torch.AcceleratorError: CUDA error: device kernel image is invalid
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2
- Forks
- 0
- Avg merge
- 9h 33m
- Merged PRs (30d)
- 52
Description
update to rocm10，When the workload reaches clip, an error is reported directly. Here is my system situation, please help me
error info：
error info.txt
win11 25h2 ，9700x，9070xt，ram 64g， Adrenalin Edition26.81.1
Python 3.12.10
python.exe -m pip list | findstr rocm
rocm 10.0.0
rocm-bootstrap 0.1.0
rocm-sdk-core 10.0.0
rocm-sdk-device-gfx1201 10.0.0
rocm-sdk-libraries 10.0.0
rocm-sdk-libraries-custom 7.2.1
torch 2.13.0+rocm10.0.0
torchaudio 2.11.0.2+rocm10.0.0
torchvision 0.28.0+rocm10.0.0
python.exe -c "import torch; print('CUDA/ROCm Available:', torch.cuda.is_available()); print('Device Name:', torch.cuda.get_device_name(0) if torch.cuda.is_available() else 'None')"
[stderr]
CUDA/ROCm Available: True
Device Name: AMD Radeon RX 9070 XT
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the attached error.info.txt and reproduce the failure when the workload reaches CLIP using the listed Windows 11, Radeon RX 9070 XT, Python 3.12.10, ROCm 10.0.0, and PyTorch 2.13.0 environment. Compare the reported device and package versions with the kernel-image error, and consider the issue complete when the cause or a verified resolution is documented.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100