-
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
22ffaf3a33
have loss for the NAR not-ignore the text prompt, I imagine this should help the NAR and explain why it's always had a bit of an issue with training
- I understood that technically this is fine tuning a model. But which model are we fine tuning exactly?
With the default backend selected (Tortoise), you're finetuning Tortoise's…
Think I got a great understanding of how it functions. It'd be better to simply "subjugate" (for lack of a better term) the XTTS model…
its very fast and the quality is pretty good also though the cloning isnt perfect ... but idk if you can figure out how to improve it or make it fine tunable..or use its inference to improve…
Don't expect any updates for a while.
Both the full sized model and double-deepend models are being retrained from scratch and not stitched and glued from existing ARs to the monolithic…
Possibly a brief issue with either HuggingFace, or something funny either with your Python install or your connection (for example, ProtonVPN under some settings will treat catbox.moe as "unsafe"…
about 700 clips (is that too much? First question haha)
The more the better. If I remember right, I've been able to get some finetune results with ~50.
I have 101 output models in the…
According to https://github.com/TimDettmers/bitsandbytes/issues/62 you should have a bunch of messages from BitsAndBytes hinting at a misconfigured system.
The solution seems to be passing in…
4aef798135
added picking final candidate based on sum of score instead of first candidate (this changes nothing).
I haven't done any tests involving training a TorToiSe model from scratch, as per the original repo's README:
These models were trained on…
[Training] [2023-09-11T23:20:33.324092] local_state[k] = v[grad_accum_step] [Training] [2023-09-11T23:20:33.324092] IndexError: list index out of range
Your batch size isn't evenly…
40ef34e1ca
this embedding class definitely works, and migrating from the previous embedding weights seems to work.
@mrq have you tried https://github.com/Plachtaa/vallex-webui ? the author say they use https://github.com/lifeiteng/vall-e for training code
I gave it a cursory glance and I find it rather…
a1f250ffac
set default max_levels for NAR to 0 and implicitly set it to max resps levels because the previous way was implicitly assuming all models were outputting at 1+7 RVQ bins.
671dca88ee
throw error when no reference audio is provided in the web UI because someone keeps doing that in the HF space