docs: add bilingual README and update configs

Author: shijiashuai

README.en.md

@@ -0,0 +1,71 @@
+# GPU SpMV (Sparse Matrix-Vector Multiplication)
+
+[![CI](https://github.com/LessUp/gpu-spmv/actions/workflows/ci.yml/badge.svg)](https://github.com/LessUp/gpu-spmv/actions/workflows/ci.yml)
+
+[简体中文](README.md) | English
+
+High-performance CUDA sparse matrix-vector multiplication library supporting CSR and ELL formats with multiple load-balancing optimization strategies.
+
+## Features
+
+- **Multiple Sparse Formats** — CSR and ELL (ELLPACK)
+- **Optimized CUDA Kernels** — Scalar CSR, Vector CSR (warp-per-row), Merge Path, ELL
+- **Auto Kernel Selection** — Based on matrix characteristics
+- **Performance Metrics** — Bandwidth utilization, GFLOPS, benchmark framework
+- **Engineering Quality** — RAII (`CudaBuffer`, `CudaTimer`, `ScopedTexture`), semantic error codes, CMake Presets, CI
+
+## Quick Start
+
+```bash
+# Using CMake Presets
+cmake --preset release
+cmake --build --preset release
+
+# Run tests
+ctest --preset default
+
+# Run benchmarks
+./build/spmv_benchmark
+```
+
+## Requirements
+
+- CUDA Toolkit 11.0+, CMake 3.18+, C++17, GPU CC 7.0+
+
+## Usage
+
+```cpp
+#include "spmv/csr_matrix.h"
+#include "spmv/spmv.h"
+#include "spmv/cuda_buffer.h"
+
+using namespace spmv;
+
+CSRMatrix* csr = csr_create(0, 0, 0);
+csr_from_dense(csr, dense.data(), 3, 3);
+csr_to_gpu(csr);
+
+SpMVConfig config = spmv_auto_config(csr);
+SpMVResult result = spmv_csr(csr, d_x.get(), d_y.get(), &config, 3);
+```
+
+## Kernel Selection Strategy
+
+- **Short rows (avg_nnz < 4)**: Scalar CSR
+- **Uniform distribution (skewness < 10)**: Vector CSR
+- **Highly skewed (skewness >= 10)**: Merge Path
+
+## Project Structure
+
+```
+├── include/spmv/       # Headers (common, cuda_buffer, csr, ell, spmv, benchmark, pagerank)
+├── src/                # Source files
+├── tests/              # Property tests + unit tests
+├── benchmarks/         # Benchmark program
+├── CMakePresets.json   # Build presets
+└── .github/workflows/  # CI
+```
+
+## License
+
+MIT License

README.md

@@ -1,6 +1,9 @@
 # GPU SpMV (稀疏矩阵向量乘法)
 
-[![CI](https://github.com/<OWNER>/gpu-spmv/actions/workflows/ci.yml/badge.svg)](https://github.com/<OWNER>/gpu-spmv/actions/workflows/ci.yml)
+[![CI](https://github.com/LessUp/gpu-spmv/actions/workflows/ci.yml/badge.svg)](https://github.com/LessUp/gpu-spmv/actions/workflows/ci.yml)
+[![Docs](https://img.shields.io/badge/Docs-GitHub%20Pages-blue?logo=github)](https://lessup.github.io/gpu-spmv/)
+
+简体中文 | [English](README.en.md)
 
 基于 CUDA 的高性能稀疏矩阵向量乘法库，支持 CSR 和 ELL 格式，包含多种负载均衡优化策略。
 
@@ -228,11 +231,11 @@ std::string json = benchmark_to_json(result);
 
 项目文档已通过 GitHub Pages 发布：
 
-> **https://\<OWNER\>.github.io/gpu-spmv/**
+> **https://lessup.github.io/gpu-spmv/**
 
 包含：
-- [API 参考](https://OWNER.github.io/gpu-spmv/api) — 头文件接口、数据结构与函数说明
-- [性能优化](https://OWNER.github.io/gpu-spmv/performance) — Kernel 选择策略、带宽优化与基准测试
+- [API 参考](https://lessup.github.io/gpu-spmv/api) — 头文件接口、数据结构与函数说明
+- [性能优化](https://lessup.github.io/gpu-spmv/performance) — Kernel 选择策略、带宽优化与基准测试
 
 ## 许可证
 

docs/index.md

@@ -90,4 +90,4 @@ csr_destroy(csr);
 
 ## 许可证
 
-[MIT License](https://github.com/OWNER/gpu-spmv/blob/main/LICENSE)
+[MIT License](https://github.com/LessUp/gpu-spmv/blob/main/LICENSE)