sbottazziunsam's picture
update model card README.md
83749b0
metadata
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - f1
model-index:
  - name: 8-classifier-finetuned-padchest
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: train
          args: default
        metrics:
          - name: F1
            type: f1
            value: 0.9325359911406422

8-classifier-finetuned-padchest

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2276
  • F1: 0.9325

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss F1
0.6321 1.0 18 0.5224 0.7896
0.4633 2.0 36 0.3809 0.7896
0.3552 3.0 54 0.3305 0.7896
0.2718 4.0 72 0.2696 0.8197
0.2345 5.0 90 0.2178 0.9149
0.211 6.0 108 0.2405 0.8861
0.2208 7.0 126 0.2713 0.8605
0.1698 8.0 144 0.1747 0.9422
0.1547 9.0 162 0.1783 0.9322
0.1697 10.0 180 0.1629 0.9350
0.1684 11.0 198 0.1740 0.9319
0.1722 12.0 216 0.1885 0.9173
0.158 13.0 234 0.1637 0.9331
0.1469 14.0 252 0.1716 0.9325
0.1271 15.0 270 0.1700 0.9384
0.131 16.0 288 0.1785 0.9409
0.1245 17.0 306 0.2124 0.9206
0.1182 18.0 324 0.1715 0.9322
0.1082 19.0 342 0.1946 0.9322
0.1274 20.0 360 0.1757 0.9379
0.1115 21.0 378 0.1908 0.9307
0.0995 22.0 396 0.2001 0.9289
0.0996 23.0 414 0.1820 0.9293
0.0993 24.0 432 0.2095 0.9355
0.1006 25.0 450 0.1973 0.9314
0.0703 26.0 468 0.1934 0.9389
0.0901 27.0 486 0.2276 0.9238
0.0827 28.0 504 0.1949 0.936
0.0701 29.0 522 0.2076 0.9317
0.0813 30.0 540 0.2001 0.9374
0.0776 31.0 558 0.2440 0.9357
0.0842 32.0 576 0.2163 0.9271
0.0872 33.0 594 0.2248 0.9332
0.0743 34.0 612 0.2007 0.9344
0.0692 35.0 630 0.1971 0.9283
0.0763 36.0 648 0.2094 0.9393
0.0714 37.0 666 0.2139 0.9271
0.0683 38.0 684 0.2065 0.9331
0.0698 39.0 702 0.2177 0.9295
0.0507 40.0 720 0.2171 0.9344
0.0523 41.0 738 0.2240 0.9344
0.0546 42.0 756 0.2083 0.9394
0.0695 43.0 774 0.2171 0.936
0.0634 44.0 792 0.2193 0.9301
0.0462 45.0 810 0.2017 0.9409
0.0581 46.0 828 0.2209 0.9350
0.0468 47.0 846 0.2335 0.9301
0.0424 48.0 864 0.2294 0.9301
0.0472 49.0 882 0.2310 0.9350
0.044 50.0 900 0.2276 0.9325

Framework versions

  • Transformers 4.28.0.dev0
  • Pytorch 2.0.0+cu117
  • Datasets 2.18.0
  • Tokenizers 0.13.3