OpenEuroLLM / OpenEuroLLM/Taskboard
sparse multilingual models scaling experiments
Open
@abhash-er is already working on this.
Since Jun 3, 2026.
T4.2 - model exploration
- Dominant language
- No language data
- Stars
- 3
- Forks
- 0
- PR merge metrics
- No merged PRs in 30d
Description
This is a parent issue to keep track of everything related to sparse (MoE) multilingual models scaling law experiments.
- Model architecture prepared in this sheet
- Abhash's dashboard for english runs: https://abhash-er.github.io/moe-analysis-dashboard/
- Updated configs on the
exp_moe_scalingbranch: models, sweeps, main configs to run with run_autoexp.
Related issues:
- #202 (caution: the models trained here show significant degradation in MoE layers, routing collapse etc., see comment below)
- #165 (small sweep limited to only 10B tokens)
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.