Spaces:
Sleeping
Sleeping
import gradio as gr | |
import spaces | |
from transformers import pipeline | |
# Initialize Model | |
get_completion = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6", device=0) | |
def summarize(input: str) -> str: | |
""" | |
Summarize the given input text using the sshleifer/distilbart-cnn-12-6 model. | |
Args: | |
input (str): The text to be summarized. | |
Returns: | |
str: The summarized version of the input text. | |
""" | |
output: List[Dict[str, str]] = get_completion(input) | |
return output[0]['summary_text'] | |
gr.close_all() | |
####### GRADIO APP ####### | |
title = """<h1 id="title"> Text Summarization </h1>""" | |
description = """ | |
Summarize any text using the `sshleifer/distilbart-cnn-12-6` model under the hood | |
- The model used for Summarizing Text [DISTILBART-12-6-CNN](https://huggingface.co/sshleifer/distilbart-cnn-12-6). | |
""" | |
css = ''' | |
h1#title { | |
text-align: center; | |
} | |
''' | |
theme = gr.themes.Soft() | |
demo = gr.Blocks(css=css, theme=theme) | |
with demo: | |
gr.Markdown(title) | |
gr.Markdown(description) | |
interface = gr.Interface(fn=summarize, | |
inputs=[gr.Textbox(label="Text to Summarize", lines=15)], | |
outputs=[gr.Textbox(label="Result", lines=7)], | |
examples=["""Artificial Intelligence (AI) has been a rapidly growing field over the past decade, transforming various industries such as healthcare, finance, and transportation. | |
One of the key areas of AI research is machine learning, which focuses on developing algorithms that allow computers to learn from and make decisions based on data. | |
Recent advancements in deep learning, a subset of machine learning, have led to significant breakthroughs in image and speech recognition, natural language processing, and autonomous systems. | |
As AI continues to evolve, it is expected to play an increasingly important role in solving complex problems, enhancing productivity, and driving innovation across different sectors.""" | |
], | |
cache_examples=True | |
) | |
demo.launch() |