resnet50_fold_3
This model is a fine-tuned version of microsoft/resnet-50 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.1685
- Accuracy: 0.9517
- F1 Score: 0.9556
- Recall: 0.9609
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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 150
- num_epochs: 100
- label_smoothing_factor: 0.15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Score | Recall |
|---|---|---|---|---|---|---|
| 2.7675 | 1.0 | 19 | 2.7627 | 0.2276 | 0.1794 | 0.2296 |
| 2.7523 | 2.0 | 38 | 2.7540 | 0.3241 | 0.2311 | 0.2797 |
| 2.7348 | 3.0 | 57 | 2.7263 | 0.3828 | 0.2281 | 0.2988 |
| 2.6936 | 4.0 | 76 | 2.6837 | 0.3793 | 0.2013 | 0.2840 |
| 2.6260 | 5.0 | 95 | 2.6374 | 0.3828 | 0.1989 | 0.2850 |
| 2.5096 | 6.0 | 114 | 2.5647 | 0.4276 | 0.2991 | 0.3443 |
| 2.3841 | 7.0 | 133 | 2.4568 | 0.5966 | 0.5553 | 0.5545 |
| 2.2120 | 8.0 | 152 | 2.2170 | 0.7172 | 0.7022 | 0.6993 |
| 1.9531 | 9.0 | 171 | 1.9750 | 0.8 | 0.8021 | 0.8009 |
| 1.8001 | 10.0 | 190 | 1.7220 | 0.8310 | 0.8415 | 0.8368 |
| 1.6636 | 11.0 | 209 | 1.5142 | 0.8759 | 0.8851 | 0.8898 |
| 1.4820 | 12.0 | 228 | 1.3768 | 0.8828 | 0.8913 | 0.8960 |
| 1.3627 | 13.0 | 247 | 1.3274 | 0.8966 | 0.9016 | 0.9069 |
| 1.2971 | 14.0 | 266 | 1.2548 | 0.9345 | 0.9376 | 0.9411 |
| 1.2295 | 15.0 | 285 | 1.2514 | 0.9276 | 0.9347 | 0.9391 |
| 1.2444 | 16.0 | 304 | 1.2267 | 0.9345 | 0.9406 | 0.9427 |
| 1.2220 | 17.0 | 323 | 1.2259 | 0.9345 | 0.9369 | 0.9397 |
| 1.1872 | 18.0 | 342 | 1.2203 | 0.9241 | 0.9260 | 0.9349 |
| 1.2176 | 19.0 | 361 | 1.1870 | 0.9310 | 0.9358 | 0.9411 |
| 1.2911 | 20.0 | 380 | 1.1895 | 0.9414 | 0.9461 | 0.9446 |
| 1.1779 | 21.0 | 399 | 1.1768 | 0.9448 | 0.9488 | 0.9534 |
| 1.1306 | 22.0 | 418 | 1.1870 | 0.9345 | 0.9395 | 0.9371 |
| 1.1647 | 23.0 | 437 | 1.1661 | 0.9448 | 0.9502 | 0.9547 |
| 1.1289 | 24.0 | 456 | 1.1722 | 0.9345 | 0.9402 | 0.9448 |
| 1.1109 | 25.0 | 475 | 1.1681 | 0.9448 | 0.9499 | 0.9521 |
| 1.1047 | 26.0 | 494 | 1.1909 | 0.9345 | 0.9401 | 0.9473 |
| 1.1087 | 27.0 | 513 | 1.1579 | 0.9483 | 0.9523 | 0.9533 |
| 1.1312 | 28.0 | 532 | 1.1588 | 0.9448 | 0.9487 | 0.9521 |
| 1.1221 | 29.0 | 551 | 1.1549 | 0.9483 | 0.9538 | 0.9572 |
| 1.1603 | 30.0 | 570 | 1.1495 | 0.9448 | 0.9498 | 0.9534 |
| 1.1105 | 31.0 | 589 | 1.1553 | 0.9379 | 0.9443 | 0.9448 |
| 1.1005 | 32.0 | 608 | 1.1685 | 0.9517 | 0.9556 | 0.9609 |
| 1.1101 | 33.0 | 627 | 1.1662 | 0.9414 | 0.9466 | 0.9484 |
| 1.0823 | 34.0 | 646 | 1.1724 | 0.9345 | 0.9403 | 0.9409 |
| 1.1197 | 35.0 | 665 | 1.1616 | 0.9448 | 0.9499 | 0.9521 |
| 1.0557 | 36.0 | 684 | 1.1553 | 0.9379 | 0.9433 | 0.9434 |
| 1.0602 | 37.0 | 703 | 1.1578 | 0.9414 | 0.9463 | 0.9484 |
| 1.0646 | 38.0 | 722 | 1.1797 | 0.9276 | 0.9316 | 0.9348 |
| 1.0431 | 39.0 | 741 | 1.1758 | 0.9483 | 0.9526 | 0.9559 |
| 1.0977 | 40.0 | 760 | 1.1542 | 0.9414 | 0.9460 | 0.9471 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2