|
--- |
|
license: cc-by-sa-4.0 |
|
base_model: nlpaueb/legal-bert-base-uncased |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
model-index: |
|
- name: legal-bert-lora |
|
results: [] |
|
--- |
|
|
|
<!-- 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-lora |
|
|
|
This model is a fine-tuned version of [nlpaueb/legal-bert-base-uncased](https://huggingface.co/nlpaueb/legal-bert-base-uncased) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.7058 |
|
- Accuracy: 0.7847 |
|
- Precision: 0.7702 |
|
- Recall: 0.7847 |
|
- Precision Macro: 0.5452 |
|
- Recall Macro: 0.5400 |
|
- Macro Fpr: 0.0199 |
|
- Weighted Fpr: 0.0192 |
|
- Weighted Specificity: 0.9737 |
|
- Macro Specificity: 0.9839 |
|
- Weighted Sensitivity: 0.7847 |
|
- Macro Sensitivity: 0.5400 |
|
- F1 Micro: 0.7847 |
|
- F1 Macro: 0.5165 |
|
- F1 Weighted: 0.7676 |
|
|
|
## 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: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 32 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | Precision Macro | Recall Macro | Macro Fpr | Weighted Fpr | Weighted Specificity | Macro Specificity | Weighted Sensitivity | Macro Sensitivity | F1 Micro | F1 Macro | F1 Weighted | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:---------------:|:------------:|:---------:|:------------:|:--------------------:|:-----------------:|:--------------------:|:-----------------:|:--------:|:--------:|:-----------:| |
|
| No log | 1.0 | 160 | 1.3252 | 0.6297 | 0.5643 | 0.6297 | 0.2865 | 0.3110 | 0.0417 | 0.0403 | 0.9455 | 0.9717 | 0.6297 | 0.3110 | 0.6297 | 0.2742 | 0.5694 | |
|
| No log | 2.0 | 321 | 0.8870 | 0.7312 | 0.6873 | 0.7312 | 0.3742 | 0.4525 | 0.0257 | 0.0256 | 0.9690 | 0.9800 | 0.7312 | 0.4525 | 0.7312 | 0.3967 | 0.6996 | |
|
| No log | 3.0 | 482 | 0.7794 | 0.7483 | 0.7169 | 0.7483 | 0.4059 | 0.4680 | 0.0239 | 0.0235 | 0.9711 | 0.9813 | 0.7483 | 0.4680 | 0.7483 | 0.4262 | 0.7282 | |
|
| 1.2835 | 4.0 | 643 | 0.7481 | 0.7444 | 0.7085 | 0.7444 | 0.3997 | 0.4588 | 0.0243 | 0.0239 | 0.9700 | 0.9810 | 0.7444 | 0.4588 | 0.7444 | 0.4100 | 0.7146 | |
|
| 1.2835 | 5.0 | 803 | 0.7360 | 0.7630 | 0.7245 | 0.7630 | 0.4176 | 0.4763 | 0.0226 | 0.0217 | 0.9702 | 0.9822 | 0.7630 | 0.4763 | 0.7630 | 0.4350 | 0.7372 | |
|
| 1.2835 | 6.0 | 964 | 0.7146 | 0.7738 | 0.7790 | 0.7738 | 0.5020 | 0.4907 | 0.0209 | 0.0205 | 0.9730 | 0.9831 | 0.7738 | 0.4907 | 0.7738 | 0.4514 | 0.7549 | |
|
| 0.6494 | 7.0 | 1125 | 0.7362 | 0.7607 | 0.7519 | 0.7607 | 0.5232 | 0.4890 | 0.0225 | 0.0220 | 0.9724 | 0.9822 | 0.7607 | 0.4890 | 0.7607 | 0.4556 | 0.7390 | |
|
| 0.6494 | 8.0 | 1286 | 0.7271 | 0.7800 | 0.7639 | 0.7800 | 0.5348 | 0.5171 | 0.0205 | 0.0197 | 0.9731 | 0.9835 | 0.7800 | 0.5171 | 0.7800 | 0.4923 | 0.7617 | |
|
| 0.6494 | 9.0 | 1446 | 0.7068 | 0.7847 | 0.7739 | 0.7847 | 0.5495 | 0.5205 | 0.0199 | 0.0192 | 0.9744 | 0.9839 | 0.7847 | 0.5205 | 0.7847 | 0.4943 | 0.7665 | |
|
| 0.5284 | 9.95 | 1600 | 0.7058 | 0.7847 | 0.7702 | 0.7847 | 0.5452 | 0.5400 | 0.0199 | 0.0192 | 0.9737 | 0.9839 | 0.7847 | 0.5400 | 0.7847 | 0.5165 | 0.7676 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.1 |
|
|
