Skip to content

[perf][WIP]: using NZ optimization for quantized GMM #906

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

Closed
wants to merge 1 commit into from
+7 −0
Closed

[perf][WIP]: using NZ optimization for quantized GMM #906

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
7 changes: 7 additions & 0 deletions vllm_ascend/quantization/w8a8_dynamic.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -663,6 +663,13 @@ def process_weights_after_loading(self, layer):
1, 2).contiguous()
layer.w2_weight.data = layer.w2_weight.data.transpose(
1, 2).contiguous()
# This optimization relies on the modifications in torch_npu, otherwise accuracy
# problem will happen. But we can evaluate the inference speed by transforming
# weights to NZ (29)
layer.w13_weight.data = torch_npu.npu_format_cast(
layer.w13_weight.data, 29)
layer.w2_weight.data = torch_npu.npu_format_cast(
layer.w2_weight.data, 29)
layer.w13_weight_scale.data = layer.w13_weight_scale.data.view(
layer.w13_weight_scale.data.shape[0], -1)
layer.w13_weight_offset.data = layer.w13_weight_offset.data.view(
Expand Down