RAG-nificent v0.0.3 - Llama 3.1 is here!

I'm excited to announce the release of RAG-nificent v0.0.3, featuring the latest Meta AI models to elevate your AI-powered chatbot experience.

What's New

Expanded Model Selection

RAG-nificent now supports the just released Llama-3.1 models from Meta AI, offering users even more flexibility and performance. The new models include:

• 🦙 llama-3.1-8b-instant: Optimized for quick and efficient responses.
• 🦙 llama-3.1-70b-versatile: Designed for versatile and complex language tasks.

These additions, powered by the Groq API, ensure lightning-fast responses, enhancing your overall user experience.

Full Changelog: v0.0.2...v0.0.3

LLama 3 is here & Groq API - Release Notes for RAG-nificent v0.0.2

I'm thrilled to announce the release of RAG-nificent v0.0.2, an upgrade packed with exciting new features designed to enhance the functionality and performance of the AI-powered chatbot system. This release focuses on expanding the model selection for retrieval and generation tasks and introducing a new user-controlled feature to adjust response creativity.

What's New

Expanded Model Selection

RAG-nificent now supports a wider range of language models, providing users with the flexibility to choose the most suitable AI for their specific needs. The new lineup includes:

• 📘 ChatGPT-3.5: Known for its engaging and human-like interactions, ideal for a variety of conversational applications.
• 🦙 Llama3-70B-8192: Delivers high-end performance for complex language tasks and deep learning insights.
• 🦙 Llama3-8B-8192: Offers robust capabilities in a more accessible form, perfect for diverse AI applications.
• 🌟 Mixtral-8x7B-32768: Utilizes ensemble modeling for nuanced understanding and superior response generation.
• 🦙 Llama2-70B-4096: A proven model for comprehensive language processing.
• 💎 Gemma-7B-IT: Tailored for IT and technical content, providing specialized interactions.

A special highlight is the new Llama70B model by Meta, added to our arsenal to ensure top-tier performance and insights. All models are now powered by the Groq API, the fastest inference LPU available, ensuring lightning-fast responses that enhance user experience.

Temperature Slider

A new "temperature" slider has been integrated into RAG-nificent, allowing users to adjust the "temperature" setting of the language models. This setting modifies the level of creativity and variability in the responses generated by the AI. A lower temperature results in more deterministic and predictable responses, while a higher temperature makes the AI's responses more diverse and potentially more creative. This feature adds a layer of customization, giving users more control over the interaction style of the chatbot.

Enhancements

• Integration of Groq API for all supported models to significantly boost response times and throughput.
• Improved user interface with a simple control for adjusting the temperature setting, enhancing interactivity and user experience.

Release Notes for RAG-nificent v0.0.1

I am excited to announce the first release of RAG-nificent, a cutting-edge chatbot system designed to revolutionize the way researchers, policy makers, and the public access and interact with key PDF documents. This initial release, version 0.0.1, introduces a robust platform built upon the latest advancements in Retrieval-Augmented Generation (RAG) and AI technologies.

Key Features

• Conversational Interface: Engage naturally with RAG-nificent to retrieve precise information from a curated collection of PDF documents. The interface is designed to understand and respond to queries with high relevance and accuracy.

• Direct Citation: Responses are not just informative—they are verifiable. Each answer provided by RAG-nificent includes a direct link to the source PDF page, ensuring that users can trace the origin of the information.

• PDF Directory: At launch, the directory includes important WHO recommendations on major health topics such as schistosomiasis and malaria, with plans to expand to other documents based on user needs and feedback.

Enhancements and Fixes

• Implemented OpenAI embeddings to convert user queries into a format suitable for deep learning models.
• Integrated Pinecone vector search to efficiently locate the most relevant sections across the document repository.
• Developed a smooth setup process to facilitate easy installation and configuration of the RAG-nificent environment on local systems.

Setup Instructions

• Clone the repository and install required dependencies to get started with ease.
• Detailed setup instructions are provided to ensure users can configure and run the application without needing extensive technical knowledge.