Skip to content

Pull requests: ROCm/TransformerEngine

New pull request New
31 Open 475 Closed
31 Open 475 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

Added worktree-friendly git commit SHA1 lookup ci-level 1 CI test level 1
#540 opened Apr 14, 2026 by Micky774 Contributor Loading…
13 tasks
NV upstream release 2.12 merge ci-level 3 CI test level 3
#538 opened Apr 13, 2026 by Micky774 Contributor Loading…
13 tasks
Full MXFP4 Training Recipe ci-level 3 CI test level 3
#537 opened Apr 13, 2026 by sarthak-amd Collaborator Loading…
4 tasks
Ipanfilo/wheel build action
#529 opened Apr 7, 2026 by ipanfilo Collaborator • Draft
1 of 13 tasks
1
CI: Refactor ROCm CI to use GPU-sized runners and build-only jobs ci-level 3 CI test level 3
#528 opened Apr 7, 2026 by leo-automation Collaborator Loading…
20
Gfx1250 changes ci-level 2 CI test level 2
#527 opened Apr 7, 2026 by ipanfilo Collaborator Loading…
1 of 13 tasks
"castonly/casttranspose HIP kernel optimization in fp8 ci-level 3 CI test level 3
#519 opened Apr 4, 2026 by alextmagro Contributor Loading…
89
[proof-of-concept] NVFP4 GEMM via BF16 dequant ci-level 1 CI test level 1
#518 opened Apr 2, 2026 by matthiasdiener Contributor • Draft
13 tasks
@matthiasdiener
[proof-of-concept] Enable NVFP4 recipe ci-level 1 CI test level 1
#517 opened Apr 2, 2026 by matthiasdiener Contributor • Draft
1 of 13 tasks
@matthiasdiener
[WIP] NVFP4: hadamard_transform_cast_fusion_columnwise ci-level 1 CI test level 1
#515 opened Apr 1, 2026 by matthiasdiener Contributor • Draft
1 of 13 tasks
@matthiasdiener
NVFP4 Random Hadamard Transform (butterfly permutation-based) ci-level 1 CI test level 1
#509 opened Mar 27, 2026 by matthiasdiener Contributor Loading…
1 of 13 tasks
@matthiasdiener
25
[TE] Enable deterministic mode for fused attention ci-level 1 CI test level 1
#508 opened Mar 27, 2026 by AllenFarcas Contributor Loading…
7 of 13 tasks
50
Full MXFP4 Training Recipe
#506 opened Mar 26, 2026 by sarthak-amd Collaborator Loading…
2 tasks
3
Add fsdp2 fp8 unit tests TE 2.10 ci-level 3 CI test level 3
#492 opened Mar 17, 2026 by sudhu2k Contributor Loading…
8 of 13 tasks
@sudhu2k
26
Add AITER fused RoPE dispatch to FusedRoPEFunc
#489 opened Mar 17, 2026 by sarthak-amd Collaborator Loading…
14
ASV-format microbenchmark suite
#487 opened Mar 16, 2026 by Micky774 Contributor Loading…
1 of 13 tasks
12
Add Claude to review PRs
#480 opened Mar 13, 2026 by wenchenvincent Collaborator Loading…
1 of 6 tasks
7
Microbenchmarking and CI performance regression test
#478 opened Mar 10, 2026 by matthiasdiener Contributor • Draft
4 of 17 tasks
@matthiasdiener
18
Utilize AITER HSA embeddings
#477 opened Mar 9, 2026 by Micky774 Contributor Loading…
1 of 13 tasks
15
Use static libraries for AITER build ci-level 3 CI test level 3
#468 opened Mar 2, 2026 by Micky774 Contributor Loading…
1 of 13 tasks
40
[NO MERGE] Integrate CK varlen cross attention for small-seq (s_q=1, s_kv<=16)
#461 opened Feb 24, 2026 by VeeraRajasekhar Contributor Loading…
13 tasks
@VeeraRajasekhar
127
Added initial AI Agent instructions and skills
#448 opened Feb 12, 2026 by Micky774 Contributor Loading…
2 of 13 tasks
42
Update README.rst
#435 opened Jan 28, 2026 by aris134 Contributor • Draft
1 of 13 tasks
Hotfix/fused ce triton
#409 opened Jan 12, 2026 by sarthak-amd Collaborator Loading…
7
CI: Switch GHA pipeline to build and test wheels
#400 opened Dec 9, 2025 by leo-automation Collaborator • Draft
1 of 6 tasks
2
ProTip! no:milestone will show everything without a milestone.