cleanup

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

CHANGELOG.md

@@ -14,9 +14,6 @@
     *note*: this is a [gated model](https://vladmandic.github.io/sdnext-docs/Gated/)  
   - [VIBE Image Edit](https://huggingface.co/iitolstykh/VIBE-Image-Edit) visual instruction based image editing model  
     combines *Sana1.5-1.6B* diffusion backbone with *Qwen3-VL-2B* multimodal conditioning  
-  - [Lumina-DiMOO](https://huggingface.co/Alpha-VLLM/Lumina-DiMOO) omni diffusion large language model for multimodal generation and understanding  
-    supports *text-to-image*, *image editing*, and *image understanding* in a unified discrete diffusion framework  
-    uses a VQ-VAE backed token space with any-to-any task coverage  
   - [Step1X-Edit v1.2](https://huggingface.co/stepfun-ai/Step1X-Edit-v1p2) image editing model using Step1X transformer  
     integrates Qwen2.5-VL text encoder and Flow Matching scheduler for high-quality in-context edits  
     enables multimodal understanding and conditional image refinement  

TODO.md

@@ -12,7 +12,8 @@
 - Test: Bria-FIBO
 - Port: ERNIE-Image (merged, unpublished) <https://github.com/huggingface/diffusers/blob/main/docs/source/en/api/pipelines/ernie_image.md>
 - Port: NucleusMoE-Image (merged, unpublished)
-- Port: JoyAI-Image-Edit (in-progress, need conversion)
+- Port: JoyAI-Image-Edit (in-progress)
+- Port: Lumina-DiMOO(in-progress)
 - Code: Bria-FIBO edit requires image handling
 - Issues: ROCm script with LoRA?
 
@@ -62,6 +63,7 @@ TODO: Investigate which models are diffusers-compatible and prioritize!
 
 ### Image
 
+- [Lumina-DiMOO](https://huggingface.co/Alpha-VLLM/Lumina-DiMOO)
 - [UltraFlux](https://huggingface.co/Owen777/UltraFlux-v1)
 - [Mugen](https://huggingface.co/CabalResearch/Mugen)
 - [Liquid](https://github.com/FoundationVision/Liquid)

data/reference.json

@@ -621,15 +621,6 @@
     "size": 20.75,
     "date": "2025 January"
   },
-  "AlphaVLLM Lumina DiMOO": {
-    "path": "Alpha-VLLM/Lumina-DiMOO",
-    "desc": "Lumina-DiMOO is an omni diffusion large language model for multimodal generation and understanding with text-to-image, image editing and understanding capabilities.",
-    "preview": "Alpha-VLLM--Lumina-DiMOO.jpg",
-    "skip": true,
-    "extras": "sampler: Default",
-    "size": 0,
-    "date": "2025 September"
-  },
 
   "HiDream-I1 Fast": {
     "path": "HiDream-ai/HiDream-I1-Fast",

modules/shared_items.py

@@ -30,7 +30,7 @@ pipelines = {
     'Chroma': getattr(diffusers, 'ChromaPipeline', None),
     'Sana': getattr(diffusers, 'SanaPipeline', None),
     'Lumina-Next': getattr(diffusers, 'LuminaText2ImgPipeline', None),
-    'Lumina-DiMOO': getattr(diffusers, 'LuminaDiMOOPipeline', None),
+    # 'Lumina-DiMOO': getattr(diffusers, 'LuminaDiMOOPipeline', None),
     'Lumina 2': getattr(diffusers, 'Lumina2Pipeline', None),
     'AuraFlow': getattr(diffusers, 'AuraFlowPipeline', None),
     'Kandinsky 2.1': getattr(diffusers, 'KandinskyCombinedPipeline', None),

pipelines/model_lumina.py

@@ -84,3 +84,15 @@ def load_lumina_dimoo(checkpoint_info, diffusers_load_config=None):
 
     devices.torch_gc(force=True, reason='load')
     return pipe
+
+""" Reference
+  "AlphaVLLM Lumina DiMOO": {
+    "path": "Alpha-VLLM/Lumina-DiMOO",
+    "desc": "Lumina-DiMOO is an omni diffusion large language model for multimodal generation and understanding with text-to-image, image editing and understanding capabilities.",
+    "preview": "Alpha-VLLM--Lumina-DiMOO.jpg",
+    "skip": true,
+    "extras": "sampler: Default",
+    "size": 0,
+    "date": "2025 September"
+  },
+"""