1580158
/
Hypernetwork-MonkeyPatch-Extension
mirror of https://github.com/aria1th/Hypernetwork-MonkeyPatch-Extension.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
Extension that provides additional training features for hypernetwork training. Also supports using multiple hypernetworks for inference.
10 Commits
8 Branches
0 Tags
240 KiB
Python 100%
 
Go to file
aria1th 42f67109af Create .gitignore 2022-11-24 14:06:24 +09:00
patches fix loss save and CUDA memory 2022-11-24 13:17:15 +09:00
scripts Add external pr 2022-11-23 21:33:31 +09:00
.gitignore Create .gitignore 2022-11-24 14:06:24 +09:00
README.md Update README.md 2022-11-24 13:51:45 +09:00

README.md

Hypernetwork-MonkeyPatch-Extension

Extension that patches Hypernetwork structures and training image

For Hypernetwork structure, see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/4334

For Variable Dropout, see https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/4288

Train_Beta tab allows some more options with improved training.

image

Features

No-Crop Training image You can train without cropping images.

Fix OSError while training

Unload Optimizer while generating Previews

Create_Beta_hypernetwork allows creating beta hypernetworks.

This extension also overrides how webui loads and finds hypernetworks, to use variable dropout rates, and etc. Thus, hypernetwork created with variable dropout rate might not work correctly in original webui.

Train_Gamma tab uses Gradient Accumulation from separate PR.

image

This also fixes some CUDA memory issues. Currently Gamma Training is working very well, as far as I could say.