trainer_state.json

{
  "best_metric": 0.5359712230215827,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-57",
  "epoch": 2.8860759493670884,
  "eval_steps": 500,
  "global_step": 57,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.51,
      "grad_norm": 3.057455539703369,
      "learning_rate": 4.607843137254902e-05,
      "loss": 1.0905,
      "step": 10
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.43884892086330934,
      "eval_loss": 1.0363357067108154,
      "eval_runtime": 7.9543,
      "eval_samples_per_second": 34.95,
      "eval_steps_per_second": 1.131,
      "step": 19
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.071553230285645,
      "learning_rate": 3.627450980392157e-05,
      "loss": 1.0368,
      "step": 20
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.5114073753356934,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.9897,
      "step": 30
    },
    {
      "epoch": 1.97,
      "eval_accuracy": 0.5071942446043165,
      "eval_loss": 0.9945195913314819,
      "eval_runtime": 5.7967,
      "eval_samples_per_second": 47.958,
      "eval_steps_per_second": 1.553,
      "step": 39
    },
    {
      "epoch": 2.03,
      "grad_norm": 6.494540214538574,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.957,
      "step": 40
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.738604784011841,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.9218,
      "step": 50
    },
    {
      "epoch": 2.89,
      "eval_accuracy": 0.5359712230215827,
      "eval_loss": 0.9638464450836182,
      "eval_runtime": 5.7057,
      "eval_samples_per_second": 48.723,
      "eval_steps_per_second": 1.577,
      "step": 57
    },
    {
      "epoch": 2.89,
      "step": 57,
      "total_flos": 1.7986296196553933e+17,
      "train_loss": 0.9899872813308448,
      "train_runtime": 160.5331,
      "train_samples_per_second": 46.682,
      "train_steps_per_second": 0.355
    }
  ],
  "logging_steps": 10,
  "max_steps": 57,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.7986296196553933e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}