app.py

import gradio as gr

# Optional: Define a wrapper function for your model
def model_wrapper(text):
    try:
        # Load the model with custom code execution allowed
        model = gr.load("models/cerebras/btlm-3b-8k-chat")
        # Use the model to predict
        return model(text)
    except Exception as e:
        # Log the exception or handle it in a user-friendly way
        return f"An error occurred: {str(e)}"

# Setup Gradio interface
iface = gr.Interface(
    fn=model_wrapper,  # your function to expose
    inputs="text",  # input type
    outputs="text"  # output type
)

# Launch the Gradio app
iface.launch()