ai4one 0.3.2

A small machine learning package

AI4One 🤖

A small, modular package for machine learning.

---

Installation

pip install ai4one

This package requires Python 3.8 or newer.

---

Usage

ai4one.config

3.9>=Python 版本>=3.7, 请使用pip isntall ai4one~=0.2安装 ai4one=0.2 版本。

The primary feature of this package is a powerful configuration system. For a comprehensive guide and examples, please see the Configuration System Guide.

from ai4one.config import BaseConfig, field
from typing import List

class TrainConfig(BaseConfig):
    learning_rate: float = 0.001
    epochs: int = 10
    optimizer: str = "Adam"
    layers: List[int] = [3, 3]

if __name__ == "__main__":
    config = TrainConfig.argument_parser()
    print(f"Using optimizer: {config.optimizer}")

You can also run the self-contained example to see it in action:

examples/example_config.py

---

ai4one.cli

ai4one gpu

Outputs:

--- CUDA Version ---
12.7 

--- PyTorch Version ---
2.1.0+cu121

--- Python Version ---
Python 3.10.12

--- Python Executable Path ---
/usr/bin/python

ai4one.tools

ai4one.tools.plt

该模块提供了一些用于绘图的工具函数。

1. setup_fonts 函数用于设置 matplotlib 字体，以支持中文字符。

   主要用途是跨平台支持中文字符的显示。

   • 默认配置为 New Times Roman 、SimSun 和 SimHei 两种字体。
   • 使用详情查看 tools.plt

from ai4one.tools.plt import setup_fonts
setup_fonts(["New Times Roman", "SimSun", "SimHei"])

Development

Interested in contributing? Set up your local development environment with uv.

1. Clone the repository:

   git clone https://github.com/bestenevoy/ai4one.git
   cd ai4one
   

2. Create a virtual environment and install dependencies: This command installs all core, optional, and development dependencies.

   uv pip install -e ".[dev]"
   

   To keep your environment in sync with the lock file, you can run uv sync.

3. Run tests:

   uv run pytest
   

---

Build and Publish

These commands are for package maintainers.

Build the package:

uv build

Publish to PyPI:

uv run twine upload dist/*