Skip to content

Pull requests: NVIDIA/TransformerEngine

New pull request New
90 Open 1,610 Closed
90 Open 1,610 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
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Don't pickle an empty dict in LayerNorm and pt base modules
#2253 opened Oct 9, 2025 by pstjohn Loading…
[PyTorch Debug] Fix issue with start_end_list logging feature
#2252 opened Oct 9, 2025 by paul-gibbons Loading…
7 tasks
[common] Split cast/gated kernels by scaling mode
#2248 opened Oct 8, 2025 by Oleg-Goncharov Loading…
4 of 13 tasks
1
Update minimum python version to 3.10 and add checks in CI
#2247 opened Oct 8, 2025 by ksivaman Loading…
7 of 13 tasks
6
[PyTorch] FSDP2 Weight Update Fix
#2245 opened Oct 7, 2025 by vthumbe1503 • Draft
13 tasks
[JAX] Improve CollectiveGEMM examples
#2244 opened Oct 7, 2025 by phu0ngng • Draft
13 tasks
[JAX] Add BRCM support for THD
#2242 opened Oct 7, 2025 by KshitijLakhani Loading…
7 of 13 tasks
@KshitijLakhani
1
Disable torch autocast context in rope forward pass
#2240 opened Oct 7, 2025 by pstjohn Loading…
5 of 13 tasks
[PyTorch] Bump minimum cuDNN version for fused attention with FP8 current scaling bug Something isn't working
#2236 opened Oct 4, 2025 by timmoon10 Loading…
7 of 13 tasks
1
Make CanonicalizeGemmInput() support non-TN layout FP8 GEMM on Blackwell with column-wise/transposed data
#2233 opened Oct 3, 2025 by denera • Draft
6 of 13 tasks
@denera
[JAX] xla_home logging during JAX build
#2232 opened Oct 3, 2025 by jberchtold-nvidia Loading…
13 tasks
[JAX] Add support for sink attention in JAX
#2225 opened Oct 1, 2025 by pggPL • Draft
8 of 13 tasks
1
1
Enable SWA with CP for THD input format
#2220 opened Sep 30, 2025 by sudhakarsingh27 Loading…
1 of 6 tasks
1
[PyTorch Debug] Custom feature tutorial.
#2216 opened Sep 30, 2025 by pggPL Loading…
8 of 13 tasks
[Draft][PyTorch][MOE] Support NVFP4 Grouped Linear
#2215 opened Sep 30, 2025 by zhongbozhu Loading…
3 of 17 tasks
@zhongbozhu
3
UBNEXT with optional add-rms fuse
#2212 opened Sep 29, 2025 by nv-akorzh Loading…
4
Test to see if SWA and Causal compute can be removed from seqlens and…
#2201 opened Sep 25, 2025 by KshitijLakhani • Draft
13 tasks
Honor COMPACT data_format for FP8 blockwise scales in MoE up-projection path to remove 5× redundant rowwise_scale_inv.T.contiguous() passes
#2199 opened Sep 24, 2025 by xiaoxi-wangfj Loading…
2 of 13 tasks
[PyTorch] Add max_score support for MuonClip 2.9.0
#2195 opened Sep 22, 2025 by cyanguwa Loading…
8 of 13 tasks
1
FSDP grad fusion support
#2191 opened Sep 21, 2025 by sanandaraj5597 Loading…
2
[Feature] Enable rope application with offsets for training 2.9.0
#2188 opened Sep 19, 2025 by sudhakarsingh27 Loading…
1 of 13 tasks
3
[Feat] Draft: support offloading activation
#2187 opened Sep 18, 2025 by lhb8125 Loading…
13 tasks
Context Parallel integration tests with a transformer layer: BSHD and THD + CP 2.9.0
#2176 opened Sep 16, 2025 by jomitchellnv Loading…
7 of 13 tasks
[PyTorch Debug] Add nvdlfw-inspect to dependencies
#2173 opened Sep 15, 2025 by pggPL • Draft
7 tasks done
1
blockwise fp8 weight memory optimization: on-demand columnwise fp8 weight creation
#2168 opened Sep 10, 2025 by skydoorkai Loading…
7 of 13 tasks
4
ProTip! Exclude everything labeled bug with -label:bug.