bart-base-legal-summarizer
This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.3425
- Rouge1: 33.2731
- Rouge2: 16.5293
- Rougel: 29.1491
- Rougelsum: 29.7726
- Gen Len: 15.5
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: 2
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 178 | 2.4487 | 27.4622 | 10.7691 | 23.652 | 24.1204 | 15.1333 |
| No log | 2.0 | 356 | 2.3498 | 34.4 | 17.7132 | 30.1947 | 30.5913 | 14.8222 |
| 2.2551 | 3.0 | 534 | 2.3425 | 33.2731 | 16.5293 | 29.1491 | 29.7726 | 15.5 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.19.1