Training failure

[sumitshrestha]

Hi everyone,

I’m encountering an issue while trying to train two separate sets of audio files. Both attempts fail with the same error, even though voice generation via image prompts works correctly (using the custom voice I added). Here’s my setup and details:

System: Windows 11

GPU: NVIDIA RTX 4080

Behavior:

✅ Voice generation from image prompts works as expected (uses my custom voice).

❌ Training fails consistently under the "Training" tab with default parameters.

Has anyone faced similar issues on newer NVIDIA GPUs or Windows 11? Could this be related to compatibility, dependencies, or VRAM allocation? I’d appreciate any troubleshooting tips or insights!

Steps I’ve tried:

Used default training parameters (no adjustments).

Verified that audio files are properly formatted and accessible.

Questions:

Are there known issues with RTX 40xx GPUs or Windows 11 in this workflow?

Should I adjust parameters like batch size or learning rate for a 4080?

Are there specific logs or error details I should share to help diagnose this?

me.txt