trainer_state.json

{
  "best_metric": 0.9178666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0546/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.1926798820495605,
      "learning_rate": 3e-05,
      "loss": 0.7826,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.3857031464576721,
      "eval_runtime": 6.4363,
      "eval_samples_per_second": 582.634,
      "eval_steps_per_second": 9.167,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.192209243774414,
      "learning_rate": 3e-05,
      "loss": 0.1704,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9074666666666666,
      "eval_loss": 0.31420040130615234,
      "eval_runtime": 5.7287,
      "eval_samples_per_second": 654.598,
      "eval_steps_per_second": 10.299,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5795260667800903,
      "learning_rate": 3e-05,
      "loss": 0.0842,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9066666666666666,
      "eval_loss": 0.3378627598285675,
      "eval_runtime": 6.1622,
      "eval_samples_per_second": 608.549,
      "eval_steps_per_second": 9.575,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.0632,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.36272990703582764,
      "eval_runtime": 5.73,
      "eval_samples_per_second": 654.451,
      "eval_steps_per_second": 10.297,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 19.6657657623291,
      "learning_rate": 3e-05,
      "loss": 0.036,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.39793407917022705,
      "eval_runtime": 5.755,
      "eval_samples_per_second": 651.608,
      "eval_steps_per_second": 10.252,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04420861601829529,
      "learning_rate": 3e-05,
      "loss": 0.0285,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9178666666666667,
      "eval_loss": 0.3625175654888153,
      "eval_runtime": 5.7349,
      "eval_samples_per_second": 653.891,
      "eval_steps_per_second": 10.288,
      "step": 1998
    }
  ],
  "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": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}