trainer_state.json

{
  "best_metric": 0.8704,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0856/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": 93.2320556640625,
      "learning_rate": 3e-05,
      "loss": 3.8184,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.464,
      "eval_loss": 3.602182626724243,
      "eval_runtime": 13.4362,
      "eval_samples_per_second": 279.097,
      "eval_steps_per_second": 4.391,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 180.6476593017578,
      "learning_rate": 3e-05,
      "loss": 2.8625,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6544,
      "eval_loss": 1.8298394680023193,
      "eval_runtime": 13.4313,
      "eval_samples_per_second": 279.198,
      "eval_steps_per_second": 4.393,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 73.93403625488281,
      "learning_rate": 3e-05,
      "loss": 1.2486,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7832,
      "eval_loss": 0.8473197221755981,
      "eval_runtime": 13.4089,
      "eval_samples_per_second": 279.666,
      "eval_steps_per_second": 4.4,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 105.59098052978516,
      "learning_rate": 3e-05,
      "loss": 0.7268,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8282666666666667,
      "eval_loss": 0.6124317646026611,
      "eval_runtime": 12.9972,
      "eval_samples_per_second": 288.524,
      "eval_steps_per_second": 4.539,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 175.02346801757812,
      "learning_rate": 3e-05,
      "loss": 0.5379,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.5061413645744324,
      "eval_runtime": 13.4496,
      "eval_samples_per_second": 278.819,
      "eval_steps_per_second": 4.387,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 172.64422607421875,
      "learning_rate": 3e-05,
      "loss": 0.4206,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8624,
      "eval_loss": 0.4536849558353424,
      "eval_runtime": 13.6057,
      "eval_samples_per_second": 275.621,
      "eval_steps_per_second": 4.336,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 191.81631469726562,
      "learning_rate": 3e-05,
      "loss": 0.3525,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.42753100395202637,
      "eval_runtime": 13.3798,
      "eval_samples_per_second": 280.274,
      "eval_steps_per_second": 4.41,
      "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
}