- 17 Jul, 2025 2 commits
-
-
hx authored
* save original input Signed-off-by:
Hongxiao Bai <hongxiaob@nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * fix input_quantizer usage in Linear bwd Signed-off-by:
pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by:
Xin Yao <xiny@nvidia.com> Co-authored-by:
Tim Moon <4406448+timmoon10@users.noreply.github.com>
-
Sudhakar Singh authored
* mxfp8 is not supported on 120+ arch yet Signed-off-by:
Sudhakar Singh <sudhakars@nvidia.com> * change the default recipe for arch 120 Signed-off-by:
-
- 16 Jul, 2025 2 commits
-
-
Paweł Gadziński authored
* some initial code Signed-off-by:
Pawel Gadzinski <pgadzinski@nvidia.com> * onnx support Signed-off-by:
Kirthi Shankar Sivamani <ksivamani@nvidia.com> Signed-off-by:
Paweł Gadziński <62263673+pggPL@users.noreply.github.com> * fix Signed-off-by:
root <root@prenyx0221.a51.clusters.nvidia.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * lint fix Signed-off-by:
Pawel Gadzinski <pgadzinski@gmail.com> * Update build.yml Signed-off-by:
-
vcherepanov-nv authored
Signed-off-by:Vladimir Cherepanov <vcherepanov@nvidia.com>
-
- 14 Jul, 2025 2 commits
-
-
Autumn1998 authored
* fix underterminsic problem in CI Signed-off-by:
tongliu <tongliu@nvidia.com> * fix bug on mbs>1 Signed-off-by:
-
hx authored
* optimize permute Signed-off-by:
-
- 12 Jul, 2025 1 commit
-
-
Jan Bielak authored
* Fix clearing tensor data in backward removing is_first_op Signed-off-by:
Jan Bielak <jbielak@nvidia.com> * Misc fixes Signed-off-by:
-
- 11 Jul, 2025 1 commit
-
-
Tim Moon authored
Make MXFP8Tensor unpickling function backward compatible Signed-off-by:Tim Moon <tmoon@nvidia.com>
-
- 10 Jul, 2025 3 commits
-
-
Autumn1998 authored
* add router fusion Signed-off-by:
-
buptzyb authored
* Reuse cudagraph input and output tensor memory Signed-off-by:
Robin Zhang <robinz@nvidia.com> * Wrap _make_graphed_callables with fp8 Signed-off-by:
-
Paweł Gadziński authored
* push Signed-off-by:
-
- 09 Jul, 2025 3 commits
-
-
Selvaraj Anandaraj authored
Signed-off-by:
Selvaraj Anandaraj <selvaraja@login-ptyche02.ptyche.clusters.nvidia.com> Co-authored-by:
-
Zhongbo Zhu authored
* functional passed Signed-off-by:
zhongboz <zhongboz@nvidia.com> * before zero padding in mxfp8 swizzle, use torch zeros to malloc for now Signed-off-by:
-
Xin Yao authored
* Fix align_size Signed-off-by:
-
- 08 Jul, 2025 2 commits
-
-
Jan Bielak authored
* Change pre_forward to pre_first_forward Signed-off-by:
-
Jan Bielak authored
Add test for `LayerNormMLP` implementation using `te.ops.Sequential` to `test_fusible_ops.py` (#1924) * Add e2e test for LayerNormMLP implemented with te.Sequential Signed-off-by:
-
- 01 Jul, 2025 3 commits
-
-
Selvaraj Anandaraj authored
* Added MCore fsdp support for TE Signed-off-by:
Selvaraj Anandaraj <selvaraja@cw-dfw-cs-001-login-01.cm.cluster> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Refactored based on new MCore FSDP Signed-off-by:
-
Jan Bielak authored
* Replace `is_float8_tensor` with `is_quantized_tensor` Replace free function `is_float8_tensor` with `is_quantized_tensor` in `_common.py` and use it throughout the `ops` codebase to check if a tensor is a (possibly internal) quantized tensor Signed-off-by:
-
guyueh1 authored
Fix layernorm output shape in LayernormLinear, make it align with input shape Signed-off-by:
Guyue Huang <guyueh@nvidia.com> Co-authored-by:
