Skip to content

Pull requests: deepspeedai/DeepSpeed

New pull request New
221 Open 3,903 Closed
221 Open 3,903 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

feat(rollout): add continuous batching profiling
#8494 opened Sep 12, 2026 by nathon-lee Contributor Loading…
4
Deprecate sparse attention
#8493 opened Sep 12, 2026 by hwchen2017 Contributor Loading…
2
Remove triton compatibility check for fp_quantizer
#8492 opened Sep 12, 2026 by hwchen2017 Contributor Loading…
4
[bugfix] _DimZeroAllToAll silently fails to send grads under torch.compile
#8491 opened Sep 11, 2026 by stas00 Collaborator Loading…
13
Refresh the whole captured input before a CUDA graph replay
#8488 opened Sep 11, 2026 by vineethsaivs Contributor Loading…
2
Make wait() idempotent on AllGatherHandle and NoGatherHandle
#8487 opened Sep 11, 2026 by pengdurice Contributor • Draft
1
tests: allow testing an installed DeepSpeed package
#8485 opened Sep 11, 2026 by arubittu Loading…
1
docs: fix precompiled ops install command
#8484 opened Sep 11, 2026 by arubittu Loading…
1
4
Give Muon's momentum the dtype of the gradient it is combined with
#8483 opened Sep 11, 2026 by alanhuangyoo Contributor Loading…
Fixed memory leak in loss backward with AC & zero3 & single-model-multi-branch network
#8482 opened Sep 11, 2026 by supermeng Loading…
1
3
Recognize a norm by its shape, not by whether its class name was listed
#8479 opened Sep 10, 2026 by alanhuangyoo Contributor Loading…
1
Count each expert once in the fp32 gradient-clipping norm
#8478 opened Sep 10, 2026 by alanhuangyoo Contributor Loading…
1
2
Fix AutoEP global L2 clipping for sharded expert ownership
#8476 opened Sep 10, 2026 by gss10282023 Loading…
1
3
Reject a non-positive out-token limit at config time
#8473 opened Sep 10, 2026 by alanhuangyoo Contributor Loading…
1
Make add_tuning_arguments' 1Cycle flags reach OneCycle
#8472 opened Sep 10, 2026 by alanhuangyoo Contributor Loading…
fix: track FusedAdam bias correction per parameter
#8471 opened Sep 10, 2026 by MrCapricornLiu Loading…
3
Support dynamic parameter offload for ZeRO-3 inference
#8468 opened Sep 10, 2026 by zupengwang Contributor Loading…
4
A zero-element parameter loses its shape when it is bound to a flat buffer
#8467 opened Sep 9, 2026 by alanhuangyoo Contributor Loading…
1
fix(inference): break circular import in ops.transformer.inference
#8465 opened Sep 9, 2026 by chakshu-dhannawat Contributor Loading…
1
4
Fix Muon optimizer under ZeRO CPU offload and bound gather buffers
#8464 opened Sep 8, 2026 by jinyouzhi Contributor • Draft
3
Checkpoint the RNG the curriculum sampler actually draws from
#8460 opened Sep 8, 2026 by alanhuangyoo Contributor Loading…
1
[AutoSP] Fix compiler correctness
#8457 opened Sep 8, 2026 by jinyouzhi Contributor Loading…
8
Say when ZenFlow will never re-select its important columns
#8456 opened Sep 8, 2026 by alanhuangyoo Contributor Loading…
5
Rebuild the full argument list when partitioning activations
#8455 opened Sep 8, 2026 by vineethsaivs Contributor Loading…
3
fix(inference): reject non-positive max_out_tokens at config validation
#8454 opened Sep 8, 2026 by chakshu-dhannawat Contributor Loading…
1
ProTip! Follow long discussions with comments:>50.