microsoft/DeepSpeed

Pull Requests Commits

Fix WarmupCosineLR multi-group initialization (#7969)

tohtana committed 211 days ago

Verified 3fd762cf

Fix DeepCompile+Z3 on PyTorch v2.9/2.10 (#7951)

tohtana committed 212 days ago

Verified ecb26a51

Fix overlap-comm buffer lifetimes (#7965)

tohtana committed 212 days ago

Verified dac1525b

Fix zero 1 and 2 CPU-offloaded gradient norm (#7967)

alek6kun committed 212 days ago

Verified 9d632f10

Fix FPQuantizer build (#7963)

Flamefire committed 214 days ago

Verified 3a5df512

fix(zero): detach flat buffer to prevent autograd inplace error on CP… (#7948)

delock committed 219 days ago

Verified 37e232fa

[SP] add SP deny list instead of allow (#7887)

kashif committed 221 days ago

Verified bf0126b5

Zero3 defragment utility (#7940)

nathon-lee committed 222 days ago

Verified 5dce124e

Fix process hang in process-group shutdown (#7941)

Flamefire committed 222 days ago

Verified 2f0924a5

Fix/fix autotp universal checkpoint ci (#7937)

tohtana committed 222 days ago

Verified 3bdebc03

Refactor consolidate transpose (#7934)

nathon-lee committed 223 days ago

Verified 89bf0d2d

Update version after latest release (v0.18.9) (#7936)

loadams committed 223 days ago

Verified 607b55f1

Update version.txt for latest incoming release 0.18.9 (#7935)

loadams committed 223 days ago

Verified 8c93851e

Remove Microsoft Corporation copyright from AGENTS.md and CLAUDE.md (#7932)

PKUWZP committed 223 days ago

Verified 9486ab75

Add fallback to full test (#7933)

tohtana committed 223 days ago

Verified 36f0b0c7

Merging AutoSP into DeepSpeed (#7860)

neeldani committed 224 days ago

Verified 5efb24ac

Update README with latest news from DeepSpeed (#7931)

PKUWZP committed 224 days ago

Verified 2bae3606

AGENTS.md: Add pre-commit command to existing CI requirements line (#7930)

delock committed 224 days ago

Verified abb88ceb

fix(superoffload) preserve multi-group updates with shared cpu buffers (#7905) (#7906)

xylian86 committed 225 days ago

Verified 729df6ca

Add news item for ASPLOS 2026 Best Paper Award (#7923)

PKUWZP committed 228 days ago

Verified 62c3e6d8

Extending Muon Optimizer Support for ZeRO Stage 3 (#7919)

PKUWZP committed 228 days ago

Verified 956ec6fb

Fix backward compatibility of torch.amp.custom_fwd for PyTorch < 2.4 (#7920)

tohtana committed 228 days ago

Verified 138f20d9

fix: handle non-existent path in is_nfs_path for Triton autotune cache (#7921)

Krishnachaitanyakc committed 228 days ago

Verified f887b982

Add HuggingFace tp_plan support for AutoTP (#7901)

delock committed 229 days ago

Verified a240c4da

Don't detect local GPU if `$DS_IGNORE_CUDA_DETECTION` is set (#7896)

Flamefire committed 229 days ago

Verified 26c954ff

fix: remove unnecessary shell=True in ROCm GPU architecture detection (#7915)

instantraaamen committed 229 days ago

Verified 5ce3abbf

Add Feature Universal Checkpoint for AutoTP (#7908)

nathon-lee committed 230 days ago

Verified f2bb1ec6

Respect `$TRITON_HOME` (#7907)

Flamefire committed 231 days ago

Verified 38bd11a7

Update version (#7903)

loadams committed 240 days ago

Verified 5f7b6870

fix(zero3): use current_stream() instead of default_stream() for grad… (#7898)

michaelroyzen committed 240 days ago

Verified be60451f

Newer Older