Langflow 0.4.0应用卡在思考阶段无响应问题求助

问题复现 #

• 已成功安装Langflow 0.4.0，构建「带提示词与历史记录的基础聊天」应用：在ChatOpenAI面板填入OpenAI密钥，设置input key为text后完成构建。
• 打开聊天界面输入「hello」，应用卡在思考阶段，界面如下：
  
• 测试其他应用存在相同问题，服务器可正常访问互联网。

关键日志信息 #

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.

Current conversation:

Human: hello
AI:
/mnt/data/yangpengcheng/miniconda3/envs/langflow/lib/python3.10/site-packages/tenacity/__init__.py:338: RuntimeWarning: coroutine 'AsyncRunManager.on_retry' was never awaited
  self.before_sleep(retry_state)
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
[2023-08-07 01:43:05 +0000] [88942] [INFO] ('10.12.186.19', 55850) - "WebSocket /api/v1/chat/6242294b-dad0-4698-a0cf-3766e80dd008" [accepted]
[2023-08-07 01:43:05 +0000] [88942] [INFO] connection open


> Entering new LLMChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.

Current conversation:


Human: hello
AI:


> Entering new LLMChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.

Current conversation:


Human: hello
AI:
[2023-08-07 01:51:10 +0000] [88940] [CRITICAL] WORKER TIMEOUT (pid:88942)
[2023-08-07 01:51:10 +0000] [89080] [INFO] Booting worker with pid: 89080
[2023-08-07 01:51:10 +0000] [89080] [INFO] Started server process [89080]
[2023-08-07 01:51:10 +0000] [89080] [INFO] Waiting for application startup.
[2023-08-07 01:51:10 +0000] [89080] [INFO] Application startup complete.

同类问题参考 #

Reddit社区有用户反馈相同问题：Langflow聊天功能无响应，症状与本次完全一致。

解决方案 #

从日志看，核心问题是协程未被正确等待引发的RuntimeWarning，最终导致Worker超时崩溃。可按以下步骤排查修复：

1. 升级Langflow版本：0.4.0存在已知的异步处理bug，升级到最新稳定版（如0.5.x及以上）可修复该问题。执行命令：

pip install --upgrade langflow

2. 调整Worker超时设置：修改Langflow的启动配置，延长超时时间。启动时添加参数：

langflow run --workers 1 --timeout 120

3. 验证OpenAI密钥与连通性：确认密钥权限（是否启用目标模型如gpt-3.5-turbo），可通过代码直接调用OpenAI API测试：

import openai
openai.api_key = "你的OpenAI密钥"
response = openai.ChatCompletion.create(model="gpt-3.5-turbo", messages=[{"role": "user", "content": "hello"}])
print(response.choices[0].message.content)

4. 强制同步执行：若异步问题仍存在，可切换为同步模式启动Langflow：

langflow run --no-async

内容的提问来源于stack exchange，提问作者pengchy