trainer_state.json

{
  "best_metric": 0.9330666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0392/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": 1.970792531967163,
      "learning_rate": 6.003003003003003e-05,
      "loss": 0.8046,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8373333333333334,
      "eval_loss": 0.5680665373802185,
      "eval_runtime": 9.4688,
      "eval_samples_per_second": 396.036,
      "eval_steps_per_second": 6.231,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0034643730614334345,
      "learning_rate": 5.0030030030030024e-05,
      "loss": 0.2349,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.37659740447998047,
      "eval_runtime": 7.5743,
      "eval_samples_per_second": 495.097,
      "eval_steps_per_second": 7.79,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0031947405077517033,
      "learning_rate": 4.003003003003003e-05,
      "loss": 0.1156,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.35936638712882996,
      "eval_runtime": 7.6659,
      "eval_samples_per_second": 489.18,
      "eval_steps_per_second": 7.696,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.13836616277694702,
      "learning_rate": 3.003003003003003e-05,
      "loss": 0.0601,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8984,
      "eval_loss": 0.3723571300506592,
      "eval_runtime": 7.6077,
      "eval_samples_per_second": 492.922,
      "eval_steps_per_second": 7.755,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.016025492921471596,
      "learning_rate": 2.003003003003003e-05,
      "loss": 0.0263,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9157333333333333,
      "eval_loss": 0.313763290643692,
      "eval_runtime": 7.7121,
      "eval_samples_per_second": 486.251,
      "eval_steps_per_second": 7.65,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4901750087738037,
      "learning_rate": 1.003003003003003e-05,
      "loss": 0.0067,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9293333333333333,
      "eval_loss": 0.2899421155452728,
      "eval_runtime": 7.4628,
      "eval_samples_per_second": 502.491,
      "eval_steps_per_second": 7.906,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0003081119793932885,
      "learning_rate": 3.0030030030030026e-08,
      "loss": 0.0029,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9330666666666667,
      "eval_loss": 0.2799321413040161,
      "eval_runtime": 8.3924,
      "eval_samples_per_second": 446.831,
      "eval_steps_per_second": 7.03,
      "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
}