hushem_5x_deit_base_adamax_001_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: 2.0054
- 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 |
---|---|---|---|---|
1.4493 | 1.0 | 28 | 1.4115 | 0.2439 |
1.4085 | 2.0 | 56 | 1.2905 | 0.2927 |
1.0193 | 3.0 | 84 | 1.4163 | 0.5366 |
1.1237 | 4.0 | 112 | 0.8304 | 0.6585 |
0.9964 | 5.0 | 140 | 0.7827 | 0.6585 |
0.9566 | 6.0 | 168 | 0.6329 | 0.7317 |
0.8689 | 7.0 | 196 | 0.6958 | 0.5854 |
0.8599 | 8.0 | 224 | 0.5797 | 0.7805 |
0.7599 | 9.0 | 252 | 0.9330 | 0.7805 |
0.7635 | 10.0 | 280 | 0.7011 | 0.6829 |
0.6603 | 11.0 | 308 | 0.9167 | 0.6829 |
0.7173 | 12.0 | 336 | 0.9009 | 0.5854 |
0.6949 | 13.0 | 364 | 0.5844 | 0.8049 |
0.5978 | 14.0 | 392 | 0.9237 | 0.7805 |
0.6548 | 15.0 | 420 | 0.5173 | 0.8049 |
0.5794 | 16.0 | 448 | 0.9750 | 0.7073 |
0.5927 | 17.0 | 476 | 0.8636 | 0.8049 |
0.4408 | 18.0 | 504 | 0.5076 | 0.8537 |
0.5047 | 19.0 | 532 | 0.9978 | 0.7073 |
0.5155 | 20.0 | 560 | 0.8993 | 0.7805 |
0.3022 | 21.0 | 588 | 1.0654 | 0.7805 |
0.3634 | 22.0 | 616 | 1.0189 | 0.8049 |
0.3346 | 23.0 | 644 | 0.9586 | 0.7805 |
0.2995 | 24.0 | 672 | 0.9302 | 0.7317 |
0.345 | 25.0 | 700 | 1.2111 | 0.7561 |
0.2746 | 26.0 | 728 | 1.7821 | 0.6585 |
0.1747 | 27.0 | 756 | 2.2030 | 0.6585 |
0.214 | 28.0 | 784 | 1.2078 | 0.6585 |
0.0609 | 29.0 | 812 | 1.3388 | 0.8049 |
0.0765 | 30.0 | 840 | 1.4109 | 0.7561 |
0.0654 | 31.0 | 868 | 1.4789 | 0.7561 |
0.0843 | 32.0 | 896 | 1.4884 | 0.7073 |
0.0165 | 33.0 | 924 | 2.0871 | 0.6341 |
0.0138 | 34.0 | 952 | 2.0174 | 0.6341 |
0.0253 | 35.0 | 980 | 2.0599 | 0.6098 |
0.0093 | 36.0 | 1008 | 1.7213 | 0.7317 |
0.0034 | 37.0 | 1036 | 1.8852 | 0.7561 |
0.0001 | 38.0 | 1064 | 1.8415 | 0.7561 |
0.0014 | 39.0 | 1092 | 1.8486 | 0.7073 |
0.001 | 40.0 | 1120 | 1.8899 | 0.7561 |
0.0001 | 41.0 | 1148 | 1.9569 | 0.7317 |
0.0001 | 42.0 | 1176 | 1.9763 | 0.7317 |
0.0001 | 43.0 | 1204 | 1.9852 | 0.7317 |
0.0001 | 44.0 | 1232 | 1.9927 | 0.7317 |
0.0001 | 45.0 | 1260 | 1.9966 | 0.7317 |
0.0 | 46.0 | 1288 | 2.0017 | 0.7317 |
0.0 | 47.0 | 1316 | 2.0041 | 0.7317 |
0.0001 | 48.0 | 1344 | 2.0053 | 0.7317 |
0.0 | 49.0 | 1372 | 2.0054 | 0.7317 |
0.0 | 50.0 | 1400 | 2.0054 | 0.7317 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 6
Inference API (serverless) is not available, repository is disabled.
Model tree for hkivancoral/hushem_5x_deit_base_adamax_001_fold5
Base model
facebook/deit-base-patch16-224
Finetuned
this model