# deploy-gradio-ml-on-gcp

A demo ML application that provides first aid information to people diagnosed with internal bleeding.A simple ML application built with HuggingFace and Gradio Deployed on GCP

## Live Demo

> https://ml-gradio-demo.ew.r.appspot.com/

## KaggleX Mentee - My Experience

> [LinkedIn Post - My experience as a KaggleX Mentee](https://www.linkedin.com/posts/emmanuel-katchy_kaggle-kagglex-bipoc-activity-7126311581159727105-QzS_?utm_source=share&utm_medium=member_desktop)

## KaggleX Program Website

> https://www.kaggle.com/kagglex

## Helpful Tutorials

 - https://medium.com/@mehmetcanfarsak/ci-cd-using-github-actions-to-deploy-on-gcp-app-engine-in-3-simple-steps-399fa9764cf2

### Notes/Todos:

  - You need to disable Gradio flagging in GCloud because the environment is read-only

  - Allow users to upload scan images, and perform live inference

  - To be able to use this template/project, create a blank `env.yaml` locally. Environment variables can then be defined as needed.

    An example:

  ```yaml

  HUGGING_FACE_TOKEN=<YOUR_API_KEY>

  ```