Pull Requests microsoft/DeepSpeed

Warn when zero.Init silently falls back to a single rank (#8084)

#8089 opened 2026-06-24 18:59 by akshansh47

fix: use local ev_values and wrap dict.values() in list()

#8087 opened 2026-06-23 04:35 by hashwnath

fix: add buffer-length check in shm.cpp

#8082 opened 2026-06-20 12:23 by orbisai0security

fix: sanitize subprocess call in ds_aio_job.py

#8081 opened 2026-06-20 06:40 by orbisai0security

ZeRO 1/2: wait on all IPG-bucket producer streams in average_tensor (#8061)

#8080 opened 2026-06-19 22:43 by arunshar

feat: add Trackio as a new experiment monitoring backend

#8065 opened 2026-06-15 14:41 by chanduripranav

Add AutoEP + AutoTP parallel folding

#8064 opened 2026-06-13 17:43 by tohtana

[DeepCompile] fix gather params in dynamo skipped frames for ZeRO3

#8059 opened 2026-06-11 14:32 by XAheli

feat(zenflow): run the overlapped CPU optimizer in a native process

#8058 opened 2026-06-10 21:34 by Antlera

Fix eigenvalue parsing for compression-only quantize configs

#8057 opened 2026-06-10 07:44 by sowndappan5

Add optional torchembed RoPE backend to apply_rotary_pos_emb

#8052 opened 2026-06-07 21:35 by py-ai-dev

Fix incorrect variable name

#8051 opened 2026-06-07 18:21 by Muneerali199

fix: log eigenvalue monitor values

#8049 opened 2026-06-05 08:29 by he-yufeng

fix: log block eigenvalue summary events

#8048 opened 2026-06-04 23:21 by he-yufeng

Fix minor comment/docstring typos in runtime and inference modules

#8046 opened 2026-06-03 08:21 by nathon-lee

zero3: defer param release during retain_graph backward #7352

#8045 opened 2026-06-03 06:55 by nathon-lee

[Draft] Add On-Policy Distillation (OPSD) Trainer in DeepSpeed

#8027 opened 2026-05-26 07:31 by PKUWZP

Add Qwen 3.5 preset to AutoTP

#7978 opened 2026-04-16 12:51 by tohtana

Refactor/torch autocast encapsulate global state

#7946 opened 2026-04-02 06:06 by nathon-lee

Fix ZeRO-3 optimizer initialization validation (#7844)

#7929 opened 2026-03-28 16:20 by amadhan882

Add torch_xla TPU support for ZeRO-1/2

#7917 opened 2026-03-21 18:43 by PKUWZP

doc: Remove suggestion to build extensions in parallel

#7899 opened 2026-03-12 15:58 by Flamefire

Fix Stage 0 + Ulysses crash: make bwc_tensor_model_parallel_rank() resilient to MP API absence

#7888 opened 2026-03-06 06:59 by nathon-lee

fix(zero): Ensure full gradient reduction for Muon optimizer with reduce_scatter

#7878 opened 2026-02-27 06:46 by nathon-lee

fix: correct DistributedAttention output shape and pad uneven sequence lengths (#7842)

#7868 opened 2026-02-22 11:00 by harshang03

fix: keep fp32-pinned parameters out of the bf16 cast path in ZeRO-3 (#7747)

#7867 opened 2026-02-22 10:52 by harshang03

Revert "fix: remove premature MPI environment variable check in OpenMPIRunner"

#7864 opened 2026-02-21 01:39 by mikloorbi-sys

Fix global .cuh ignore and enforce tracked CUDA headers

#7858 opened 2026-02-18 04:38 by harshang03

Fix ZeRO legacy grad-hook crash when next_functions is missing

#7857 opened 2026-02-17 22:07 by harshang03

Reject non-finite fp16 loss_scale across config and ZeRO paths

#7856 opened 2026-02-17 18:13 by harshang03