-
Notifications
You must be signed in to change notification settings - Fork 3.9k
Pull requests: microsoft/DeepSpeed
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
WA for Torch-compile-Z3-act-apt accuracy issue from the Pytorch repo
#5590
opened May 30, 2024 by
NirSonnenschein
Loading…
FastGen H100 MoE support: Add PyTorch multi-gemm MOE implementation
#5586
opened May 29, 2024 by
HeyangQin
Loading…
Remove compile wrapper to simplify access to model attributes
#5581
opened May 29, 2024 by
tohtana
Loading…
_exec_forward_pass: place zeros(1) on the same device as the param
#5576
opened May 28, 2024 by
nelyahu
Loading…
[CPU] SHM based allreduce improvement for small message size
#5571
opened May 27, 2024 by
delock
Loading…
assumption of initial_seed function accepting seed arg in DeepSpeedAccelerator abstract class is incorrect
#5569
opened May 26, 2024 by
polisettyvarma
Loading…
Add support for Microsoft Phi-3 model to DeepSpeed-FastGen
#5559
opened May 21, 2024 by
adk9
Loading…
create mininal universal checkpoint info for client state
#5526
opened May 13, 2024 by
xylian86
Loading…
Z3: optimizations for grad norm calculation and gradient clipping
#5504
opened May 7, 2024 by
nelyahu
Loading…
Make the quantized data shape compatible with original tensor shape
#5483
opened Apr 30, 2024 by
sfc-gh-reyazda
Loading…
Previous Next
ProTip!
Type g p on any issue or pull request to go back to the pull request listing page.