The Wayback Machine - https://web.archive.org/web/20220824175406/https://github.com/microsoft/DeepSpeed/pulls
Skip to content

Pull requests: microsoft/DeepSpeed

New pull request New
61 Open 1,129 Closed
61 Open 1,129 Closed
Author
Filter by author
Label
Filter by label
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Milestones
Filter by milestone
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

testing pydantic v1.10.0a2
#2256 opened Aug 24, 2022 by samuelcolvin Loading…
3
Refactor universal checkpointing and tensor fragments
#2253 opened Aug 23, 2022 by tjruwase Loading…
1
DeepSpeed CPU Support
#2245 opened Aug 20, 2022 by Quentin-Anthony • Draft
4 tasks
1
fix #2240: wrong time unit in flops_profiler
#2241 opened Aug 19, 2022 by yzs981130 Loading…
1
5
Moe pipelining
#2238 opened Aug 18, 2022 by siddharth9820 Loading…
memory reallocation for bigger batch size
#2236 opened Aug 18, 2022 by reymondzzzz Loading…
2
MoE - Token dropping for Full Tensor Paralellism
#2235 opened Aug 18, 2022 by siddharth9820 Loading…
4
Add tensor parallel inference unit tests
#2232 opened Aug 17, 2022 by mrwyattii Loading…
[RFC] add device abstraction to allow other device than CUDA be used
#2221 opened Aug 16, 2022 by delock • Draft
Ds-inference Int8 support through ZeroQuant technology
#2217 opened Aug 13, 2022 by RezaYazdaniAminabadi Loading…
5
Refactor remaining distributed tests
#2216 opened Aug 12, 2022 by mrwyattii • Draft
Type checking in zero-3 to avoid silent bugs
#2214 opened Aug 12, 2022 by awan-10 Loading…
Extend scratch buffer for long prompts
#2212 opened Aug 12, 2022 by cmikeh2 Loading…
3
Pydantify sub-configs
#2210 opened Aug 11, 2022 by mrwyattii • Draft
Trajepl: nebula load fix
#2182 opened Aug 4, 2022 by trajepl Loading…
3
Simplifying inference API
#2181 opened Aug 4, 2022 by RezaYazdaniAminabadi Loading…
[deepspeed/autotuner] Bug fix for skipping mbs on gas
#2171 opened Aug 2, 2022 by rahilbathwal5 Loading…
[deepspeed/autotuner] Fix for timer for gas > 1
#2165 opened Aug 2, 2022 by rahilbathwal5 Loading…
2
[deepspeed/autotuner] Bug fix for binary search for batch size
#2162 opened Aug 1, 2022 by rahilbathwal5 Loading…
5
some fix in flops_profiler
#2068 opened Jul 5, 2022 by lucasleesw Loading…
5
Small fix in injection module to replace transformer
#2064 opened Jun 28, 2022 by RezaYazdaniAminabadi Loading…
Add MuP optimizers
#2043 opened Jun 22, 2022 by mrwyattii Loading…
Share a list of weight attributes instead of a single one in TiedLayerSpec API
#2035 opened Jun 21, 2022 by thomasw21 Loading…
4
Fixing several issue in API and kernels to run inference with model-parallelism
#2028 opened Jun 17, 2022 by RezaYazdaniAminabadi Loading…
3
feat: Support for training with MoE module in PipelineEngine
#1942 opened May 7, 2022 by shjwudp Loading…
10
ProTip! Add no:assignee to see everything that’s not assigned.