ContextQMD
Back to Sglang

README

benchmark/kernels/deepseek/README.md

0.5.11941 B
Original Source

DeepSeek kernels benchmark

Prerequisites

  • You should install DeepGemm from source before run benchmark_deepgemm_fp8_gemm.py and benchmark_deepgemm_fp8_group_gemm.py.

Benchmark

  • benchmark_deepgemm_fp8_gemm.py

    bash
    python benchmark_deepgemm_fp8_gemm.py --run_correctness --tp_size 1

  • benchmark_deepgemm_fp8_group_gemm.py

    bash
    python benchmark_deepgemm_fp8_group_gemm.py --run_correctness --tp_size 1

  • You can use the --run_correctness parameter to verify all kernels results's correctness.

    • You can use the --tp_size parameter to benchmark all FP8 w8a8 block-wise matrix multiplications involved in DeepSeek V3/R1 under the current tensor parallelism (TP) setting. This benchmark compares DeepSeek's open-source DeepGemm implementation with SGLang's and VLLM Triton implementation.