The published v2.6.4 release body (English) is unchanged; only the language-switch button was added, linking to zhmd/v2.6.4.md.
zhmd/v2.6.4.md is the Chinese translation of the same notes (switch button back to the release).
- patches/rgthree_logo_route_guard.py (new): validate the markup rgthree caches from tool.comfy.Icon and fall back to rgthree's bundled SVG. Without validation a proxy/CDN error page (HTML with braces) made routes_config.get_logo raise ValueError: unexpected '{' in field name, and any naive suppression served that page as the logo. Applied on import and again via PromptServer on_startup (the UI asks for the logo before any prompt).
- patches/sam3_seg_features_scalp_patch.py (removed): Comfy-Org/ComfyUI#15979 is merged and present in ComfyUI v0.35.0+ (verified: merge commit is an ancestor of the installed HEAD; detector.py takes seg_features after the scalp trim), so the parity patch was inert. Removal is scoped to that patch; restoration condition (ComfyUI < v0.35.0) is documented.
- md/RGTHREE_LOGO_GUARD_AND_SAM3_PATCH_REMOVAL.md (new): technical guide (problem, files, full code, meaning).
Krea2ControlNetLoraLoader with an SDXL LLLite/ControlNet file (e.g.
kohya_controllllite_xl_openpose_anime_v2.safetensors) produced a generic
'No block LoRA patches matched' RuntimeError because the file has no
*.lora_A/*.lora_B pairs at all. Add a fast sanity check in the openpose
route that raises an actionable message naming the file type mismatch and
the correct Krea2 openpose LoRA to load.
The Chinese changelog entry now links to the published Chinese release
notes (v2.6.1.md) instead of the GitHub release page, and the
'(release notes to be published)' note is removed from both EN and ZH.
Documents the LoRA-type ControlNet investigation: full rank-64 LoRA over
all 60 blocks with no injection machinery, ref-concat double-structure
artifact, kv_cache no-steering, not Nunchaku-specific (bf16 also draws
the inner picture). Recommends a standard ControlNet (Alibaba
Qwen-Image-2512-Fun-Controlnet-Union) for Nunchaku Qwen Image. Release
notes link included (to be published).
Expand the Known Limitations entry for LoRA-type ControlNet with the
full background and test results: full-rank-64 LoRA over all 60 blocks
with no injection machinery (Comfy-Org Qwen-Image-DiffSynth-ControlNets /
DiffSynth In-Context Control Union), ref-concat index (double-structure
with inner canny picture), index_timestep_zero (inner picture becomes a
source copy), isolated kv_cache (no steering at all), and the note that
the double structure is not Nunchaku-specific (bf16 also draws the inner
picture; only the outer rendering differs).
Expand the Known Limitations entry for model-patch type ControlNet with
the full technical background: SVDQW4A4Linear hidden-state amplification
(~29x to_qkv, ~760x MLP, block output mean ~1.4e5), per-channel scale
non-uniformity (up to ~1500x), the ~0.3% residual burying, and why
Z-Image (NextDiT) is unaffected. Nunchaku Qwen Image limitation, not a
ControlNet bug.
- Model-patch type ControlNet (DiffSynth block-wise, control_block) is
not usable: W4A4 hidden-state amplification (~29x to_qkv, ~760x MLP,
block output mean ~1.4e5 vs ~0.3-5 standard) buries the standard-scale
residual; naive rescaling breaks channel structure
- LoRA-type ControlNet (qwen_image_union_diffsynth_lora) is not usable:
ref-concat produces the double-structure (inner-picture) artifact on
Nunchaku; isolated kv_cache injection does not steer generation at all
- Recommended: standard (non-model-patch) ControlNet, e.g. Alibaba's
Qwen-Image-2512-Fun-Controlnet-Union via the normal Apply ControlNet flow
The DiffSynth QI route does not support Nunchaku Qwen Image (quantized
hidden-state scale mismatch). Point users to Alibaba's official
Qwen-Image-2512-Fun-Controlnet-Union ControlNet instead.
ussoewwin