NVIDIA-NeMo / NVIDIA-NeMo/RL

Inconsistent DTensorPolicyWorker logits with TP>1 and bf16

Open
#1,227 0 comments 0 reactions 1 assignee Assigned to @joyang-nv View on GitHub
bug t-pytdensor
Dominant language
Python
Stars
2k
Forks
561
Avg merge
4d 5h
Merged PRs (30d)
145

Description

**Describe the bug**

When using TP>1 and bf16 policy.precision, we observed inconsistent logits computation during the forward pass of DTensorPolicyWorker, impacting RM training.
- Identical rows yield different outputs depending on their position in the microbatch
- Consecutive forward passes produce different outputs

Using fp32 for policy.precision seems to mitigate this issue.

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.