<!-- 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. -->
legal_bert_sm_cv_defined_summarized_4
This model is a fine-tuned version of nlpaueb/legal-bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.7595
- Accuracy: 0.811
- Precision: 0.5385
- Recall: 0.2154
- F1: 0.3077
- D-index: 1.5216
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: 5e-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
- lr_scheduler_warmup_steps: 8000
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | D-index |
|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 250 | 0.4882 | 0.805 | 0.0 | 0.0 | 0.0 | 1.4370 |
| 0.5662 | 2.0 | 500 | 0.4681 | 0.805 | 0.0 | 0.0 | 0.0 | 1.4370 |
| 0.5662 | 3.0 | 750 | 0.4649 | 0.807 | 0.625 | 0.0256 | 0.0493 | 1.4491 |
| 0.4397 | 4.0 | 1000 | 0.4675 | 0.819 | 0.7692 | 0.1026 | 0.1810 | 1.4931 |
| 0.4397 | 5.0 | 1250 | 0.5234 | 0.816 | 0.7391 | 0.0872 | 0.1560 | 1.4836 |
| 0.3492 | 6.0 | 1500 | 0.5137 | 0.825 | 0.6562 | 0.2154 | 0.3243 | 1.5406 |
| 0.3492 | 7.0 | 1750 | 0.5490 | 0.81 | 0.5490 | 0.1436 | 0.2276 | 1.4952 |
| 0.2409 | 8.0 | 2000 | 0.6896 | 0.82 | 0.5882 | 0.2564 | 0.3571 | 1.5478 |
| 0.2409 | 9.0 | 2250 | 0.7600 | 0.808 | 0.5155 | 0.2564 | 0.3425 | 1.5316 |
| 0.1506 | 10.0 | 2500 | 1.0232 | 0.813 | 0.5714 | 0.1641 | 0.2550 | 1.5065 |
| 0.1506 | 11.0 | 2750 | 1.0855 | 0.823 | 0.6731 | 0.1795 | 0.2834 | 1.5255 |
| 0.0851 | 12.0 | 3000 | 1.1956 | 0.797 | 0.4655 | 0.2769 | 0.3473 | 1.5236 |
| 0.0851 | 13.0 | 3250 | 1.2379 | 0.808 | 0.5190 | 0.2103 | 0.2993 | 1.5157 |
| 0.0538 | 14.0 | 3500 | 1.4613 | 0.807 | 0.5143 | 0.1846 | 0.2717 | 1.5055 |
| 0.0538 | 15.0 | 3750 | 1.4960 | 0.815 | 0.5658 | 0.2205 | 0.3173 | 1.5288 |
| 0.0334 | 16.0 | 4000 | 1.6423 | 0.806 | 0.5067 | 0.1949 | 0.2815 | 1.5076 |
| 0.0334 | 17.0 | 4250 | 1.6386 | 0.804 | 0.4958 | 0.3026 | 0.3758 | 1.5419 |
| 0.0364 | 18.0 | 4500 | 1.6520 | 0.797 | 0.45 | 0.1846 | 0.2618 | 1.4917 |
| 0.0364 | 19.0 | 4750 | 1.6842 | 0.804 | 0.4953 | 0.2718 | 0.3510 | 1.5314 |
| 0.0167 | 20.0 | 5000 | 1.7595 | 0.811 | 0.5385 | 0.2154 | 0.3077 | 1.5216 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
