OpenEuroLLM / OpenEuroLLM/Taskboard

baby run post decay performance via weight averaging

Open
#336 0 comments 0 reactions 1 assignee View on GitHub

@Niccolo-Ajroldi is already working on this.

Since Jun 26, 2026.

T4.2 - model exploration
Dominant language
No language data
Stars
3
Forks
0
No merged PRs in 30d

Description

Goal
Estimate and compute the performance of the baby model after LR decay.

Description
We train at stable LR, what's the performance after LR decay?

Deliverable scope

  • Decay the 1T checkpoint
  • Predict the performance of the baby model after LR decay using Weight Averaging.
  • Compare the performance of the two weight average checkpoint with the lr-decayed one

Dependencies
Needs the implementation of a weight averaging routine that works on distributed checkpoints.
Needs transferring of checkpoint from Leo to LUMI [in progress].

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.