Merge commit '734bcedd' of...

Merge commit '734bcedd' of https://github.com/NVIDIA/TransformerEngine

Merge commit '734bcedd' of...
Merge commit '734bcedd' of https://github.com/NVIDIA/TransformerEngine
87e3e56e · yuguo · 2f11bd2e · 734bcedd · 87e3e56e · 87e3e56e
Commit 87e3e56e authored Aug 27, 2025 by yuguo
20 changed files
--- a/tests/pytorch/test_numerics.py
+++ b/tests/pytorch/test_numerics.py
--- a/tests/pytorch/test_onnx_export.py
+++ b/tests/pytorch/test_onnx_export.py
--- a/tests/pytorch/test_parallel_cross_entropy.py
+++ b/tests/pytorch/test_parallel_cross_entropy.py
@@ -3,7 +3,6 @@
 # See LICENSE for license information.

 import random
-import pytest
 import torch
 from transformer_engine.pytorch.cross_entropy import parallel_cross_entropy


--- a/tests/pytorch/test_qk_norm.py
+++ b/tests/pytorch/test_qk_norm.py
--- a/tests/pytorch/test_recipe.py
+++ b/tests/pytorch/test_recipe.py
--- a/tests/pytorch/test_sanity.py
+++ b/tests/pytorch/test_sanity.py
--- a/tests/pytorch/utils.py
+++ b/tests/pytorch/utils.py
--- a/transformer_engine/common/CMakeLists.txt
+++ b/transformer_engine/common/CMakeLists.txt
--- a/transformer_engine/common/__init__.py
+++ b/transformer_engine/common/__init__.py
--- a/transformer_engine/common/comm_gemm_overlap/comm_gemm_overlap.cpp
+++ b/transformer_engine/common/comm_gemm_overlap/comm_gemm_overlap.cpp
--- a/transformer_engine/common/comm_gemm_overlap/userbuffers/userbuffers.cu
+++ b/transformer_engine/common/comm_gemm_overlap/userbuffers/userbuffers.cu
--- a/transformer_engine/common/comm_gemm_overlap/userbuffers/userbuffers.h
+++ b/transformer_engine/common/comm_gemm_overlap/userbuffers/userbuffers.h
--- a/transformer_engine/common/common.cu
+++ b/transformer_engine/common/common.cu
--- a/transformer_engine/common/common.h
+++ b/transformer_engine/common/common.h
--- a/transformer_engine/common/fused_attn/fused_attn.cpp
+++ b/transformer_engine/common/fused_attn/fused_attn.cpp
--- a/transformer_engine/common/fused_router/fused_moe_aux_loss.cu
+++ b/transformer_engine/common/fused_router/fused_moe_aux_loss.cu
--- a/transformer_engine/common/fused_router/fused_score_for_moe_aux_loss.cu
+++ b/transformer_engine/common/fused_router/fused_score_for_moe_aux_loss.cu
--- a/transformer_engine/common/fused_router/fused_topk_with_score_function.cu
+++ b/transformer_engine/common/fused_router/fused_topk_with_score_function.cu
--- a/transformer_engine/common/fused_router/utils.h
+++ b/transformer_engine/common/fused_router/utils.h
--- a/transformer_engine/common/gemm/cublaslt_gemm.cu
+++ b/transformer_engine/common/gemm/cublaslt_gemm.cu