.484,
      "eval_steps_per_second": 3.847,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 146.9651336669922,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 1.998,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.704,
      "eval_loss": 1.502894401550293,
      "eval_runtime": 15.3387,
      "eval_samples_per_second": 244.479,
      "eval_steps_per_second": 3.846,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 118.64312744140625,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 1.318,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7493333333333333,
      "eval_loss": 1.1313848495483398,
      "eval_runtime": 15.2374,
      "eval_samples_per_second": 246.104,
      "eval_steps_per_second": 3.872,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 100.35333251953125,
      "learning_rate": 5.647652972118998e-06,
      "loss": 1.0491,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7666666666666667,
      "eval_loss": 0.9399846196174622,
      "eval_runtime": 15.6605,
      "eval_samples_per_second": 239.456,
      "eval_steps_per_second": 3.767,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 97.02046203613281,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.962,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7693333333333333,
      "eval_loss": 0.9117268919944763,
      "eval_runtime": 15.2643,
      "eval_samples_per_second": 245.671,
      "eval_steps_per_second": 3.865,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 229.87448120117188,
      "learning_rate": 0.0,
      "loss": 0.9302,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7736,
      "eval_loss": 0.8885847926139832,
      "eval_runtime": 15.4453,
      "eval_samples_per_second": 242.792,
      "eval_steps_per_second": 3.82,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}