Megatron cannot save checkpoint?

Repository metrics

Stars

23,003 stars (23,003 stars)

PR merge metrics

Avg merge 3d 6h (Avg merge 3d 6h)125 merged PRs in 30d (125 merged PRs in 30d)

Description

Contributor guide

Research direction

Investigate the missing `save checkpoint` method in `verl/workers/megatron workers.py` at line 428. Examine existing checkpoint implementations in other VERL workers (e.g., `fsdp workers.py`) and Megatron LM's checkpointing utilities. Implement a method that saves model and optimizer states in a format compatible with `load checkpoint`. Review the 4 comments on the issue for any additional context or proposed approaches.

Tech stack

pythonpytorch

Domain

backendmachine learning

Issue type

Feature

DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.

3

Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.

1-3 hours

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.

Clear

Prerequisites

PythonPyTorchMegatron (distributed training)

Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.

55