NVIDIA-NeMo / NVIDIA-NeMo/Gym

NOOA post-training readiness

Open
#3,458 0 comments 0 reactions 1 assignee View on GitHub

@init-nikhil is already working on this.

Since Sep 17, 2026.

area:agent area:training tracking training
Dominant language
Python
Stars
1.2k
Forks
349
Avg merge
1d 23h
Merged PRs (30d)
324

Description

Goal

Define and close the remaining gaps required to use NOOA agents for Gym-based post-training.

This work builds on the object-oriented agent integration in #2096 and the CLU evaluations for NOOA agents in #3453.

Scope

The initial scope is an RFC describing the target post-training architecture, required rollout contracts, integration boundaries, validation plan, and staged delivery milestones.

Features

  • #3459

Success criteria

  • The RFC defines how NOOA rollouts become training-ready data.
  • Required model, trajectory, reward, token, and failure contracts are explicit.
  • Dependencies on the existing NOOA integration and CLU evaluation work are identified.
  • Reviewers can approve a concrete implementation plan with measurable readiness gates.

Contributor guide

Open the contributing guide

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.