-
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
Also was a consideration, similar to the X/Y plots thst Voldy's Stable Diffusion Web UI has.
Outside of trying to figure out an intuitive way of implementing it, the big hurdle that comes to…
On my hypothetical backlog (that kept getting refractored trying to get training to work on lower VRAM), it should be a piece of cake to add.
I honestly need to clean up updating dropdowns,…
oh dear :(
Not a huge deal, only the 5 hours I slept while it trained.
I have no idea what could be causing it, and if occurs every time, but I thought I'd let you know.
I'll look…
Settings are updated when anything in the page is changed. The button was actually only for force-changing the models because of a quirk with firing change events from Dropdowns, but that's been…
Remedied in commit 487f2ebf321570f0e0c9b34d311339dfde7228c0. Outputted units are in terms of epochs, rather than the incestuous blend of iterations and steps.
Funny, since I was intending to…
It should mostly be right, but I need to re-validate my units and conversion. I made a similar "mistake" and set a comically large epoch count of 1500 and, due to my low batch size, it would…
thankfully it never OOMed and it was able to complete its training.
Good to hear.
if prompt.strip() != "": AttributeError: 'NoneType' object has no attribute 'strip'
Looks like…
941a27d2b3
removed the logic to toggle BNB capabilities, since I guess I can't do that from outside the module
Success!
I was able to successfully train a model on my 2060. I didn't have anyone come and break in and CBT me, nor are my outputs from a model finetuned with this optimization degraded at…
I've had a breakthrough with being able to train even on my 2060.
Refer to mrq/ai-voice-cloning#25.