Skip to content
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

ControlNet training for SDXL #1655

Merged
merged 10 commits into from
Oct 11, 2024
Merged

ControlNet training for SDXL #1655

merged 10 commits into from
Oct 11, 2024

Conversation

kohya-ss
Copy link
Owner

WIP Sample image generation is not implemented yet.

sdbds and others added 9 commits September 29, 2024 10:11
* Create sdxl_train_controlnet.py

* add fuse_background_pass

* Update sdxl_train_controlnet.py

* add fuse and fix error

* update

* Update sdxl_train_controlnet.py

* Update sdxl_train_controlnet.py

* Update sdxl_train_controlnet.py

* update

* Update sdxl_train_controlnet.py
@sdbds
Copy link
Contributor

sdbds commented Oct 11, 2024

Is it necessary to set strategy_controlnet for different architectures later?
For example, xinsir trained controlnet++,
others like controlnext and controlnetllite.

@kohya-ss
Copy link
Owner Author

Modules such as strategy_sdxl etc. only abstract the VAE and Text Encoder, so we need to prepare training scripts for each ControlNet with a different architecture, such as train_controlnet.py, sdxl_train_control_net.py and sdxl_train_control_net_lllite.py.

@kohya-ss kohya-ss marked this pull request as ready for review October 11, 2024 13:24
@kohya-ss kohya-ss merged commit 43bfeea into sd3 Oct 11, 2024
2 checks passed
@kohya-ss kohya-ss deleted the sdxl-ctrl-net branch October 11, 2024 13:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants