- 30 Oct, 2025 1 commit
-
-
Kirthi Shankar Sivamani authored
* Fix attention backend and tests for sm120 Signed-off-by:
Kirthi Shankar Sivamani <ksivamani@nvidia.com> * Disable MLA only for backward Signed-off-by:
-
- 29 Oct, 2025 1 commit
-
-
vthumbe1503 authored
* changes working Signed-off-by:
Varun Thumbe <vthumbe@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add support for onnx, minor comments Signed-off-by:
Przemyslaw Tredak <ptrendx@gmail.com> Signed-off-by:
pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by:
-
- 25 Oct, 2025 1 commit
-
-
Charlene Yang authored
* add max_score for fused/unfused F16 non-CP Signed-off-by:
Charlene Yang <8636796+cyanguwa@users.noreply.github.com> * calculate max per head instead of max over all heads Signed-off-by:
-
- 23 Oct, 2025 1 commit
-
-
Paweł Gadziński authored
* fix perf issue Signed-off-by:
Pawel Gadzinski <pgadzinski@nvidia.com> * fix Signed-off-by:
-
- 22 Oct, 2025 1 commit
-
-
Evgeny Tsykunov authored
* rename experimental -> custom_recipes Signed-off-by:
Evgeny <etsykunov@nvidia.com> * Decouple python base classes (api) Signed-off-by:
Evgeny Tsykunov <e.tsykunov@gmail.com> * Update tests/pytorch/test_custom_recipe.py Co-authored-by:
-
- 21 Oct, 2025 2 commits
-
-
Kunlun Li authored
* Add post-processing API for FP8 primary weights to support CUDA Graph Signed-off-by:
kunlunl <kunlunl@nvidia.com> * Add post-processing support for plain pytorch tensors Signed-off-by:
Tim Moon <4406448+timmoon10@users.noreply.github.com> --------- Signed-off-by:
-
Zhongbo Zhu authored
* pipeclean, fix nvfp4 padding of 32 alignment Signed-off-by:
Zhongbo Zhu <zhongboz@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * numerical test passed Signed-off-by:
Tim Moon <tmoon@nvidia.com> * Fix invalid conversion from tensor to int Signed-off-by:
-
- 20 Oct, 2025 1 commit
-
-
Kirthi Shankar Sivamani authored
* Fix CI failures due to deterministic attention Signed-off-by:
-
- 17 Oct, 2025 1 commit
-
-
Neil Tenenholtz authored
* Fix test of FSDP2 by correcting init logic and applying autocast This fixes multiple issues in the FSDP2 test, namely 1. Previously fp8 init was performed when `args.fp8_init == False`. I have updated the logic to match what I presume was intended by leveraging the nullcontext context manager. 2. `te.fp8_autocast` was previously not called; the recipe was created but was unused. The autocast context manager now wraps the model's computation. Signed-off-by:
Neil Tenenholtz <ntenenz@users.noreply.github.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix typo Signed-off-by:
-
- 14 Oct, 2025 2 commits
-
-
Kirthi Shankar Sivamani authored
* Initial API change Signed-off-by:
-
Evgeny Tsykunov authored
* Fix update_quantized in ref nvfp4 quantizer Signed-off-by:
-
- 13 Oct, 2025 1 commit
-
-
Peter St. John authored
Signed-off-by:
Peter St. John <pstjohn@nvidia.com> Co-authored-by:
-
- 03 Oct, 2025 1 commit
-
-
Jan Bielak authored
* Update to_string(NVTEScalingMode) to include block scaling Signed-off-by:
Jan Bielak <jbielak@nvidia.com> * Add `nvte_swizzle_block_scaling_to_mxfp8_scaling_factors` Signed-off-by:
Copilot <175728472+Copilot@users.noreply.github.com> Signed-off-by:
-
- 02 Oct, 2025 1 commit
-
-
Tim Moon authored
* Make sure to set usages for linear op quantizers before forward Signed-off-by:
-
- 01 Oct, 2025 1 commit
-
-
Evgeny Tsykunov authored
* Introduce QuantizerBase Signed-off-by:
-
- 30 Sep, 2025 2 commits
-
-
vthumbe1503 authored
* Test working as I think it should work Signed-off-by:
Vladimir Cherepanov <vcherepanov@nvidia.com> * Saving... Signed-off-by:
Vladimir Cherepanov <56651474+mk-61@users.noreply.github.com> * Add license Signed-off-by:
Ming Huang <mingh@nvidia.com> * Slightly refactor Signed-off-by:
Phuong Nguyen <phuonguyen@nvidia.com> Signed-off-by:
Jeremy Berchtold <jberchtold@nvidia.com> Co-authored-by:
jberchtold-nvidia <158520091+jberchtold-nvidia@users.noreply.github.com> Signed-off-by:
Kshitij Lakhani <klakhani@nvidia.com> Signed-off-by:
oliver könig <okoenig@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update setup.py Signed-off-by:
Xin Yao <xiny@nvidia.com> * Remove exceptions from destructors Signed-off-by:
Robin Zhang <robinz@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * update Signed-off-by:
Vasudevan Rengasamy <vrengasamy@nvidia.com> Co-authored-by:
Selvaraj Anandaraj <selvaraja@login-ptyche01.ptyche.clusters.nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fixed typo Signed-off-by:
Selvaraj Anandaraj <selvaraja@login-preos01.a51.clusters.nvidia.com> * Fixed typo Signed-off-by:
