This release includes support for Nvidia 50xx GPUs, a way to relate models (e.g. LoRAs with a specific main model), and new IP Adapter methods.
Changes
- Bumped PyTorch dependency to v2.7.0, which means Invoke now supports Nvidia 50xx GPUs.
- New model relationship feature. In the model manager tab, you may "link" two models. At this time, the primary intended use case is to link LoRAs to main models. When you have the main model selected, the linked LoRAs will be at the top of the LoRA list. Thanks @xiaden!
- New IP Adapter methods
Style (Strong)andStyle (Precise). The previous style method is renamed toStyle (Simple). Thanks @cubiq! - Fixed GGUF quantization on MPS. Thanks @Vargol!
- Internal invocation model changes, which aim to reduce occurrences of
ValidationErrorerrors. - Addressed pydantic deprecation warning.
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
- fix(nodes): pydantic field type massaging improvements by @psychedelicious in #7984
- ui: translations update from weblate by @weblate in #7938
- Jazzhaiku/stats by @jazzhaiku in #8006
- feat(ui): model relationship management by @xiaden in #7963
- Add to overload for GGMLTensor, so calling to on the model moves the quantized data. by @Vargol in #7949
- fix(app): address pydantic deprecation warning for accessing
BaseModel.model_fieldsby @psychedelicious in #8012 - chore: bump torch to 2.7.0 by @psychedelicious in #8013
- Expanded IP Adapter modes. by @hipsterusername in #8011
- ui: translations update from weblate by @weblate in #8015
- chore: prep for v5.12.0rc1 by @psychedelicious in #8014
New Contributors
Full Changelog: v5.11.0...v5.12.0rc1