This project is a demo of interpreting text classification model using Tensorflow 2 and Streamlit.
IMDB dataset contains 50k movie reviews and each of them is labeled as either "positive" or "negative".
Data source: Kaggle
Build the docker container of this app by
docker build -t x-text-classification .
Run the container by
docker run -p 8501:8501 x-text-classification
# the app should be up and run on localhost:8501
Open the GCP cloud shell or local terminal with gcloud installed, clone this repo and cd into the project root directory, then
# build and push to Container Registry
gcloud builds submit --tag gcr.io/<PROJECT_ID>/x-text-classification:v1.0
# deploy it to Cloud Run
gcloud run deploy --image=gcr.io/<PROJECT_ID>/x-text-classification:v1.0 --port 8501 --memory 2Gi