fix: qwen-image-edit-2509 docs and examples (#731)

* fix: add missing scheduler to qwen-image-edit-2509 lightning example

* docs: add documentation for qwen-image-edit-2509 lightning

* docs: add Distilled Qwen-Image-Edit-2509-Lightning section and fix Hugging Face link refs

  - Add a new "Distilled Qwen-Image-Edit-2509 (Qwen-Image-Edit-2509-Lightning)" section to
    docs/source/usage/qwen-image-edit.rst, referencing
    examples/v1/qwen-image-edit-2509-lightning.py.
  - Fix incorrect reStructuredText named-reference syntax that produced relative links:
    changed occurrences like <hf_qwen-image>/<hf_qwen-image-edit>/<hf_qwen-image-lightning>
    to the correct named-reference form with trailing underscore (e.g. <hf_qwen-image_>).
    This makes those links resolve to the external URLs defined in
    docs/source/links/huggingface.txt instead of becoming relative links.
  - Files modified:
      - docs/source/usage/qwen-image-edit.rst
      - docs/source/usage/qwen-image.rst

* style: make linter happy

* fix the repo

Author: amchii

docs/source/usage/qwen-image-edit.rst

@@ -4,8 +4,8 @@ Qwen-Image-Edit
 Original Qwen-Image-Edit
 ------------------------
 
-`Qwen-Image-Edit <hf_qwen-image-edit>`_ is the image editing version of Qwen-Image.
-Below is a minimal example for running the 4-bit quantized `Qwen-Image-Edit <hf_qwen-image-edit>`_ model with Nunchaku.
+`Qwen-Image-Edit <hf_qwen-image-edit_>`_ is the image editing version of Qwen-Image.
+Below is a minimal example for running the 4-bit quantized `Qwen-Image-Edit <hf_qwen-image-edit_>`_ model with Nunchaku.
 Nunchaku offers an API compatible with `Diffusers <github_diffusers_>`_, allowing for a familiar user experience.
 
 .. literalinclude:: ../../../examples/v1/qwen-image-edit.py
@@ -26,7 +26,7 @@ The :meth:`~nunchaku.models.transformers.transformer_qwenimage.NunchakuQwenImage
 Distilled Qwen-Image-Edit (Qwen-Image-Lightning)
 ------------------------------------------------
 
-For faster inference, we provide pre-quantized 4-step and 8-step Qwen-Image-Edit models by integrating `Qwen-Image-Lightning LoRAs <hf_qwen-image-lightning>`_.
+For faster inference, we provide pre-quantized 4-step and 8-step Qwen-Image-Edit models by integrating `Qwen-Image-Lightning LoRAs <hf_qwen-image-lightning_>`_.
 See the example script below:
 
 .. literalinclude:: ../../../examples/v1/qwen-image-edit-lightning.py
@@ -41,7 +41,7 @@ Qwen-Image-Edit-2509
    :alt: Nunchaku-Qwen-Image-Edit-2509
 
 Qwen-Image-Edit-2509 is an monthly iteration of Qwen-Image-Edit.
-Below is a minimal example for running the 4-bit quantized `Qwen-Image-Edit-2509 <hf_qwen-image-edit-2509>`_ model with Nunchaku.
+Below is a minimal example for running the 4-bit quantized `Qwen-Image-Edit-2509 <hf_qwen-image-edit-2509_>`_ model with Nunchaku.
 
 .. literalinclude:: ../../../examples/v1/qwen-image-edit-2509.py
     :language: python
@@ -52,3 +52,14 @@ Below is a minimal example for running the 4-bit quantized `Qwen-Image-Edit-2509
    This example requires ``diffusers`` version 0.36.0 or higher.
 
 Custom LoRA support is under development.
+
+Distilled Qwen-Image-Edit-2509 (Qwen-Image-Edit-2509-Lightning)
+---------------------------------------------------------------
+
+For faster inference of the 2509 branch, we provide pre-quantized Lightning variants by integrating `Qwen-Image-Lightning LoRAs <hf_qwen-image-lightning_>`_.
+See the example script below:
+
+.. literalinclude:: ../../../examples/v1/qwen-image-edit-2509-lightning.py
+    :language: python
+    :caption: Running Qwen-Image-Edit-2509-Lightning (`examples/v1/qwen-image-edit-2509-lightning.py <https://github.com/nunchaku-tech/nunchaku/blob/main/examples/v1/qwen-image-edit-2509-lightning.py>`__)
+    :linenos:

docs/source/usage/qwen-image.rst

@@ -7,8 +7,8 @@ Original Qwen-Image
 .. image:: https://huggingface.co/datasets/nunchaku-tech/cdn/resolve/main/nunchaku/assets/qwen-image.jpg
    :alt: Qwen-Image with Nunchaku
 
-`Qwen-Image <hf_qwen-image>`_ is an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering.
-Below is a minimal example for running the 4-bit quantized `Qwen-Image <hf_qwen-image>`_ model with Nunchaku.
+`Qwen-Image <hf_qwen-image_>`_ is an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering.
+Below is a minimal example for running the 4-bit quantized `Qwen-Image <hf_qwen-image_>`_ model with Nunchaku.
 Nunchaku offers an API compatible with `Diffusers <github_diffusers_>`_, allowing for a familiar user experience.
 
 .. literalinclude:: ../../../examples/v1/qwen-image.py
@@ -29,7 +29,7 @@ The :meth:`~nunchaku.models.transformers.transformer_qwenimage.NunchakuQwenImage
 Distilled Qwen-Image (Qwen-Image-Lightning)
 -------------------------------------------
 
-For faster inference, we provide pre-quantized 4-step and 8-step Qwen-Image models by integrating `Qwen-Image-Lightning LoRAs <hf_qwen-image-lightning>`_.
+For faster inference, we provide pre-quantized 4-step and 8-step Qwen-Image models by integrating `Qwen-Image-Lightning LoRAs <hf_qwen-image-lightning_>`_.
 See the example script below:
 
 .. literalinclude:: ../../../examples/v1/qwen-image-lightning.py

examples/v1/qwen-image-edit-2509-lightning.py

@@ -28,13 +28,13 @@
 
 num_inference_steps = 4  # you can also use the 8-step model to improve the quality
 rank = 32  # you can also use the rank=128 model to improve the quality
-model_path = f"nunchaku-tech/nunchaku-qwen-image-edit-2509-lightning/svdq-{get_precision()}_r{rank}-qwen-image-edit-2509-lightningv2.0-{num_inference_steps}steps.safetensors"
+model_path = f"nunchaku-tech/nunchaku-qwen-image-edit-2509/svdq-{get_precision()}_r{rank}-qwen-image-edit-2509-lightningv2.0-{num_inference_steps}steps.safetensors"
 
 # Load the model
 transformer = NunchakuQwenImageTransformer2DModel.from_pretrained(model_path)
 
 pipeline = QwenImageEditPlusPipeline.from_pretrained(
-    "Qwen/Qwen-Image-Edit-2509", transformer=transformer, torch_dtype=torch.bfloat16
+    "Qwen/Qwen-Image-Edit-2509", transformer=transformer, scheduler=scheduler, torch_dtype=torch.bfloat16
 )
 
 if get_gpu_memory() > 18: