Popular GitHub Repositories for Avx 512

Fast inference engine for Transformer models

oneAPI Deep Neural Network Library (oneDNN)

C++ image processing and machine learning library with using of SIMD: SSE, AVX, AVX-512, AMX for x86/x64, NEON for ARM.

TensorFlow binaries supporting AVX, FMA, SSE

Library for specialized dense and sparse matrix operations, and deep learning primitives.

Performance-optimized wheels for TensorFlow (SSE, AVX, FMA, XLA, MPI)

FlyCV is a high-performance library for processing computer visual tasks.

BERT classification model for processing texts longer than 512 tokens. Text is first divided into smaller chunks and after feeding them to BERT, intermediate results are pooled. The implementation allows fine-tuning.

deepfake-detector-model-v1 is a vision-language encoder model fine-tuned from siglip2-base-patch16-512 for binary deepfake image classification. It is trained to detect whether an image is real or generated using synthetic media techniques. The model uses the SiglipForImageClassification architecture.

An optimized Tensor/Matrix library for HPC applications an numerical relativity (AVX2/AVX512 and thread safe) with a Symbolic parser and Python wrapper

Document Level Sentiment Analysis is an End-to-End deep learning workflow using Hugging Face transformers API to do a "classification" task at document level, to analyze the sentiment of input document containing English sentences or paragraphs.

Tensorflow for MacOS compiled without AVX.

This repository contains a Master's Thesis on enhancing LiDAR-based 3D Object Detection in autonomous vehicles using synthetic data. The research explores the use of the Ansys AVxcelerate Sensors Simulator (AVX) to create synthetic point clouds. It includes a hybrid training approach, experimental design, results, and related scripts.

A simple SQLite extension that allows direct GPU and/or SIMD accelerated vector operations

siglip2-mini-explicit-content is an image classification vision-language encoder model fine-tuned from siglip2-base-patch16-512 for a single-label classification task. It is designed to classify images into categories related to explicit, sensual, or safe-for-work content using the SiglipForImageClassification architecture.

Yet Another Machine Inference framework

A simple backend for LLM inference written in C++17, Metal (for GPU) and AVX/NEON (for CPU).

This repository contains supplementary materials for a Master Thesis project on efficient object detection. It includes a series of video demonstrations showcasing the performance of networks trained and tested on synthetic (AVX) and real-world (KITTI) data sets.

DepthLens: This is a demo for DPT Beit-Large-512 used to estimate the depth of objects in images.

open-scene-detection is a vision-language encoder model fine-tuned from siglip2-base-patch16-512 for multi-class scene classification. It is trained to recognize and categorize natural and urban scenes using a curated visual dataset. The model uses the SiglipForImageClassification architecture.