<!-- 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_5_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: 1.7395
- F1: 0.8206
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 | 288 | 0.4246 | 0.8154 |
| 0.4211 | 2.0 | 576 | 0.5181 | 0.8063 |
| 0.4211 | 3.0 | 864 | 0.4939 | 0.8149 |
| 0.2483 | 4.0 | 1152 | 0.6181 | 0.8227 |
| 0.2483 | 5.0 | 1440 | 0.9251 | 0.8006 |
| 0.1512 | 6.0 | 1728 | 0.9639 | 0.8082 |
| 0.0858 | 7.0 | 2016 | 1.1315 | 0.8074 |
| 0.0858 | 8.0 | 2304 | 1.1322 | 0.8303 |
| 0.053 | 9.0 | 2592 | 1.3171 | 0.8017 |
| 0.053 | 10.0 | 2880 | 1.3729 | 0.8100 |
| 0.0325 | 11.0 | 3168 | 1.2708 | 0.8252 |
| 0.0325 | 12.0 | 3456 | 1.5105 | 0.8242 |
| 0.0203 | 13.0 | 3744 | 1.4902 | 0.8233 |
| 0.0179 | 14.0 | 4032 | 1.5874 | 0.8194 |
| 0.0179 | 15.0 | 4320 | 1.5933 | 0.8135 |
| 0.0174 | 16.0 | 4608 | 1.5908 | 0.8088 |
| 0.0174 | 17.0 | 4896 | 1.5692 | 0.8249 |
| 0.0129 | 18.0 | 5184 | 1.6597 | 0.8167 |
| 0.0129 | 19.0 | 5472 | 1.6009 | 0.8218 |
| 0.0095 | 20.0 | 5760 | 1.6962 | 0.8225 |
| 0.0062 | 21.0 | 6048 | 1.7075 | 0.8182 |
| 0.0062 | 22.0 | 6336 | 1.7335 | 0.8181 |
| 0.0077 | 23.0 | 6624 | 1.7175 | 0.8204 |
| 0.0077 | 24.0 | 6912 | 1.7680 | 0.8187 |
| 0.0024 | 25.0 | 7200 | 1.7395 | 0.8206 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
