yewentao256 9f04a6cf57 Revert "Merge commit '6e8d8c4afbddf725b34ef938616701869f5b3462' into sage/dbo-full-cudagraphsh" ...

This reverts commit 5215c80a4988e81d2f5971e02d50d3785cab5ae8, reversing
changes made to dd2a94fd9d90d0c04772380c220b5ec81bd0b61e.

2025-08-19 12:17:31 -07:00

..

marlin_moe_wna16

remove unused variables in marlin_template.h (#20236)

2025-07-02 00:51:52 +00:00

permute_unpermute_kernels

Fix CUDA permute/unpermute for use with DeepGemm Moe (#17934)

2025-07-27 07:08:00 -07:00

moe_align_sum_kernels.cu

[perf] Speed up align sum kernels (#21079)

2025-07-21 11:19:23 -07:00

moe_ops.h

[Perf] Optimize moe_align_block_size CUDA kernel (#19572)

2025-06-17 11:49:26 -07:00

moe_permute_unpermute_op.cu

Fix CUDA permute/unpermute for use with DeepGemm Moe (#17934)

2025-07-27 07:08:00 -07:00

moe_wna16_utils.h

pre-commit autoupdate (#17380)

2025-04-29 06:46:55 -07:00

moe_wna16.cu

[BugFix] Accuracy fix for llama4 int4 - improperly casted scales (#16801)

2025-04-17 22:13:29 -07:00

topk_softmax_kernels.cu

Revert "Merge commit '6e8d8c4afbddf725b34ef938616701869f5b3462' into sage/dbo-full-cudagraphsh"

2025-08-19 12:17:31 -07:00

torch_bindings.cpp

Fix CUDA permute/unpermute for use with DeepGemm Moe (#17934)

2025-07-27 07:08:00 -07:00