Issue

[IEM-Regular-User]

Hello,
Im using windows 11,
AMD as recommended 25.20.01.17
GPU 7900 XTX
I ran through your setup which i believe worked correctly, but when trying to generate an image it seems to default CPU then error at the end see logs.
I'm not very familiar with ComfyUI so i was just loading a template then hitting run.
Any suggestions are welcome :)
rocm_output.txt

[OrsoEric]

I see you tried Zimage and Qwen Edit, there seems to be a few issues with APIs that should not be needed as this is fully local, make sure you use local in the tutorial with the workflows examples.

Zimage works for me, , Qwen Edit for me is 90s per iteration.

Try a SD1.5 workflow and a Flux workflow for sanity check.

Use the workflow SD1.5 in the tutorials in ComfyUI and download their reccomanded models.

The following lines check that ROCm is working properly

.venv\Scripts\activate

python -c "import torch" 2>nul && echo Success || echo Failure

python -c "import torch; print(torch.cuda.is_available())"

python -c "import torch; print(f'device name [0]:', torch.cuda.get_device_name(0))"

python -m torch.utils.collect_env