smids_1x_deit_small_adamax_00001_fold2
This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.7991
- Accuracy: 0.8569
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: 1e-05
- 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 |
|---|---|---|---|---|
| 0.5561 | 1.0 | 75 | 0.5421 | 0.7804 |
| 0.3586 | 2.0 | 150 | 0.4304 | 0.8286 |
| 0.3328 | 3.0 | 225 | 0.3782 | 0.8586 |
| 0.266 | 4.0 | 300 | 0.3735 | 0.8536 |
| 0.2451 | 5.0 | 375 | 0.3596 | 0.8536 |
| 0.2001 | 6.0 | 450 | 0.3408 | 0.8669 |
| 0.1658 | 7.0 | 525 | 0.3450 | 0.8552 |
| 0.1239 | 8.0 | 600 | 0.3523 | 0.8519 |
| 0.0994 | 9.0 | 675 | 0.3736 | 0.8602 |
| 0.0972 | 10.0 | 750 | 0.3715 | 0.8636 |
| 0.0323 | 11.0 | 825 | 0.4031 | 0.8569 |
| 0.0293 | 12.0 | 900 | 0.4332 | 0.8586 |
| 0.0328 | 13.0 | 975 | 0.4440 | 0.8636 |
| 0.0162 | 14.0 | 1050 | 0.4947 | 0.8552 |
| 0.02 | 15.0 | 1125 | 0.5267 | 0.8586 |
| 0.014 | 16.0 | 1200 | 0.5586 | 0.8586 |
| 0.0039 | 17.0 | 1275 | 0.5843 | 0.8536 |
| 0.0174 | 18.0 | 1350 | 0.6241 | 0.8586 |
| 0.003 | 19.0 | 1425 | 0.6332 | 0.8502 |
| 0.0035 | 20.0 | 1500 | 0.6412 | 0.8569 |
| 0.0013 | 21.0 | 1575 | 0.6514 | 0.8519 |
| 0.0012 | 22.0 | 1650 | 0.6772 | 0.8502 |
| 0.0007 | 23.0 | 1725 | 0.6821 | 0.8502 |
| 0.0007 | 24.0 | 1800 | 0.7040 | 0.8569 |
| 0.0006 | 25.0 | 1875 | 0.6982 | 0.8569 |
| 0.0005 | 26.0 | 1950 | 0.7116 | 0.8519 |
| 0.0162 | 27.0 | 2025 | 0.7266 | 0.8602 |
| 0.0086 | 28.0 | 2100 | 0.7241 | 0.8536 |
| 0.0081 | 29.0 | 2175 | 0.7251 | 0.8602 |
| 0.0111 | 30.0 | 2250 | 0.7386 | 0.8602 |
| 0.0161 | 31.0 | 2325 | 0.7520 | 0.8586 |
| 0.0003 | 32.0 | 2400 | 0.7496 | 0.8519 |
| 0.0003 | 33.0 | 2475 | 0.7540 | 0.8552 |
| 0.0003 | 34.0 | 2550 | 0.7601 | 0.8602 |
| 0.0123 | 35.0 | 2625 | 0.7663 | 0.8519 |
| 0.0002 | 36.0 | 2700 | 0.7752 | 0.8586 |
| 0.0002 | 37.0 | 2775 | 0.7743 | 0.8502 |
| 0.0002 | 38.0 | 2850 | 0.7784 | 0.8519 |
| 0.0004 | 39.0 | 2925 | 0.7826 | 0.8536 |
| 0.0002 | 40.0 | 3000 | 0.7838 | 0.8536 |
| 0.0044 | 41.0 | 3075 | 0.7838 | 0.8586 |
| 0.0002 | 42.0 | 3150 | 0.7911 | 0.8552 |
| 0.001 | 43.0 | 3225 | 0.7929 | 0.8569 |
| 0.0106 | 44.0 | 3300 | 0.7934 | 0.8552 |
| 0.0002 | 45.0 | 3375 | 0.7947 | 0.8552 |
| 0.0002 | 46.0 | 3450 | 0.7979 | 0.8536 |
| 0.001 | 47.0 | 3525 | 0.7983 | 0.8536 |
| 0.0002 | 48.0 | 3600 | 0.7989 | 0.8569 |
| 0.0022 | 49.0 | 3675 | 0.7991 | 0.8569 |
| 0.0004 | 50.0 | 3750 | 0.7991 | 0.8569 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0