trainer_state.json

{
  "best_metric": 0.9266666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0195/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": 10.755352973937988,
      "learning_rate": 2.85232922088503e-05,
      "loss": 0.8125,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.39098718762397766,
      "eval_runtime": 6.8077,
      "eval_samples_per_second": 550.846,
      "eval_steps_per_second": 8.667,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.4383924288793503e-05,
      "loss": 0.2119,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.3683837950229645,
      "eval_runtime": 6.5446,
      "eval_samples_per_second": 572.996,
      "eval_steps_per_second": 9.015,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.249573826789856,
      "learning_rate": 1.837722045066209e-05,
      "loss": 0.0969,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9010666666666667,
      "eval_loss": 0.3237946629524231,
      "eval_runtime": 5.9398,
      "eval_samples_per_second": 631.332,
      "eval_steps_per_second": 9.933,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9867069721221924,
      "learning_rate": 1.1701616683386336e-05,
      "loss": 0.0441,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9136,
      "eval_loss": 0.3085401654243469,
      "eval_runtime": 5.928,
      "eval_samples_per_second": 632.588,
      "eval_steps_per_second": 9.953,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0073767416179180145,
      "learning_rate": 5.679298183982636e-06,
      "loss": 0.0174,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9176,
      "eval_loss": 0.3177186846733093,
      "eval_runtime": 7.0957,
      "eval_samples_per_second": 528.487,
      "eval_steps_per_second": 8.315,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.024422673508524895,
      "learning_rate": 1.5030589901472829e-06,
      "loss": 0.0068,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9234666666666667,
      "eval_loss": 0.29711079597473145,
      "eval_runtime": 5.916,
      "eval_samples_per_second": 633.872,
      "eval_steps_per_second": 9.973,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.10078700631856918,
      "learning_rate": 5.4492431898478526e-11,
      "loss": 0.003,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9266666666666666,
      "eval_loss": 0.3016139268875122,
      "eval_runtime": 6.0867,
      "eval_samples_per_second": 616.099,
      "eval_steps_per_second": 9.693,
      "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}