A: 0.3.3-preview2

Change log

• update to latest avaloniaUI-0.9.2 -> avaloniaUI-0.9.4
• update to latest netcoreapp2.2 -> netcoreapp3.1
• one binary for cpu and gpu inference
• runtime unification for Linux distributions
• fix multiple bugs with osx
• fix multiple bugs Linux
• better performance
• better ui with Citrus.Avalonia (thanks @worldbeater for awesome styles)
• add wizard mode
• add ability to change ml-model configurations
• close #8 #14 #17 #38 #40

Ussage

system requirements
CPU support for Windows/Linux/CentOS/MacOS X
GPU support ONLY for linux, only nvidia graphic support

OS: Windows 10 / Lunux / MacOS X (x64 only!).
CPU: 2x+ Core CPU (with AVX, SSE SSE2 SSE3 SSE4 SSE4.1) Intel Core i3/i5/i7/xeon Sandy Bridge / AMD Buldozer and higher (Note: intel celeron does not support!)
GPU (Optional): GTX, 4Gb vRAM, CUDA 9.1+ comparable (including CUDA 10 and higher), no CUDA driver required, e.g. GTX 950m and higher
RAM: 4096 mb RAM and higher
Storage: 5gb free disk space (10gb free disk space for GPU version)

1. Installation

CPU

• install docker and docker service (and add user to docker group for linux)
• unzip archive with your runtime

GPU

• install docker and docker service (and add user to docker group for linux)
• install nvidia-docker and run it
• unzip archive with your runtime

2. Usage

Linux\CentOS\OSX

cd /directory/with/app/
./ReacuerLaApp

Windows

go to directory with app
run ReacuerLaApp.exe

Supported Runtimes (x64 only!)

• win10 - windows 10 pro (x64 only!)
• linux - most computer linux distributions (x64 only!), such as Ubuntu, CentOS, Debian, Fedora, Arch, Gentoo and their derivatives
• osx - macOS 10.12 Sierra or higher (x64 only!)