-
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
I'll see about cramming those in. I know Voldy's Stable Diffusion Web UI has them, some of the time but not all the time. Gradio's documentation doesn't seem to have an easy way to expose them, so…
Added in aa96edde2fa9072a4a56a89459675a24b655319e. I did a few rudimentary tests (that involved nuking my userdata, but it's not a huge deal since it's an old model) and works fine, no userdata nukes.
aa96edde2f
Updated notebook to put userdata under a dedicated folder (and some safeties to not nuke them if you double run the script like I did thinking rm -r [symlink] would just remove the symlink
Right right, almost forgot. Give me a few to figure out an elegant way to use setups as I'm not well versed in bash scripts.
It seems to at least be training somewhat (albeit slowly with that 20s/it) before OOMing, try lowering the batch size to 2.
If not, then use this copium in place of your train.bat:
call…For sure really fixed in 526a430c2adf2287d22697b8e2800128e7a42060 or I'm blowing my brains out. I don't know how it reverted.
2aa70532e8
added '''suggested''' voice chunk size (it just updates it to how many files you have, not based on combined voice length, like it should
Tried to allocate 1.61 GiB
Depending on if your sources are too large when combined, you'll need to play around with the "voice chunk" slider, which I might as well try and adjust how it…
Fixed in commit 93b061fb4d85fb4c26bf88cf47de879a7ab187ad. It's because I did my testing while still having a results folder, and when I did incidentally empty my results folder, the UI was already…
Oops, don't know how I managed to get it working for me, fixed in commit 93b061fb4d85fb4c26bf88cf47de879a7ab187ad.