Skip to content

Pull requests: NVIDIA/TransformerEngine

New pull request New
185 Open 2,532 Closed
185 Open 2,532 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

[PyTorch] Add head-parallel FA4 backward for cuDNN CP attention attention community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3510 opened Sep 11, 2026 by bzantium Loading…
8 of 13 tasks
4
[PyTorch] Fuse MoE chunk sorting and padding community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3509 opened Sep 11, 2026 by bzantium Loading…
7 of 13 tasks
3
[Draft] Port cuDNN frontend attention to Python API
#3508 opened Sep 11, 2026 by vcherepanov-nv Collaborator • Draft
13 tasks
@cyanguwa
12
[PyTorch] Avoid temporary state casts when loading FusedAdam checkpoints community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3507 opened Sep 11, 2026 by bzantium • Draft
5 of 13 tasks
Fix THD P2P pad detection and tail-zero gating community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3506 opened Sep 10, 2026 by RPalmr Loading…
1
3
[PyTorch] Support FP8 weight caching in compiled Linear
#3505 opened Sep 10, 2026 by pggPL Collaborator Loading…
6 of 8 tasks
1
[Pytorch][Attention] Fix FP8 THD backward workspace scaling
#3504 opened Sep 10, 2026 by sudhakarsingh27 Member Loading…
4 of 13 tasks
3
MOE Sequential Block with Dispatch and Combine as Basic Ops 2.20
#3503 opened Sep 10, 2026 by vthumbe1503 Collaborator Loading…
13 tasks
5
[PyTorch] Add timestep-conditioned AdaptiveLayerNorm to op fuser community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3501 opened Sep 10, 2026 by zupengwang Loading…
8 of 9 tasks
1
Fix hybrid extra state size mismatch
#3499 opened Sep 9, 2026 by negvet Collaborator • Draft
13 tasks
[JAX] Fix undefined sr_rng_state that breaks --dry-run in two encoder examples community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3498 opened Sep 9, 2026 by Anai-Guo Contributor Loading…
1
[PyTorch] torch.compile for an OperationFuser group holding one operation
#3496 opened Sep 8, 2026 by pggPL Collaborator Loading…
7 of 13 tasks
14
[Docs] Add Mixture of Experts guide documentation Improvements or additions to documentation
#3494 opened Sep 7, 2026 by pggPL Collaborator Loading…
7
[PyTorch] [torch.compile] Prepare LayerNormLinear and LayerNormMLP for torch.compile
#3493 opened Sep 7, 2026 by pggPL Collaborator Loading…
6 of 8 tasks
5
[PyT] Disable FA3 for training when head_dim_qk != head_dim_v community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3490 opened Sep 7, 2026 by yuweih205 Loading…
6 tasks done
1
3
Add native transport for dynamic context parallelism community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3489 opened Sep 6, 2026 by xiaoyao0115 • Draft
Support row-only MXFP8 distributed master-weight casts community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3488 opened Sep 6, 2026 by xiuhu17 Contributor Loading…
5 tasks done
2
[PyTorch] Split FusedAttnFunc into single-argument forward/backward helpers 2.20
#3480 opened Sep 4, 2026 by pggPL Collaborator Loading…
6 of 13 tasks
9
Drop the unsupported window_size argument from attention backend queries community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3479 opened Sep 4, 2026 by Anai-Guo Contributor Loading…
1
[PyT] Linear Attention API 2.20
#3477 opened Sep 4, 2026 by KshitijLakhani Collaborator • Draft
13 tasks
@KshitijLakhani
1
[PyTorch] Enable fused activation recompute for ScaledTanhSReLU community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3473 opened Sep 3, 2026 by wanyingw Contributor • Draft
13 tasks
5
[PyTorch] torch.compile support for FusedAttention
#3472 opened Sep 3, 2026 by pggPL Collaborator • Draft
7 of 13 tasks
[PyTorch] DeepSeekV3Layer: full MoE transformer layer (MLA + DeepSeek MoE)
#3471 opened Sep 3, 2026 by pggPL Collaborator Loading…
7 of 13 tasks
6
[All] Guard THD learnable dSink on older cuDNN attention bug Something isn't working
#3470 opened Sep 3, 2026 by KshitijLakhani Collaborator Loading…
2 of 13 tasks
1
@KshitijLakhani
9
[PyTorch] Fix FP8 illegal memory access in single-process multi-GPU execution community-contribution PRs from external contributor outside the core maintainers, representing community-driven work.
#3469 opened Sep 3, 2026 by SuperGoodGame Loading…
1
3
ProTip! Exclude everything labeled bug with -label:bug.