sglang/benchmark/kernels/deepseek

DeepSeek kernels benchmark

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