trainer_state.json

{
  "best_metric": 0.9017857142857143,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-Lego-v1-3ep/checkpoint-147",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 147,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20408163265306123,
      "grad_norm": 3.9616575241088867,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 2.7824,
      "step": 10
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 5.216483116149902,
      "learning_rate": 4.810606060606061e-05,
      "loss": 2.4565,
      "step": 20
    },
    {
      "epoch": 0.6122448979591837,
      "grad_norm": 8.999798774719238,
      "learning_rate": 4.431818181818182e-05,
      "loss": 1.7257,
      "step": 30
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 13.705337524414062,
      "learning_rate": 4.053030303030303e-05,
      "loss": 1.1779,
      "step": 40
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.7035489082336426,
      "eval_runtime": 103.5,
      "eval_samples_per_second": 1.082,
      "eval_steps_per_second": 0.039,
      "step": 49
    },
    {
      "epoch": 1.0204081632653061,
      "grad_norm": 10.42480754852295,
      "learning_rate": 3.6742424242424246e-05,
      "loss": 0.9019,
      "step": 50
    },
    {
      "epoch": 1.2244897959183674,
      "grad_norm": 11.608043670654297,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.7082,
      "step": 60
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 9.341365814208984,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.6998,
      "step": 70
    },
    {
      "epoch": 1.6326530612244898,
      "grad_norm": 9.492186546325684,
      "learning_rate": 2.537878787878788e-05,
      "loss": 0.6349,
      "step": 80
    },
    {
      "epoch": 1.836734693877551,
      "grad_norm": 10.111123085021973,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.5591,
      "step": 90
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8482142857142857,
      "eval_loss": 0.38913723826408386,
      "eval_runtime": 0.7084,
      "eval_samples_per_second": 158.099,
      "eval_steps_per_second": 5.646,
      "step": 98
    },
    {
      "epoch": 2.0408163265306123,
      "grad_norm": 9.370523452758789,
      "learning_rate": 1.7803030303030303e-05,
      "loss": 0.4968,
      "step": 100
    },
    {
      "epoch": 2.2448979591836733,
      "grad_norm": 8.256933212280273,
      "learning_rate": 1.4015151515151515e-05,
      "loss": 0.4329,
      "step": 110
    },
    {
      "epoch": 2.4489795918367347,
      "grad_norm": 11.488386154174805,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.4378,
      "step": 120
    },
    {
      "epoch": 2.6530612244897958,
      "grad_norm": 9.060242652893066,
      "learning_rate": 6.43939393939394e-06,
      "loss": 0.4341,
      "step": 130
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 8.625679016113281,
      "learning_rate": 2.651515151515152e-06,
      "loss": 0.4446,
      "step": 140
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9017857142857143,
      "eval_loss": 0.2857852876186371,
      "eval_runtime": 0.7016,
      "eval_samples_per_second": 159.624,
      "eval_steps_per_second": 5.701,
      "step": 147
    },
    {
      "epoch": 3.0,
      "step": 147,
      "total_flos": 4.6750026057231974e+17,
      "train_loss": 0.9658059327780795,
      "train_runtime": 6347.4478,
      "train_samples_per_second": 2.962,
      "train_steps_per_second": 0.023
    }
  ],
  "logging_steps": 10,
  "max_steps": 147,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 4.6750026057231974e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}