verl-project/verl

Support Generative Reward Model (GenRM)

Closed

#229 opened on Feb 9, 2025

View on GitHubContributor guide
 (17 comments) (0 reactions) (1 assignee)Python (4,413 forks)auto 404
enhancementgood first issue

Repository metrics

Stars
 (22,985 stars)
PR merge metrics
 (Avg merge 3d 6h) (125 merged PRs in 30d)

Description

According to the documentation, veRL only supports AutoModelForSequenceClassification. What would be the best way to implement generative reward model (GenRM) for veRL? I tried looking at FSDP Workers and Megatron-LM Workers but they no longer exist.

Contributor guide

Research direction
Examine the current reward model class in veRL that uses AutoModelForSequenceClassification and adapt it to support generative models like a language model that outputs a token representing the reward. Explore how FSDP workers and Megatron workers were previously structured, and consider integrating a generative model head that produces a reward token instead of a scalar.
Tech stack
python
Domain
machine learningai
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Mostly clear
Prerequisites
PythonPyTorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
30