Getting Started: Share Your First DeepX Build & Python Run

[miaobyte]

如果你已经顺利完成了 deepx 的编译和 Python 示例的运行，说明你已经迈出了参与 deepx 开发的第一步！

🎊 恭喜你！，📢 请在本 Issue 下展示你的成果

[LJXjean]

What i did for successfully run an example:

1. System update & essential dev stack

sudo apt update &&
sudo apt upgrade -y 
sudo apt install -y build-essential cmake clang git wget pkg-config ninja-build libopenblas-dev libyaml-cpp-dev libjemalloc-dev libomp-dev

---

2. Highway SIMD: build from source (same as Dockerfile):

git clone https://github.com/google/highway.git cd highway
cmake -S . -B build -GNinja \
      -DCMAKE_BUILD_TYPE=Release \
      -DCMAKE_INSTALL_PREFIX=/usr/local
cmake --build build -j$(nproc)
sudo cmake --install build cd ~

---

4. Configure & build with CMake/Ninja

deepx/excuter/
 ├─ cpp-common/
 ├─ op-mem-ompsimd/
 │   ├─ CMakeLists.txt (the one you pasted)
 │   └─ src/…
 └─ … other modules

cd op-mem-ompsimd mkdir -p build && cd build

cmake .. \
  -GNinja \
  -DCMAKE_BUILD_TYPE=Debug \
  -DHWY_ENABLE=ON \
  -DCMAKE_CXX_COMPILER=clang++ # or g++ if you prefer  # Append SIMD flags only once—CMakeLists already sets them 
cmake --build . -j$(nproc)

---

I failed in cmake --build . -j$(nproc),
I met a problem:

/home/venus/deepx/excuter/op-mem-ompsimd/src/deepx/tensorfunc/elementwise_miaobyte.hpp:104:17: error: static declaration of 'todtype' follows non-static declaration
104 | static void todtype(const Tensor &A, Tensor &C)
| ^
/home/venus/deepx/excuter/cpp-common/src/deepx/tensorfunc/elementwise.hpp:11:10: note: previous declaration is here
11 | void todtype(const Tensor &input, Tensor &output);
| ^
1 warning and 1 error generated.

I managed it by following modification:

[zhaziqwe]

build successfully, for python front you need to install pyyaml numpy triton torch