Skip to content

Pull requests: deepspeedai/DeepSpeed

New pull request New
147 Open 3,775 Closed
147 Open 3,775 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

Add fused triton kernel for swiglu
#8244 opened Aug 11, 2026 by hwchen2017 Contributor Loading…
3
Update version.txt after 0.19.5 release
#8243 opened Aug 10, 2026 by loadams Collaborator Loading…
@loadams
2
Drop the documented grad_hooks ZeRO option, which does not exist
#8242 opened Aug 10, 2026 by vineethsaivs Contributor Loading…
[AutoTP] Replace tp_shard process-wide globals with per-model AutoTPMeta
#8241 opened Aug 10, 2026 by delock Collaborator • Draft
1
1
[DeepCompile] Fix staticmethod handling on Python 3.9
#8240 opened Aug 9, 2026 by baremetaldevx86 Loading…
1
Filter --include against the real slots, not against itself
#8239 opened Aug 9, 2026 by vineethsaivs Contributor Loading…
4
Keep the elastic batch size within max_train_batch_size
#8237 opened Aug 9, 2026 by vineethsaivs Contributor Loading…
2
Add adaptive gather-buffer pooling for DeepCompile ZeRO-3
#8234 opened Aug 7, 2026 by tohtana Collaborator Loading…
3
Split DeepCompile ZeRO-3 memory scheduler
#8233 opened Aug 7, 2026 by tohtana Collaborator Loading…
3
Add configurable sum gradient reduction
#8232 opened Aug 7, 2026 by zupengwang Contributor Loading…
7
Fix docstring Args entries that name a parameter the function does not take
#8223 opened Aug 6, 2026 by darkdi Loading…
2
Update pytest version per latest dependabot
#8222 opened Aug 6, 2026 by loadams Collaborator • Draft
Update versions of steps in workflows
#8221 opened Aug 6, 2026 by loadams Collaborator • Draft
Keep parameter dtype through ZeRO-3 weight quantization
#8215 opened Aug 5, 2026 by adityasingh2400 Contributor Loading…
1
3
[DeepCompile] Fix KeyError on frozen parameters in ZeRO-3
#8214 opened Aug 5, 2026 by roycho96 Contributor Loading…
2
Add an opt-in DeepEP transport for the AutoEP expert all-to-all
#8213 opened Aug 4, 2026 by yh0903 Contributor Loading…
38
Add native (DeepNVMe) host-memory pinning backend for accelerators
#8211 opened Aug 4, 2026 by sfc-gh-truwase Collaborator Loading…
4 tasks done
24
(1/2) Implementing Compiler Pass for AutoTP
#8204 opened Aug 2, 2026 by therealnaveenkamal Contributor Loading…
12
Implement the documented per-param-group lists in OneCycle
#8201 opened Aug 1, 2026 by vineethsaivs Contributor Loading…
8
[AutoTP] Complete uneven sharding and universal checkpoint support
#8185 opened Jul 28, 2026 by jinyouzhi Contributor Loading…
41
Non-reentrant activation checkpoint CPU offload
#8181 opened Jul 27, 2026 by winglian Loading…
11
fix: avoid ZeroDivisionError in compute_elastic_config return_microbatch path
#8162 opened Jul 21, 2026 by Solaris-star Loading…
1
[Pipeline] Add Ray-backed stage placement and execution with pluggable transportsFeature/ray pipeline parallelism
#8161 opened Jul 21, 2026 by SisPiao Loading…
7
Add governance process page to deepspeed.ai
#8138 opened Jul 13, 2026 by PKUWZP Collaborator Loading…
3 tasks
5
Feature/zero multi loss separate backward
#8135 opened Jul 12, 2026 by nathon-lee Contributor Loading…
14
ProTip! Exclude everything labeled bug with -label:bug.