<!-- 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. -->
xlnet-base-cased_fold_4_binary_v1
This model is a fine-tuned version of xlnet-base-cased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.5724
- F1: 0.8315
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 25
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|---|---|---|---|---|
| No log | 1.0 | 289 | 0.4043 | 0.8009 |
| 0.4373 | 2.0 | 578 | 0.4093 | 0.8260 |
| 0.4373 | 3.0 | 867 | 0.5084 | 0.8206 |
| 0.2707 | 4.0 | 1156 | 0.5945 | 0.8087 |
| 0.2707 | 5.0 | 1445 | 0.6389 | 0.8251 |
| 0.1691 | 6.0 | 1734 | 0.8131 | 0.8156 |
| 0.1012 | 7.0 | 2023 | 0.9865 | 0.8190 |
| 0.1012 | 8.0 | 2312 | 1.1356 | 0.8342 |
| 0.0506 | 9.0 | 2601 | 1.0624 | 0.8369 |
| 0.0506 | 10.0 | 2890 | 1.2604 | 0.8255 |
| 0.0384 | 11.0 | 3179 | 1.2648 | 0.8183 |
| 0.0384 | 12.0 | 3468 | 1.3763 | 0.8158 |
| 0.0318 | 13.0 | 3757 | 1.4966 | 0.8217 |
| 0.0221 | 14.0 | 4046 | 1.3889 | 0.8250 |
| 0.0221 | 15.0 | 4335 | 1.4014 | 0.8284 |
| 0.0145 | 16.0 | 4624 | 1.5321 | 0.8289 |
| 0.0145 | 17.0 | 4913 | 1.4914 | 0.8233 |
| 0.0172 | 18.0 | 5202 | 1.3946 | 0.8314 |
| 0.0172 | 19.0 | 5491 | 1.5032 | 0.8269 |
| 0.0135 | 20.0 | 5780 | 1.5111 | 0.8328 |
| 0.0087 | 21.0 | 6069 | 1.4899 | 0.8318 |
| 0.0087 | 22.0 | 6358 | 1.5562 | 0.8311 |
| 0.0061 | 23.0 | 6647 | 1.5384 | 0.8327 |
| 0.0061 | 24.0 | 6936 | 1.5798 | 0.8304 |
| 0.0052 | 25.0 | 7225 | 1.5724 | 0.8315 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
