trainer_state.json

{
  "best_metric": 0.9448,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0522/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 2.9106113962682575e-05,
      "loss": 0.7977,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.330710232257843,
      "eval_runtime": 7.6099,
      "eval_samples_per_second": 492.776,
      "eval_steps_per_second": 7.753,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.784917831420898,
      "learning_rate": 2.6510855361961218e-05,
      "loss": 0.1762,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9152,
      "eval_loss": 0.30422356724739075,
      "eval_runtime": 6.7532,
      "eval_samples_per_second": 555.296,
      "eval_steps_per_second": 8.737,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.366439819335938,
      "learning_rate": 2.2527217722455156e-05,
      "loss": 0.0744,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9234666666666667,
      "eval_loss": 0.3060300052165985,
      "eval_runtime": 6.7694,
      "eval_samples_per_second": 553.967,
      "eval_steps_per_second": 8.716,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.491821825504303,
      "learning_rate": 1.763568653571882e-05,
      "loss": 0.0337,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9293333333333333,
      "eval_loss": 0.28583094477653503,
      "eval_runtime": 7.5083,
      "eval_samples_per_second": 499.444,
      "eval_steps_per_second": 7.858,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1851516216993332,
      "learning_rate": 1.2426252654184347e-05,
      "loss": 0.016,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9341333333333334,
      "eval_loss": 0.277398943901062,
      "eval_runtime": 7.6382,
      "eval_samples_per_second": 490.951,
      "eval_steps_per_second": 7.724,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4485439658164978,
      "learning_rate": 7.5272506870206094e-06,
      "loss": 0.004,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9392,
      "eval_loss": 0.2645826041698456,
      "eval_runtime": 7.3342,
      "eval_samples_per_second": 511.304,
      "eval_steps_per_second": 8.045,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.01871308870613575,
      "learning_rate": 3.5295725730362483e-06,
      "loss": 0.0015,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.94,
      "eval_loss": 0.2699105739593506,
      "eval_runtime": 6.8307,
      "eval_samples_per_second": 548.99,
      "eval_steps_per_second": 8.637,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.03199732303619385,
      "learning_rate": 9.153972922231291e-07,
      "loss": 0.0008,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9448,
      "eval_loss": 0.2536012828350067,
      "eval_runtime": 7.0602,
      "eval_samples_per_second": 531.145,
      "eval_steps_per_second": 8.357,
      "step": 2664
    }
  ],
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}