ailiteall 0.0.7

A simplistic ai for Huggingface llms free hosted models.

AILiteLLM

AILiteLLM is a lightweight Python wrapper around the Hugging Face Inference API that provides OpenAI-like interface for various open-source language models. It makes it easy to use powerful open-source models with an interface familiar to OpenAI developers.

📊 Package Stats

• PyPI: https://pypi.org/project/ailitellm/
• Documentation: https://ailitellm.readthedocs.io/
• Source Code: https://github.com/yourusername/ailitellm
• Issue Tracker: https://github.com/yourusername/ailitellm/issues

Features

• OpenAI-compatible interface
• Support for multiple Hugging Face models
• Stream responses support
• Function calling capabilities
• Full typing support
• Easy model switching

Installation

From PyPI:

pip install ailitellm

From source:

git clone https://github.com/yourusername/ailitellm.git
cd ailitellm
pip install -e .

Development installation:

pip install -e ".[dev]"

Quick Start

from ailitellm import ai, ailite_model

# Simple completion
response = ai("What is the capital of France?")
print(response.choices[0].message.content)

# Using a specific model
response = ai(
    "Explain quantum computing",
    model=ailite_model("Qwen/Qwen2.5-72B-Instruct")
)

Available Models

AILiteLLM supports the following models:

• Qwen/Qwen2.5-72B-Instruct - Large general purpose model
• Qwen/QwQ-32B-Preview - Preview version of QwQ model
• Qwen/Qwen2.5-Coder-32B-Instruct - Specialized for coding tasks
• NousResearch/Hermes-3-Llama-3.1-8B - Efficient general purpose model
• microsoft/Phi-3.5-mini-instruct - Lightweight instruction-following model

Advanced Usage

Chat Completions

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Tell me a joke about programming."}
]

response = ai(messages)

Streaming Responses

for chunk in ai("Write a poem about AI", stream=True):
    print(chunk.choices[0].delta.content, end="")

Function Calling

def get_weather(location: str, unit: str = "celsius") -> str:
    """Get the weather for a location"""
    pass

response = ai(
    "What's the weather in London?",
    tools=[get_weather],
    tool_choice="auto"
)

Advanced Parameters

response = ai(
    "Generate a creative story",
    temperature=0.8,
    max_tokens=500,
    top_p=0.9,
    presence_penalty=0.6
)

Custom Client

from ailitellm import AILite

custom_client = AILite(
    base_url="your_custom_endpoint",
    api_key="your_api_key"
)

Error Handling

try:
    response = ai("Your prompt here")
except Exception as e:
    print(f"An error occurred: {e}")

API Reference

Main Functions

ai(messages_or_prompt, **kwargs)

Main interface for generating completions.

Key parameters:

• messages_or_prompt: List of messages or string prompt
• model: Model to use (default: "Qwen/Qwen2.5-72B-Instruct")
• temperature: Sampling temperature (default: 0)
• max_tokens: Maximum tokens to generate
• stream: Enable streaming responses
• tools: List of functions for tool calling
• See source code for full list of parameters

ailite_model(model: HFModelType)

Helper function to specify model type with proper type checking.

Classes

AILite

Custom client class extending OpenAI's base client.

🧑‍💻 Development

To set up the development environment:

# Clone the repository
git clone https://github.com/yourusername/ailitellm.git
cd ailitellm

# Create a virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/

📦 Dependencies

• Python >= 3.8
• openai >= 1.0.0
• httpx
• typing-extensions

Contributing

Contributions are welcome! Please feel free to submit a Pull Request. See CONTRIBUTING.md for guidelines.

📝 Citation

If you use AILiteLLM in your research, please cite:

@software{ailitellm2024,
  author = {Your Name},
  title = {AILiteLLM: OpenAI-compatible Interface for Hugging Face Models},
  year = {2024},
  publisher = {GitHub},
  url = {https://github.com/yourusername/ailitellm}
}

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• OpenAI for the API interface design
• Hugging Face for model hosting and inference API
• All model creators and contributors

📫 Contact

• GitHub Issues: https://github.com/yourusername/ailitellm/issues
• Email: [e-mail]
• Twitter: @yourusername