trainer_state.json

{
  "best_metric": 0.4613333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0981/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": Infinity,
      "learning_rate": 0.00028540736881871185,
      "loss": 3.2989,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2392,
      "eval_loss": 2.893644332885742,
      "eval_runtime": 13.6288,
      "eval_samples_per_second": 275.152,
      "eval_steps_per_second": 4.329,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 32.474910736083984,
      "learning_rate": 0.00024399687365927005,
      "loss": 2.7528,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2888,
      "eval_loss": 2.6541078090667725,
      "eval_runtime": 13.8025,
      "eval_samples_per_second": 271.69,
      "eval_steps_per_second": 4.275,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.79929542541504,
      "learning_rate": 0.00018396914487560668,
      "loss": 2.4776,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.35333333333333333,
      "eval_loss": 2.3974287509918213,
      "eval_runtime": 13.2356,
      "eval_samples_per_second": 283.327,
      "eval_steps_per_second": 4.458,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00011741071341041153,
      "loss": 2.2146,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.40586666666666665,
      "eval_loss": 2.20223331451416,
      "eval_runtime": 13.7019,
      "eval_samples_per_second": 273.684,
      "eval_steps_per_second": 4.306,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.612735748291016,
      "learning_rate": 5.711011116874128e-05,
      "loss": 1.9274,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.42213333333333336,
      "eval_loss": 2.119164228439331,
      "eval_runtime": 13.1429,
      "eval_samples_per_second": 285.324,
      "eval_steps_per_second": 4.489,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 35.321510314941406,
      "learning_rate": 1.5207490629826985e-05,
      "loss": 1.648,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.45386666666666664,
      "eval_loss": 2.020270586013794,
      "eval_runtime": 13.4271,
      "eval_samples_per_second": 279.286,
      "eval_steps_per_second": 4.394,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 23.364456176757812,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.4448,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4613333333333333,
      "eval_loss": 2.0609421730041504,
      "eval_runtime": 13.5818,
      "eval_samples_per_second": 276.105,
      "eval_steps_per_second": 4.344,
      "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
}