dlaima/summarisation

1import gradio as gr2from transformers import pipeline3 4# Load the summarization pipeline5pipe = pipeline("summarization", model="facebook/bart-large-cnn")6 7# Define the summarization function8def summarize_text(text, max_length=130, min_length=30, length_penalty=2.0):9    response = pipe(10        text,11        max_length=max_length,12        min_length=min_length,13        length_penalty=length_penalty,14        truncation=True15    )16    return response[0]['summary_text']17 18# Create the Gradio app interface19with gr.Blocks() as app:20    gr.Markdown("## Text Summarization App")21    gr.Markdown(22        "Enter a long text below, and the model will generate a concise summary. "23        "This app uses the `facebook/bart-large-cnn` model."24    )25 26    with gr.Row():27        input_text = gr.Textbox(28            label="Input Text",29            placeholder="Paste your text here...",30            lines=1031        )32        output_summary = gr.Textbox(label="Summary", lines=5)33 34    max_length = gr.Slider(35        label="Max Length", 36        minimum=50, 37        maximum=200, 38        step=10, 39        value=13040    )41    min_length = gr.Slider(42        label="Min Length", 43        minimum=10, 44        maximum=100, 45        step=10, 46        value=3047    )48    length_penalty = gr.Slider(49        label="Length Penalty", 50        minimum=0.5, 51        maximum=3.0, 52        step=0.1, 53        value=2.054    )55 56    submit_button = gr.Button("Summarize")57    58    submit_button.click(59        fn=summarize_text,60        inputs=[input_text, max_length, min_length, length_penalty],61        outputs=output_summary62    )63 64# Launch the app65if __name__ == "__main__":66    app.launch()67