Paweł Gadziński <62263673+pggPL@users.noreply.github.com> Signed-off-by:
djns99 <40156487+djns99@users.noreply.github.com> Signed-off-by:
Casper <casperbh.96@gmail.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci Signed-off-by:
Hongbin Liu <hongbinl@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix CI Signed-off-by:
tongliu <tongliu@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Signed-off-by:
Jonathan Mitchell <jomitchell@login-eos02.eos.clusters.nvidia.com> * assert line change Signed-off-by:
Sudhakar Singh <sudhakars@nvidia.com> * address review comments Signed-off-by:
Yuzhong Wang <yuzhongw@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update transformer_engine/pytorch/tensor/_internal/float8_blockwise_tensor_base.py Signed-off-by:
Min Yang <min.yang@shopee.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * refactor: refactor multi tensor gemm interface Signed-off-by:
alan yang <89962857+cassiewilliam@users.noreply.github.com> Co-authored-by:
zhujian <zhujian.whu.cs@gmail.com> Fix cuDNN version checks when getting backend and for sm89 kv cache (#2185) * Fix cudnn version checks for kv cache for sm89. Add cudnn version check in preparation for 9.14 when getting backend Signed-off-by:
-
Charlene Yang authored
* debug existing usage Signed-off-by:
-
- 29 Sep, 2025 1 commit
-
-
Kirthi Shankar Sivamani authored
* Add NVFP4 recipe Signed-off-by:
Frank Sun <frsun@nvidia.com> Co-authored-by:
Oleg Goncharov <ogoncharov@nvidia.com> Co-authored-by:
Teddy Do <tdophung@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Add MathDx dependency to GitHub builds Signed-off-by:
Przemek Tredak <ptredak@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Fix linter warning Signed-off-by:
-
- 22 Sep, 2025 1 commit
-
-
Charlene Yang authored
* first draft; debug plan failure Signed-off-by:
Chen Cui <chcui@nvidia.com> Signed-off-by:
-
- 18 Sep, 2025 2 commits
-
-
zhujian authored
feature(FA3,MLA,CP): 1. Update FA3 to commit-id 3ba6f82 (tag 2.8.0.post2 with compile error fixed), PR-1604 support hdimQK != hdimV backward 2. Update get_attention_backend method because FA3 support MLA now 3. Add CP MLA support for FA3 4. Add unit tests for FA3 MLA CP 5. Update attention doc Signed-off-by: -
alan yang authored
* feat: add cutlass group gemm support Signed-off-by:
-
- 17 Sep, 2025 1 commit
-
-
vthumbe1503 authored
* Test working as I think it should work Signed-off-by:
-
- 15 Sep, 2025 2 commits
-
-
Tim Moon authored
* Add CPU offloading logic to ops. Fix test to compute dgrad. Signed-off-by:
-
Paweł Gadziński authored
* fix underflows log issue Signed-off-by:
-
- 10 Sep, 2025 1 commit
-
-
jomitchellnv authored
Adds context parallelism utilities: moving cp shards to diff ranks and pad sequence to divisibility factory (#2129) * test - adds unit test for cp utilities and the utilites Signed-off-by:
-
- 08 Sep, 2025 1 commit
-
-
vasunvidia authored
* Fused RoPE with combined QKV input. Initial commit for Dropout with 8-bit RNG Fix documentation Initial commit for Fused QKV RoPE WIP Initial tests passing Enable rotary percent and margin Enable CP2, start_positions, interleaved Cleanup test Revert "Fix documentation" This reverts commit 53df10044e7769982bd4af2ae2628e6b7717e715. Revert "Initial commit for Dropout with 8-bit RNG" This reverts commit 301505e24031cbcd679069e1c2cd4d00eedf2dca. Cleanup. Minor cleanup Signed-off-by:
-
- 04 Sep, 2025 1 commit
-
-
Casper authored
* fix cross entropy Signed-off-by:
-
- 03 Sep, 2025 1 commit
-
-
Daniel Stokes authored
Signed-off-by:
-
- 31 Aug, 2025 1 commit
-
-
vasunvidia authored
* Add dropout kernel with 8-bit RNG Co-authored-by:
-
- 29 Aug, 2025 1 commit
-
-
Daniel Stokes authored
-
- 28 Aug, 2025 2 commits
-
-
Paweł Gadziński authored
* Compute amax in normalization forward in current scaling in untuned kernels Signed-off-by:
-
Charlene Yang authored
* disable determinism for sm100+ and cudnn<9.14 Signed-off-by:
-
- 26 Aug, 2025 3 commits
-
-
Paweł Gadziński authored
fix test on old device Signed-off-by: -
vcherepanov-nv authored
* Bump cuDNN FE to 1.14.0 Signed-off-by:
-
Xin Yao authored
expose more activation functions Signed-off-by:
-
- 20 Aug, 2025 2 commits
-
-
Paweł Gadziński authored
* code drop Signed-off-by:
-
Jan Bielak authored
* Add rmsnorm_bwd_add Signed-off-by:
-
- 15 Aug, 2025 1 commit
-
-
Jan Bielak authored
* Add `nvte_cublas_gemm_scaled` Signed-off-by:
-
- 14 Aug, 2025 2 commits
-
-
Tim Moon authored
* Unfused impl for dbias-quantize Signed-off-by:
-
Xin Yao authored
* reduce driver calls Signed-off-by:
-
