[Bug]: docker安装版本适配spark api报错

[lluvialuo]

Installation Method | 安装方法与平台

Docker-Compose（Windows/Mac）

Version | 版本

Latest | 最新版

OS | 操作系统

Mac

Describe the bug | 简述

在mac本地使用conda安装方式，适配spark api可以正常运行。但是通过docker compose方式安装之后通过spark api会出现报错，不过千帆api则可以正常使用

Screen Shot | 有帮助的截图

Terminal Traceback & Material to Help Reproduce Bugs | 终端traceback（如有） + 帮助我们复现的测试材料样本（如有）

gpt_academic_nolocalllms-1 | error: Connection to remote host was lost. gpt_academic_nolocalllms-1 | Exception ignored in thread started by: <function SparkRequestInstance.create_blocking_request..run at 0x2aaaf7fdfa60> gpt_academic_nolocalllms-1 | Traceback (most recent call last): gpt_academic_nolocalllms-1 | File "/gpt/request_llms/com_sparkapi.py", line 113, in run gpt_academic_nolocalllms-1 | ws.send(data) gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/websocket/_app.py", line 284, in send gpt_academic_nolocalllms-1 | raise WebSocketConnectionClosedException("Connection is already closed.") gpt_academic_nolocalllms-1 | websocket._exceptions.WebSocketConnectionClosedException: Connection is already closed. gpt_academic_nolocalllms-1 | Traceback (most recent call last): gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/routes.py", line 422, in run_predict gpt_academic_nolocalllms-1 | output = await app.get_blocks().process_api( gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/blocks.py", line 1323, in process_api gpt_academic_nolocalllms-1 | result = await self.call_function( gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/blocks.py", line 1067, in call_function gpt_academic_nolocalllms-1 | prediction = await utils.async_iteration(iterator) gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/utils.py", line 336, in async_iteration gpt_academic_nolocalllms-1 | return await iterator.anext() gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/utils.py", line 329, in anext gpt_academic_nolocalllms-1 | return await anyio.to_thread.run_sync( gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/anyio/to_thread.py", line 56, in run_sync gpt_academic_nolocalllms-1 | return await get_async_backend().run_sync_in_worker_thread( gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 2134, in run_sync_in_worker_thread gpt_academic_nolocalllms-1 | return await future gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/anyio/_backends/_asyncio.py", line 851, in run gpt_academic_nolocalllms-1 | result = context.run(func, *args) gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/usr/local/lib/python3.11/site-packages/gradio/utils.py", line 312, in run_sync_iterator_async gpt_academic_nolocalllms-1 | return next(iterator) gpt_academic_nolocalllms-1 | ^^^^^^^^^^^^^^ gpt_academic_nolocalllms-1 | File "/gpt/toolbox.py", line 115, in decorated gpt_academic_nolocalllms-1 | yield from f(txt_passon, llm_kwargs, plugin_kwargs, chatbot_with_cookie, history, system_prompt, *args) gpt_academic_nolocalllms-1 | File "/gpt/request_llms/bridge_all.py", line 765, in predict gpt_academic_nolocalllms-1 | yield from method(inputs, llm_kwargs, *args, **kwargs) gpt_academic_nolocalllms-1 | File "/gpt/request_llms/bridge_spark.py", line 60, in predict gpt_academic_nolocalllms-1 | if response == f"[Local Message] 等待{model_name}响应中 ...": gpt_academic_nolocalllms-1 | ^^^^^^^^ gpt_academic_nolocalllms-1 | UnboundLocalError: cannot access local variable 'response' where it is not associated with a value

[binary-husky]

我修复了response的问题：https://github.com/binary-husky/gpt_academic/commit/e359fff0405c4cb865b809b4ecfc0a95a54d2512

但是具体到这个问题的话，大概是星火的key失效导致的

[lluvialuo]

我修复了response的问题：e359fff

但是具体到这个问题的话，大概是星火的key失效导致的

重新尝试之后还是会报错，但是报错信息变成了：“error: Connection to remote host was lost.” 我比较疑惑为啥本地运行没问题，但是docker容器运行会出问题，如果是key有问题理论上本地运行的版本也会报错呀

[binary-husky]

我修复了response的问题：e359fff 但是具体到这个问题的话，大概是星火的key失效导致的

重新尝试之后还是会报错，但是报错信息变成了：“error: Connection to remote host was lost.” 我比较疑惑为啥本地运行没问题，但是docker容器运行会出问题，如果是key有问题理论上本地运行的版本也会报错呀

会不会因为配置格式有点问题呢？