This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
Integrate Horovod training API as part of MXNet native distributed training API #17531
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@ChaokunChang FYI |
apeforest
requested review from
aaronmarkham,
anirudh2290,
gigasquid,
iblislin,
leezu,
marcoabreu,
nswamy,
sergeykolychev and
yzhliu
as code owners
apeforest
force-pushed
the
dev/unified_api_hvd
branch
2 times, most recently
from
fa6ea1a
to
a0f5848
Compare
apeforest
force-pushed
the
dev/unified_api_hvd
branch
2 times, most recently
from
8ac5a82
to
fc22ff1
Compare
apeforest
changed the title
apeforest
force-pushed
the
dev/unified_api_hvd
branch
2 times, most recently
from
040370c
to
08d28a3
Compare
apeforest
force-pushed
the
dev/unified_api_hvd
branch
from
08d28a3
to
9fd7493
Compare
…cubator-mxnet into dev/unified_api_hvd
eric-haibin-lin
approved these changes
Apr 14, 2020
AntiZpvoh
pushed a commit
to AntiZpvoh/incubator-mxnet
that referenced
this pull request
Jul 6, 2020
…aining API (apache#17531) * implement pushpull for horovod * add local_rank function * add tests * Remove in-place broadcast API * Add kvstore horovod example * Fix the list to singlton conversion * Add horood test to CI * Remove test horovod from unit test * Add docstring * Add horovod in test * sync with master * Fix horovod dependency in CI * Fix merge conflict with byteps * Update __init__.py * Resolve conflict * Remove openib warning message * Add log message in test * Remove tmp file * Fix lint Co-authored-by: Haibin Lin <linhaibin.eric@gmail.com>
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Integrate Horovod training API as part of MXNet native distributed training API by making Horovod one special type of KVStore as proposed in #16795
Part of MXNet 2.0 project tracked by #17111
Checklist
Essentials
Still need to debug the integration tests core dump.