trainer_state.json

{
  "best_metric": 0.9764705882352941,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-24",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.83,
      "grad_norm": 4.032827854156494,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 1.2376,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7470588235294118,
      "eval_loss": 0.5182495713233948,
      "eval_runtime": 3.913,
      "eval_samples_per_second": 43.445,
      "eval_steps_per_second": 1.533,
      "step": 12
    },
    {
      "epoch": 1.67,
      "grad_norm": 3.3620471954345703,
      "learning_rate": 2.5e-05,
      "loss": 0.4743,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9764705882352941,
      "eval_loss": 0.16659174859523773,
      "eval_runtime": 3.9553,
      "eval_samples_per_second": 42.98,
      "eval_steps_per_second": 1.517,
      "step": 24
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.355973720550537,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.2261,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9764705882352941,
      "eval_loss": 0.11346375942230225,
      "eval_runtime": 4.6585,
      "eval_samples_per_second": 36.492,
      "eval_steps_per_second": 1.288,
      "step": 36
    },
    {
      "epoch": 3.0,
      "step": 36,
      "total_flos": 1.137256821195264e+17,
      "train_loss": 0.5669683615366617,
      "train_runtime": 135.6407,
      "train_samples_per_second": 33.729,
      "train_steps_per_second": 0.265
    }
  ],
  "logging_steps": 10,
  "max_steps": 36,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.137256821195264e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}