smids_10x_deit_tiny_rms_0001_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: 1.2831
- Accuracy: 0.745
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 |
|---|---|---|---|---|
| 0.8666 | 1.0 | 750 | 0.8178 | 0.5867 |
| 0.793 | 2.0 | 1500 | 0.8394 | 0.5383 |
| 0.7315 | 3.0 | 2250 | 0.8051 | 0.6133 |
| 0.6465 | 4.0 | 3000 | 0.7374 | 0.65 |
| 0.703 | 5.0 | 3750 | 0.7241 | 0.6517 |
| 0.6607 | 6.0 | 4500 | 0.6935 | 0.6617 |
| 0.6906 | 7.0 | 5250 | 0.6781 | 0.675 |
| 0.673 | 8.0 | 6000 | 0.6701 | 0.7 |
| 0.5876 | 9.0 | 6750 | 0.6156 | 0.715 |
| 0.5761 | 10.0 | 7500 | 0.6686 | 0.6883 |
| 0.695 | 11.0 | 8250 | 0.6673 | 0.675 |
| 0.5527 | 12.0 | 9000 | 0.6193 | 0.7183 |
| 0.5532 | 13.0 | 9750 | 0.6407 | 0.6983 |
| 0.6398 | 14.0 | 10500 | 0.6327 | 0.7267 |
| 0.5686 | 15.0 | 11250 | 0.6250 | 0.71 |
| 0.6507 | 16.0 | 12000 | 0.6131 | 0.7183 |
| 0.586 | 17.0 | 12750 | 0.5959 | 0.7367 |
| 0.6263 | 18.0 | 13500 | 0.6433 | 0.7083 |
| 0.5943 | 19.0 | 14250 | 0.5766 | 0.7467 |
| 0.6095 | 20.0 | 15000 | 0.5801 | 0.7383 |
| 0.4915 | 21.0 | 15750 | 0.5843 | 0.7467 |
| 0.5994 | 22.0 | 16500 | 0.5711 | 0.74 |
| 0.4915 | 23.0 | 17250 | 0.5881 | 0.7367 |
| 0.5455 | 24.0 | 18000 | 0.5829 | 0.73 |
| 0.5646 | 25.0 | 18750 | 0.6056 | 0.73 |
| 0.4802 | 26.0 | 19500 | 0.5993 | 0.73 |
| 0.4066 | 27.0 | 20250 | 0.5797 | 0.7617 |
| 0.5295 | 28.0 | 21000 | 0.6131 | 0.7433 |
| 0.4838 | 29.0 | 21750 | 0.5976 | 0.7533 |
| 0.454 | 30.0 | 22500 | 0.5851 | 0.755 |
| 0.3428 | 31.0 | 23250 | 0.6240 | 0.745 |
| 0.3934 | 32.0 | 24000 | 0.6108 | 0.755 |
| 0.3564 | 33.0 | 24750 | 0.6563 | 0.755 |
| 0.4234 | 34.0 | 25500 | 0.6360 | 0.7633 |
| 0.3741 | 35.0 | 26250 | 0.6145 | 0.765 |
| 0.3785 | 36.0 | 27000 | 0.6637 | 0.7583 |
| 0.3282 | 37.0 | 27750 | 0.6548 | 0.7817 |
| 0.3768 | 38.0 | 28500 | 0.7250 | 0.7483 |
| 0.3263 | 39.0 | 29250 | 0.6603 | 0.7633 |
| 0.3862 | 40.0 | 30000 | 0.6936 | 0.7617 |
| 0.2705 | 41.0 | 30750 | 0.7486 | 0.7733 |
| 0.2694 | 42.0 | 31500 | 0.8322 | 0.7683 |
| 0.2855 | 43.0 | 32250 | 0.8068 | 0.7733 |
| 0.2669 | 44.0 | 33000 | 0.9199 | 0.755 |
| 0.2143 | 45.0 | 33750 | 0.9335 | 0.7667 |
| 0.1925 | 46.0 | 34500 | 1.0133 | 0.76 |
| 0.1987 | 47.0 | 35250 | 1.0665 | 0.745 |
| 0.1978 | 48.0 | 36000 | 1.1590 | 0.75 |
| 0.1441 | 49.0 | 36750 | 1.2474 | 0.7517 |
| 0.1372 | 50.0 | 37500 | 1.2831 | 0.745 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2