Skip to content

Fix the bug in realquant #301

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Fix the bug in realquant #301

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
760c583
catch input.numel()<1 in _RealQuantMegatronParallelLinear
yeyu-nvidia Sep 8, 2025
b4cb2d5
fix fp8 gemm bug
yeyu-nvidia Sep 9, 2025
fc44595
fix a typo
yeyu-nvidia Sep 9, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions modelopt/torch/quantization/plugins/megatron.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -418,8 +418,10 @@ class forward(). This is not desired since _forward_impl introduces much more ar
while the original forward only takes 1 positional argument. We must above the fallback path
in RealQuantLinear.forward().
"""
if self._should_run_real_quant_gemm and self.get_real_quant_gemm_impl(
input, *args, **kwargs
if (
self._should_run_real_quant_gemm
and self.get_real_quant_gemm_impl(input, *args, **kwargs)
and input.numel() > 1
):
allreduce_dgrad = kwargs.get("allreduce_dgrad", False)
tp_group = kwargs.get("tp_group")
Expand Down
Loading