SPLAA is an innovative AI assistant framework designed for a conversational experience. With capabilities like voice recognition, text-to-speech, and tool integration, it harnesses LLMs through Ollama to offer a customizable and interactive user interface, enabling seamless interactions with external tools.
SPLAA is a cutting-edge AI assistant framework designed for an engaging and interactive conversational experience. Leveraging advanced voice recognition and text-to-speech technologies, SPLAA allows users to communicate naturally and effortlessly. Powered by Ollama's LLMs, SPLAA stands out with its customizable modular tool system, enabling developers to extend its functionalities easily.
Key Features
- Voice Input: Utilizes Whisper for high-accuracy speech-to-text transcription, allowing for smooth user interaction.
- Voice Output: Incorporates XTTS v2 to generate clear, natural-sounding speech, making conversations feel human-like.
- Tool Calling: Seamlessly integrates with external tools and APIs, enabling the assistant to perform various actions and retrieve information on demand.
- Customizable Options: Personalize the assistant's name, system prompts, and models using simple command-line arguments.
Example Usage
To initiate SPLAA, simply run:
splaa --options [option] ...
Once activated, you can communicate with SPLAA by saying commands that include its name, such as:
"Athena, what's the weather like in London?"
Available Tools
Enhance your experience with built-in tools:
getWeather: Access real-time weather updates.wikipediaSearch: Retrieve summaries from Wikipedia.getNews: Fetch news articles based on specific topics.getStockPrice: Get the latest stock prices.todoList: Manage your tasks effortlessly.
Extending Functionality
Develop your custom tools by defining functions in the helperFunctions.py file. Each function should have a clear purpose, accept necessary parameters, and return appropriate data formats.
Voice Cloning Feature
Easily clone voices from YouTube clips to personalize your assistant. For optimal results, capture at least six seconds of audio with minimal background noise.
Contributing to the Project
Contributions are encouraged! Create pull requests with clear descriptions of your proposed changes to enhance SPLAA.
FAQs
Common questions about performance and troubleshooting are addressed to assist users in maximizing their SPLAA experience.
Acknowledgments
Special thanks to the Coqui TTS project for their contributions to the SPLAA framework, particularly for the Python 3.12 adaptations.
SPLAA is fundamentally changing how we interact with AI, making communication effortless and dynamic. Explore the full potential of your AI assistant with SPLAA today!