v0.9.0

## What's Changed
* Revert "Force sequential execution in CI test jobs" by @yoland68 in https://github.com/Comfy-Org/ComfyUI/pull/11725
* bump comfyui_manager version to the 4.0.5 by @ltdrdata in https://github.com/Comfy-Org/ComfyUI/pull/11732
* Better detection if AMD torch compiled with efficient attention. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11745
* Fix import issue. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11746
* Add node: JoinAudioChannels by @kijai in https://github.com/Comfy-Org/ComfyUI/pull/11728
* Fix csp error in frontend when forcing offline. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11749
* Bump comfyui-frontend-package to 1.36.13 by @comfy-pr-bot in https://github.com/Comfy-Org/ComfyUI/pull/11645
* Add most basic Asset support for models by @Kosinkadink in https://github.com/Comfy-Org/ComfyUI/pull/11315
* Fix VAEEncodeForInpaint to support WAN VAE tuple downscale_ratio by @rattus128 in https://github.com/Comfy-Org/ComfyUI/pull/11572
* add node - image compare by @jtydhr88 in https://github.com/Comfy-Org/ComfyUI/pull/11343
* feat: add 'cancelled' status for interrupted jobs in /api/jobs by @ric-yu in https://github.com/Comfy-Org/ComfyUI/pull/11680
* Add workaround for a hacky nodepack existing by @Kosinkadink in https://github.com/Comfy-Org/ComfyUI/pull/11755
* Be less strict when loading mixed ops weights. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11769
* fix(api-nodes): do not downscale the input image for Topaz Enhance by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/11768
* [API Nodes] add nodes for Vidu2 by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/11760
* Update embedded docs to v0.4.0 by @comfyui-wiki in https://github.com/Comfy-Org/ComfyUI/pull/11776
* Properly save mixed ops. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11772
* Enable PyTorch Attention by default on gfx1200  by @0xDELUXA in https://github.com/Comfy-Org/ComfyUI/pull/11793
* Fix issue with t5 text encoder in fp4. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11794
* Fix chroma fp8 te being treated as fp16. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11795
* fix(api-nodes): use a unique name for uploading audio files by @bigcat88 in https://github.com/Comfy-Org/ComfyUI/pull/11778
* Put more details about portable in readme. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11816
* Support ModelScope-Trainer/DiffSynth LoRA format for Z-Image models by @kelseyee in https://github.com/Comfy-Org/ComfyUI/pull/11805
* Support the siglip 2 naflex model as a clip vision model. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11831
* Reduce LTX2 VRAM use by more efficient timestep embed handling by @kijai in https://github.com/Comfy-Org/ComfyUI/pull/11829
* Update workflow templates to v0.8.0 by @comfyui-wiki in https://github.com/Comfy-Org/ComfyUI/pull/11828
* Update workflow templates to v0.8.4 by @comfyui-wiki in https://github.com/Comfy-Org/ComfyUI/pull/11835
* Make loras work on nvfp4 models. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11837
* Refactor  to try to lower mem usage. by @comfyanonymous in https://github.com/Comfy-Org/ComfyUI/pull/11840
* Make database bulk_ops.py more widely compatible across sqlalchemy+python versions by @Kosinkadink in https://github.com/Comfy-Org/ComfyUI/pull/11839
* Patch frontend to 1.36.14 (from 1.36.13) by @christian-byrne in https://github.com/Comfy-Org/ComfyUI/pull/11841

## New Contributors
* @kelseyee made their first contribution in https://github.com/Comfy-Org/ComfyUI/pull/11805

**Full Changelog**: https://github.com/Comfy-Org/ComfyUI/compare/v0.8.2...v0.9.0