-
-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
learning_rate_te2 must be given as 0 #1992
Comments
yep i verified this bug exists and breaks my config :/ |
Hummm, this should be an easy fix. I will address it in the dev branch. |
I am loking at the code in the master branch and it should work. Have you tried pulling the latest version I published this morning? |
I just did a git pull on master and no still not showing v 22.6.2 there is also new high vram option you plan to add? thank you |
Can you share a json config file that trigger the issue? And that is is (Lora, dreambooth, fine tune) |
sure here it is it is dreambooth sdxl |
Found the source of the issue. Should be fixed in next release. For now you can text in the dev-pure branch. |
Currently when we set Text Encoder rate as 0, the GUI is not providing it
Kohya said it has to be provided
SDXL training separate text encoder Learning rates for TE1 and TE2
kohya-ss/sd-scripts#1131 (comment)
The text was updated successfully, but these errors were encountered: