judgeabook's Introduction

Judge a Book

First of all, you will copy the env.sample.yaml to env.yaml and substitute all variables as needed.

For the file .env.sample, copy to .env and substitute the values as needed.

The images for the interface are located in $ASSETS_PATH, by default ./judgeabook/interface/assets.

Running locally

# start the api:
make run_api

# open local
streamlit run judgeabook/interface/interface.py --server.port $FRONTEND_PORT --theme.backgroundColor $BACKGROUND_COLOR

Docker Build

There is only one docker images for this project, ${GCR_REGION}/${GCP_PROJECT}/${GCR_IMAGE}:0.0.1. To build it to run on GCP, use the following command:

docker build --platform linux/amd64 \
  -t "${GCR_REGION}/${GCP_PROJECT}/${GCR_IMAGE}:0.0.1" .
docker push "${GCR_REGION}/${GCP_PROJECT}/${GCR_IMAGE}:0.0.1"

Deploying on Google Cloud

Deploying backend:

gcloud run deploy judgeabook-backend --image ${GCR_REGION}/${GCP_PROJECT}/${GCR_IMAGE}:0.0.1 \
  --memory 8Gi --cpu 2 --region ${GCP_REGION} --env-vars-file env.yaml

After the command above is completed, the output will show the service url, that must be added to the env.yaml file on the variable SERVICE_URL.

Deploying frontend:

gcloud run deploy judgeabook-frontend --image ${GCR_REGION}/${GCP_PROJECT}/${GCR_IMAGE}:0.0.1 \
  --memory 2Gi --region ${GCP_REGION} --env-vars-file env.yaml \
  --command=streamlit,run,--server.port,8080,--theme.backgroundColor,"${BACKGROUND_COLOR}",judgeabook/interface/interface.py

The output will show you the service url, copy the service url and paste on your browser. The streamlit page should appear, and you can upload a picture to test the application.

You can see your cloud run instances in https://console.cloud.google.com/run. Click on the frontend instance to get the URL if needed.

Recommend Projects