Skip to content

Pull requests: microsoft/DeepSpeed

New pull request New
124 Open 2,877 Closed
124 Open 2,877 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

fix memcpy issue on backward for zero-infinity
#6670 opened Oct 26, 2024 by xylian86 Loading…
Add support for H100/sm_90 arch compilation
#6669 opened Oct 25, 2024 by loadams Loading…
1
Use one param coordinator for both train/inference scenarios
#6662 opened Oct 23, 2024 by tohtana Loading…
A faster and more memory-efficient implementation of zero_to_fp32
#6658 opened Oct 23, 2024 by xu-song Loading…
19
Support the parallel conversion from ZeRO checkpoints to FP32/FP16/BF16 param weight
#6655 opened Oct 23, 2024 by xylian86 Loading…
5 tasks done
2
add zero3 coalesced parameters fetch to zero optimization.
#6649 opened Oct 21, 2024 by inkcherry Loading…
2
Update gaudi2 docker version to latest release
#6648 opened Oct 21, 2024 by raza-sikander Loading…
3
AIO File Offsets
#6641 opened Oct 18, 2024 by jomayeri Loading…
1
[DO NOT MERGE] Log test results to file
#6627 opened Oct 15, 2024 by tohtana • Draft
6
modify_load_save_model
#6626 opened Oct 15, 2024 by ssklzx Loading…
3
[Bug Fix] Support threads_per_head < 64 for wavefront size of 64
#6622 opened Oct 11, 2024 by jagadish-amd Loading…
11
Improve consistency of zero_grad
#6554 opened Sep 18, 2024 by tohtana • Draft
Enabled configurable auto Tensor Parallelism (TP) for the inference of diverse models
#6553 opened Sep 18, 2024 by gyou2021 Loading…
4
Set shuffle=True by default in data_sampler
#6531 opened Sep 13, 2024 by ranzhejiang Loading…
1
Change compile for pipeline module torch.compile
#6478 opened Sep 2, 2024 by NirSonnenschein Loading…
2
Adding the new feature of FPDT
#6462 opened Aug 29, 2024 by YJHMITWEB Loading…
14
Unpin tests that previously used a pinned version of transformers
#6387 opened Aug 20, 2024 by loadams Loading…
1
Add weights_only=True in torch.load
#6094 opened Aug 17, 2024 by terry-for-github Loading…
8
[NaN check] Add NaN check to support bfloat16.
#5879 opened Aug 8, 2024 by ys950902 Loading…
2
Fix circular import in ds_transformer.py
#5804 opened Jul 28, 2024 by sznmelvin Loading…
3
Add DataStates-LLM: Asynchronous Checkpointing Engine Support
#5763 opened Jul 10, 2024 by mauryaavinash95 • Draft
2
Switch what versions of python are supported
#5676 opened Jun 17, 2024 by loadams • Draft
1
Hybrid Offloading for ZeRO3
#5625 opened Jun 7, 2024 by tohtana • Draft
Add support for Phi-3 small to FastGen
#5614 opened Jun 4, 2024 by adk9 • Draft
state_dict_factory: llama checkpoint - support SWIGLU
#5601 opened Jun 2, 2024 by nelyahu Loading…
@tohtana
3
ProTip! Adding no:label will show everything without a label.