NVIDIA-NeMo / NVIDIA-NeMo/RL

Async data collector is single-epoch

Open
#1,814 2 comments 0 reactions 1 assignee Claimed by @parthchadha View on GitHub
bug
Dominant language
Python
Stars
2k
Forks
561
Avg merge
4d 5h
Merged PRs (30d)
145

Description

**Describe the bug**

The async data collector (in the `_collection_loop` function) will stop data collection when the dataloader is exhausted (which forces it to be single-epoch). I think it needs to continue forever until it is stopped from the main grpo loop.

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.