問題1: AssertionError: Torch not compiled with CUDA enabled？

解決辦法：修改代碼以 CPU 運行

第一步：找到 /ComfyUI/custom_nodes/ComfyUI-MMAudio/mmaudio/ext/autoencoder/vae.py文件中的下面這兩行代碼

self.data_mean = nn.Buffer(torch.tensor(DATA_MEAN_128D, dtype=torch.float32).cuda())
self.data_std = nn.Buffer(torch.tensor(DATA_STD_128D, dtype=torch.float32).cuda())

第二步：將其修改為：

self.data_mean = nn.Buffer(torch.tensor(DATA_MEAN_128D, dtype=torch.float32).cpu())
self.data_std = nn.Buffer(torch.tensor(DATA_STD_128D, dtype=torch.float32).cpu())

問題2: NotImplementedError: The operator ‘aten::_upsample_bicubic2d_aa.out’ is not currently implemented for the MPS device. If you want this op to be added in priority during the prototype phase of this feature, please comment on https://github.com/pytorch/pytorch/issues/77764. As a temporary fix, you can set the environment variable PYTORCH_ENABLE_MPS_FALLBACK=1 to use the CPU as a fallback for this op. WARNING: this will be slower than running natively on MPS？

解決辦法：臨時解決辦法：啟用 MPS 回退機制

在main.py中 import os 后面添加如下代碼：

import os
os.environ["PYTORCH_ENABLE_MPS_FALLBACK"] = "1"

問題3: ComfyUI/custom_nodes/ComfyUI-MMAudio/mmaudio/model/utils/features_utils.py", line 112, in encode_video_with_sync x = torch.stack(segments, dim=1) # (B, S, T, C, H, W) RuntimeError: stack expects a non-empty TensorList？

解決辦法：上傳的視頻時長太短，重新選擇時長稍微長些的視頻。

問題4: NotImplementedError: Output channels > 65536 not supported at the MPS device？

解決辦法：更新 PyTorch 版本

pip install --upgrade torch torchvision torchaudio

問題5: ComfyUI/custom_nodes/ComfyUI-VideoHelperSuite/videohelpersuite/server.py", line 86, in view_video if int(query.get(‘force_rate’,0)) != 0:^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ValueError: invalid literal for int() with base 10: ‘7.87890625’？

解決辦法：在VideoHelperSuite節點的format格式animatediff 修改為Hunyuan