The long text audio file was successfully generated by web_inference.py, but the connection of discord(on_message) was terminated with some exceptions. Then the voice channel connection seems to be disconnected so the following say / say_jp commands won't be processed.
A possible way to fix is that: when on_message is called, it just create a async task to a task queue instead of starting generating the audio file, and also have another worker to process the tasks in the newly added task queue to process tasks.
The long text audio file was successfully generated by
web_inference.py
, but the connection of discord(on_message
) was terminated with some exceptions. Then the voice channel connection seems to be disconnected so the followingsay / say_jp
commands won't be processed.A possible way to fix is that: when
on_message
is called, it just create a async task to a task queue instead of starting generating the audio file, and also have another worker to process the tasks in the newly added task queue to process tasks.