From 0bd67ba2bd52a4d8fd2298bc5476d191cac47773 Mon Sep 17 00:00:00 2001 From: Zhijian Liu <5782437+zhijian-liu@users.noreply.github.com> Date: Sat, 14 Jun 2025 16:54:40 -0700 Subject: [PATCH] Fix a minor bug related to DeepGEMM upgrade (#7191) --- .../srt/layers/quantization/deep_gemm_wrapper/entrypoint.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/python/sglang/srt/layers/quantization/deep_gemm_wrapper/entrypoint.py b/python/sglang/srt/layers/quantization/deep_gemm_wrapper/entrypoint.py index b2551471d..e407a3598 100644 --- a/python/sglang/srt/layers/quantization/deep_gemm_wrapper/entrypoint.py +++ b/python/sglang/srt/layers/quantization/deep_gemm_wrapper/entrypoint.py @@ -52,7 +52,8 @@ def grouped_gemm_nt_f8f8bf16_masked( expected_m, n, k, num_groups, kernel_type ): _grouped_gemm_nt_f8f8bf16_masked_raw( - lhs, rhs, out, masked_m, expected_m, recipe=recipe + lhs, rhs, out, masked_m, expected_m, + **({"recipe": recipe} if DEEPGEMM_V202506 else {}) )