Streamlit Starter Template

Quick Start (Codespaces)

1. Fork this repository
2. Open in GitHub Codespaces (or press . for github.dev)
3. In terminal: pip install -r requirements.txt
4. Run: streamlit run app.py

Quick Start (Local + VSCode)

1. Install Python
2. Install VSCode
3. Install Git (and connect to Github)
4. Clone the repo

   git clone https://github.com/x/streamlit-starter.git

5. Install the requirements

   cd streamlit-starter
   pip install -r requirements.txt

6. Run the server

   streamlit run app.py

7. Your browser will open to http://localhost:8501
8. Edit by opening the streamlit-starter folder in VSCode

Structure

There are three sample applications.

app.py

This is just your basic "hello world" but with a simple input. Start here.

game.py

Here's an example of how you can turn your text adventure game into a streamlit app.

model_eval.py

Here's an example of how to take a pickled model and make an interactive evaluator using Streamlit.

Caution

In order for the data/model.pkl to load, I had to make sure the scikit-learn and numpy versions matched the Collab environment where we trained the model. Be aware that, if you are from the future, you may need to update these in requirements.txt to match the new versions of Collab (use !pip show numpy and !pip show scikit-learn to check).