<!-- 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. -->
vit-base_rvl-cdip_r2_32
This model is a fine-tuned version of jordyvl/vit-base_rvl-cdip on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6372
- Accuracy: 0.8985
- Brier Loss: 0.1792
- Nll: 1.1736
- F1 Micro: 0.8985
- F1 Macro: 0.8987
- Ece: 0.0847
- Aurc: 0.0201
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: 96
- eval_batch_size: 96
- 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: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Brier Loss | Nll | F1 Micro | F1 Macro | Ece | Aurc |
---|---|---|---|---|---|---|---|---|---|---|
0.1647 | 1.0 | 3334 | 0.4024 | 0.8887 | 0.1682 | 1.2086 | 0.8887 | 0.8891 | 0.0457 | 0.0178 |
0.1418 | 2.0 | 6668 | 0.4075 | 0.8941 | 0.1646 | 1.2066 | 0.8941 | 0.8942 | 0.0522 | 0.0177 |
0.0989 | 3.0 | 10002 | 0.4409 | 0.8932 | 0.1690 | 1.1966 | 0.8932 | 0.8932 | 0.0647 | 0.0175 |
0.0614 | 4.0 | 13336 | 0.4781 | 0.8944 | 0.1730 | 1.2083 | 0.8944 | 0.8951 | 0.0694 | 0.0181 |
0.0392 | 5.0 | 16670 | 0.5329 | 0.8959 | 0.1761 | 1.1777 | 0.8959 | 0.8958 | 0.0776 | 0.0187 |
0.0231 | 6.0 | 20004 | 0.5714 | 0.8957 | 0.1799 | 1.2083 | 0.8957 | 0.8958 | 0.0813 | 0.0198 |
0.0126 | 7.0 | 23338 | 0.6002 | 0.8966 | 0.1802 | 1.1732 | 0.8966 | 0.8972 | 0.0839 | 0.0197 |
0.0079 | 8.0 | 26672 | 0.6193 | 0.8984 | 0.1789 | 1.1849 | 0.8984 | 0.8985 | 0.0833 | 0.0200 |
0.0049 | 9.0 | 30006 | 0.6333 | 0.8976 | 0.1798 | 1.1906 | 0.8976 | 0.8978 | 0.0851 | 0.0205 |
0.0034 | 10.0 | 33340 | 0.6372 | 0.8985 | 0.1792 | 1.1736 | 0.8985 | 0.8987 | 0.0847 | 0.0201 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2