DeepSpeed
Pipeline parallel training engine.
#392
Merged

Pipeline parallel training engine. #392

ShadenSmith
improved pipe test + merge
a1c0aee7
fixed model ckpt naming
e77704eb
Improves checkpoint config and tensorboard.
28487b9a
less noise
d29f45e4
tb directory creation
e26bd33b
Improved activation checkpointing error checkpoing
7cad50f4
preparing for large perf tests
a1d4b6af
jeffra Pipeline parallelism (squash) (#82)
6704ebdd
completing renames
2eafdd39
code formatter
5c1449b6
moving over updates from staging v1 review
4e0f140b
Tied module indexing bugfix.
a59c2775
Train and inference pipeline schedules.
5c677218
adds check for no_grad()
0b3b42a1
doctring improvements
03a460ff
documentation improvements
59e132db
engine documentation
fc45bf64
Move code quality tests to Azure-hosted agents. (#368)
ed716a61
documentation
71afb3fd
verbose pipeline unit test
211d23b4
extends module to support torch.nn.Sequential
42a3c56e
docstring tweak
191fe4b9
catching up unit tests
917afebc
Pipeline staging v2 PR #1 - scheduler (#83)
13c64416
merge conflicts
d3546be7
outlining tutorial
1b7a0b85
Support torch.nn.Sequential (#84)
10c1af46
checking in tutorial
fa9d0871
edits for PP tutorial
01ff1a4c
checkpointing tutorial
689b1e6b
Merge branch 'shaden/pp-tutorial' into shaden/pipe-tutorial
55b55f33
Merge branch 'staging-pp-v2' into shaden/pp-tutorial
9ea0d398
init exports
81251406
Merge branch 'shaden/pp-tutorial' of github.com:microsoft/DeepSpeed-i…
51f91671
tutorial edits
51d3f4b0
section title
ff6b1b32
property to method
67d0463b
Merge branch 'shaden/pp-tutorial' of github.com:microsoft/DeepSpeed-i…
e1017bef
exporting module types from pipeline
2917ca0d
better pipe schedule figure
908b9b48
relative path fix
f840ddff
Preparing pipeline parallelism pull request.
0b9bf2a5
cleaning up merge
0bec1052
finishing merge
8a84062e
tutorial edits
25acb5cf
ShadenSmith ShadenSmith added enhancement
ShadenSmith ShadenSmith requested a review from arashashari arashashari 5 years ago
ShadenSmith ShadenSmith requested a review from awan-10 awan-10 5 years ago
ShadenSmith ShadenSmith requested a review from cli99 cli99 5 years ago
ShadenSmith ShadenSmith requested a review from conglongli conglongli 5 years ago
ShadenSmith ShadenSmith requested a review from eltonzheng eltonzheng 5 years ago
ShadenSmith ShadenSmith requested a review from jeffra jeffra 5 years ago
ShadenSmith ShadenSmith requested a review from minjiaz minjiaz 5 years ago
ShadenSmith ShadenSmith requested a review from niumanar niumanar 5 years ago
ShadenSmith ShadenSmith requested a review from RezaYazdaniAminabadi RezaYazdaniAminabadi 5 years ago
ShadenSmith ShadenSmith requested a review from samyam samyam 5 years ago
ShadenSmith ShadenSmith requested a review from tjruwase tjruwase 5 years ago
ShadenSmith ShadenSmith added documentation
ShadenSmith ShadenSmith added website
jeffra
jeffra approved these changes on 2020-09-10
ShadenSmith ShadenSmith merged 65c2f974 into master 5 years ago
tjruwase
tjruwase approved these changes on 2020-09-10

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone