PR #5478 Improve checkpointing for Zero stage 1

Improve checkpointing for Zero stage 1 #5478

ashbhandare merged 18 commits into master from aibhanda/zero_1_ckpt

ashbhandare requested a review from

BowenBao 5 years ago

ashbhandare requested a review from

liqunfu 5 years ago

ashbhandare requested a review from

spandantiwari 5 years ago

ashbhandare requested a review from

thiagocrepaldi 5 years ago

ashbhandare requested a review 5 years ago

ashbhandare requested a review from

jessebenson 5 years ago

thiagocrepaldi dismissed these changes on 2020-10-13

thiagocrepaldi commented on 2020-10-13

thiagocrepaldi requested a review from

baijumeswani 5 years ago

jessebenson commented on 2020-10-15

jessebenson dismissed these changes on 2020-10-15

ashbhandare dismissed their stale review via df4070cb 5 years ago

ashbhandare force pushed from c27b5295 to df4070cb 5 years ago

ashbhandare force pushed from df4070cb 5 years ago

ashbhandare force pushed to a1ca6e9b 5 years ago

baijumeswani commented on 2020-10-29

baijumeswani dismissed these changes on 2020-10-30

ashbhandare dismissed their stale review 5 years ago

change has been addressed, dismissing to unblock

ashbhandare dismissed their stale review via 5a453d00 5 years ago

thiagocrepaldi requested changes on 2020-11-02

ashbhandare force pushed 5 years ago

ashbhandare force pushed to ca498c7b 5 years ago

baijumeswani commented on 2020-11-03

ashbhandare force pushed 5 years ago

ashbhandare force pushed to 20b7fead 5 years ago

ashbhandare force pushed from 20b7fead to 705d25e3 5 years ago

ashbhandare force pushed from 31f1ca8b to dd2db4c3 5 years ago

baijumeswani commented on 2020-11-25

thiagocrepaldi commented on 2020-11-25

ashbhandare force pushed from 0d286b1e to 372e3724 5 years ago

ashbhandare force pushed from 372e3724 5 years ago

ashbhandare force pushed to 59a2698d 5 years ago

Initial running changes

d71857d8

Checkpointing aggregation changes

2cbe436c

compare with older version

903a44aa

initial cleanup

b31e833e

Add zero test, minor fix

340ef43e

Fix zero test, transform, formatting

3d776787

Review comments

1757e5cc

add more unit tests

bd402344

review comments

0146849e

Try fix CI

29510b78

Add additional check on just aggregation code

9055d292

Try fix ckpt gen

82fef12d

Add pregenerated ckpt for CI, enable zero test in e2e

30d2ce78

Moving test to nightly, removing ckpt files

a928d873

Add tests to dist GPU CI

6adec1b4

Fix dist test

99ca90dd

Review comments

5bc59119

ashbhandare force pushed from 59a2698d to 5bc59119 5 years ago

thiagocrepaldi dismissed these changes on 2020-12-03

ashbhandare dismissed their stale review 5 years ago

thiagocrepaldi commented on 2020-12-04

Fix test

eac63e6c

ashbhandare force pushed to eac63e6c 5 years ago

thiagocrepaldi approved these changes on 2020-12-04

ashbhandare merged 7cebf76a into master 5 years ago

ashbhandare deleted the aibhanda/zero_1_ckpt branch 5 years ago

Reviewers

thiagocrepaldi

baijumeswani

jessebenson

BowenBao

liqunfu

spandantiwari

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

onnxruntime Improve checkpointing for Zero stage 1 #5478 Merged

Improve checkpointing for Zero stage 1 #5478

onnxruntime
Improve checkpointing for Zero stage 1
#5478

Merged