This release adds FLUX Fill support in Workflows and Canvas, support for the LLaVA OneVision VLLM family of models, and a selection of minor fixes and enhancements.
FLUX Fill
FLUX Fill provides high quality inpainting and outpainting, improving on these tasks over the other FLUX models. It's a "main" model, like FLUX dev or schnell.
To use it, download it from Starter Models and then select it from the main model drop-down on Canvas. It's not compatible* with Text to Image or Image to Image - you'll get an error if you try to Invoke without an inpaint mask or some empty regions in your bbox.
*Technically, it can do Text to Image and Image to Image - but the quality is very poor. We've opted to disallow this on Canvas.
LLaVA OneVision VLLM
This multimodal model generates text from text, image and/or video* inputs. You can use it to generate prompts and and describe images. You can use it in Workflows with the LLaVA OneVision VLLM node.
The 0.5B variant of the model is available for download from Starter Models.
*Invoke does not support video inputs.
Other Changes
- Better error messages when scanning models with
picklescan. - Fixed issue where shift-clicking to draw on Canvas ignored
Clip to Bboxsetting. - Improved UX for missing or unexpected fields in Workflows.
- Internal: First iteration of improved model probing API.
- Internal: Improved testing system for model-related tests.
- Updated translations. Thanks @rikublock @Harvester62 @Linos1391!
Installing and Updating
The new Invoke Launcher is the recommended way to install, update and run Invoke. It takes care of a lot of details for you - like installing the right version of python - and runs Invoke as a desktop application.
Follow the Quick Start guide to get started with the launcher.
If you don't want to use the launcher, or need a headless install, you can follow the manual install guide.
What's Changed
- Add LLaVA OneVision model support by @RyanJDick in #7693
- Model classification api by @jazzhaiku in #7742
- feat(app): better errors when scanning models with picklescan by @psychedelicious in #7803
- Stripped models by @jazzhaiku in #7797
- Use git-lfs for larger assets by @jazzhaiku in #7804
- Add FLUX Fill support by @RyanJDick in #7780
- Customized workflow browsing by @maryhipp in #7810
- ui: workflows recommended followup by @psychedelicious in #7812
- fix(ui): clip shift-draw strokes to bbox when clip to bbox enabled by @psychedelicious in #7813
- feat(ui): support FLUX Fill on Canvas by @psychedelicious in #7811
- feat(ui): better missing fields ux in workflows & builder by @psychedelicious in #7816
- ui: translations update from weblate by @weblate in #7805
- chore: prep for v5.9.0rc1 by @psychedelicious in #7815
Full Changelog: v5.8.1...v5.9.0rc1