[BUG/Help] 为什么max_length指定到非常大了，输出还是不完整？

[pipidoudou]

Is there an existing issue for this?

• [X] I have searched the existing issues

Current Behavior

运行web_demo2.py，指定了max_length到最大了，top_p=0.8，temperature=0.1。 只问一个问题，input大概500个字，输出每次都停在同样一个字上。 再输入“继续”又能继续输出。

Expected Behavior

能一次性完整输出答案

Steps To Reproduce

1.配置本地目录 2.streamlit run web_demo2.py

Environment

- OS:Windows10
- Python:3.10
- Transformers:4.30.2
- PyTorch:2.0.1
- CUDA Support (`python -c "import torch; print(torch.cuda.is_available())"`) :True

Anything else?

请问是哪里还需要再配置？