trainer_state.json

{
  "best_metric": 0.8912,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0428/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": 37.71211242675781,
      "learning_rate": 5e-05,
      "loss": 0.7563,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8792,
      "eval_loss": 0.38425272703170776,
      "eval_runtime": 6.1372,
      "eval_samples_per_second": 611.023,
      "eval_steps_per_second": 9.613,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.03153418377041817,
      "learning_rate": 5e-05,
      "loss": 0.2022,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8848,
      "eval_loss": 0.41262128949165344,
      "eval_runtime": 6.1774,
      "eval_samples_per_second": 607.048,
      "eval_steps_per_second": 9.551,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4014420211315155,
      "learning_rate": 5e-05,
      "loss": 0.1207,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8794666666666666,
      "eval_loss": 0.4336581826210022,
      "eval_runtime": 6.1196,
      "eval_samples_per_second": 612.781,
      "eval_steps_per_second": 9.641,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3182053565979004,
      "learning_rate": 5e-05,
      "loss": 0.078,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.43774276971817017,
      "eval_runtime": 6.8411,
      "eval_samples_per_second": 548.155,
      "eval_steps_per_second": 8.624,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.0682,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.510521411895752,
      "eval_runtime": 6.6061,
      "eval_samples_per_second": 567.658,
      "eval_steps_per_second": 8.931,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.003923167008906603,
      "learning_rate": 5e-05,
      "loss": 0.0552,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.4639873802661896,
      "eval_runtime": 6.3949,
      "eval_samples_per_second": 586.407,
      "eval_steps_per_second": 9.226,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}