Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/renmada/t5-pegasus-pytorch
https://github.com/renmada/t5-pegasus-pytorch
Last synced: 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/renmada/t5-pegasus-pytorch
- Owner: renmada
- Created: 2021-03-08T05:10:53.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2024-03-04T09:02:24.000Z (3 months ago)
- Last Synced: 2024-03-04T10:27:45.908Z (3 months ago)
- Language: Python
- Size: 54.7 KB
- Stars: 380
- Watchers: 3
- Forks: 59
- Open Issues: 35
-
Metadata Files:
- Readme: README.md
Lists
- awesome-pretrained-chinese-nlp-models - t5-pegasus-pytorch
README
# t5-pegasus pytorch
## 最新更新
- 重构代码,支持更多模型
- 支持transformers最新版本
[老版代码点这里](https://github.com/renmada/t5-pegasus-pytorch/tree/legacy)
## 模型效果对比
数据集:[LCSTS_new](https://www.luge.ai/#/luge/dataDetail?id=10)
训练集取前一万条,验证集取前一千条| model | bleu | rouge-1 | rouge-2 | rouge-l |
|----------------------|-------------|---------------|--------------|--------------|
| t5-pegasus-base | 0.1276 | 0.3490 | 0.2123 | 0.3155 |
| t5-copy | 0.0938 | 0.3369 | 0.1955 | 0.3086 |
| Pegasus-238M-Chinese | 0.1200 | 0.3252 | 0.1957 | 0.2924 |
| Pegasus-523M-Chinese | 0.1233 | 0.3313 | 0.2032 | 0.2996 |
| cpt-large | **0.1366** | **0.3550** | **0.2242** | **0.3220** |
| prophet-zh | 0.1240 | 0.3419 | 0.2109 | 0.3107 |## 数据格式
[样例数据](https://github.com/renmada/t5-pegasus-pytorch/blob/legacy/examples/sample_data.json)
## huggingface模型| model_type | model_type |
|-------------|----------------------------------------|
| t5-pegasus | imxly/t5-pegasus |
| t5copy | imxly/t5-copy |
| Pegasus | IDEA-CCNL/Randeng-Pegasus-238M-Chinese |
| Pegasus | IDEA-CCNL/Randeng-Pegasus-523M-Chinese |
| cpt | fnlp/cpt-large |
| prophet | imxly/prophetnet-zh |## 训练命令
### requirements
环境可以参考这个[issue](https://github.com/renmada/t5-pegasus-pytorch/issues/58)
```
torch >=1.10.0
transformers
pytorch_lightning==1.4.9
torchmetrics==0.5.0
```
model_type见上方表格
```shell
python train.py \
--train_file train.json \
--dev_file dev.json \
--batch_size 6 \
--max_epochs 10 \
--max_source_length 512 \
--max_target_length 300 \
--model_path imxly/t5-pegasus \
--gpus 4 \
--lr 5e-5 \
--model_type t5-pegasus
```
## 参考
https://github.com/ZhuiyiTechnology/t5-pegasus
https://github.com/fastnlp/CPT
https://github.com/IDEA-CCNL/Fengshenbang-LM
https://github.com/microsoft/ProphetNet