metadata
license: cc-by-sa-4.0
base_model: frisibeli/legal-bert-finetuned-1985
tags:
- generated_from_trainer
model-index:
- name: legal-bert-finetuned-1985-classifier
results: []
legal-bert-finetuned-1985-classifier
This model is a fine-tuned version of frisibeli/legal-bert-finetuned-1985 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.3472
- F1 Score: 0.5128
- F1 Combined: 0.604040404040404
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.02
- num_epochs: 7
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Score |
|---|---|---|---|---|
| 0.5817 | 0.99 | 82 | 0.5112 | 0.4419 |
| 0.4875 | 2.0 | 165 | 0.5126 | 0.4921 |
| 0.2992 | 2.99 | 247 | 0.6876 | 0.5171 |
| 0.1575 | 4.0 | 330 | 0.9381 | 0.5216 |
| 0.0727 | 4.99 | 412 | 1.1702 | 0.5308 |
| 0.0453 | 6.0 | 495 | 1.2749 | 0.5084 |
| 0.0258 | 6.96 | 574 | 1.3472 | 0.5128 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.0
- Tokenizers 0.15.0