diffusers

diffusers copied to clipboard

Any Wan version crashes

13

### Describe the bug Running any diffusers Wan example code crashes for me, when running from inside Blender python (everything else in Diffusers has run fine from Blender Python over...

tin2tin

how to quantization wan 2.2 vace after loading lora?

1

```python3 diffusers.WanVACEPipeline.from_pretrained('linoyts/Wan2.2-VACE-Fun-14B-diffusers', vae=diffusers.AutoencoderKLWan.from_pretrained('linoyts/Wan2.2-VACE-Fun-14B-diffusers', subfolder='vae', torch_dtype=torch.float32), torch_dtype=torch.bfloat16, quantization_config=diffusers.PipelineQuantizationConfig(quant_backend='bitsandbytes_8bit', quant_kwargs={'load_in_8bit':True}, components_to_quantize=['transformer', 'transformer_2'])).save_pretrained('wan') ``` normally I can save the quantization model in this way But now I want to merge lora and the...

chaowenguo

Z-Image Turbo Controlnet Union is out

3

Worth considering? https://huggingface.co/alibaba-pai/Z-Image-Turbo-Fun-Controlnet-Union

tin2tin

cannot import name 'export_to_video' from 'diffusers

2

### Describe the bug tried different versions of diffusers but getting this error, trying to run the https://github.com/SkyworkAI/SkyReels-V2 ### Reproduction tried different versions of diffusers but getting this error, trying...

Arslan-Mehmood1

perf: optimize CausalConv3d for wan autoencoders

7

# What does this PR do? By optimizing CausalConv3d, this patch improves the overall performance of wan autoencoders by 5-10%. ## Before submitting - [ ] This PR fixes a...

c8ef

Fix `QwenImageEditPlusPipeline` not using the passed in width/height for the VAE

4

# What does this PR do? Fixes `QwenImageEditPlusPipeline` not using the passed in width/height for the VAE. Currently if you set the width/height to anything other than 1024x1024, for example...

cakedan

# What does this PR do? Following the plan outlined for Diffusers 1.0.0, this PR introduces changes to our model testing approach in order to reduce the overhead involved in...

DN6

How to use quantizer after pipeline loaded?

1

How to use quantizer after pipeline loaded? - Currently ```python # Quantization occurs at load time. pipe = QwenImagePipeline.from_pretrained( ( args.model_path if args.model_path is not None else os.environ.get( "QWEN_IMAGE_DIR", "Qwen/Qwen-Image",...

DefTruth

[wip] [core] introduce pipeline-specific mixins to hold common methods

1

# What does this PR do? As discussed internally, this PR introduces pipeline-specific Mixin classes that hold common methods shared across different task-specific pipelines. The actual pipelines (such as `StableDiffusionPipeline`,...

sayakpaul

[core] support device type device_maps to work with offloading.

1

# What does this PR do? This PR allows users to pass a `device_map="cpu"` while initializing a pipeline and then enable model CPU offloading. This is beneficial when users want...

sayakpaul