add Nunchaku-Qwen-Image-Lightning

Signed-off-by: Vladimir Mandic <mandic00@live.com>

CHANGELOG.md

@@ -6,6 +6,9 @@
   - **Chroma** final versions: [Chroma1-HD](https://huggingface.co/lodestones/Chroma1-HD), [Chroma1-Base](https://huggingface.co/lodestones/Chroma1-Base) and [Chroma1-Flash](https://huggingface.co/lodestones/Chroma1-Flash)  
   - **Qwen-Image** [InstantX ControlNet Union](https://huggingface.co/InstantX/Qwen-Image-ControlNet-Union) support  
     *note* qwen-image is already a very large model and controlnet adds 3.5GB on top of that so quantization and offloading are highly recommended!  
+  - [Nunchaku-Qwen-Image-Lightning](https://huggingface.co/nunchaku-tech/nunchaku-qwen-image)  
+    if you have a compatible nVidia GPU, Nunchaku is the fastest quantization engine, currently available for Flux.1, SANA and Qwen-Image models  
+    *note*: release version of `nunchaku==0.3.2` does NOT include support, so you need to build [nunchaku](https://nunchaku.tech/docs/nunchaku/installation/installation.html) from source  
   - updated [SD.Next Model Samples Gallery](https://vladmandic.github.io/sd-samples/compare.html)  
 - **Core**
   - enable offload during pre-forward by default  

html/reference.json

@@ -196,7 +196,7 @@
     "skip": true,
     "extras": ""
   },
-  "Qwen-Lightning": {
+  "Qwen-Image-Lightning": {
     "path": "vladmandic/Qwen-Lightning",
     "preview": "Qwen-Lightning.jpg",
     "desc": " Qwen-Lightning is step-distilled from Qwen-Image to allow for generation in 8 steps.",

pipelines/qwen/qwen_nunchaku.py

@@ -12,7 +12,9 @@ def load_qwen_nunchaku(repo_id):
         shared.log.error(f'Load module: quant=Nunchaku module=transformer repo="{repo_id}" low nunchaku version')
         return None
     if repo_id.lower().endswith('qwen-image'):
-        nunchaku_repo = f"nunchaku-tech/nunchaku-qwen-image/svdq-{nunchaku_precision}_r128-qwen-image.safetensors" # r32 vs R128
+        nunchaku_repo = f"nunchaku-tech/nunchaku-qwen-image/svdq-{nunchaku_precision}_r128-qwen-image.safetensors" # r32 vs r128
+    elif repo_id.lower().endswith('qwen-lightning'):
+        nunchaku_repo = f"nunchaku-tech/nunchaku-qwen-image/svdq-{nunchaku_precision}_r128-qwen-image-lightningv1.1-8steps.safetensors" # 8-step variant
     else:
         shared.log.error(f'Load module: quant=Nunchaku module=transformer repo="{repo_id}" unsupported')
     if nunchaku_repo is not None: