smids_1x_deit_tiny_rms_001_fold3
This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.5663
- Accuracy: 0.7317
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: 0.001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- 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 | Accuracy |
|---|---|---|---|---|
| 1.1316 | 1.0 | 75 | 1.0924 | 0.485 |
| 0.926 | 2.0 | 150 | 0.9324 | 0.4883 |
| 0.9183 | 3.0 | 225 | 0.9377 | 0.5067 |
| 0.8817 | 4.0 | 300 | 0.8909 | 0.5233 |
| 0.9151 | 5.0 | 375 | 0.8809 | 0.5467 |
| 0.8442 | 6.0 | 450 | 0.8675 | 0.5433 |
| 0.8079 | 7.0 | 525 | 0.8557 | 0.56 |
| 0.7894 | 8.0 | 600 | 0.8225 | 0.56 |
| 0.8216 | 9.0 | 675 | 0.7829 | 0.5967 |
| 0.8087 | 10.0 | 750 | 0.8146 | 0.6033 |
| 0.748 | 11.0 | 825 | 1.0004 | 0.5133 |
| 0.7666 | 12.0 | 900 | 0.8100 | 0.62 |
| 0.6897 | 13.0 | 975 | 0.7991 | 0.5483 |
| 0.7137 | 14.0 | 1050 | 0.8839 | 0.6233 |
| 0.5912 | 15.0 | 1125 | 0.8794 | 0.5883 |
| 0.6325 | 16.0 | 1200 | 0.7543 | 0.6117 |
| 0.6753 | 17.0 | 1275 | 0.6833 | 0.6983 |
| 0.6515 | 18.0 | 1350 | 0.7976 | 0.6667 |
| 0.646 | 19.0 | 1425 | 0.7002 | 0.6983 |
| 0.6399 | 20.0 | 1500 | 0.6939 | 0.6783 |
| 0.6505 | 21.0 | 1575 | 0.7405 | 0.68 |
| 0.576 | 22.0 | 1650 | 0.6489 | 0.7283 |
| 0.6111 | 23.0 | 1725 | 0.7398 | 0.685 |
| 0.5464 | 24.0 | 1800 | 0.6157 | 0.7283 |
| 0.5394 | 25.0 | 1875 | 0.6999 | 0.6983 |
| 0.5477 | 26.0 | 1950 | 0.8201 | 0.6383 |
| 0.4763 | 27.0 | 2025 | 0.6302 | 0.725 |
| 0.4373 | 28.0 | 2100 | 0.6307 | 0.7217 |
| 0.5016 | 29.0 | 2175 | 0.6401 | 0.7283 |
| 0.4888 | 30.0 | 2250 | 0.6903 | 0.73 |
| 0.4977 | 31.0 | 2325 | 0.6449 | 0.7233 |
| 0.4672 | 32.0 | 2400 | 0.6248 | 0.735 |
| 0.378 | 33.0 | 2475 | 0.7223 | 0.7133 |
| 0.4146 | 34.0 | 2550 | 0.6409 | 0.73 |
| 0.3898 | 35.0 | 2625 | 0.6376 | 0.7517 |
| 0.3685 | 36.0 | 2700 | 0.6638 | 0.75 |
| 0.3996 | 37.0 | 2775 | 0.6101 | 0.7633 |
| 0.3384 | 38.0 | 2850 | 0.7105 | 0.745 |
| 0.3173 | 39.0 | 2925 | 0.6903 | 0.75 |
| 0.2666 | 40.0 | 3000 | 0.7489 | 0.7383 |
| 0.217 | 41.0 | 3075 | 0.7934 | 0.7267 |
| 0.2624 | 42.0 | 3150 | 0.8366 | 0.73 |
| 0.2541 | 43.0 | 3225 | 0.9217 | 0.7317 |
| 0.1857 | 44.0 | 3300 | 0.9118 | 0.7333 |
| 0.1064 | 45.0 | 3375 | 1.0537 | 0.7383 |
| 0.1268 | 46.0 | 3450 | 1.1266 | 0.7317 |
| 0.092 | 47.0 | 3525 | 1.2383 | 0.7433 |
| 0.0334 | 48.0 | 3600 | 1.4560 | 0.74 |
| 0.0452 | 49.0 | 3675 | 1.5075 | 0.7367 |
| 0.0233 | 50.0 | 3750 | 1.5663 | 0.7317 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0