4.1.0

What's Changed

• Alpha support for sd v2.x in auto in #244
• Auto: clean up interrogators in #265
• Smaller xformers wheel in #271

Full Changelog: 4.0.0...4.1.0

This release will be shortly followed up by 5.0.0 (with breaking changes)

4.0.0

Breaking Changes

• hlky has been renamed to sygil
  • a new entry has been added: sygil-sl which runs the streamlit version of the UI
  • note: the streamlit version is still in dev and has bugs!
• lstein has been renamed to invoke
• the structure of the output folder has changed to
  • img2img, txt2img, extras used by auto & sygill
  • invoke is subfolder exclusively for invoke since the output folders of this UI are not customizable.

Migration Guide

Make sure you use the new names when running the corresponding UIs

For the output folder:

• create a backup
• create a new folder in output with any name you want and move everything in the output folder to it
• you can use the script migratev3tov4.sh to do this for you:

bash ./.devscripts/migratev3tov4.sh

What's Changed

• Enable extensions by default in #243
• Update versions in #246
• Invoke 2.2.3 in #250
• Fix mounting in Lstein in #253
• [BREAKING] Rename UIs in #254

Full Changelog: 3.2.0...4.0.0

3.2.0

What's Changed

• Update versions in #189 #193 #195 #198 #213 #230 #234
• Deepdanbooru Support in #194
• Fix UI Layout in #196
• Hack to allow installing extensions in #200
• Add Samplers in #205
• Use cuda 11.6 for auto in #220
• v1.5 Inpainting in #221
• Update xformers in #241

Full Changelog: 3.1.0...3.2.0

Even though Stable Diffusion 2.0 was released, it is unfortunately not backwards compatible, and will require some time to be implemented by the different UI's...

3.1.0

What's Changed

• Tags for CLIP interrogation by @ProducerMatt in #166
• Aria fixes by @DevilaN in #170
• Update versions by @AbdBarho in #185
• Scripts support by @AbdBarho in #187
• Auto Extensions by @AbdBarho in #176

New Contributors

• @ProducerMatt made their first contribution in #166

Full Changelog: 3.0.0...3.1.0

The attached xformers compiled binary was created for python3.10 and the following cuda archs: 5.0 5.2 5.3 6.0 6.1 6.2 7.0 7.2 7.5 8.0 8.6

3.0.0 Support SD v1.5

SD v1.5 is now available and enabled per default in auto and hlky.

Update versions

• auto: AUTOMATIC1111/stable-diffusion-webui@f49c08e
• hlky: Sygil-Dev/sygil-webui@8d1e42b
• lstein: invoke-ai/InvokeAI@554445a

What's Changed

• Expose auto ui config in #149
• New Url by in #161
• SD 1.5 by in #164

Full Changelog: 2.1.0...3.0.0

Migration Guide

Running

docker compose --profile download  up --build

should suffice

2.1.0

What's Changed

• Fix&Update to use lstein (PRELOAD=true) by @moulai in #131
• xformers by @AbdBarho in #136

Update versions

• auto: AUTOMATIC1111/stable-diffusion-webui@36a0ba3
  • with xformers support for faster GPU generation (20% - 100% speed up)
• hlky: Sygil-Dev/sygil-webui@bd57d22
• lstein: invoke-ai/InvokeAI@fe2a2cf

New Contributors

• @moulai made their first contribution in #131

Full Changelog: 2.0.1...2.1.0

The attached xformers compiled binary was created for python3.10 and the following cuda archs: "5.0;5.2;5.3;6.0;6.1;6.2;7.0;7.2;7.5;8.0;8.6"

2.0.1

What's Changed

• auto: AUTOMATIC1111/stable-diffusion-webui@2995107
  • More samplers
  • Textual inversion training
• hlky: Sygil-Dev/sygil-webui@1a9c053
  • Build times are SLOW
• lstein: invoke-ai/InvokeAI@4f247a3
  • Prepare for 2.0 release
  • very cool new UI!

• Fix chmod on non-existing dir by @AbdBarho in #113

Full Changelog: 2.0.0...2.0.1

2.0.0

What's Changed

• Docker compose override example file + .gitignore update by @DevilaN in #100
• Update / Disable lstein Temporarily by @AbdBarho in #106
• v2 in #108

Update versions

• auto: AUTOMATIC1111/stable-diffusion-webui@3f41756
• hlky Sygil-Dev/sygil-webui@aaa3be1

Breaking changes

• renamed automatic-1111 service to auto
• lstein is disabled temporarily
• the cache folder is now deprecated, replaced with data (see migration guide below)
• embeddings folder has been moved to data/embeddings
• use GFPGAN 1.4

Migration Guide

Note: in theory, running the command

docker compose --profile download up --build

is all you need to use the new version, however, this means you will also have to download everything again.

If you want to save time: a new script is available under scripts/migratev1tov2.sh that will copy files to the new structure and should get you most of the way there, run

./scripts/migratev1tov2.sh

in the root of the repo, or you can manually inspect the script and copy the files.

After that, run

docker compose --profile download up --build

to validate everything.

New Contributors

• @DevilaN made their first contribution in #100

Full Changelog: 1.2.0...2.0.0

1.2.0

Versions:

• auto: AUTOMATIC1111/stable-diffusion-webui@ca3e551
• hlky: Sygil-Dev/sygil-webui@1fd28ee
• lstein: invoke-ai/InvokeAI@b40bfb5

What's Changed

• Fix CLIP model caching #88 in #95
• Adding embeddings volume for auto textual inversion by @rgriffogoes in #98

New Contributors

• @rgriffogoes made their first contribution in #98

Full Changelog: 1.1.1...1.2.0

1.1.1

What's Changed

• auto: AUTOMATIC1111/stable-diffusion-webui@dd911a4
• hlky: Sygil-Dev/sygil-webui@17748cb
• lstein: invoke-ai/InvokeAI@50d607f

Full Changelog: 1.1.0...1.1.1