trainer_state.json

{
  "best_metric": 0.43306666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0275/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": 0.00028549422332891283,
      "loss": 3.3629,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.232,
      "eval_loss": 2.9062697887420654,
      "eval_runtime": 7.2758,
      "eval_samples_per_second": 515.41,
      "eval_steps_per_second": 8.109,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 35.41544723510742,
      "learning_rate": 0.00024415433369325305,
      "loss": 2.7922,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2949333333333333,
      "eval_loss": 2.6466166973114014,
      "eval_runtime": 7.713,
      "eval_samples_per_second": 486.194,
      "eval_steps_per_second": 7.649,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 36.29129409790039,
      "learning_rate": 0.0001841660235425207,
      "loss": 2.5119,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3258666666666667,
      "eval_loss": 2.518737554550171,
      "eval_runtime": 7.5274,
      "eval_samples_per_second": 498.181,
      "eval_steps_per_second": 7.838,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 34.028350830078125,
      "learning_rate": 0.00011741071341041153,
      "loss": 2.2479,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3672,
      "eval_loss": 2.3124959468841553,
      "eval_runtime": 7.4169,
      "eval_samples_per_second": 505.6,
      "eval_steps_per_second": 7.955,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 22.832992553710938,
      "learning_rate": 5.711011116874128e-05,
      "loss": 1.9544,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.40213333333333334,
      "eval_loss": 2.169468879699707,
      "eval_runtime": 8.2494,
      "eval_samples_per_second": 454.578,
      "eval_steps_per_second": 7.152,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 41.48280334472656,
      "learning_rate": 1.5207490629826985e-05,
      "loss": 1.6244,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4312,
      "eval_loss": 2.1382734775543213,
      "eval_runtime": 7.6406,
      "eval_samples_per_second": 490.796,
      "eval_steps_per_second": 7.722,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 23.477861404418945,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.3592,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.43306666666666666,
      "eval_loss": 2.1966373920440674,
      "eval_runtime": 7.2192,
      "eval_samples_per_second": 519.45,
      "eval_steps_per_second": 8.173,
      "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
}