Skip to content

Reduce memory usage for fp4 moe #8413

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

Merged
zhyncs merged 4 commits into from
Jul 29, 2025
Merged

Reduce memory usage for fp4 moe #8413

Changes from 3 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
05708e8
more
fzyzcjy Jul 27, 2025
cb9a14f
more
fzyzcjy Jul 27, 2025
ae6ceeb
Revert "more"
fzyzcjy Jul 27, 2025
49aad4c
Merge branch 'main' into feat/rm_scale
zhyncs Jul 27, 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
2 changes: 2 additions & 0 deletions python/sglang/srt/layers/quantization/modelopt_quant.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -885,6 +885,7 @@ def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
layer.w13_blockscale_swizzled = Parameter(
w13_blockscale_swizzled, requires_grad=False
)
del layer.w13_weight_scale

# This is for quantization, so we need to invert it.
layer.w13_input_scale_quant = Parameter(
Expand Down Expand Up @@ -920,6 +921,7 @@ def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
layer.w2_blockscale_swizzled = Parameter(
w2_blockscale_swizzled, requires_grad=False
)
del layer.w2_weight_scale
layer.w2_weight = Parameter(layer.w2_weight.data, requires_grad=False)

device = layer.w13_weight.device
Expand Down
Loading