| title: ChatBOT | |
| emoji: 🌍 | |
| colorFrom: blue | |
| colorTo: gray | |
| sdk: gradio | |
| sdk_version: 3.33.1 | |
| app_file: app.py | |
| pinned: false | |
| license: unknown | |
| # Document Retrieval Augmented Language Model version 2.0 with LangChain and Meta's LLaMA-2.0 Chat. | |
| ## Description | |
| This project involves the creation of a vector database using OpenAI embeddings and Chroma DB, followed by the retrieval of document snippets through a similarity search with LangChain's retrieval system. Upon retrieval of relevant snippets, the system uses LLaMA-2.0 to generate responses to input questions using the retrieved snippets as context. The system also incorporates a ConversationBufferMemory to store the memory of the chat, enhancing the quality of the conversational context and the relevance of generated responses. | |
| ## Contents | |
| 1. **OpenAI Embeddings and Chroma DB**: Utilizes the rich semantic information in OpenAI embeddings and the efficient storage and retrieval capabilities of Chroma DB to create a performant and effective vector database. | |
| 2. **Document Retrieval**: Uses LangChain's retrieval system to perform similarity search and retrieve relevant snippets from documents based on input queries. | |
| 3. **Response Generation with LLaMA-2.0**: Leverages the advanced language understanding and generation capabilities of LLaMA-2.0 to generate responses to input questions using Langchain's `RetrievalQA`. | |
| 4. **ConversationBufferMemory**: Stores the history of the conversation to ensure context continuity and enhance the relevance of the responses generated. | |
| ## Getting Started | |
| ### Prerequisites | |
| Before you begin, ensure you have met the following requirements: | |
| - You have installed Python 3.x. | |
| - You have access to Meta's LLaMA-2.0 and relevant API credentials. | |
| - You have set up Chroma DB on your server/machine, and the documents in the database. | |
| - You have access to LangChain's retrieval system. | |
| ### Usage | |
| After installation, you can use the system via command line or GUI through **gradio** `app.py`. | |