trainer_state.json

{
  "best_metric": 0.8944,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0516/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": 149.2913360595703,
      "learning_rate": 6.788924172750679e-05,
      "loss": 3.1794,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6584,
      "eval_loss": 1.4563989639282227,
      "eval_runtime": 7.0037,
      "eval_samples_per_second": 535.431,
      "eval_steps_per_second": 8.424,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 108.93872833251953,
      "learning_rate": 6.181155550916422e-05,
      "loss": 0.8935,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8261333333333334,
      "eval_loss": 0.5942649841308594,
      "eval_runtime": 6.9712,
      "eval_samples_per_second": 537.929,
      "eval_steps_per_second": 8.463,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 163.96160888671875,
      "learning_rate": 5.2499999999999995e-05,
      "loss": 0.4989,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.45362240076065063,
      "eval_runtime": 7.6381,
      "eval_samples_per_second": 490.963,
      "eval_steps_per_second": 7.724,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 266.4144592285156,
      "learning_rate": 4.1077686218342566e-05,
      "loss": 0.35,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8722666666666666,
      "eval_loss": 0.41766557097435,
      "eval_runtime": 6.9775,
      "eval_samples_per_second": 537.439,
      "eval_steps_per_second": 8.456,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 291.2462463378906,
      "learning_rate": 2.8922313781657438e-05,
      "loss": 0.2622,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8826666666666667,
      "eval_loss": 0.38817986845970154,
      "eval_runtime": 7.3613,
      "eval_samples_per_second": 509.422,
      "eval_steps_per_second": 8.015,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 234.2328643798828,
      "learning_rate": 1.7500000000000005e-05,
      "loss": 0.2064,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.36369648575782776,
      "eval_runtime": 6.9428,
      "eval_samples_per_second": 540.125,
      "eval_steps_per_second": 8.498,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 330.6797180175781,
      "learning_rate": 8.188444490835772e-06,
      "loss": 0.1685,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.3588147759437561,
      "eval_runtime": 6.9427,
      "eval_samples_per_second": 540.134,
      "eval_steps_per_second": 8.498,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}