microsoft/DeepSpeed

Pull Requests Commits

support higher sequence length in softmax backward (>2048 & <8192)

Reza Yazdani committed 4 years ago

a0bd66c6

Avoid partitioning small activations (#1154)

tjruwase committed 4 years ago

Verified b1669c0d

add assert to make sure zero isn't enabled with 1bit-* (#1169)

jeffra committed 4 years ago

Verified 1ba3e8e3

patch in ompi local rank if local_rank isn't set (#1164)

jeffra committed 4 years ago

Verified 4c114a27

Remove hard torch dependency at install (#1166)

jeffra committed 4 years ago

Verified da1fe2f8

[Doc] Fix steps_per_print description (#1163)

tjruwase committed 4 years ago

Verified fa7921e2

Samyamr/largest partitioned params calculation fix (#1150)

samyam committed 4 years ago

Verified 4eaf9106

Fix bugs in the tutorial documentation (#1157)

hyunwoongko committed 4 years ago

Verified aa16828c

Add Windows support in README, use c++17 on Windows to support latest VC & cuda build tool (#1151)

eltonzheng committed 4 years ago

Verified 71ecf7e6

unit test for bugfix #1135 (#1148)

Shaden Smith committed 4 years ago

Verified 8e48756c

jeffra committed 4 years ago

5f16b172

Update index.md

jeffra committed 4 years ago

Verified 2d302d6a

Update README.md

jeffra committed 4 years ago

Verified 25f5f87a

Add local attention for GPT-Neo model architecture (#1114)

RezaYazdaniAminabadi committed 4 years ago

Verified aca7fc54

remove old unit test, should have been removed in rebase (#1097)

jeffra committed 4 years ago

Verified 8def3cb3

adjust azureml-examples links (#1015)

lostmygithubaccount committed 4 years ago

Verified 6924d840

correct cpu_offload deprecation (#1140)

stas00 committed 4 years ago

Verified a8d6dfe8

missing f-string (#1134)

stas00 committed 4 years ago

Verified 1b187191

[zero] fix missed subclasses partitioning bug (#1135)

stas00 committed 4 years ago

Verified 5ca81678

[builder] allow minor cuda version differences (#1139)

stas00 committed 4 years ago

Verified c5d5ce68

fix config name (#1103)

stas00 committed 4 years ago

Verified c697d7ae

fix empty entry in gcc cmd (#1129)

stas00 committed 4 years ago

Verified ccc522c2

remove triton version to use the latest stable pypi version (#1128)

RezaYazdaniAminabadi committed 4 years ago

Verified cf82168c

Change the sparse attention API to be compatible with latest changes of triton (#902)

RezaYazdaniAminabadi committed 4 years ago

Verified 26e3841c

fix links for inference tutorial (#1113)

RezaYazdaniAminabadi committed 4 years ago

Verified 11e94e6c

release inference quantized kernels (#1104)

RezaYazdaniAminabadi committed 4 years ago

Verified d2cf66a6

Create symlinks on Windows setup (#1099)

eltonzheng committed 4 years ago

Verified 10104284

delay imports for replace policies and fix missing req (#1100)

jeffra committed 4 years ago

Verified 96eb5b12

jeffra committed 4 years ago

0449cbd3

Add inference API to RTD (#1096)

jeffra committed 4 years ago

Verified 0d701ec4

Older