-
https://git.ecker.tech/ aims to provide a place to share my efforts while maintaining true ownership of my code, as I do not trust GitHub.
XMR: 4B9TQdkAkBFYrbj5ztvTx89e5LpucPeTSPzemCihdDi9EBnx7btn8RDNZTBz2zihWsjMnDkzn5As1LU6gLv3KQy8BLsZ8SG
- Joined on
2022-10-10
If I did botch that then I'm going to scream, but that's what consistently staying up until 2AM gets. I'll check when I get a moment.
Anyways, the compat fix simply just makes the unified_voice2…
optimizer: ${optimizer}
Yeah, you'll need to regenerate your training configuration. I'd say you can manually edit that as "adamw", but I'm certain I had other minor bugs that were around at…
I can't really tell exactly the issue since you neglected to include the full training output (the provided configuration is very important), but it's safe to assume you need to remake your…
Yeah, that actually seemed to have fixed shitty output for some voices.
https://vocaroo.com/101fw74rhfgz
Where last week it was…
This might be related to mrq/ai-voice-cloning#70, where the other DLAS repo had a compatibility fix for models that are going to be used for TorToiSe (I will admit I…
Edit: I created a louise belcher model with 152334H's DL-Art-School fork with the learning rate set to 1e-20. I'm not sure if it's because of the learning rate or if it's the implementation of…
7c71f7239c
expose options for CosineAnnealingLR_Restart (seems to be able to train very quickly due to the restarts
5460e191b0
added loss graph, because I'm going to experiment with cosine annealing LR and I need to view my loss
Ah, I think I understand what causes it.
The latents and seed used most definitely persist between lines, but any slight variation in the input lines will cause it to diverge pretty fast, even…
Yeah, it requires a lower version librosa (I'd imagine whisperx is to blame, again). I'll add it to update-force.bat for the reinstall line, but you can run this yourself to fix it:
.\ven…