microsoft/DeepSpeed

Pull Requests Commits

Relax tolerances for FP8 unit test only for ROCm + FP16 (#7373)

rraminen committed 193 days ago

Verified d33baf00

Flops profiler support for F.interpolate (#7353)

sfc-gh-truwase committed 193 days ago

Verified 25da6fc1

add ALST paper reference (#7372)

stas00 committed 194 days ago

Verified 86097872

Enable torch.autocast with ZeRO (#6993)

tohtana committed 194 days ago

Verified ed5f7375

sequence parallel default dtype (#7364)

stas00 committed 194 days ago

Verified d3b9cb8c

Fix(scheduler): WarmupLR inherits optimizer lr when not specified (#7360)

Flink-ddd committed 194 days ago

Verified 22cf1a44

Restore real inputs for recompilation (#7356)

tohtana committed 194 days ago

Verified 6f1a1c04

Fix tutorial title (#7365)

stas00 committed 196 days ago

Verified f394e780

Fix 404s (#7363)

tjruwase committed 197 days ago

Verified 9ac94414

Improve padding util for compile (#7355)

tohtana committed 199 days ago

Verified 600d280f

Fix error of <glog/logging.h> (#7351)

Freed-Wu committed 201 days ago

Verified 76631215

Don't break set_start_method (#7349)

tjruwase committed 202 days ago

Verified 10b10661

s/UlyssesPlus/Arctic Long Sequence Training (ALST)/ (#7348)

stas00 committed 203 days ago

Verified d7e60fd0

Update version after 0.17.1 release (#7345)

loadams committed 204 days ago

Verified 4686d5ef

Move pytest pinning from individual tests to requirements-dev.txt until fixed. (#7327)

loadams committed 204 days ago

Verified 2ce55057

Fix docs that are rendering Incorrectly (#7344)

felixgondwe committed 204 days ago

Verified 4d0c1596

Improve overflow handling in ZeRO (#6976)

tjruwase committed 204 days ago

Verified e440506b

Update folder name (#7343)

sfc-gh-truwase committed 204 days ago

Verified bb293aea

Fix LoRA arxiv reference (#7340)

emmanuel-ferdman committed 206 days ago

Verified 05818e90

fixed: Modified the topkgating function and modified the test_moe file for testing (#7163)

xiongjyu committed 207 days ago

Verified 770967f5

DeepNVMe update (#7215)

tjruwase committed 207 days ago

Verified 24a1d8f9

fp16 optimizer timers fix - TypeError: 'NoneType' object is not callable (#7330)

rraminen committed 207 days ago

Verified cb3ad0c1

Fix issue with symint input (#7243)

tohtana committed 208 days ago

Verified 7d0c3f78

Fix pytest version to 8.3.5 in hpu-gaudi actions (#7337)

raza-sikander committed 208 days ago

Verified 2ad2011c

Update config_utils.py (#7333)

qgallouedec committed 208 days ago

Verified d0f7091a

Improve Ulysses Plus Docs (#7335)

cynricfu committed 208 days ago

Verified b8d4b842

UlyssesPlus Docs take 2 (#7332)

stas00 committed 210 days ago

Verified 097f0637

Ulysses Plus Docs (#7331)

stas00 committed 210 days ago

Verified 81a47408

Update version.txt after v0.17.0 release (#7326)

loadams committed 211 days ago

Verified 8f3c3e78

Bump to v0.17.0 (#7324)

sfc-gh-mwyatt committed 211 days ago

Verified 720787e7

Newer Older