main.py

import gradio as gr
from fastcore.all import *
from fastai.vision.all import *

# Load the FastAI model
learn = load_learner('export.pkl')
labels = learn.dls.vocab
# Define a function to classify an image
def classify_image(file):

    # Run the model to get a prediction
    pred_class, pred_idx, outputs = learn.predict(file)
    # Return the predicted class
    return {labels[i]: float(outputs[i]) for i in range(len(labels))}

# Create a Gradio interface
iface = gr.Interface(
    fn=classify_image,
    inputs=gr.components.Image(label="Image"),
    outputs=gr.components.Label(num_top_classes=3),
    title="Llamalpaca-tron 5000",
    description="The llama-alpaca image classifier is a machine learning model designed to accurately identify whether an image contains a llama or an alpaca. Trained on a large dataset of llama and alpaca images, the model uses deep learning algorithms to analyze various features of the animals, such as their fur, body shape, and facial characteristics, and then makes a prediction based on those features. With high accuracy, this model can help identify llamas and alpacas in images, which can be useful for various applications, such as wildlife conservation, agriculture, and animal research.",
)

# Run the interface
iface.launch(share=True)