auto-round
Optimize mtp tensor process by compiling layer names once for all
#1554

Merged

Optimize mtp tensor process by compiling layer names once for all #1554

xin3he merged 4 commits into main from xinhe/3-17

Optimize mtp tensor process by compiling layer names once for all

d344fd5e

xin3he requested a review from

copilot-pull-request-reviewer 4 days ago

xin3he requested a review from

wenhuach21 4 days ago

xin3he requested a review from

yiliu30 4 days ago

xin3he requested a review from

WeiweiZhang1 4 days ago

xin3he requested a review from

n1ck-guo 4 days ago

copilot-pull-request-reviewer commented on 2026-03-17

wenhuach21 commented on 2026-03-17

wenhuach21 requested a review from

wenhuach21 4 days ago

wenhuach21 approved these changes on 2026-03-17

Potential fix for pull request finding

16692ade

Potential fix for pull request finding

191b3e6a

apply compress_layer_names for all cases

d1089436

chensuyue added this to the 0.12.0 milestone 4 days ago

xin3he merged 1111bcd4 into main 4 days ago

xin3he deleted the xinhe/3-17 branch 4 days ago

Reviewers

wenhuach21

yiliu30

WeiweiZhang1

n1ck-guo

Assignees

No one assigned

Labels

None yet

Milestone

0.12.0