Unexpected output from WanVideoPostUnit_AnimateFacePixelValues

Since support of Wan2.2-Animate-14B, the example inference scripts for wan no longer work. The issue is the new WanVideoPostUnit_AnimateFacePixelValues does not return expected number of values.

python3 examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py                                          

...

Loading models from: ./models/Wan-AI/Wan2.1-T2V-1.3B/Wan2.1_VAE.pth
    model_name: wan_video_vae model_class: WanVideoVAE
    The following models are loaded: ['wan_video_vae'].
Using wan_video_text_encoder from ./models/Wan-AI/Wan2.1-T2V-1.3B/models_t5_umt5-xxl-enc-bf16.pth.
Using wan_video_dit from ./models/Wan-AI/Wan2.1-T2V-1.3B/diffusion_pytorch_model.safetensors.
Using wan_video_vae from ./models/Wan-AI/Wan2.1-T2V-1.3B/Wan2.1_VAE.pth.
No wan_video_image_encoder models available.
No wan_video_motion_controller models available.
No wan_video_vace models available.
No wans2v_audio_encoder models available.
No wan_video_animate_adapter models available.
Downloading Model from https://www.modelscope.cn to directory: DiffSynth-Studio/models/Wan-AI/Wan2.1-T2V-1.3B
2025-10-01 23:16:16,283 - modelscope - INFO - Target directory already exists, skipping creation.
Traceback (most recent call last):
  File "DiffSynth-Studio/examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py", line 19, in <module>
    video = pipe(
  File ".local/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
    return func(*args, **kwargs)
  File "DiffSynth-Studio/diffsynth/pipelines/wan_video_new.py", line 515, in __call__
    inputs_shared, inputs_posi, inputs_nega = self.unit_runner(unit, self, inputs_shared, inputs_posi, inputs_nega)
  File "DiffSynth-Studio/diffsynth/utils/__init__.py", line 252, in __call__
    inputs_shared, inputs_posi, inputs_nega = unit.process(pipe, inputs_shared=inputs_shared, inputs_posi=inputs_posi, inputs_nega=inputs_nega)
ValueError: not enough values to unpack (expected 3, got 0)

Commit ed256ef is the last commit that works.

Start with examples/wanvideo/model_inference/Wan2.1-T2V-1.3B.py and reproduce the failure. Trace the unit call through diffsynth/pipelines/wan_video_new.py and diffsynth/utils/init.py, comparing behavior with commit ed256ef. Done means the example inference scripts run without the unpacking error and the post-unit output has the expected shape.