hushem_5x_deit_base_adamax_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: 0.4553
- Accuracy: 0.9024
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 |
---|---|---|---|---|
0.7557 | 1.0 | 28 | 0.4830 | 0.7805 |
0.1392 | 2.0 | 56 | 0.3138 | 0.8293 |
0.0115 | 3.0 | 84 | 0.3481 | 0.8537 |
0.0078 | 4.0 | 112 | 0.3098 | 0.8537 |
0.0013 | 5.0 | 140 | 0.3283 | 0.9024 |
0.0023 | 6.0 | 168 | 0.5654 | 0.8537 |
0.0004 | 7.0 | 196 | 0.4129 | 0.9024 |
0.0003 | 8.0 | 224 | 0.4041 | 0.9024 |
0.0002 | 9.0 | 252 | 0.4192 | 0.9024 |
0.0002 | 10.0 | 280 | 0.4257 | 0.9024 |
0.0002 | 11.0 | 308 | 0.4271 | 0.9024 |
0.0002 | 12.0 | 336 | 0.4272 | 0.9024 |
0.0001 | 13.0 | 364 | 0.4303 | 0.9024 |
0.0002 | 14.0 | 392 | 0.4309 | 0.9024 |
0.0001 | 15.0 | 420 | 0.4302 | 0.9024 |
0.0001 | 16.0 | 448 | 0.4300 | 0.9024 |
0.0001 | 17.0 | 476 | 0.4319 | 0.9024 |
0.0001 | 18.0 | 504 | 0.4342 | 0.9024 |
0.0001 | 19.0 | 532 | 0.4349 | 0.9024 |
0.0001 | 20.0 | 560 | 0.4354 | 0.9024 |
0.0001 | 21.0 | 588 | 0.4378 | 0.9024 |
0.0001 | 22.0 | 616 | 0.4393 | 0.9024 |
0.0001 | 23.0 | 644 | 0.4414 | 0.9024 |
0.0001 | 24.0 | 672 | 0.4417 | 0.9024 |
0.0001 | 25.0 | 700 | 0.4428 | 0.9024 |
0.0001 | 26.0 | 728 | 0.4429 | 0.9024 |
0.0001 | 27.0 | 756 | 0.4437 | 0.9024 |
0.0001 | 28.0 | 784 | 0.4437 | 0.9024 |
0.0001 | 29.0 | 812 | 0.4449 | 0.9024 |
0.0001 | 30.0 | 840 | 0.4459 | 0.9024 |
0.0001 | 31.0 | 868 | 0.4470 | 0.9024 |
0.0001 | 32.0 | 896 | 0.4471 | 0.9024 |
0.0001 | 33.0 | 924 | 0.4499 | 0.9024 |
0.0001 | 34.0 | 952 | 0.4499 | 0.9024 |
0.0001 | 35.0 | 980 | 0.4504 | 0.9024 |
0.0001 | 36.0 | 1008 | 0.4504 | 0.9024 |
0.0001 | 37.0 | 1036 | 0.4513 | 0.9024 |
0.0001 | 38.0 | 1064 | 0.4525 | 0.9024 |
0.0001 | 39.0 | 1092 | 0.4530 | 0.9024 |
0.0001 | 40.0 | 1120 | 0.4533 | 0.9024 |
0.0001 | 41.0 | 1148 | 0.4538 | 0.9024 |
0.0001 | 42.0 | 1176 | 0.4539 | 0.9024 |
0.0001 | 43.0 | 1204 | 0.4547 | 0.9024 |
0.0001 | 44.0 | 1232 | 0.4551 | 0.9024 |
0.0001 | 45.0 | 1260 | 0.4551 | 0.9024 |
0.0001 | 46.0 | 1288 | 0.4551 | 0.9024 |
0.0001 | 47.0 | 1316 | 0.4553 | 0.9024 |
0.0001 | 48.0 | 1344 | 0.4553 | 0.9024 |
0.0001 | 49.0 | 1372 | 0.4553 | 0.9024 |
0.0001 | 50.0 | 1400 | 0.4553 | 0.9024 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 5
Inference API (serverless) is not available, repository is disabled.
Model tree for hkivancoral/hushem_5x_deit_base_adamax_0001_fold5
Base model
facebook/deit-base-patch16-224
Finetuned
this model