README.md

metadata

title: Llm Project
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 5.42.0
app_file: app.py
pinned: false
hf_oauth: true
hf_oauth_scopes:
  - inference-api

An example chatbot using Gradio, huggingface_hub, and the Hugging Face Inference API.

                                                AI Tutor Chatbot 🤖

AI Tutor is an intelligent chatbot application built with LLamaIndex and RAG (Retrieval-Augmented Generation) techniques. It is designed to answer questions strictly related to AI, ML, and their subfields. Questions outside of AI will not be answered.

                                               Functionalities

Metadata Filtering: Filter retrievals by keywords or metadata for more precise results.

Reranking for Accuracy: Uses Cohere Reranker to improve the quality of retrieved nodes.

Context Augmentation via Perplexity API: Enhances answers by providing additional context before generating a final response.

RAG Pipeline Evaluation: Evaluates retrieval and answer quality using Faithfulness and Answer Relevance metrics.

Additional Data Sources: Incorporates external sources beyond the course material to improve answer coverage.

                                                Cost Estimation

The chatbot is optimized for low-cost experimentation. Estimated cost to try all features: ~$0.50 or less This includes all functionalities: LLM responses, reranking, and context augmentation.

                                              Evaluation Results

The RAG system has been evaluated using faithfulness and answer relevance metrics: Metric Score Faithfulness 0.93 Answer Relevance 0.93 This indicates highly accurate, grounded, and relevant answers for AI-related queries.