WanToDance: hardcoded `device='cuda'` in music encoder construction crashes model loading on Ascend NPU

Vấn đề nằm ở diffsynth/models/wan_video_dit.py dòng 462 và diffsynth/models/wantodance.py dòng 156. Hãy xóa đối số được mã hóa cứng device='cuda' khỏi hàm tạo WanToDanceMusicEncoderLayer. Kiểm tra bản sửa lỗi bằng cách chạy một bài kiểm tra tải mô hình trên một thiết bị không phải CUDA, chẳng hạn như CPU hoặc NPU, để đảm bảo không xảy ra lỗi khẳng định.

Summary

WanModel.__init__ builds the WanToDance music encoder with a hardcoded CUDA device (diffsynth/models/wan_video_dit.py:462):

self.music_encoder.append(
    WanToDanceMusicEncoderLayer(
        d_model=latent_dim,
        nhead=nhead,
        dim_feedforward=ff_size,
        dropout=dropout,
        activation=activation,
        batch_first=True,
        rotary=rotary,
        device='cuda',
    )
)

WanToDanceMusicEncoderLayer forwards device to nn.MultiheadAttention(..., device=device) (diffsynth/models/wantodance.py:156), so the attention parameters are allocated on CUDA at construction time.

Impact

On any build without CUDA (Ascend NPU, CPU-only, ...) the model cannot be constructed at all, before the surrounding pipeline ever gets a chance to move it to the target device:

AssertionError: Torch not compiled with CUDA enabled

This block is gated by wantodance_enable_global / wantodance_enable_dynamicfps / wantodance_enable_unimodel, i.e. loading a WanToDance model on Ascend fails outright. Every sibling layer in the same block (nn.Linear, nn.Sequential) leaves the device to the caller's .to(device), so the hardcoded argument is also inconsistent with the rest of the file — and with diffsynth.core.device.npu_compatible_device, which the project already ships for NPU support.

Reproduction

# Ascend 910B4, torch 2.15.0.dev + torch_npu 2.15.0.dev
from wantodance import WanToDanceMusicEncoderLayer
WanToDanceMusicEncoderLayer(d_model=256, nhead=4, dim_feedforward=1024,
                            dropout=0.1, batch_first=True, rotary=None,
                            device='cuda')
# AssertionError: Torch not compiled with CUDA enabled

Proposed fix

Drop the hardcoded device='cuda'; the module is moved to the real device by the pipeline afterwards, exactly like its sibling layers. Verified on real Ascend 910B4 hardware (construction + forward pass on NPU); I will open a PR referencing this issue.