<!-- 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_small
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: 2.0455
- Accuracy: 0.815
- Precision: 0.5
- Recall: 0.3784
- F1: 0.4308
- D-index: 1.5791
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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1600
- 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 | 200 | 0.4205 | 0.84 | 0.7778 | 0.1892 | 0.3043 | 1.5473 |
| No log | 2.0 | 400 | 0.5287 | 0.785 | 0.425 | 0.4595 | 0.4416 | 1.5664 |
| 0.4788 | 3.0 | 600 | 0.8663 | 0.78 | 0.4146 | 0.4595 | 0.4359 | 1.5597 |
| 0.4788 | 4.0 | 800 | 1.0432 | 0.8 | 0.4681 | 0.5946 | 0.5238 | 1.6309 |
| 0.2168 | 5.0 | 1000 | 1.2325 | 0.795 | 0.375 | 0.1622 | 0.2264 | 1.4766 |
| 0.2168 | 6.0 | 1200 | 1.3369 | 0.815 | 0.5 | 0.2432 | 0.3273 | 1.5326 |
| 0.2168 | 7.0 | 1400 | 1.4949 | 0.785 | 0.4286 | 0.4865 | 0.4557 | 1.5754 |
| 0.0682 | 8.0 | 1600 | 1.4499 | 0.815 | 0.5 | 0.3514 | 0.4127 | 1.5700 |
| 0.0682 | 9.0 | 1800 | 1.7761 | 0.8 | 0.4348 | 0.2703 | 0.3333 | 1.5218 |
| 0.0154 | 10.0 | 2000 | 1.8939 | 0.805 | 0.4375 | 0.1892 | 0.2642 | 1.5000 |
| 0.0154 | 11.0 | 2200 | 1.9630 | 0.8 | 0.4211 | 0.2162 | 0.2857 | 1.5028 |
| 0.0154 | 12.0 | 2400 | 1.9712 | 0.805 | 0.4545 | 0.2703 | 0.3390 | 1.5286 |
| 0.0132 | 13.0 | 2600 | 1.9184 | 0.805 | 0.4737 | 0.4865 | 0.4800 | 1.6021 |
| 0.0132 | 14.0 | 2800 | 1.9261 | 0.805 | 0.4706 | 0.4324 | 0.4507 | 1.5841 |
| 0.0 | 15.0 | 3000 | 1.9619 | 0.815 | 0.5 | 0.4054 | 0.4478 | 1.5883 |
| 0.0 | 16.0 | 3200 | 1.9798 | 0.82 | 0.5172 | 0.4054 | 0.4545 | 1.5949 |
| 0.0 | 17.0 | 3400 | 2.0126 | 0.815 | 0.5 | 0.3784 | 0.4308 | 1.5791 |
| 0.0 | 18.0 | 3600 | 2.0203 | 0.82 | 0.5185 | 0.3784 | 0.4375 | 1.5858 |
| 0.0 | 19.0 | 3800 | 2.0286 | 0.82 | 0.5185 | 0.3784 | 0.4375 | 1.5858 |
| 0.0 | 20.0 | 4000 | 2.0455 | 0.815 | 0.5 | 0.3784 | 0.4308 | 1.5791 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
