onnxruntime
Added a tool to quantize Gather to GatherBlockQuantized
#21697
Merged

Added a tool to quantize Gather to GatherBlockQuantized #21697

fajin-corp merged 19 commits into main from fajin/gatherblockquantizetool
fajin-corp
github-advanced-security
github-advanced-security commented on 2024-08-10
fajin-corp fajin-corp requested a review from yufenglee yufenglee 2 years ago
yufenglee
yufenglee commented on 2024-08-14
yufenglee
yufenglee commented on 2024-08-14
fajin-corp updated interface to add quantize ops argument
154785ee
fajin-corp updated quantize op to be set
d26d329b
fajin-corp added per op quant axis argument
7af17419
fajin-corp added quantize gather framework
7f4d545f
fajin-corp not supporting hqq
37be006a
fajin-corp finished tool code
e11ff161
fajin-corp added ut
8ddd272f
fajin-corp update quantize.py
ec838aed
fajin-corp fixing ut
d169ccd2
fajin-corp fix linting
a141af44
fajin-corp fix type hinting
aa190625
fajin-corp fix union in ut
41f3537e
fajin-corp fix type hinting in ut
7e63c4c5
fajin-corp fix training package import error
cc467cd7
fajin-corp fix linting
68622888
fajin-corp add nodes to include
95a712fa
fajin-corp fajin-corp force pushed from d54a2e2b to 95a712fa 2 years ago
fajin-corp fix ut
8379d684
yufenglee
yufenglee commented on 2024-08-16
fajin-corp change quantize predicate order
2a536e07
yufenglee
yufenglee dismissed these changes on 2024-08-16
fajin-corp fix linting
457321d3
fajin-corp fajin-corp dismissed their stale review via 457321d3 2 years ago
yufenglee
yufenglee approved these changes on 2024-08-17
fajin-corp fajin-corp merged 64674c50 into main 2 years ago
fajin-corp fajin-corp deleted the fajin/gatherblockquantizetool branch 2 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone