Skip to content

Pull requests: NVIDIA/TransformerEngine

New pull request New
123 Open 1,869 Closed
123 Open 1,869 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] Make grouped weights opt-in
#2678 opened Feb 13, 2026 by ksivaman Loading…
7 of 13 tasks
4
[PyTorch][Fused Attn] Add support for cuDNN to return Softmax Stats always and Max when return_max_logit=True
#2677 opened Feb 12, 2026 by sudhakarsingh27 Loading…
7 of 13 tasks
4
[PyTorch] Add dtype information to QuantizedTensorStorage class
#2676 opened Feb 12, 2026 by ptrendx Loading…
1 of 13 tasks
4
[Common] MOE Split dBias cpu_overhead enhancement New feature or request MoE
#2674 opened Feb 11, 2026 by Oleg-Goncharov Loading…
8 of 13 tasks
9
Cherry pick "Adds dst.dtype information in copy_ method of quantized tensors. #2120"
#2673 opened Feb 11, 2026 by ptrendx Loading…
13 tasks
4
[JAX] TE Permutation integration to Maxtext
#2672 opened Feb 11, 2026 by tdophung Loading…
6 of 13 tasks
22
Fix and harden attention benchmark script
#2671 opened Feb 10, 2026 by KshitijLakhani • Draft
13 tasks
@KshitijLakhani
Fix nvfp4 convert_and_update_tensor shape check
#2670 opened Feb 10, 2026 by skydoorkai Loading…
13 tasks
1
9
Pytorch binding for cublas gemm + Grouped Linear integration
#2669 opened Feb 10, 2026 by vthumbe1503 • Draft
13 tasks
[pyTorch] Replace the make_empty implementation to use C++ implementation
#2666 opened Feb 10, 2026 by ptrendx • Draft
1 of 13 tasks
3
Improve handling of the NVTE_CUDA_ARCHS
#2665 opened Feb 9, 2026 by ptrendx Loading…
1 of 13 tasks
12
Add multi-precision training support to FSDP script
#2662 opened Feb 9, 2026 by aagallo Loading…
6 of 13 tasks
26
[JAX] Debugging inspect utility
#2651 opened Feb 4, 2026 by jberchtold-nvidia Loading…
8 of 13 tasks
19
[JAX] Fix FSDP when FSDP+EP is active
#2649 opened Feb 3, 2026 by jberchtold-nvidia Loading…
8 of 13 tasks
6
start devcontainer for transformerEngine
#2646 opened Feb 3, 2026 by pstjohn • Draft
[PyTorch Debug] Support tensor dump
#2645 opened Feb 3, 2026 by pggPL • Draft
13 tasks
Add NVTE_KEEP_BACKWARD_UNQUANTIZED
#2644 opened Feb 3, 2026 by zianglih Loading…
6 of 13 tasks
100
ci(fix): NGC build
#2643 opened Feb 2, 2026 by ko3n1g Loading…
13 tasks
13
Add examples for MoE models - Mixtral in TE
#2642 opened Feb 2, 2026 by faradawn Loading…
2 of 13 tasks
1
4
docs(readme): update convergence table, latest news, and outdated links
#2638 opened Feb 1, 2026 by sbhavani Loading…
5 of 13 tasks
12
Fix FP8 block scaling with sequence parallel
#2637 opened Jan 31, 2026 by cuichenx Loading…
1 of 13 tasks
22
Fix Github workflows issues
#2636 opened Jan 30, 2026 by pggPL Loading…
8 of 13 tasks
43
Add 2d quant for mxfp8
#2634 opened Jan 29, 2026 by kunlunl Loading…
13 tasks
10
[Common][PyTorch] Add a new score func sqrtsoftplus to the fused router MoE
#2633 opened Jan 29, 2026 by yaox12 Loading…
4 of 16 tasks
@yaox12
18
[Common][PyTorch] Fuse scaling and unscaling of bf16 momentums into kernels 2.13.0
#2632 opened Jan 29, 2026 by yaox12 Loading…
13 tasks
9
ProTip! Exclude everything labeled bug with -label:bug.