{
  "best_metric": 0.9772727272727273,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-batch8-nocrop2/checkpoint-147",
  "epoch": 4.974619289340102,
  "eval_steps": 500,
  "global_step": 245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20304568527918782,
      "grad_norm": 0.18894483149051666,
      "learning_rate": 2e-05,
      "loss": 0.0212,
      "step": 10
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 24.582429885864258,
      "learning_rate": 4e-05,
      "loss": 0.0305,
      "step": 20
    },
    {
      "epoch": 0.6091370558375635,
      "grad_norm": 0.053805481642484665,
      "learning_rate": 4.886363636363637e-05,
      "loss": 0.0858,
      "step": 30
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 0.27398234605789185,
      "learning_rate": 4.659090909090909e-05,
      "loss": 0.0433,
      "step": 40
    },
    {
      "epoch": 0.9949238578680203,
      "eval_accuracy": 0.9545454545454546,
      "eval_loss": 0.2792539596557617,
      "eval_runtime": 1.4174,
      "eval_samples_per_second": 124.175,
      "eval_steps_per_second": 15.522,
      "step": 49
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 45.7299690246582,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.1388,
      "step": 50
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 6.678280830383301,
      "learning_rate": 4.204545454545455e-05,
      "loss": 0.0434,
      "step": 60
    },
    {
      "epoch": 1.4213197969543148,
      "grad_norm": 22.982635498046875,
      "learning_rate": 3.9772727272727275e-05,
      "loss": 0.0735,
      "step": 70
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 21.329736709594727,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0162,
      "step": 80
    },
    {
      "epoch": 1.8274111675126905,
      "grad_norm": 27.677804946899414,
      "learning_rate": 3.522727272727273e-05,
      "loss": 0.0419,
      "step": 90
    },
    {
      "epoch": 1.9898477157360406,
      "eval_accuracy": 0.9715909090909091,
      "eval_loss": 0.1445770561695099,
      "eval_runtime": 1.4427,
      "eval_samples_per_second": 121.992,
      "eval_steps_per_second": 15.249,
      "step": 98
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 0.007859536446630955,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.0197,
      "step": 100
    },
    {
      "epoch": 2.233502538071066,
      "grad_norm": 0.011098887771368027,
      "learning_rate": 3.068181818181818e-05,
      "loss": 0.0097,
      "step": 110
    },
    {
      "epoch": 2.436548223350254,
      "grad_norm": 0.08474420756101608,
      "learning_rate": 2.8409090909090912e-05,
      "loss": 0.0037,
      "step": 120
    },
    {
      "epoch": 2.6395939086294415,
      "grad_norm": 0.013242855668067932,
      "learning_rate": 2.6136363636363637e-05,
      "loss": 0.0192,
      "step": 130
    },
    {
      "epoch": 2.8426395939086295,
      "grad_norm": 0.17217466235160828,
      "learning_rate": 2.3863636363636365e-05,
      "loss": 0.0155,
      "step": 140
    },
    {
      "epoch": 2.984771573604061,
      "eval_accuracy": 0.9772727272727273,
      "eval_loss": 0.14674212038516998,
      "eval_runtime": 1.4158,
      "eval_samples_per_second": 124.312,
      "eval_steps_per_second": 15.539,
      "step": 147
    },
    {
      "epoch": 3.045685279187817,
      "grad_norm": 0.01039437111467123,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.0271,
      "step": 150
    },
    {
      "epoch": 3.248730964467005,
      "grad_norm": 0.007275668904185295,
      "learning_rate": 1.9318181818181818e-05,
      "loss": 0.0058,
      "step": 160
    },
    {
      "epoch": 3.451776649746193,
      "grad_norm": 0.002001961227506399,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.006,
      "step": 170
    },
    {
      "epoch": 3.6548223350253806,
      "grad_norm": 0.0007674383814446628,
      "learning_rate": 1.4772727272727274e-05,
      "loss": 0.0005,
      "step": 180
    },
    {
      "epoch": 3.8578680203045685,
      "grad_norm": 0.08923467248678207,
      "learning_rate": 1.25e-05,
      "loss": 0.0011,
      "step": 190
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9602272727272727,
      "eval_loss": 0.17827549576759338,
      "eval_runtime": 1.4363,
      "eval_samples_per_second": 122.539,
      "eval_steps_per_second": 15.317,
      "step": 197
    },
    {
      "epoch": 4.060913705583756,
      "grad_norm": 0.014368077740073204,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.0376,
      "step": 200
    },
    {
      "epoch": 4.2639593908629445,
      "grad_norm": 0.07785768061876297,
      "learning_rate": 7.954545454545455e-06,
      "loss": 0.0148,
      "step": 210
    },
    {
      "epoch": 4.467005076142132,
      "grad_norm": 4.838421821594238,
      "learning_rate": 5.681818181818182e-06,
      "loss": 0.001,
      "step": 220
    },
    {
      "epoch": 4.67005076142132,
      "grad_norm": 0.0007783414330333471,
      "learning_rate": 3.409090909090909e-06,
      "loss": 0.0001,
      "step": 230
    },
    {
      "epoch": 4.873096446700508,
      "grad_norm": 6.467840194702148,
      "learning_rate": 1.1363636363636364e-06,
      "loss": 0.001,
      "step": 240
    },
    {
      "epoch": 4.974619289340102,
      "eval_accuracy": 0.9659090909090909,
      "eval_loss": 0.17341464757919312,
      "eval_runtime": 1.4121,
      "eval_samples_per_second": 124.639,
      "eval_steps_per_second": 15.58,
      "step": 245
    },
    {
      "epoch": 4.974619289340102,
      "step": 245,
      "total_flos": 1.9487095364911104e+17,
      "train_loss": 0.026832528186638897,
      "train_runtime": 142.496,
      "train_samples_per_second": 55.3,
      "train_steps_per_second": 1.719
    }
  ],
  "logging_steps": 10,
  "max_steps": 245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.9487095364911104e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}