|
--- |
|
base_model: google/pegasus-large |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- eur-lex-sum |
|
model-index: |
|
- name: Pegasus_no_extraction_V1 |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# Pegasus_no_extraction_V1 |
|
|
|
This model is a fine-tuned version of [google/pegasus-large](https://huggingface.co/google/pegasus-large) on the eur-lex-sum dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 2.1652 |
|
|
|
## 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: 5e-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: 0.1 |
|
- num_epochs: 40 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-------:|:----:|:---------------:| |
|
| 5.0055 | 0.9927 | 68 | 3.8532 | |
|
| 3.5972 | 2.0 | 137 | 3.1452 | |
|
| 3.0381 | 2.9927 | 205 | 2.6382 | |
|
| 2.6047 | 4.0 | 274 | 2.4301 | |
|
| 2.4507 | 4.9927 | 342 | 2.3523 | |
|
| 2.3115 | 6.0 | 411 | 2.3044 | |
|
| 2.2714 | 6.9927 | 479 | 2.2742 | |
|
| 2.1854 | 8.0 | 548 | 2.2519 | |
|
| 2.1754 | 8.9927 | 616 | 2.2373 | |
|
| 2.1119 | 10.0 | 685 | 2.2238 | |
|
| 2.114 | 10.9927 | 753 | 2.2155 | |
|
| 2.0527 | 12.0 | 822 | 2.2069 | |
|
| 2.0622 | 12.9927 | 890 | 2.2004 | |
|
| 2.0085 | 14.0 | 959 | 2.1929 | |
|
| 2.0133 | 14.9927 | 1027 | 2.1902 | |
|
| 1.9695 | 16.0 | 1096 | 2.1857 | |
|
| 1.9827 | 16.9927 | 1164 | 2.1826 | |
|
| 1.9389 | 18.0 | 1233 | 2.1789 | |
|
| 1.9541 | 18.9927 | 1301 | 2.1758 | |
|
| 1.9105 | 20.0 | 1370 | 2.1749 | |
|
| 1.9244 | 20.9927 | 1438 | 2.1731 | |
|
| 1.8858 | 22.0 | 1507 | 2.1716 | |
|
| 1.9028 | 22.9927 | 1575 | 2.1712 | |
|
| 1.8625 | 24.0 | 1644 | 2.1698 | |
|
| 1.8798 | 24.9927 | 1712 | 2.1695 | |
|
| 1.847 | 26.0 | 1781 | 2.1678 | |
|
| 1.8648 | 26.9927 | 1849 | 2.1698 | |
|
| 1.8283 | 28.0 | 1918 | 2.1671 | |
|
| 1.8517 | 28.9927 | 1986 | 2.1656 | |
|
| 1.8166 | 30.0 | 2055 | 2.1658 | |
|
| 1.8391 | 30.9927 | 2123 | 2.1646 | |
|
| 1.8068 | 32.0 | 2192 | 2.1670 | |
|
| 1.8307 | 32.9927 | 2260 | 2.1654 | |
|
| 1.7956 | 34.0 | 2329 | 2.1656 | |
|
| 1.8219 | 34.9927 | 2397 | 2.1647 | |
|
| 1.7936 | 36.0 | 2466 | 2.1652 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.1 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.17.1 |
|
- Tokenizers 0.19.1 |
|
|
