|
--- |
|
license: mit |
|
base_model: facebook/bart-large-cnn |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: bart-large-cnn-finetuned-small-context-news-1000 |
|
results: [] |
|
pipeline_tag: summarization |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# bart-large-cnn-finetuned-small-context-news-1000 |
|
|
|
This model is a fine-tuned version of [facebook/bart-large-cnn](https://huggingface.co/facebook/bart-large-cnn) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.9930 |
|
- Rouge1: 65.1207 |
|
- Rouge2: 55.5654 |
|
- Rougel: 60.1703 |
|
- Rougelsum: 61.6717 |
|
- Gen Len: 66.6529 |
|
|
|
## 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 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 20 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len | |
|
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:| |
|
| No log | 1.0 | 85 | 0.4915 | 61.0185 | 47.1863 | 53.5499 | 55.4476 | 66.2824 | |
|
| No log | 2.0 | 170 | 0.5558 | 63.1675 | 51.7011 | 57.0742 | 58.1801 | 67.2235 | |
|
| No log | 3.0 | 255 | 0.5447 | 64.6201 | 54.8904 | 59.8669 | 60.7456 | 67.4529 | |
|
| No log | 4.0 | 340 | 0.5770 | 65.2542 | 54.571 | 59.89 | 61.0988 | 65.0941 | |
|
| No log | 5.0 | 425 | 0.6406 | 64.8868 | 54.2641 | 59.2758 | 60.4861 | 67.4118 | |
|
| 0.2062 | 6.0 | 510 | 0.6468 | 65.1216 | 54.5784 | 59.3594 | 60.3826 | 66.7529 | |
|
| 0.2062 | 7.0 | 595 | 0.6828 | 64.162 | 54.1786 | 59.1392 | 60.2517 | 67.4412 | |
|
| 0.2062 | 8.0 | 680 | 0.7481 | 64.6093 | 54.4423 | 59.9194 | 61.1767 | 66.2647 | |
|
| 0.2062 | 9.0 | 765 | 0.7916 | 65.0347 | 55.2975 | 60.3007 | 61.4619 | 67.8471 | |
|
| 0.2062 | 10.0 | 850 | 0.7699 | 65.672 | 55.5276 | 60.3711 | 61.5138 | 66.9529 | |
|
| 0.2062 | 11.0 | 935 | 0.7712 | 65.7327 | 55.9363 | 61.0215 | 62.1639 | 65.7294 | |
|
| 0.0273 | 12.0 | 1020 | 0.9920 | 65.2328 | 55.3817 | 60.0671 | 61.4812 | 66.3588 | |
|
| 0.0273 | 13.0 | 1105 | 0.8023 | 65.2372 | 55.2458 | 60.2251 | 61.5193 | 65.4824 | |
|
| 0.0273 | 14.0 | 1190 | 0.8660 | 65.0369 | 55.2548 | 59.8089 | 61.3785 | 68.0353 | |
|
| 0.0273 | 15.0 | 1275 | 0.9539 | 65.4251 | 55.1068 | 60.2355 | 61.6598 | 66.7765 | |
|
| 0.0273 | 16.0 | 1360 | 0.8840 | 65.544 | 55.951 | 59.9112 | 61.6029 | 66.7529 | |
|
| 0.0273 | 17.0 | 1445 | 0.9141 | 65.7685 | 55.4981 | 60.575 | 62.2381 | 66.4882 | |
|
| 0.009 | 18.0 | 1530 | 1.0024 | 65.4152 | 55.7546 | 60.5256 | 62.0985 | 67.2412 | |
|
| 0.009 | 19.0 | 1615 | 0.9997 | 65.0153 | 55.1772 | 60.103 | 61.4286 | 66.3529 | |
|
| 0.009 | 20.0 | 1700 | 0.9930 | 65.1207 | 55.5654 | 60.1703 | 61.6717 | 66.6529 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.1 |
|
- Pytorch 2.1.2 |
|
- Datasets 2.1.0 |
|
- Tokenizers 0.15.2 |