-
- 28 Jun, 2025 1 commit
-
-
yuzhongw-nvidia authored
* fix: (1) UT ignores MLA; (2) bshd format runtime error. Ban fp8 mla attn + cp due to correctness problem Signed-off-by:
Yuzhong Wang <yuzhongw@nvidia.com> * only disable FP8 CP for MLA Signed-off-by:
Charlene Yang <8636796+cyanguwa@users.noreply.github.com> --------- Signed-off-by:
-
- 26 Jun, 2025 4 commits
-
-
Charlene Yang authored
* skip kv cache for sm89, cudnn < 9.12 Signed-off-by:
-
Jan Bielak authored
* Optimize _common.reshape by removing redundant operations Signed-off-by:
-
xiaoxi-wangfj authored
* [PyTorch|common] Implement unpadding kernel for FP8 1. Add multi-tensor unpadding kernel 2. Replace split+cat with unpadding kernel in Fp8Padding and Fp8Unpadding 3. Add unpadding with padding unit tests Signed-off-by:
xiaoxi-wangfj <690912414@qq.com> * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * add license Signed-off-by:
-
Zhongbo Zhu authored
* finish python ref impl for bulk alloc Signed-off-by:
-
- 16 Jun, 2025 1 commit
-
-
Li Tao authored
* Fix an issue when mcore uses te fusion ce implementation Signed-off-by:
lit <lit@nvidia.com> * simplify unit test code Signed-off-by:
-
- 13 Jun, 2025 6 commits
-
-
Charlene Yang authored
* add support for head dim > 128 Signed-off-by:
Charlene Yang <charleney@nvidia.com> * remove debugging Signed-off-by:
-
Oleg Goncharov authored
* Added support of FP4 data type Signed-off-by:
Oleg Goncharov <ogoncharov@nvidia.com> * Refactoring to BitsNum in progress Signed-off-by:
Oleg Goncharov <64355998+Oleg-Goncharov@users.noreply.github.com> * Pass correct num bits to create_2D_tensor_map; fixes CI Signed-off-by:
-
Tim Moon authored
* Add FP8 current scaling to te.Sequential tests Signed-off-by:
-
Tim Moon authored
* Do not initialize quantized weights with column-wise usage in inference mode Signed-off-by:
-
Jan Bielak authored
* Flatten basic op params during fuser init Signed-off-by:
-
Daniel Stokes authored
* Add support for overlapping wgrad NCCL AG with dgrad GEMM Signed-off-by:
djns99 <40156487+djns99@users.noreply.github.com> * Remove unused wait on memcpy API from UB Signed-off-by:
dastokes <dastokes@dastokes-dvt-01.nvidia.com>
-
- 12 Jun, 2025 4 commits
-
-
Evgeny Tsykunov authored
* Support L2Norm basic op Signed-off-by:
Evgeny <etsykunov@nvidia.com> * Add L2Norm module wrapper Signed-off-by:
-
Selvaraj Anandaraj authored
* Added double buffering support initial commit Signed-off-by:
Selvaraj Anandaraj <selvaraja@login-preos01.a51.clusters.nvidia.com> * Fixed bugs Signed-off-by:
Selvaraj Anandaraj <anandaraj@wisc.edu> Signed-off-by:
-
Paweł Gadziński authored
typo fix Signed-off-by: -
Kirthi Shankar Sivamani authored
Fix for loading old ckpt formats Signed-off-by:
-
- 10 Jun, 2025 2 commits
-
-
Kirthi Shankar Sivamani authored
* Initial basic setup Signed-off-by:
-
yuzhongw-nvidia authored
* Support MLA (qk_dim != v_dim) for AttnFuncWithCPAndKVP2P Signed-off-by:
Xiaowei Ren <103958965+xrennvidia@users.noreply.github.com>
-
- 09 Jun, 2025 2 commits
-
-
Kirthi Shankar Sivamani authored
* Manage deps and add einops Signed-off-by:
-
Selvaraj Anandaraj authored
* Lora spike Signed-off-by:
Selvaraj Anandaraj <selvaraja@login-preos02.a51.clusters.nvidia.com> * Added FP8 param support Signed-off-by:
-
- 07 Jun, 2025 1 commit
-
-
Xin Yao authored
only get skip_fp8_weight_update in fp8_graph_capturing Signed-off-by:
-
