onnxruntime
Utilize ext data location to reduce qd matmul memory usage
#21451
Merged

Utilize ext data location to reduce qd matmul memory usage #21451

fajin-corp merged 10 commits into main from fajin/dqmatmulmemoryusage
fajin-corp
fajin-corp fajin-corp requested a review from skottmckay skottmckay 1 year ago
fajin-corp fajin-corp requested a review from snnn snnn 1 year ago
skottmckay skottmckay requested a review from pranavsharma pranavsharma 1 year ago
skottmckay
fajin-corp
fajin-corp hack ext data location to reduce qd matmul memory usage
cecf225e
fajin-corp fixing gls::narrow
ae4fffb0
fajin-corp fix linting
d397d577
fajin-corp fix linting
4ac79aae
fajin-corp fajin-corp force pushed from dbc29037 to 4ac79aae 1 year ago
fajin-corp try fixing memory sanitize errro
2a10552a
fajin-corp fixing optional
21f63fbe
fajin-corp fix memory leak when tensor is too small to become external buffer
a6e4ae47
fajin-corp add memory leak guardian and fix linting
5a9d15c8
yufenglee
yufenglee
yufenglee dismissed these changes on 2024-07-30
fajin-corp fajin-corp changed the title hack ext data location to reduce qd matmul memory usage Utilize ext data location to reduce qd matmul memory usage 1 year ago
fajin-corp
azure-pipelines
azure-pipelines
fajin-corp
azure-pipelines
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
snnn
snnn commented on 2024-07-30
fajin-corp resolve comments
20f6cea7
fajin-corp fajin-corp dismissed their stale review via 20f6cea7 1 year ago
fajin-corp resolve comments
f3a456f7
snnn
snnn approved these changes on 2024-07-30
fajin-corp fajin-corp merged e7aa1160 into main 1 year ago
fajin-corp fajin-corp deleted the fajin/dqmatmulmemoryusage branch 1 year ago
fajin-corp fajin-corp added release:1.19.0
prathikr prathikr added cherry-picked

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone