<!-- 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. -->
twhin-bert-base-finetuned-twhin-epoch
This model is a fine-tuned version of Twitter/twhin-bert-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8164
- Precision: 0.8381
- Recall: 0.8347
- F1: 0.8360
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: 1e-05
- train_batch_size: 36
- eval_batch_size: 36
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 |
|---|---|---|---|---|---|---|
| No log | 1.0 | 52 | 0.5851 | 0.7441 | 0.7796 | 0.7613 |
| No log | 2.0 | 104 | 0.5324 | 0.7630 | 0.7861 | 0.7700 |
| No log | 3.0 | 156 | 0.4893 | 0.7774 | 0.8104 | 0.7923 |
| No log | 4.0 | 208 | 0.5204 | 0.7862 | 0.8104 | 0.7936 |
| No log | 5.0 | 260 | 0.5753 | 0.7728 | 0.8120 | 0.7907 |
| No log | 6.0 | 312 | 0.5552 | 0.7729 | 0.8071 | 0.7889 |
| No log | 7.0 | 364 | 0.5975 | 0.7768 | 0.8136 | 0.7946 |
| No log | 8.0 | 416 | 0.6527 | 0.8015 | 0.8055 | 0.7915 |
| No log | 9.0 | 468 | 0.6521 | 0.8285 | 0.8233 | 0.8252 |
| 0.3755 | 10.0 | 520 | 0.6629 | 0.8315 | 0.8104 | 0.8175 |
| 0.3755 | 11.0 | 572 | 0.7238 | 0.8260 | 0.8266 | 0.8263 |
| 0.3755 | 12.0 | 624 | 0.7782 | 0.8318 | 0.8201 | 0.8239 |
| 0.3755 | 13.0 | 676 | 0.7788 | 0.8263 | 0.8266 | 0.8260 |
| 0.3755 | 14.0 | 728 | 0.8164 | 0.8381 | 0.8347 | 0.8360 |
| 0.3755 | 15.0 | 780 | 0.8701 | 0.8238 | 0.8201 | 0.8212 |
| 0.3755 | 16.0 | 832 | 0.8774 | 0.8295 | 0.8282 | 0.8288 |
| 0.3755 | 17.0 | 884 | 0.9193 | 0.8311 | 0.8233 | 0.8259 |
| 0.3755 | 18.0 | 936 | 0.9321 | 0.8339 | 0.8282 | 0.8299 |
| 0.3755 | 19.0 | 988 | 0.9350 | 0.8307 | 0.8233 | 0.8261 |
| 0.0554 | 20.0 | 1040 | 0.9344 | 0.8256 | 0.8185 | 0.8213 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu117
- Datasets 2.10.1
- Tokenizers 0.13.2
