Commits · 72ba94a17e87320c04c5f0b6bd43e4c27a27cdb0 · OpenDAS / TransformerEngine

19 Dec, 2025 1 commit
- Add bias fwd/bwd at group gemm · 72ba94a1
  wenjh authored Dec 19, 2025
```
Signed-off-by: wenjh <wenjh@sugon.com>
```
  72ba94a1
15 Dec, 2025 2 commits

Fix blaslt group gemm crush · e698a0a7
wenjh authored Dec 15, 2025
```
Signed-off-by: wenjh <wenjh@sugon.com>
```
e698a0a7

Merge branch 'TE_ta_develop2.9' into 'develop_v2.9' · 4086a4cc

wenjh authored Dec 15, 2025

[DCU]Fix compilation unable to find nvte-extract_ded_and_offset

See merge request dcutoolkit/deeplearing/TransformerEngine!64

4086a4cc

10 Dec, 2025 2 commits
- [DCU]fix mip and compile issues · 121d9224
  tabuchixiangcai3 authored Dec 10, 2025
```
Signed-off-by: Tangao <2205747538@qq.com>
```
  121d9224
- [DCU]Fix compilation unable to find nvte-extract_ded_and_offset · ba058648
  tabuchixiangcai3 authored Dec 10, 2025
```
Signed-off-by: Tangao <2205747538@qq.com>
```
  ba058648
04 Dec, 2025 2 commits
- Merge branch 'develop_v2.9' into 'develop_v2.9' · 177291ac
  wenjh authored Dec 04, 2025
```
grouped_linear supports activation offloading

See merge request dcutoolkit/deeplearing/TransformerEngine!62
```
  177291ac
- grouped_linear supports activation offloading · db0ad945
  dongcl authored Dec 04, 2025
  
  db0ad945
03 Dec, 2025 5 commits
- Make release_v2.9 compile pass · 99e60246
  wenjh authored Dec 03, 2025
  
  99e60246
- Fix build error · cbb14a5f
  wenjh authored Dec 03, 2025
```
Signed-off-by: wenjh <wenjh@sugon.com>
```
  cbb14a5f
- Fix build error · b3dcfc28
  wenjh authored Dec 03, 2025
  
  b3dcfc28
- Update build and install command · 1e3c6a25
  wenjh authored Dec 03, 2025
```
Signed-off-by: wenjh <wenjh@sugon.com>
```
  1e3c6a25
- Merge nv main up to v2.10.0.dev0 · 063ef88d
  wenjh authored Dec 03, 2025
```
Signed-off-by: wenjh <wenjh@sugon.com>
```
  063ef88d
26 Nov, 2025 2 commits

Merge branch 'develop_v2.8' into 'main' · 91670b05

wenjh authored Nov 26, 2025

[DCU] Skip some tests in test_sanity.py

See merge request dcutoolkit/deeplearing/TransformerEngine!61

91670b05

Merge branch 'fix_develop2.8_zc' into 'develop_v2.8' · 3a040217
wenjh authored Nov 26, 2025
```
[DCU]Fix some bugs

See merge request dcutoolkit/deeplearing/TransformerEngine!56
```
3a040217

12 Nov, 2025 4 commits
- Merge branch 'develop_v2.8' into 'main' · e3780e3a
  wenjh authored Nov 12, 2025
```
Fix build error

See merge request dcutoolkit/deeplearing/TransformerEngine!60
```
  e3780e3a
- Fix build error · a622988a
  wenjh authored Nov 12, 2025
  
  a622988a
- Merge branch 'develop_v2.8' into 'main' · a145a62a
  wenjh authored Nov 12, 2025
```
Fix hipblaslt handle manage

See merge request dcutoolkit/deeplearing/TransformerEngine!59
```
  a145a62a
- Fix hipblaslt handle manage · f4bd89eb
  wenjh authored Nov 12, 2025
  
  f4bd89eb
08 Nov, 2025 2 commits
- Merge branch 'develop_v2.8' into 'main' · e32965ff
  wenjh authored Nov 08, 2025
```
Fix user args core dump in mt

See merge request dcutoolkit/deeplearing/TransformerEngine!57
```
  e32965ff
- Fix user args core dump in mt · a13c52ad
  wenjh authored Nov 08, 2025
  
  a13c52ad
03 Nov, 2025 8 commits
- [DCU] fix some bugs in test_numerics.py · f7c66e28
  zhaochao authored Nov 03, 2025
  
