Self-hosted, local only NVR and AI Computer Vision software. With features such as object detection, motion detection, face recognition and more, it gives you the power to keep an eye on your home, office or any other place you want to monitor.

Brevitas: neural network quantization in PyTorch

QKeras: a quantization deep learning library for Tensorflow Keras

Algebraic enhancements for GEMM & AI accelerators

Object detection for video surveillance

Awesome Mojo🔥

Hashed Lookup Table based Matrix Multiplication (halutmatmul) - Stella Nera accelerator

Mayo: Auto-generation of hardware-friendly deep neural networks. Dynamic Channel Pruning: Feature Boosting and Suppression.

An open-source parameterizable NPU generator with full-stack multi-target compilation stack for intelligent workloads.

Installing hardware-accelerated PyTorch with Poetry on different hardware using the same `pyproject.toml`

Implementation of NEWMA: a new method for scalable model-free online change-point detection

Open source RTL simulation acceleration on commodity hardware

Imperative deep learning framework with customized GPU and CPU backend

ML benchmarks performance featuring LightOn's Optical Processing Unit (OPU) vs CPU and GPU.

Client for the Jellyfin Hardware Survey Server (https://hwa.jellyfin.org/). Benchmarks hardware performance for simultaneous ffmpeg transcoding, enabling detailed comparisons and insights for optimizing Jellyfin setups.

MobileNet trained with VoxCeleb dataset and used for voice verification

Model LLM inference on single-core dataflow accelerators

Code used in the paper “Nonideality-Aware Training for Accurate and Robust Low-Power Memristive Neural Networks”