smids_5x_deit_small_rms_001_fold5
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: 0.9266
- Accuracy: 0.7933
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 |
---|---|---|---|---|
0.8478 | 1.0 | 375 | 0.8288 | 0.5567 |
0.8902 | 2.0 | 750 | 0.8202 | 0.5483 |
0.8343 | 3.0 | 1125 | 0.7711 | 0.65 |
0.8476 | 4.0 | 1500 | 0.8002 | 0.5683 |
0.7349 | 5.0 | 1875 | 0.7526 | 0.62 |
0.7146 | 6.0 | 2250 | 0.7401 | 0.645 |
0.6924 | 7.0 | 2625 | 0.7402 | 0.6467 |
0.8092 | 8.0 | 3000 | 0.7366 | 0.6267 |
0.7302 | 9.0 | 3375 | 0.7094 | 0.67 |
0.6674 | 10.0 | 3750 | 0.6785 | 0.6733 |
0.7108 | 11.0 | 4125 | 0.6584 | 0.6967 |
0.5797 | 12.0 | 4500 | 0.7184 | 0.68 |
0.6078 | 13.0 | 4875 | 0.6814 | 0.6767 |
0.6457 | 14.0 | 5250 | 0.6261 | 0.72 |
0.6638 | 15.0 | 5625 | 0.5980 | 0.7217 |
0.5829 | 16.0 | 6000 | 0.5841 | 0.7517 |
0.5785 | 17.0 | 6375 | 0.5759 | 0.7283 |
0.5879 | 18.0 | 6750 | 0.5909 | 0.7233 |
0.5859 | 19.0 | 7125 | 0.6185 | 0.7183 |
0.5569 | 20.0 | 7500 | 0.5506 | 0.745 |
0.5537 | 21.0 | 7875 | 0.5606 | 0.7617 |
0.5092 | 22.0 | 8250 | 0.5522 | 0.7483 |
0.61 | 23.0 | 8625 | 0.6297 | 0.7383 |
0.5833 | 24.0 | 9000 | 0.5399 | 0.7667 |
0.5315 | 25.0 | 9375 | 0.5551 | 0.7517 |
0.5313 | 26.0 | 9750 | 0.5176 | 0.7717 |
0.5327 | 27.0 | 10125 | 0.5547 | 0.775 |
0.4821 | 28.0 | 10500 | 0.5221 | 0.77 |
0.5852 | 29.0 | 10875 | 0.4983 | 0.7717 |
0.4777 | 30.0 | 11250 | 0.5766 | 0.75 |
0.4511 | 31.0 | 11625 | 0.5104 | 0.7733 |
0.5002 | 32.0 | 12000 | 0.5870 | 0.76 |
0.465 | 33.0 | 12375 | 0.4942 | 0.7917 |
0.4934 | 34.0 | 12750 | 0.5302 | 0.7783 |
0.4217 | 35.0 | 13125 | 0.5314 | 0.7883 |
0.3994 | 36.0 | 13500 | 0.5461 | 0.7917 |
0.3823 | 37.0 | 13875 | 0.5187 | 0.7933 |
0.3965 | 38.0 | 14250 | 0.5803 | 0.7917 |
0.3576 | 39.0 | 14625 | 0.5564 | 0.79 |
0.3853 | 40.0 | 15000 | 0.5425 | 0.8033 |
0.3694 | 41.0 | 15375 | 0.5885 | 0.7967 |
0.3496 | 42.0 | 15750 | 0.6131 | 0.7967 |
0.3293 | 43.0 | 16125 | 0.6330 | 0.8033 |
0.2565 | 44.0 | 16500 | 0.6562 | 0.795 |
0.3188 | 45.0 | 16875 | 0.7306 | 0.7933 |
0.2833 | 46.0 | 17250 | 0.8042 | 0.7917 |
0.2208 | 47.0 | 17625 | 0.7887 | 0.79 |
0.1436 | 48.0 | 18000 | 0.8206 | 0.7933 |
0.1521 | 49.0 | 18375 | 0.8762 | 0.8083 |
0.1603 | 50.0 | 18750 | 0.9266 | 0.7933 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2
- Downloads last month
- 1