Fix: TypeError on AMD GPUs by moving audio tensor to CPU before numpy conversion #11787
+1
−1
Conversation
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
TheephopWS
requested review from
Kosinkadink,
comfyanonymous and
guill
as code owners
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Summary
Fixes TypeError: can't convert cuda:0 device type tensor to numpy when using SaveVideo with LTX-2 audio generation on AMD hardware.
This change resolves issue #11785 for AMD GPU users experiencing crashes during video generation.
Problem
In
comfy_api/latest/_input_impl/video_types.py, the audio tensor generated by LTXVAudioVAEDecode remains on the GPU device. The code attempts to call .numpy() directly on this CUDA tensor. While NVIDIA drivers often handle this implicitly or differently in some contexts, PyTorch explicitly throws an error on AMD ROCm (and strict CUDA environments) requiring tensors to be on the host memory first.This crashes the workflow with:
TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.
Fix
Added .cpu() before .numpy() in the tensor conversion chain.
Changed .float().numpy() to .float().cpu().numpy().