NVIDIA-NeMo / NVIDIA-NeMo/Megatron-Bridge
Add test to check loss on checkpoint resume
Open
@maanug-nv is already working on this.
Since Aug 20, 2025.
area:ci
ci
tracking
- Dominant language
- Python
- Stars
- 921
- Forks
- 506
- Avg merge
- 1d 16h
- Merged PRs (30d)
- 236
Description
Compare loss curve on checkpoint resume with a continuous run. In other words:
- Do 10 steps continuous run. get loss from TB file.
- Start 10 step run with ckpt and exit at step 5. Resume and complete. Get loss from TB file.
- Compare loss from 1 and 2
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.