trainer_state.json

{
  "best_metric": 0.42533333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0136/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": Infinity,
      "learning_rate": 0.00028549422332891283,
      "loss": 3.3107,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20266666666666666,
      "eval_loss": 2.9514989852905273,
      "eval_runtime": 6.6468,
      "eval_samples_per_second": 564.178,
      "eval_steps_per_second": 8.876,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 31.737279891967773,
      "learning_rate": 0.00024415433369325305,
      "loss": 2.7508,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2632,
      "eval_loss": 2.6903791427612305,
      "eval_runtime": 6.4843,
      "eval_samples_per_second": 578.319,
      "eval_steps_per_second": 9.099,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 32.269134521484375,
      "learning_rate": 0.0001841660235425207,
      "loss": 2.4788,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3085333333333333,
      "eval_loss": 2.4743919372558594,
      "eval_runtime": 6.4805,
      "eval_samples_per_second": 578.662,
      "eval_steps_per_second": 9.104,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 38.75617980957031,
      "learning_rate": 0.00011741071341041153,
      "loss": 2.2054,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3608,
      "eval_loss": 2.2801549434661865,
      "eval_runtime": 6.5462,
      "eval_samples_per_second": 572.85,
      "eval_steps_per_second": 9.013,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 36.15947723388672,
      "learning_rate": 5.711011116874128e-05,
      "loss": 1.9027,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4106666666666667,
      "eval_loss": 2.132503032684326,
      "eval_runtime": 6.5521,
      "eval_samples_per_second": 572.335,
      "eval_steps_per_second": 9.005,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 50.548038482666016,
      "learning_rate": 1.5207490629826985e-05,
      "loss": 1.6007,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.41946666666666665,
      "eval_loss": 2.0828237533569336,
      "eval_runtime": 7.3725,
      "eval_samples_per_second": 508.647,
      "eval_steps_per_second": 8.003,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 39.729331970214844,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.3636,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.42533333333333334,
      "eval_loss": 2.1433849334716797,
      "eval_runtime": 6.5315,
      "eval_samples_per_second": 574.144,
      "eval_steps_per_second": 9.033,
      "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
}