From ae957014ed52ddf76798780e686dab39d51c391e Mon Sep 17 00:00:00 2001 From: Baizhou Zhang Date: Tue, 2 Jun 2026 19:03:05 -0700 Subject: [PATCH] Update instructions for SGLang deepgemm fork development --- README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/README.md b/README.md index 6ef705ffce..a253f97b3a 100644 --- a/README.md +++ b/README.md @@ -1,5 +1,13 @@ # DeepGEMM +## Notice for SGLang DeepGemm Fork Developers + +- `main` tracks upstream updates and should stay clean. +- `dev` is the active SGL integration branch and can be force-updated during rebases. Please open pull requests to this branch +- `release/*` branches are archives for past wheel releases and shouldn't be modified. + +For more instructions, please check this [document](https://github.com/sgl-project/DeepGEMM/blob/dev/sgl_deep_gemm/README.md). + DeepGEMM is a unified, high-performance tensor core kernel library that brings together the key computation primitives of modern large language models — GEMMs (FP8, FP4, BF16), fused MoE with overlapped communication (Mega MoE), MQA scoring for the lightning indexer, HyperConnection (HC), and more — into a single, cohesive CUDA codebase. All kernels are compiled at runtime via a lightweight Just-In-Time (JIT) module, requiring no CUDA compilation during installation. DeepGEMM leverages some concepts from [CUTLASS](https://github.com/nvidia/cutlass) and [CuTe](https://github.com/NVIDIA/cutlass/tree/main/include/cute), but avoids heavy reliance on their templates or algebras. The library is designed for simplicity, with only a limited number of core kernel functions, making it a clean and accessible resource for learning NVIDIA GPU kernel optimization techniques.