Navin Kumar M's Projects
AI-Learning-Platform, a LLM-RAG pipeline which behaves like a guide and able to solve doubts. Deployed on-premise IBM ppc64le architecture. vLLM for model inference & Qdrant with Langchain for RAG Pipeline. Server written in django, postgres & cassandra as the sql & nosql databases.
Fast API Endpoint!. This project examines, analyses the malware statically & dynamically using conventional strategies and also apply machine learning algorithms lke lightgbm, svm and deep learning algorithms like CoAtNet, LSTM. FrontEnd App is Antivirus built on Tauri
Face Recognition on Edge and Cloud GPU. yolov7 to detect & arcface to recognize the faces. Built on deepstream with Azure IOT support
Real Time Detection of Anomalous Activity From Videos (mainly crime actvity). Images of the video is trained using AutoEncoder to get the imtermediate feature representation of image & applied svm model for the bag of such features to detect the anomaly & LSTM to detect the type of Anomaly.
Benetech Kaggle Competition Work. Fine Tuning Matcha (Multi Modal Transformer) on Line, Scatter, Dot, Horizontal and Vertical bar dataset.
Face Recognition from Crowd by using yolov7 .Extracting the faces from a video/image/live source, which is then passed to the custom facenet network in order to recognize the peoples
Mistral 7b - SFT on Alpaca + PEFT + DPO on HH-RLHF.
Malware Analysis using Deep Learning & Machine Learning deployed on AWS cloud. ML & DL algorithms was written in Python, Server Part written in node.js
Everything I practised in 2024 FY
My Skills : ā¢ Frameworks: PyTorch, Tensorflow, AWS, MLFlow, Ray, Tauri, RAPIDS ā¢ Languages: Python, TypeScript, C++, Rust, Node.js, MongoDB, MySQL, Tailwind CSS
IamAlwaysMega āØ. AI Engineer. Student at VIT University (Chennai Campus). Working Hard to build My Own AI Company; Interested in Humanity & Security fields of AI. "Open for Startups". Data Science, Deep Learning & MLOps Engineer
Documentation of Setting up Parallel & Distributed ML Workspace in your systems. And to work seamlessly without error. Package to easily setup a environment in the group of systems
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
Development repository for the Triton language and compiler
A high-throughput and memory-efficient inference and serving engine for LLMs