|
--- |
|
license: apache-2.0 |
|
base_model: google/vit-base-patch16-224-in21k |
|
tags: |
|
- image-classification |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: test-hasy-6 |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# test-hasy-6 |
|
|
|
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co./google/vit-base-patch16-224-in21k) on the HASY dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6506 |
|
- Accuracy: 0.8025 |
|
|
|
## 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: 2e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 1787 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 50 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:-----:|:---------------:|:--------:| |
|
| 3.0857 | 1.0 | 541 | 2.4484 | 0.5572 | |
|
| 2.3006 | 2.0 | 1082 | 2.1588 | 0.5904 | |
|
| 2.4406 | 3.0 | 1623 | 1.8879 | 0.6445 | |
|
| 2.342 | 4.0 | 2164 | 1.7122 | 0.6674 | |
|
| 2.1852 | 5.0 | 2705 | 1.5467 | 0.6923 | |
|
| 1.9535 | 6.0 | 3246 | 1.4113 | 0.7048 | |
|
| 1.9061 | 7.0 | 3787 | 1.3136 | 0.6881 | |
|
| 1.5934 | 8.0 | 4328 | 1.2059 | 0.7089 | |
|
| 1.8755 | 9.0 | 4869 | 1.1638 | 0.7173 | |
|
| 1.6319 | 10.0 | 5410 | 1.1024 | 0.7235 | |
|
| 1.5899 | 11.0 | 5951 | 1.0375 | 0.7339 | |
|
| 1.6427 | 12.0 | 6492 | 0.9656 | 0.7526 | |
|
| 1.8022 | 13.0 | 7033 | 0.9760 | 0.7422 | |
|
| 1.7161 | 14.0 | 7574 | 0.8952 | 0.7609 | |
|
| 1.2123 | 15.0 | 8115 | 0.8750 | 0.7692 | |
|
| 1.5721 | 16.0 | 8656 | 0.8586 | 0.7755 | |
|
| 1.7482 | 17.0 | 9197 | 0.8279 | 0.7755 | |
|
| 1.5992 | 18.0 | 9738 | 0.8321 | 0.7547 | |
|
| 1.8179 | 19.0 | 10279 | 0.7898 | 0.7817 | |
|
| 1.2744 | 20.0 | 10820 | 0.7984 | 0.7672 | |
|
| 1.2221 | 21.0 | 11361 | 0.7757 | 0.7734 | |
|
| 1.4893 | 22.0 | 11902 | 0.7512 | 0.7817 | |
|
| 1.5184 | 23.0 | 12443 | 0.7512 | 0.7817 | |
|
| 1.6562 | 24.0 | 12984 | 0.7514 | 0.7796 | |
|
| 1.4148 | 25.0 | 13525 | 0.7241 | 0.7817 | |
|
| 1.2765 | 26.0 | 14066 | 0.6907 | 0.8046 | |
|
| 1.3378 | 27.0 | 14607 | 0.7132 | 0.7900 | |
|
| 1.5446 | 28.0 | 15148 | 0.6973 | 0.7963 | |
|
| 1.1969 | 29.0 | 15689 | 0.7010 | 0.7921 | |
|
| 1.3721 | 30.0 | 16230 | 0.6928 | 0.8004 | |
|
| 1.4051 | 31.0 | 16771 | 0.6976 | 0.7921 | |
|
| 1.1004 | 32.0 | 17312 | 0.6785 | 0.8004 | |
|
| 1.2668 | 33.0 | 17853 | 0.6883 | 0.7817 | |
|
| 1.0728 | 34.0 | 18394 | 0.6924 | 0.7859 | |
|
| 1.1856 | 35.0 | 18935 | 0.6840 | 0.7921 | |
|
| 1.2387 | 36.0 | 19476 | 0.6739 | 0.8025 | |
|
| 1.5242 | 37.0 | 20017 | 0.6554 | 0.7963 | |
|
| 1.351 | 38.0 | 20558 | 0.6736 | 0.7942 | |
|
| 1.2441 | 39.0 | 21099 | 0.6659 | 0.8046 | |
|
| 1.2113 | 40.0 | 21640 | 0.6709 | 0.7983 | |
|
| 1.1608 | 41.0 | 22181 | 0.6630 | 0.7983 | |
|
| 1.266 | 42.0 | 22722 | 0.6693 | 0.8004 | |
|
| 0.9426 | 43.0 | 23263 | 0.6639 | 0.8046 | |
|
| 1.0066 | 44.0 | 23804 | 0.6636 | 0.8025 | |
|
| 1.0856 | 45.0 | 24345 | 0.6530 | 0.8004 | |
|
| 1.0128 | 46.0 | 24886 | 0.6506 | 0.8025 | |
|
| 1.0369 | 47.0 | 25427 | 0.6617 | 0.8025 | |
|
| 1.1458 | 48.0 | 25968 | 0.6546 | 0.8004 | |
|
| 1.0696 | 49.0 | 26509 | 0.6597 | 0.7942 | |
|
| 1.2227 | 50.0 | 27050 | 0.6566 | 0.7942 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
