Skip to content

Pull requests: NVIDIA/TransformerEngine

New pull request New
29 Open 607 Closed
29 Open 607 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

[C] Allow bias support for sm80/86/89 for cuDNN 9+
#863 opened May 23, 2024 by cyanguwa Loading…
2 tasks
Avoid framework specific import from top level
enhancement New feature or request
#862 opened May 22, 2024 by ksivaman • Draft
6 of 11 tasks
@ksivaman
1
[C/PyTorch/JAX] Build system improvements for rpath and C++11 ABI
build Build system enhancement New feature or request
#858 opened May 20, 2024 by denera Loading…
9 of 11 tasks
@denera
17
[Common/PyTorch] Grouped GEMM via multi-stream cuBLAS
#853 opened May 17, 2024 by yaox12 Loading…
8 of 11 tasks
Use correct FP8 group in multi-GPU docs
1.7.0 documentation Improvements or additions to documentation
#852 opened May 16, 2024 by timmoon10 Loading…
4 of 11 tasks
1
[JAX] Rewrite the Format of FP8 Meta and Remove unused ShardingTypes.
#842 opened May 13, 2024 by mingxu1067 Loading…
8 of 11 tasks
25
Different dimension for attention
#833 opened May 3, 2024 by pggPL Loading…
8 of 11 tasks
13
[C/PyTorch] Add THD support for cuDNN attention
#832 opened May 2, 2024 by cyanguwa Loading…
8 of 11 tasks
10
Added comments about Llama3 weights to Llama tutorial
1.7.0 documentation Improvements or additions to documentation
#830 opened May 1, 2024 by pggPL Loading…
7 of 11 tasks
1
Draft of generation for Gemma
#829 opened May 1, 2024 by pggPL • Draft
3 of 11 tasks
Find CXX component for MPI, fortran and C are not needed
#828 opened May 1, 2024 by aurianer Loading…
1 of 5 tasks
3
[Pytorch] Implement fp32 accumulation for attention with context parallel in both forward and backward pass.
#821 opened Apr 28, 2024 by Yuxin-CV Loading…
2
[PyTorch] Refactor FP8 workspaces in linear modules
bug Something isn't working enhancement New feature or request
#820 opened Apr 27, 2024 by timmoon10 Loading…
13
[UB] Adding support for multinode nvlink [WIP]
#815 opened Apr 26, 2024 by shamisp Loading…
Bug fix in DGRAD->RS overlap
#802 opened Apr 23, 2024 by vasunvidia • Draft
[JAX] POC of custom call selection
#797 opened Apr 22, 2024 by zlsh80826 • Draft
3
[PyTorch] Fix minor bug in computing num_gqa_groups_per_partition
bug Something isn't working
#777 opened Apr 13, 2024 by knowlsie Loading…
4
[C/PyTorch] Refactor and move userbuffers into TE/common
#760 opened Apr 8, 2024 by denera Loading…
6 of 13 tasks
1
@denera
43
Fix bhss bias format before sm90
#736 opened Mar 27, 2024 by zlsh80826 Loading…
3
[PyTorch] Sequential fuser
enhancement New feature or request
#707 opened Mar 9, 2024 by timmoon10 Loading…
2 of 6 tasks
1
[PyTorch] Distributed intermediate/activation tensors for FSDP
1.7.0
#687 opened Feb 28, 2024 by denera Loading…
@denera
10
[PyTorch] Re-enable bias+GELU fusion for non-reentrant checkpointing -- WIP
#686 opened Feb 27, 2024 by denera • Draft
@denera
1
Remove now useless padding as it is now down automatically.
jax
#680 opened Feb 25, 2024 by nouiz Loading…
12
[Draft] show how to not use a global mesh.
jax
#549 opened Dec 1, 2023 by nouiz Loading…
1
Add the examples and the tests in the installed packages.
#514 opened Nov 10, 2023 by nouiz Loading…
12
ProTip! Filter pull requests by the default branch with base:main.