Conversation
The `get_i2v_mask` method had a hardcoded `device="cuda"` default, which crashes on non-CUDA accelerators (Ascend NPU, etc.) with "Torch not compiled with CUDA enabled" when called without an explicit device argument. Change the default to None and resolve via `self._execution_device`, matching the pattern used across other pipeline methods. Verified on Ascend 910B NPU: torch.zeros(device="cuda") crashes, fix with device-agnostic resolution creates tensors on the correct device.
…v_mask` `get_i2v_mask` in `modular_pipelines/wan_animate_2/encoders.py` defaulted its `device` argument to `"cuda"`, so any call that omitted it allocated the mask on CUDA and failed on non-CUDA accelerators (NPU/XPU/MPS/CPU) with `AssertionError: Torch not compiled with CUDA enabled`. The default is now `None` and raises a clear `ValueError` instead of silently allocating on CUDA. All in-tree call sites already pass `device=` explicitly (`wan_animate_2/denoise.py` lines 128 and 220), so behaviour for existing callers is unchanged. Same fix as the one applied to `WanAnimatePipeline.get_i2v_mask` in this PR.
Author
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #14881
Description: Fix the
get_i2v_maskmethod'sdevice="cuda"default parameter which causes a crash on non-CUDA accelerators (Ascend NPU, Intel XPU, Apple MPS, etc.) when called without an explicit device argument.Change:
devicefrom"cuda"toNonedevice = device or self._execution_deviceat the function topprepare_reference_image_latents,encode_image, and other pipeline methodsVerification on Ascend 910B (torch
2.15.0.dev20260917+cpu+ torch_npu2.15.0.dev20260917+gitec69335, CANN 9.2.0-beta.2):torch.zeros(1, device="cuda")crashes withAssertionError: Torch not compiled with CUDA enableddevice = device or self._execution_deviceresolves tonpu:0, tensor created successfully on NPUSecond occurrence (included in this PR): the module-level helper
get_i2v_maskatsrc/diffusers/modular_pipelines/wan_animate_2/encoders.py:84carried the samedevice="cuda"default. It is aplain function (no
self._execution_deviceto fall back to), so its default is nowNoneand it raises aclear
ValueErrorwhendeviceis omitted instead of silently allocating on CUDA. Both in-tree call sitesalready pass
device=explicitly (wan_animate_2/denoise.py:128and:220), so existing callers areunaffected.
Scope note (see #14881): all in-tree call sites in both files already pass
deviceexplicitly, so this is alatent default / API-hygiene fix rather than a fix for a currently-failing in-tree path.
(Replaces #14765, which the issue-link auto-close bot closed on 2026-09-25 and which GitHub refused to reopen - same commit, unchanged content.)