pytc-client

Docker installation instructions

# Pull the image (tag 0.0.1 -- subject to change in future versions)
docker pull xbalbinus/pytc-client:0.0.1
# Expose ports to run the backend servers on Docker
docker run -it -p 4242:4242 -p 4243:4243 -p 4244:4244 -p 6006:6006 --shm-size=8g xbalbinus/pytc-client:0.0.1

Installation

0. Create a Virtual Environment via. Conda

conda create -n pytc python=3.9
conda activate pytc
conda install pytorch torchvision cudatoolkit=11.3 -c pytorch

Alternatively, dependencies can be installed with native Python via. the following:

# Create a venv
python -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install torch torchvision cuda-python

1. Client

cd client
npm install

2. API Server:

cd server_api
pip install -r requirements.txt

3. Pytc-connectomics:

In root folder,

git clone https://github.com/zudi-lin/pytorch_connectomics.git
cd pytorch_connectomics
pip install --editable .

Run Project

To run

./start.sh

In a separate terminal

cd client
npm run electron

Next, please move the image and labels that you'd like to train your models off of into the samples_pytc folder. Afterwards, upload the images as per the prompts on the applicaation.

Below is a link to a video demo: showing how to set up and run the app: https://www.loom.com/share/45c09b36bf37408fb3e5a9172e427deb?sid=2777bf8f-a705-4d47-b17a-adf882994168