Last updated:

Streamlit Applications

What is Streamlit

Streamlit is an open-source Python framework designed for building data applications and interactive dashboards. Without any frontend development experience, you can quickly create web applications with charts, tables, and interactive components using Python scripts.

Create a Streamlit Space

Follow the steps in Create Space to open the creation form.
Select Streamlit as the SDK Type.
If you selected a GPU compute resource, also choose a Driver Version (11.8.0 or 12.1.0).
Fill in the remaining parameters and click Create Application Space to submit.

Initialize the Application

After creation, push application code to the repository to initialize the Streamlit space.

Step 1: Clone the Repository

git clone https://<platform-host>/<namespace>/<space-name>
cd <space-name>

Step 2: Create the Application File

Create an app.py file with your Streamlit application code. Here is a simple data display example:

import streamlit as st

st.set_page_config(page_title="AI Platform Demo", layout="wide")

st.title("Streamlit Demo Application")
st.write("This is a simple Streamlit application deployed on the platform.")

name = st.text_input("Enter your name:")
if name:
    st.success(f"Hello, {name}! Welcome to the AI platform.")

st.subheader("Sample Data")
import pandas as pd

data = pd.DataFrame({
    "Model": ["GPT-4", "LLaMA-3", "Qwen-2"],
    "Parameters": ["1.8T", "70B", "72B"],
    "Type": ["Closed", "Open", "Open"]
})
st.dataframe(data, use_container_width=True)

st.subheader("Interactive Chart")
chart_data = pd.DataFrame({
    "Category": ["Training", "Inference", "Fine-tuning"],
    "Hours": [120, 45, 80]
})
st.bar_chart(chart_data.set_index("Category"))

If additional Python dependencies are needed, create a requirements.txt file:

pandas

Step 3: Push the Code

git add app.py
git commit -m "Initialize Streamlit application"
git push origin main

Automatic Build and Deployment

After the code is pushed, the platform automatically triggers the build and deployment process:

Installs dependencies (if requirements.txt is present).
Starts the app.py application.
Once the build is complete, the space page displays the Streamlit interface.

You can view build logs and running status on the space detail page.