NVIDIA-NeMo / NVIDIA-NeMo/Curator
Fault Tolerance for all pipelines
Open
@oyilmaz-nvidia is already working on this.
Since Mar 21, 2026.
enhancement
- Dominant language
- Python
- Stars
- 1.8k
- Forks
- 328
- Avg merge
- 4d 5h
- Merged PRs (30d)
- 30
Description
Is your feature request related to a problem? Please describe.
For long pipelines, it's possible that the tasks might fail and the pipeline needs to restart from scratch. In a cluster environment with lots of GPUs, it's usually hard to manage this.
Describe the solution you'd like
Use https://github.com/NVIDIA/nvidia-resiliency-ext to add fault tolerance. The easy way to add this support is that using this library, hardware failures are detected automatically and the whole process is started automatically.
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.
Assessment
This issue has not been assessed yet.