Stability-AI / Stability-AI/stable-audio-3
TensorRT plugin samel::diff_attn_swa missing
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 763
- Forks
- 96
- Avg merge
- 3h 4m
- Merged PRs (30d)
- 4
Description
The provided TensorRT support is limited to H100/H200. I tried converting the ONNX model to TensorRT for my hardware and succeeded with the DiT, but Medium’s decoder is still blocked because it depends on a missing custom plugin.
Contributor guide
No contributing guide indexed for this repository
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 by tracing the TensorRT conversion path for the Medium decoder and the missing samel::diff_attn_swa custom plugin described in the issue. Compare the working DiT conversion with the decoder failure, then verify that the Medium decoder converts successfully on hardware beyond H100/H200.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100