bart-base_fine_tuned_on_billsum_by_hydra_1751809688_
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: 2.4507
- Rouge1: 16.5702
- Rouge2: 7.1924
- Rougel: 13.6612
- Rougelsum: 14.9246
- Gen Len: 21.0
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.9409 | 1.0 | 1185 | 2.4586 | 16.2787 | 7.0658 | 13.5385 | 14.7208 | 21.0 |
| 1.7808 | 2.0 | 2370 | 2.4507 | 16.5702 | 7.1924 | 13.6612 | 14.9246 | 21.0 |
Framework versions
- Transformers 4.53.1
- Pytorch 2.7.1+cu128
- Datasets 3.6.0
- Tokenizers 0.21.2