trainer_state.json

{
  "best_metric": 0.8728,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0932/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": 122.31808471679688,
      "learning_rate": 8.657457896300791e-05,
      "loss": 3.0283,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6888,
      "eval_loss": 1.1498730182647705,
      "eval_runtime": 9.3384,
      "eval_samples_per_second": 401.569,
      "eval_steps_per_second": 6.318,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 186.98080444335938,
      "learning_rate": 7.681980515339464e-05,
      "loss": 0.8392,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7944,
      "eval_loss": 0.6562054753303528,
      "eval_runtime": 9.3771,
      "eval_samples_per_second": 399.909,
      "eval_steps_per_second": 6.292,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 66.69602966308594,
      "learning_rate": 6.222075445642905e-05,
      "loss": 0.5462,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8349333333333333,
      "eval_loss": 0.5329174995422363,
      "eval_runtime": 9.2331,
      "eval_samples_per_second": 406.148,
      "eval_steps_per_second": 6.39,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 173.43260192871094,
      "learning_rate": 4.5e-05,
      "loss": 0.4388,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.4594840407371521,
      "eval_runtime": 9.2278,
      "eval_samples_per_second": 406.383,
      "eval_steps_per_second": 6.394,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 109.69825744628906,
      "learning_rate": 2.7779245543570963e-05,
      "loss": 0.373,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8589333333333333,
      "eval_loss": 0.4340592324733734,
      "eval_runtime": 9.3831,
      "eval_samples_per_second": 399.656,
      "eval_steps_per_second": 6.288,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 104.96597290039062,
      "learning_rate": 1.3180194846605365e-05,
      "loss": 0.3255,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8605333333333334,
      "eval_loss": 0.4257410168647766,
      "eval_runtime": 9.155,
      "eval_samples_per_second": 409.614,
      "eval_steps_per_second": 6.445,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 73.51651000976562,
      "learning_rate": 3.425421036992097e-06,
      "loss": 0.3025,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.41869258880615234,
      "eval_runtime": 9.3395,
      "eval_samples_per_second": 401.518,
      "eval_steps_per_second": 6.317,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}