Views
No views yet
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
|---|---|---|---|---|---|
| 通用 General | 自然语言生成 NLG | 闻仲 Wenzhong | GPT2 | 110M | 中文 Chinese |
1from transformers import BertTokenizer,GPT2LMHeadModel
2hf_model_path = 'IDEA-CCNL/Wenzhong2.0-GPT2-110M-BertTokenizer-chinese'
3tokenizer = BertTokenizer.from_pretrained(hf_model_path)
4model = GPT2LMHeadModel.from_pretrained(hf_model_path)1def generate_word_level(input_text,n_return=5,max_length=128,top_p=0.9):
2 inputs = tokenizer(input_text,return_tensors='pt',add_special_tokens=False).to(model.device)
3 gen = model.generate(
4 inputs=inputs['input_ids'],
5 max_length=max_length,
6 do_sample=True,
7 top_p=top_p,
8 eos_token_id=21133,
9 pad_token_id=0,
10 num_return_sequences=n_return)
11
12 sentences = tokenizer.batch_decode(gen)
13 for idx,sentence in enumerate(sentences):
14 print(f'sentence {idx}: {sentence}')
15 print('*'*20)
16 return gen
17
18outputs = generate_word_level('西湖的景色',n_return=5,max_length=128)1@article{fengshenbang,
2 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},
3 title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
4 journal = {CoRR},
5 volume = {abs/2209.02970},
6 year = {2022}
7}1@misc{Fengshenbang-LM,
2 title={Fengshenbang-LM},
3 author={IDEA-CCNL},
4 year={2021},
5 howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
6}