Benchmarks, compose files, and findings for running Qwen3.6 (27B dense + 35B-A3B MoE) on 4× RTX 3090. vLLM/AWQ vs llama.cpp/GGUF, MTP speculative decoding, single-GPU vs TP=2/TP=4 trade-offs.

Tiny, no-nonsense, self-contained, Tensorflow and ONNX inference

The best JavaScript Data Table for building Enterprise Applications. Supports React / Angular / Vue / Plain JavaScript.

The repository provides code for running inference with the SAM 3D Body Model (3DB), links for downloading the trained model checkpoints and datasets, and example notebooks that show how to use the…

SAM 3D Objects

Browser MCP is a Model Context Provider (MCP) server that allows AI applications to control your browser

Rust multi‑backend OCR/VLM engine (DeepSeek‑OCR-1/2, PaddleOCR‑VL, DotsOCR) with DSQ quantization and an OpenAI‑compatible server & CLI – run locally without Python.

A fast and secure browser for standalone virtual-reality and augmented-reality headsets.

A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization.

Official implementation code of the paper <AnyText: Multilingual Visual Text Generation And Editing>

State-of-the-art TTS model under 25MB 😻

Kitten TTS web demo using tansformers.js

WiLoR: End-to-end 3D hand localization and reconstruction in-the-wild

HaMeR: Reconstructing Hands in 3D with Transformers

🚀2.3x faster than MinIO for 4KB object payloads. RustFS is an open-source, S3-compatible high-performance object storage system supporting migration and coexistence with other S3-compatible platfor…

BoT-SORT: Robust Associations Multi-Pedestrian Tracking

MOT using deepsort and yolov3 with pytorch

Fast ML inference & training for ONNX models in Rust

Pytorch Implementation For LPRNet, A High Performance And Lightweight License Plate Recognition Framework.

Simple Online Realtime Tracking with a Deep Association Metric

Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set (CVPRW 2019)

Pre-trained Deep Learning models and demos (high quality and extremely fast)

High-Resolution 3D Assets Generation with Large Scale Hunyuan3D Diffusion Models.

FFSVM stands for "Really Fast Support Vector Machine"

"Effective Whole-body Pose Estimation with Two-stages Distillation" (ICCV 2023, CV4Metaverse Workshop)

Empowering everyone to build reliable and efficient software.

Very low latency speech to text, intent recognition, and text to speech, for building voice agents and interfaces

Bring portraits to life!