psychedelicious e4a640f0a7 feat(ui): optimized rendering of selected layer ...

Instead of caching on every stroke, we can use a compositing rect when the layer is being drawn to improve performance.

2024-05-04 12:03:28 -04:00

..

app

fix(nodes): fix constraints in cnet processors

2024-05-02 12:24:04 +10:00

assets

feat(api): chore: pydantic & fastapi upgrade

2023-10-17 14:59:25 +11:00

backend

Merge branch 'main' into pr/6086

2024-05-01 00:37:06 +05:30

configs

feat(mm): support sdxl ckpt inpainting models

2024-04-28 12:57:27 +10:00

frontend

feat(ui): optimized rendering of selected layer

2024-05-04 12:03:28 -04:00

invocation_api

Remove support for Prompt-to-Prompt cross-attention control (aka .swap()). This feature is not widely used. It does not work with SDXL and is incompatible with IP-Adapter and regional prompting. The implementation is also intertwined with both text embedding and the UNet attention layers, resulting in a high maintenance burden. For all of these reasons, we have decided to drop support.

2024-04-09 10:57:02 -04:00

version

Update invokeai_version.py

2024-05-03 13:23:19 +10:00

__init__.py

Various fixes

2023-01-30 18:42:17 -05:00