microsoft / DeepSpeed

Watch 64
Star 2.8k
Fork 254

Code
Issues 33
Pull requests 12
Actions
Projects
Security
Insights

Code
Issues
Pull requests
Actions
Projects
Security
Insights

Labels 12 Milestones 0

New pull request New

12 Open 195 Closed

Author

Filter by author

Label

Filter by label

Use alt + click/return to exclude labels.

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

Switches from tensorboardX to torch.utils.tensorboard

dependencies

#330 opened Aug 21, 2020 by ShadenSmith • Approved

Build hip extensions for ROCm

#323 opened Aug 20, 2020 by jithunnair-amd • Review required

Apex rocm submodule

dependencies enhancement

#321 opened Aug 17, 2020 by jithunnair-amd • Review required

support dynamic sequence length

#320 opened Aug 15, 2020 by RezaYazdaniAminabadi • Approved

Fix zero-1 bug related to stitching reduced grads across comm partitions

bug

#318 opened Aug 14, 2020 by jeffra • Review required

ZeRO-2 enable CPU offload

#313 opened Aug 11, 2020 by jren73 • Changes requested

Update deepspeed_zero_optimizer.py

#293 opened Jul 15, 2020 by MikalaiDrabovich • Approved

Module replacement support

#277 opened Jun 25, 2020 by jeffra • Draft

Support distributed init outside runtime init

#276 opened Jun 25, 2020 by jeffra • Approved

remove transformer layer ID from the top module

#263 opened Jun 12, 2020 by RezaYazdaniAminabadi • Approved

fixing the dataloader original len issue

#178 opened Mar 31, 2020 by tomekrut • Review required

Catching up CPU- and Python-only modes.

#159 opened Mar 20, 2020 by ShadenSmith • Draft

ProTip! Type g i on any issue or pull request to go back to the issue listing page.