Open alexlnkp opened 1 month ago
but i fail to understand why even do that if you can create a job in the python that executes code
The initial author do not know how to do that. He just want a subprocess to run some code asynchronously.
Also, the code will look just more appealing without all of the subprocess creation
Yes. And it will make it possible to add a button break
in gradio.
Also, the code will look just more appealing without all of the subprocess creation
Yes. And it will make it possible to add a button
break
in gradio.
That reminds me of my Mangio-RVC-Fork
and how painful it was to add the stop training button :joy:
how painful it was to add the stop training button
Yes 😂. So very important to change the code structure. It is the base to the further improvement.
The title might be a bit confusing, but what i'm referring to is how some python scripts are ran in a subshell, even though they could be used as regular python modules.
Is something that is present A LOT in the web.py file, but i fail to understand why even do that if you can create a job in the python that executes code, instead of opening a shell as a subprocess.
I assume this will be fixed when the RVC becomes a module, however I'll prefer fixing this sooner than later.
Opening a subshell and executing a python script adds a huge amount of overhead and introduces extreme amounts of security vulnerabilities,
shell=True
is one of the culprits for that in particular.Also, the code will look just more appealing without all of the subprocess creation within shells and all the yields of the
stdout
as a generator for gradio interface to update properly