Exception in callback _ProactorBasePipeTransport._call_connection_lost(None) asyncio (Windows)

[erew123]

Describe the bug

Pretty sure this is a windows only issue (from any research on the internet)

2023-12-02 03:15:36 ERROR:Exception in callback _ProactorBasePipeTransport._call_connection_lost(None)
handle: <Handle _ProactorBasePipeTransport._call_connection_lost(None)>
Traceback (most recent call last):
File "C:\AI\text-generation-webui\installer_files\env\Lib\asyncio\events.py", line 80, in _run
self._context.run(self._callback, *self._args)
File "C:\AI\text-generation-webui\installer_files\env\Lib\asyncio\proactor_events.py", line 165, in _call_connection_lost
self._sock.shutdown(socket.SHUT_RDWR)
ConnectionResetError: [WinError 10054] An existing connection was forcibly closed by the remote host

Is there an existing issue for this?

• I have searched the existing issues

Reproduction

Load up the text-generation-webui on windows

Screenshot

Logs

2023-12-02 03:15:36 ERROR:Exception in callback _ProactorBasePipeTransport._call_connection_lost(None)
handle: <Handle _ProactorBasePipeTransport._call_connection_lost(None)>
Traceback (most recent call last):
  File "C:\AI\text-generation-webui\installer_files\env\Lib\asyncio\events.py", line 80, in _run
    self._context.run(self._callback, *self._args)
  File "C:\AI\text-generation-webui\installer_files\env\Lib\asyncio\proactor_events.py", line 165, in _call_connection_lost
    self._sock.shutdown(socket.SHUT_RDWR)
ConnectionResetError: [WinError 10054] An existing connection was forcibly closed by the remote host

System Info

Windows 11. Nothing special about it. Tested with being fully up to date as of 2nd December. Google Chrome too (which may be somehow related, based on reports I saw on the internet that Edge doesnt cause this issue).

I however found someone listing a possible solution to this:

https://stackoverflow.com/questions/62412754/python-asyncio-errors-oserror-winerror-6-the-handle-is-invalid-and-runtim

Which seems to be related to this https://docs.python.org/3/library/asyncio-policy.html

I thought it could have been related to the extension I wrote, but I disabled the extension (as you can see in the screenshot) 

It doesn't seem to cause any problems, its just annoying. I think it started maybe 1-2 months ago (at a guess).

[Desm0nt]

Same bug with mixtral model. First call to model with long (about 3k+) context in dialogue produce this problem. If I regenerate response after this - it generates fine, but when I send next request (with context increased by previous response) - it's again fail with same error.