Installation

Prerequisites

• Python 3.9 or later
• CMake 3.18 or later
• C++17 compatible compiler

Installing via Pip

CPU Version (PyPI)

You can install the stable CPU version directly from PyPI:

pip install tensorplay

CUDA Version (Custom Index)

For CUDA-enabled versions, use our custom wheel index. For example, for CUDA 13.0:

pip install tensorplay --extra-index-url https://download.tensorplay.cn/whl/cu130

Note: Replace cu130 with your supported CUDA version (e.g., cu121, cu118).

Permanent Configuration

To avoid typing the extra index URL every time, you can add it to your pip configuration:

Windows: Create or edit %APPDATA%\pip\pip.ini:

[global]
extra-index-url = https://download.tensorplay.cn/whl/cu130

Linux/macOS: Create or edit ~/.pip/pip.conf:

[global]
extra-index-url = https://download.tensorplay.cn/whl/cu130

Installing from Source

1. Clone the repository:

   git clone https://github.com/bluemoon-o2/tensorplay.git
   cd tensorplay

2. Install dependencies:

   pip install -r requirements.txt

3. Build and install:

   python setup.py install

   or for development:

   pip install -e .