Installtion

Pre-Install

1. Clone the repo

2. Copy the weight file into data directory inside this repo, weights files will be mounted inside container at /data/<your_weights_name>.hdf5

3. For annotation, copy the required folders

Using Linux OS

apt install docker.io
sudo docker build -t asksina_app:1.0 . 
sudo docker run --name asksina -p 8080:8080 -d asksina_app:1.0

then access application on localhost:8080/ai/login

using Mac / windows

Install Docker as follows:

MacOS: https://docs.docker.com/docker-for-mac/install/

Microsoft Windows: https://docs.docker.com/docker-for-windows/install/

On Windows, make sure the docker is running with Linux containers and Settings>Daemon>Experimental features is checked. Finally restart the docker

Run the following commands from Powershell/Terminal

docker build -t asksina_app:1.1 . 
docker run --name asksina -p 8080:8080 -d asksina_app:1.1

then access application on localhost:8080/ai/login

Mount volumes

if you want to mount another weights file then ( otherwise you don't need this step if you loaded weights earlier in docker build"

make sure weights are at /data on your machine so it can be mounted inside container

on Linux/MacOS sh sudo docker run --name asksina -v /data:/data -p 8080:8080 -d asksina_app:1.0

on Windows

if it is on "C:/data then docker run --name test1 -v c:/data:/data -p 8080:8080 -d asksina_app:1.0

otherwise replace "c:/data" with your other drive like "d:/data"

Using GPU on Linux

Setup CUDA and cudaNN

apt-get install -y nvidia-container-runtime-hook <--- this is run time by nvidia

sudo docker run --name asksina --runtime=nvidia -v /data:/data -p 8080:8080 -d app:1.0