app.py

# This Gradio app uses the Hugging Face model `google/derm-foundation` to classify skin conditions.
import gradio as gr
from transformers_js_py import pipeline  # Use transformers_js_py instead of transformers

# Load the Hugging Face model for skin condition classification
model = pipeline("image-classification", model="google/derm-foundation")

# Define a function to classify an image using the model
def classify_skin_condition(image):
    # Run the image through the model
    result = model(image)
    # Extract the top prediction
    top_prediction = result[0]
    # Return the label and confidence score
    return f"Condition: {top_prediction['label']}, Confidence: {top_prediction['score']:.2f}"

# Create a Gradio interface that takes an image input, runs it through the classify_skin_condition function, and returns the output to a textbox.
demo = gr.Interface(fn=classify_skin_condition, inputs="image", outputs="textbox")

# Launch the interface.
if __name__ == "__main__":
    demo.launch(show_error=True)