I created this repository as an interest to how well apple metal will do in a sample deep learning models opposed to Nvidia Cuda

Post-Training Quantization for MeshGraphNet Physics-Based ML Model: Cardiovascular Flow Simulation Implementation

The Turing Board is a concept autonomous longboard that is able to exhibit self-driving capabilities using computer vision.

Finetuning a Local LLM Gemma 2 2B using Unsloth and your own custom dataset for Custom Attribute extraction from an unstructured content

This script collects some informations about NVLink and PCI bus traffic of NVidia GPUs. Results are published as prometheus metrics via a websocket.

Realtime Monitor of Nvidia GPU Metrics with NVML Library

This project analyzes GPU performance metrics from AI training and inference tasks, highlighting the optimal power settings for efficient operation. It generates visualizations to help identify the best max power settings for training and inference based on energy consumption.

Real-time Embedded Deep Learning for Autonomous Lane Change Systems of Autonomous Driving

my thesis works on mri image segmentation of brain tumour using deep learning models

AlphAI is a versatile Python toolkit for GPU profiling and analytics, supporting various tensor models. It enhances GPU server operations and serves as a client for American Data Science's notebook servers.

Automatic transcriber made with the Nvidia NeMo AI toolkit. Used to transcribe speech to text in real-time from any source. Requires CUDA capable GPU to run on the local machine, if setup using virtual audio cables can transcribe the audio that is being played in real-time without any other requirements.

This system tracks artifacts in museum and triggers alarm if artifact goes missing from the frame.

Hand/Face detection model using python (No hardware(Arduino, sensors required)

🔥 Unleash the Power of Computer Vision!

A pre-configured instant-ngp workspace that includes helpful scripts for getting started with NeRF training.

GenAI playground maximizing the use of open-source software and models

for nvidia graphic metrics on datadog

Large scale GPU accelerated k-means algorithm

Python script for collectd, wrapping system calls to nvidia-smi.

Detecting vehicles and warn them if they get close to another vehicle.