善于处理自然语言转换任务，是中文版的BART-base模型。
Good at solving NLT tasks, Chinese BART-base.
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
|---|---|---|---|---|---|
| 通用 General | 自然语言转换 NLT | 燃灯 Randeng | BART | 139M | 中文-Chinese |
为了得到一个中文版的BART-base，我们用悟道语料库(180G版本)进行预训练。具体地，我们在预训练阶段中使用了封神框架，大约需要8张A100显卡，耗时3天左右。
To get a Chinese BART-base, we use WuDao Corpora (180 GB version) for pre-training. Specifically, we use the fengshen framework in the pre-training phase which cost about 3 days with 8 A100 GPUs.
from transformers import BartForConditionalGeneration, AutoTokenizer, Text2TextGenerationPipeline
import torch
tokenizer=AutoTokenizer.from_pretrained('IDEA-CCNL/Randeng-BART-139M', use_fast=false)
model=BartForConditionalGeneration.from_pretrained('IDEA-CCNL/Randeng-BART-139M')
text = '桂林市是世界闻名<mask> ，它有悠久的<mask>'
text2text_generator = Text2TextGenerationPipeline(model, tokenizer)
print(text2text_generator(text, max_length=50, do_sample=False))如果您在您的工作中使用了我们的模型，可以引用我们的论文：
如果您在工作中使用了该资源，请引用我们的论文：
@article{fengshenbang,
author = {Jiaxing Zhang and Ruyi Gan and Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}也可以引用我们的网站:
您也可以引用我们的网站:
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}