-
Notifications
You must be signed in to change notification settings - Fork 489
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
Skipping verses and sound noise and volume. #73
Comments
May I look at your txts? It Could be too many verses. |
Sorry, I missed this notification earlier. In the sample below it misses the entire Outro section and the lines (in italics below) Here is the cmd line
genre.txt lyrics2.txt
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When the number of verses increase to 10+, the model starts skipping verses. No Errors. Each verse is around 130 to 180 characters.
Also, some times the song has a lot of noise only for certain words and the volume is much higher. Data corruption?
I am running with --new tokens at 1500 or 2000 to make sure the model does not run out of context size. Is there a way to increase the context?
I am running the YuE-s1-7B-anneal-en-icl model with audio prompt on a H200 NVL (144GB VRAM) with python 3.10 and torch 2.6.0+cu124
Thanks,
Ash
The text was updated successfully, but these errors were encountered: