OpenEuroLLM / OpenEuroLLM/Taskboard
Throughput scaling - Jupiter - Qwen3-32B Dense model
Open
@tvosch is already working on this.
Since Jun 3, 2026.
T4.1 - Optimization HPC
- Dominant language
- No language data
- Stars
- 3
- Forks
- 0
- PR merge metrics
- No merged PRs in 30d
Description
In preparation for the flagship model training, we need to ensure satisfactory throughput with a dense model. This issue shall get updated as the details get figured out.
Goal
Using the https://huggingface.co/Qwen/Qwen3-32B configuration get preliminary results for throughput scaling up to X amount of nodes.
Target
What can be considered "satisfactory" throughput? Look for previous scaling results done by OpenSci folks and others.
Considerations
- What batch size to use?
- Best model parallelism configuration
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.
Assessment
This issue has not been assessed yet.