-
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
The above solution works, but this issue has been documented on the wiki…
d58b67004a
colab notebook uses venv and normal scripts to keep it on parity with a local install (and it literally just works stop creating issues for someething inconsistent with known solutions)
Just did a git pull
Either update through the update script, or run
git submodule update --remote
as you also need to update tortoise-tts.
I do not have a better answer than what I've explained on the wiki.
>but I can't…
ff07f707cb
disable validation if validation dataset not found, clamp validation batch size to validation dataset size instead of simply reusing batch size, switch to adamw_zero optimizier when training with multi-gpus (because the yaml comment said to and I think it might be why I'm absolutely having garbage luck training this japanese dataset)
Forgot to mention the cache clearing has been implemented to mrq/tortoise-tts in commit cc36c0997c on grabbing the…
...or at least just bandaid the problem until it happens to me again. Bandaided in commit f1788a5639f749d05f3f63eba673a7d428cba170.
f1788a5639
lazy wrap around the voicefixer block because sometimes it just an heros itself despite having a specific block to load it beforehand
Right, forgot to figure that out. I can't remember if I had to enable Voicefixer or disable it, but the web UI gets in a gummy state and I haven't been able to replicate it again.
I'll try and…
Added my paperspace notebook in commit 83b512585410469ec573011f32d71ce9b0b092d4. It's much cleaner than the colab, as it can leverage being able to use a venv and whatnot.
The notebook is practically incompatible with paperspace. I have one brewed up for paperspace that I just haven't gotten off my ass to clean up and push to the repo.
Although, that reminds me I…