  f7c66e28
- [DCU]Skip configurations that FlashAttention does not support · 87682fe2
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  87682fe2
- [DCU]Resolve the issue of checkpoint test weights not existing. · 9d34e27a
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  9d34e27a
- [DCU] Fix the bug in test_onnx_export.py under L0 · d5cd815f
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  d5cd815f
- [DCU] Skip alpha non-1 tests · ef65dd33
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  ef65dd33
- [DCU] fix bug with cannot import name 'use_lightop_w8a8' from 'transformer_engine.pytorch.utils' · 3d36696b
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  3d36696b
- [DCU] Skip some tests in test_cuda_graphs.py under L0 · 2fc4b10c
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  2fc4b10c
- [DCU] Skip some tests in test_sanity.py · 6af7b77d
  zhaochao authored Nov 03, 2025
```
Signed-off-by: zhaochao <zhaochao1@sugon.com>
```
  6af7b77d
31 Oct, 2025 1 commit

Merge branch 'TE_develop2.8' into 'develop_v2.8' · 3a5755b1

wenjh authored Oct 31, 2025

[DCU]Fix memory overflow and test-didistributed in L1_pytorch_istributed_unittest

See merge request dcutoolkit/deeplearing/TransformerEngine!49

3a5755b1

17 Oct, 2025 3 commits
- [DCU]Fix the original code · b11d6fca
  tabuchixiangcai3 authored Oct 17, 2025
```
Signed-off-by: Tangao <2205747538@qq.com>
```
  b11d6fca
- Merge branch 'develop_v2.8' into 'develop_v2.8' · 4b65dfa3
  yuguo authored Oct 17, 2025
```
Update activation offload code to align with the official version

See merge request dcutoolkit/deeplearing/TransformerEngine!52
```
  4b65dfa3
- Update activation offload code to align with the official version · 9711d439
  dongcl authored Oct 17, 2025
  
  9711d439
16 Oct, 2025 6 commits

Changed VERSION to 2.10.0.dev0 · 5624dbb4
Przemek Tredak authored Oct 16, 2025
```
Signed-off-by: Przemek Tredak <ptredak@nvidia.com>
```
5624dbb4

[PyTorch] Add record_stream and untyped_storage func op in QuantizedTensor (#2144) · 81c363bf

xiaoxi-wangfj authored Oct 17, 2025



* [PyTorch] Add record_stream and untyped_storage func op in QuantizedTensor
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>

* Update transformer_engine/pytorch/tensor/float8_blockwise_tensor.py
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>

* Update transformer_engine/pytorch/tensor/float8_blockwise_tensor.py
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>

---------
Signed-off-by: xiaoxi-wangfj <690912414@qq.com>
Co-authored-by: Tim Moon <4406448+timmoon10@users.noreply.github.com>

81c363bf

Added support for DistOpt with offloading with MoE's (#2264) · 452c7374

Selvaraj Anandaraj authored Oct 16, 2025

Signed-off-by: Selvaraj Anandaraj <selvaraja@login-ptyche02.ptyche.clusters.nvidia.com>
Co-authored-by: Selvaraj Anandaraj <selvaraja@login-ptyche02.ptyche.clusters.nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

452c7374

Merge branch 'develop_v2.8' of... · 712d526a

yuguo authored Oct 16, 2025

Merge branch 'develop_v2.8' of http://10.16.6.30/dcutoolkit/deeplearing/TransformerEngine into develop_v2.8

712d526a

[DCU] remove redundant gemm · 47077129
yuguo authored Oct 16, 2025

47077129
[DCU]Fix memory overflow and test-didistributed in L1_pytorch_istributed_unittest · 2a64c9a6
tabuchixiangcai3 authored Oct 16, 2025
```
Signed-off-by: Tangao <2205747538@qq.com>
```
2a64c9a6

15 Oct, 2025 2 commits

README - latest news update (#2273) · 88564d59

Santosh Bhavani authored Oct 15, 2025



* Enhance Latest News section with recent TE and FP8 developments

- Adds NVFP4 pretraining research paper with PR #2177 reference
Signed-off-by: Santosh Bhavani <santosh.bhavani@live.com>

* update nvfp4 reference
Signed-off-by: Santosh Bhavani <santosh.bhavani@live.com>

* Update README.rst
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

---------
Signed-off-by: Santosh Bhavani <santosh.bhavani@live.com>
Signed-off-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

88564d59

[PyTorch Debug] Fix issue with start_end_list logging feature (#2252) · 4c572f04

Paul Gibbons authored Oct 15, 2025



* fixes for start_end_list usage in TE debug
Signed-off-by: Paul Gibbons <pgibbons@nvidia.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Signed-off-by: Paul Gibbons <pgibbons@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Paweł Gadziński <62263673+pggPL@users.noreply.github.com>

4c572f04