trainer_state.json

{
  "best_metric": 0.4992,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0578/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": 20.714725494384766,
      "learning_rate": 0.0002911145304096916,
      "loss": 3.1644,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2496,
      "eval_loss": 2.7316410541534424,
      "eval_runtime": 10.7676,
      "eval_samples_per_second": 348.267,
      "eval_steps_per_second": 5.479,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 30.714384078979492,
      "learning_rate": 0.00026520930752569493,
      "loss": 2.5302,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3304,
      "eval_loss": 2.3724915981292725,
      "eval_runtime": 10.7041,
      "eval_samples_per_second": 350.333,
      "eval_steps_per_second": 5.512,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 38.29639434814453,
      "learning_rate": 0.00022540814184580826,
      "loss": 2.2008,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3933333333333333,
      "eval_loss": 2.175440549850464,
      "eval_runtime": 10.7667,
      "eval_samples_per_second": 348.295,
      "eval_steps_per_second": 5.48,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 32.88588333129883,
      "learning_rate": 0.00017651164135367125,
      "loss": 1.8935,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.42106666666666664,
      "eval_loss": 2.035309314727783,
      "eval_runtime": 10.8561,
      "eval_samples_per_second": 345.427,
      "eval_steps_per_second": 5.435,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 22.12914276123047,
      "learning_rate": 0.00012441744564412654,
      "loss": 1.5859,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4730666666666667,
      "eval_loss": 1.8860554695129395,
      "eval_runtime": 10.9521,
      "eval_samples_per_second": 342.4,
      "eval_steps_per_second": 5.387,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.934654235839844,
      "learning_rate": 7.540888354819138e-05,
      "loss": 1.2108,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.49733333333333335,
      "eval_loss": 1.8317880630493164,
      "eval_runtime": 10.8585,
      "eval_samples_per_second": 345.351,
      "eval_steps_per_second": 5.434,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 3.5498622085445925e-05,
      "loss": 0.7834,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4992,
      "eval_loss": 2.0777945518493652,
      "eval_runtime": 11.341,
      "eval_samples_per_second": 330.658,
      "eval_steps_per_second": 5.202,
      "step": 2331
    }
  ],
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}