bart-base-finetuned-samsum
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: 1.5273
- Rouge1: 46.8865
- Rouge2: 23.8976
- Rougel: 39.8604
- Rougelsum: 43.0185
- Gen Len: 18.0659
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.0008 | 1.0 | 921 | 1.6050 | 45.4152 | 21.5898 | 38.2192 | 41.5283 | 18.3272 |
| 1.6741 | 2.0 | 1842 | 1.5611 | 45.6316 | 22.7331 | 38.6353 | 42.0206 | 17.9963 |
| 1.547 | 3.0 | 2763 | 1.5362 | 46.4511 | 23.218 | 39.1461 | 42.4645 | 17.9255 |
| 1.4668 | 4.0 | 3684 | 1.5338 | 46.8899 | 23.7554 | 39.7789 | 43.0769 | 18.3553 |
| 1.4218 | 5.0 | 4605 | 1.5273 | 46.8865 | 23.8976 | 39.8604 | 43.0185 | 18.0659 |
Framework versions
- Transformers 4.40.1
- Pytorch 1.13.1+cu117
- Datasets 2.19.0
- Tokenizers 0.19.1