Advanced debugging

This guide explains how to debug your application

Introduction

In this tutorial we are going to show you how to run application in debug mode and interact with it from Supervisely platform


Prepare environment

Step 1. Prepare ~/supervisely.env file with credentials. Learn more here.

Step 2. Clone repository with source code and demo data and create Virtual Environment.

git clone https://github.com/supervisely-ecosystem/integrate-inst-seg-model
cd integrate-inst-seg-model
./create_venv.sh

Step 3. Install wireguard-tools

sudo apt-get install wireguard iproute2

Step 4. Open repository directory in Visual Studio Code.

code -r .

Step 5. Change TEAM_ID of "Advanced mode for Supervisely Team" configuration in .vscode/launch.json file by copying the ID from the context menu of the team. A new debug app session will be created for the team you define:

{
    "name": "Advanced mode for Supervisely Team",
    ***
    "env": {
        ***
        "TEAM_ID": "7", // ⬅️ change value
        ***
    }
}

Step 6. Copy contents of my_model directory to results and download model weights.

mkdir -p results/model
cp -R my_model/* results/model
cd results/model
./init.sh
cd ../..

Run debug

Step 1. Go to Run and Debug section (Ctrl+Shift+D).

Step 2. Select Advanced mode for Supervisely Team from configuration dropdown

Step 3. Press green triangle or F5 to start debugging. You should see Debug task has been successfully created: 12345 or Debug task already exists: 12345 message in console logs.

If you need to debug SDK, you can clone Supervisely repository and create a Symbolic Link to it inside your project.

cd *your_desired_dir*
git clone git@github.com:supervisely/supervisely.git
cd back to project
ln -s *your_desired_dir*/supervisely/supervisely ./supervisely

Interact with app from Supervisely platform

Step 1. Go to Supervisely platform

Step 2. Select App sessions from start menu. You will see a list of applications. Find app named Develop and Debug and click on sessions button. There you should see your debug app session with the same ID as in Step 3 of previous section.

Step 3. Now you can try the app. Choose any project or create new one e.g. Demo Images. Open dataset in Basic image labeling toolbox.

Step 4. Run NN Image Labeling app from the list and connect to your app session. Press Apply model to image button (Ctrl+m). Predictions are loaded and added to image automatically without need to refresh the page.

Step 5. You can control the execution of the application with any debugging instrument such as breakpoints.

Last updated