trainer_state.json

{
  "best_metric": 0.8736,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0418/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": 240.02720642089844,
      "learning_rate": 5e-05,
      "loss": 3.6775,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.47973333333333334,
      "eval_loss": 3.029951572418213,
      "eval_runtime": 14.9425,
      "eval_samples_per_second": 250.962,
      "eval_steps_per_second": 3.948,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 136.63430786132812,
      "learning_rate": 5e-05,
      "loss": 1.8019,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_loss": 0.9993602633476257,
      "eval_runtime": 14.4183,
      "eval_samples_per_second": 260.086,
      "eval_steps_per_second": 4.092,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 202.58168029785156,
      "learning_rate": 5e-05,
      "loss": 0.8125,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.796,
      "eval_loss": 0.6769412159919739,
      "eval_runtime": 14.7904,
      "eval_samples_per_second": 253.543,
      "eval_steps_per_second": 3.989,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 114.95118713378906,
      "learning_rate": 5e-05,
      "loss": 0.5904,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8309333333333333,
      "eval_loss": 0.5518678426742554,
      "eval_runtime": 14.6818,
      "eval_samples_per_second": 255.417,
      "eval_steps_per_second": 4.019,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 140.66746520996094,
      "learning_rate": 5e-05,
      "loss": 0.4857,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5000717043876648,
      "eval_runtime": 14.2445,
      "eval_samples_per_second": 263.259,
      "eval_steps_per_second": 4.142,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 72.94957733154297,
      "learning_rate": 5e-05,
      "loss": 0.4043,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.4375742971897125,
      "eval_runtime": 14.2547,
      "eval_samples_per_second": 263.072,
      "eval_steps_per_second": 4.139,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 142.59872436523438,
      "learning_rate": 5e-05,
      "loss": 0.3323,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.42849433422088623,
      "eval_runtime": 14.4042,
      "eval_samples_per_second": 260.341,
      "eval_steps_per_second": 4.096,
      "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
}