From 50f28f65a0c086dbbde5fb58b16932852a08b6e3 Mon Sep 17 00:00:00 2001 From: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Date: Sun, 2 Mar 2025 16:34:00 +0800 Subject: [PATCH] fix typo in deep gemm benchmarking(#3991) --- benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py | 1 - 1 file changed, 1 deletion(-) diff --git a/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py b/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py index db210b3a5..bcfd133f1 100644 --- a/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py +++ b/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py @@ -211,7 +211,6 @@ def get_benchmark(tp_size): x = torch.randn((m, k), device="cuda", dtype=torch.bfloat16) y = torch.randn((n, k), device="cuda", dtype=torch.bfloat16) - # 预处理数据，在计时之前完成 x_fp8, x_scale = per_token_cast_to_fp8(x) y_fp8, y_scale = per_block_cast_to_fp8(y) x_scale_col_major = get_col_major_tma_aligned_tensor(x_scale.clone())