Skip to content

Pull requests: microsoft/DeepSpeed

New pull request New
129 Open 2,602 Closed
129 Open 2,602 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 Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add support for Microsoft Phi-3 model to DeepSpeed-FastGen
#5559 opened May 21, 2024 by adk9 Loading…
fix sequence parallel(Ulysses) grad scale for zero0
#5555 opened May 21, 2024 by inkcherry Loading…
unscale_and_clip_grads: use clamp on device instead of on host
#5547 opened May 20, 2024 by nelyahu Loading…
Add chatglm2 & chatglm3 autotp
#5540 opened May 16, 2024 by Yejing-Lai Loading…
1
create mininal universal checkpoint info for client state
#5526 opened May 13, 2024 by xylian86 Loading…
4
Fix deadlock in PipeEngine._exec_recv_grads
#5518 opened May 10, 2024 by i4never Loading…
3
DeepSpeedCheckpoint: support custom final ln idx
#5506 opened May 8, 2024 by nelyahu Loading…
2
inference: remove unused _validate_args function
#5505 opened May 8, 2024 by nelyahu Loading…
2
z3 scaled_global_grad_norm: repalce get_global_norm with torch.norm
#5504 opened May 7, 2024 by nelyahu Loading…
4
Update to ROCm6
#5491 opened May 1, 2024 by loadams Loading…
1
Rearrange inference OPS and stop using builder.load
#5490 opened May 1, 2024 by oelayan7 Loading…
2
Make the quantized data shape compatible with original tensor shape
#5483 opened Apr 30, 2024 by sfc-gh-reyazda Loading…
2
Enable dynamic shapes for pipeline parallel engine inputs
#5481 opened Apr 30, 2024 by tohtana • Draft
Fix training of pipeline based peft's lora model
#5477 opened Apr 29, 2024 by xuanhua Loading…
1
Universal checkpoint for zero stage 3
#5475 opened Apr 29, 2024 by xylian86 Loading…
12
Add Compressedbackend for Onebit optimizers
#5473 opened Apr 28, 2024 by Liangliang-Ma Loading…
6
[Draft][Demo] auto tp training
#5445 opened Apr 22, 2024 by inkcherry Loading…
4
enable yuan autotp & add conv tp
#5428 opened Apr 17, 2024 by Yejing-Lai Loading…
10
[XPU] support op builder from intel_extension_for_pytorch kernel path
#5425 opened Apr 17, 2024 by YizhouZ Loading…
7
uniform deepspeed overflow check
#5424 opened Apr 16, 2024 by GuanhuaWang Loading…
4
Adding DS Feature API in accelerator
#5423 opened Apr 16, 2024 by duli2012 Loading…
8
Add fp16 support of Qwen1.5MoE models (A2.7B) to DeepSpeed-FastGen
#5403 opened Apr 12, 2024 by ZonePG Loading…
1
Update names of CPU Adam/Adagrad/Lion params to better match torch/GPU ops.
#5382 opened Apr 8, 2024 by loadams • Draft
1 task
2
fix dist attn reshape error
#5366 opened Apr 5, 2024 by tkdcjf159 Loading…
1
Fix redundant seq data parallel grp argument in Z3/MiCS
#5352 opened Apr 2, 2024 by samadejacobs Loading…
2
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.