-
-
Notifications
You must be signed in to change notification settings - Fork 640
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bad Fix function for sentence splitting((Draft for experimentation) #472
base: v25
Are you sure you want to change the base?
Conversation
Workflow is running right now We shall see how this modification does https://github.com/DrewThomasson/ebook2audiobook/actions/runs/13792236665 |
This other workflow is also running which made the Long_sentence test shorter, but a single giant paragraph with no punctuation for sentence splitting. https://github.com/DrewThomasson/ebook2audiobook/actions/runs/13792248967 It's Just using a smaller test file that also produces the same error So should be Faster in testing out the workflow |
I think I fixed it but well see later if those workflows show my modification to the |
Ran on the Alice in wonderland test that broke the last one here Wait it keeps on saying pause.... did I accidentally add in the string "pause" into the sentence in function? |
It keeps saying pause.... Is that because of this in the functions.py? tts_manager.params['sentence'] = sentence.replace('.', '<pause>').replace(',', '<pause>') |
In other news this other attempt did not work tho lol |
Potential fix for custom models tho with the being spoken If a custom xtts model swap "." -> "|" instead of the current "." -> "" This is just me talking to myself tho The Works well for the default xtts model But for custom fine-tuned xtts models "." -> "|" works (Confirmed on the fine-tuned David attenborough model for english) and also potentially "." - > ";\n" or "." - > "\n" could be fixes, But those may be language specific... Unknown at this exact moment Findings derived from this source |
I already explored this since long time... not a solution for all languages and can break at anytime. |
No description provided.