smids_5x_deit_base_sgd_0001_fold5
This model is a fine-tuned version of facebook/deit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.5006
- Accuracy: 0.8133
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: 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 |
---|---|---|---|---|
1.0891 | 1.0 | 375 | 1.0922 | 0.36 |
1.0632 | 2.0 | 750 | 1.0728 | 0.4133 |
1.0234 | 3.0 | 1125 | 1.0519 | 0.4667 |
1.0095 | 4.0 | 1500 | 1.0279 | 0.505 |
0.9691 | 5.0 | 1875 | 1.0014 | 0.54 |
0.9521 | 6.0 | 2250 | 0.9722 | 0.5683 |
0.9099 | 7.0 | 2625 | 0.9405 | 0.6033 |
0.8832 | 8.0 | 3000 | 0.9085 | 0.6267 |
0.8563 | 9.0 | 3375 | 0.8771 | 0.6533 |
0.8097 | 10.0 | 3750 | 0.8470 | 0.685 |
0.7629 | 11.0 | 4125 | 0.8186 | 0.705 |
0.7531 | 12.0 | 4500 | 0.7923 | 0.715 |
0.7082 | 13.0 | 4875 | 0.7677 | 0.7333 |
0.7318 | 14.0 | 5250 | 0.7449 | 0.7433 |
0.7243 | 15.0 | 5625 | 0.7237 | 0.7533 |
0.6668 | 16.0 | 6000 | 0.7041 | 0.7567 |
0.6939 | 17.0 | 6375 | 0.6860 | 0.76 |
0.6736 | 18.0 | 6750 | 0.6692 | 0.77 |
0.6795 | 19.0 | 7125 | 0.6538 | 0.78 |
0.6094 | 20.0 | 7500 | 0.6398 | 0.7833 |
0.5982 | 21.0 | 7875 | 0.6269 | 0.7817 |
0.5784 | 22.0 | 8250 | 0.6150 | 0.7867 |
0.6034 | 23.0 | 8625 | 0.6042 | 0.7933 |
0.6235 | 24.0 | 9000 | 0.5942 | 0.7967 |
0.5888 | 25.0 | 9375 | 0.5851 | 0.7933 |
0.5892 | 26.0 | 9750 | 0.5766 | 0.7933 |
0.5908 | 27.0 | 10125 | 0.5688 | 0.7983 |
0.5781 | 28.0 | 10500 | 0.5616 | 0.7983 |
0.5631 | 29.0 | 10875 | 0.5551 | 0.8 |
0.5055 | 30.0 | 11250 | 0.5492 | 0.8017 |
0.5168 | 31.0 | 11625 | 0.5436 | 0.805 |
0.5659 | 32.0 | 12000 | 0.5386 | 0.81 |
0.568 | 33.0 | 12375 | 0.5339 | 0.8083 |
0.5472 | 34.0 | 12750 | 0.5295 | 0.8117 |
0.5227 | 35.0 | 13125 | 0.5256 | 0.81 |
0.4679 | 36.0 | 13500 | 0.5220 | 0.81 |
0.5236 | 37.0 | 13875 | 0.5188 | 0.8117 |
0.5206 | 38.0 | 14250 | 0.5158 | 0.8117 |
0.5047 | 39.0 | 14625 | 0.5132 | 0.8133 |
0.5461 | 40.0 | 15000 | 0.5108 | 0.8133 |
0.495 | 41.0 | 15375 | 0.5087 | 0.8133 |
0.508 | 42.0 | 15750 | 0.5069 | 0.8133 |
0.5153 | 43.0 | 16125 | 0.5053 | 0.8133 |
0.4846 | 44.0 | 16500 | 0.5040 | 0.8133 |
0.5055 | 45.0 | 16875 | 0.5029 | 0.8133 |
0.5156 | 46.0 | 17250 | 0.5020 | 0.8133 |
0.525 | 47.0 | 17625 | 0.5013 | 0.8133 |
0.4795 | 48.0 | 18000 | 0.5009 | 0.8133 |
0.4888 | 49.0 | 18375 | 0.5006 | 0.8133 |
0.4989 | 50.0 | 18750 | 0.5006 | 0.8133 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2
- Downloads last month
- 14