trainer_state.json

{
  "best_metric": 0.9112,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0159/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 6.226893560226894e-05,
      "loss": 0.9241,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8397333333333333,
      "eval_loss": 0.5351495146751404,
      "eval_runtime": 7.2791,
      "eval_samples_per_second": 515.172,
      "eval_steps_per_second": 8.105,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.400972366333008,
      "learning_rate": 5.449115782449115e-05,
      "loss": 0.2863,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8576,
      "eval_loss": 0.5137800574302673,
      "eval_runtime": 6.8533,
      "eval_samples_per_second": 547.18,
      "eval_steps_per_second": 8.609,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.06381368637085,
      "learning_rate": 4.671338004671337e-05,
      "loss": 0.1511,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4477166533470154,
      "eval_runtime": 7.2466,
      "eval_samples_per_second": 517.485,
      "eval_steps_per_second": 8.142,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.76527214050293,
      "learning_rate": 3.89356022689356e-05,
      "loss": 0.0865,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.45343017578125,
      "eval_runtime": 6.7667,
      "eval_samples_per_second": 554.186,
      "eval_steps_per_second": 8.719,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5575488805770874,
      "learning_rate": 3.115782449115782e-05,
      "loss": 0.0497,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.47427546977996826,
      "eval_runtime": 6.7001,
      "eval_samples_per_second": 559.697,
      "eval_steps_per_second": 8.806,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.009685559198260307,
      "learning_rate": 2.3380046713380048e-05,
      "loss": 0.0305,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8962666666666667,
      "eval_loss": 0.44841891527175903,
      "eval_runtime": 6.7911,
      "eval_samples_per_second": 552.195,
      "eval_steps_per_second": 8.688,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03579683229327202,
      "learning_rate": 1.5602268935602267e-05,
      "loss": 0.0121,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.4202936291694641,
      "eval_runtime": 6.721,
      "eval_samples_per_second": 557.953,
      "eval_steps_per_second": 8.778,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.003137853927910328,
      "learning_rate": 7.824491157824491e-06,
      "loss": 0.0038,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9048,
      "eval_loss": 0.4306701421737671,
      "eval_runtime": 7.9299,
      "eval_samples_per_second": 472.895,
      "eval_steps_per_second": 7.44,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.5305806398391724,
      "learning_rate": 4.6713380046713376e-08,
      "loss": 0.0008,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9112,
      "eval_loss": 0.398320734500885,
      "eval_runtime": 6.7203,
      "eval_samples_per_second": 558.01,
      "eval_steps_per_second": 8.779,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}