trainer_state.json

{
  "best_metric": 0.9136,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0935/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": 10.337950706481934,
      "learning_rate": 7.878378378378379e-05,
      "loss": 0.9403,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.836,
      "eval_loss": 0.5556292533874512,
      "eval_runtime": 12.3982,
      "eval_samples_per_second": 302.463,
      "eval_steps_per_second": 4.759,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.207980632781982,
      "learning_rate": 6.753378378378378e-05,
      "loss": 0.3337,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.47973623871803284,
      "eval_runtime": 12.375,
      "eval_samples_per_second": 303.029,
      "eval_steps_per_second": 4.768,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.35305729508399963,
      "learning_rate": 5.628378378378378e-05,
      "loss": 0.175,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.868,
      "eval_loss": 0.4881253242492676,
      "eval_runtime": 12.4691,
      "eval_samples_per_second": 300.745,
      "eval_steps_per_second": 4.732,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3227742910385132,
      "learning_rate": 4.503378378378379e-05,
      "loss": 0.1013,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8688,
      "eval_loss": 0.4908599853515625,
      "eval_runtime": 12.4831,
      "eval_samples_per_second": 300.407,
      "eval_steps_per_second": 4.726,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.005405168980360031,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.0566,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.45199447870254517,
      "eval_runtime": 12.3508,
      "eval_samples_per_second": 303.624,
      "eval_steps_per_second": 4.777,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7097229957580566,
      "learning_rate": 2.2533783783783782e-05,
      "loss": 0.0268,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9013333333333333,
      "eval_loss": 0.4186803102493286,
      "eval_runtime": 12.3862,
      "eval_samples_per_second": 302.757,
      "eval_steps_per_second": 4.763,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0007228588801808655,
      "learning_rate": 1.1283783783783785e-05,
      "loss": 0.0086,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9122666666666667,
      "eval_loss": 0.3661345839500427,
      "eval_runtime": 12.3222,
      "eval_samples_per_second": 304.33,
      "eval_steps_per_second": 4.788,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.03358548879623413,
      "learning_rate": 3.3783783783783786e-08,
      "loss": 0.002,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9136,
      "eval_loss": 0.35262197256088257,
      "eval_runtime": 12.7893,
      "eval_samples_per_second": 293.215,
      "eval_steps_per_second": 4.613,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}