An Android App recreating the Simon Says game. Uses MediaPipe to run an LLM on device

An Android App recreating the Simon Says game. Uses MediaPipe to run an LLM on device using RAG

FileChat-RAG is a simple Retrieval-Augmented Generation (RAG) system that allows users to ask questions about the contents of various file formats. It extracts text from PDFs, JSON, text files(.txt, .docx, .odt, .md), and code files, then enables interactive conversations using an LLM powered by Ollama.

🎮 Build your own Simon Says game on Android, leveraging MediaPipe for on-device LLM and RAG capabilities, enhancing interactivity and fun.

Educational benchmark suite for NVIDIA V100: FP16/FP32 compute analysis, precision quantization testing, and transformer inference profiling.

A Streamlit-based AI application that analyzes resumes, compares them with job descriptions, and provides smart, practical advice for improving job fit and highlighting relevant skills using AI-powered language models.

A sleek web-based chat interface powered by your local Ollama models. includes a professional model selector and per-request model control. Supports streaming responses, stop generation, markdown rendering, and code syntax highlighting.

A powerful AI chat application built with Streamlit, LangChain, and Ollama, powered by the Gemma3:1B model. Engage in real-time conversations or ask questions about PDF documents with embedded OCR and semantic search capabilities.

A simple chatbot built with Ollama running the gemma3:1b model and a Streamlit UI. This project lets you run a local chatbot with conversation history, multi-chat support, and an interactive interface.

Gemma3-1B fine-tuned to provide emotional support for Tamil and English mental- health queries.

Cogno 1B is a 1.1 billion-parameter decoder-only transformer in the Gemma 3 family, fine-tuned on over 3 million tokens drawn from University of the Western Cape (UWC) materials.