trainer_state.json

{
  "best_metric": 0.9256,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0621/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": 37.06629943847656,
      "learning_rate": 6.130255255255254e-05,
      "loss": 0.7796,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.46837982535362244,
      "eval_runtime": 12.8031,
      "eval_samples_per_second": 292.897,
      "eval_steps_per_second": 4.608,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8047108054161072,
      "learning_rate": 5.255255255255254e-05,
      "loss": 0.253,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8744,
      "eval_loss": 0.42132025957107544,
      "eval_runtime": 12.3841,
      "eval_samples_per_second": 302.808,
      "eval_steps_per_second": 4.764,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.19875949621200562,
      "learning_rate": 4.3802552552552546e-05,
      "loss": 0.123,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.45923346281051636,
      "eval_runtime": 11.8529,
      "eval_samples_per_second": 316.378,
      "eval_steps_per_second": 4.978,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5894108414649963,
      "learning_rate": 3.5052552552552544e-05,
      "loss": 0.0652,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8938666666666667,
      "eval_loss": 0.3990130126476288,
      "eval_runtime": 11.6995,
      "eval_samples_per_second": 320.526,
      "eval_steps_per_second": 5.043,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.008539936505258083,
      "learning_rate": 2.630255255255255e-05,
      "loss": 0.0309,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.40568670630455017,
      "eval_runtime": 11.7819,
      "eval_samples_per_second": 318.286,
      "eval_steps_per_second": 5.008,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.004536112304776907,
      "learning_rate": 1.7552552552552552e-05,
      "loss": 0.0104,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9149333333333334,
      "eval_loss": 0.3493638038635254,
      "eval_runtime": 11.7371,
      "eval_samples_per_second": 319.499,
      "eval_steps_per_second": 5.027,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.013375037349760532,
      "learning_rate": 8.802552552552553e-06,
      "loss": 0.0045,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9213333333333333,
      "eval_loss": 0.3391200304031372,
      "eval_runtime": 11.7482,
      "eval_samples_per_second": 319.199,
      "eval_steps_per_second": 5.022,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.15855228900909424,
      "learning_rate": 5.255255255255255e-08,
      "loss": 0.0006,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9256,
      "eval_loss": 0.32132023572921753,
      "eval_runtime": 11.875,
      "eval_samples_per_second": 315.789,
      "eval_steps_per_second": 4.968,
      "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
}