app.py

import torch
import gradio as gr
from accelerate import Accelerator
from transformers import pipeline

# Initialize the accelerator
accelerator = Accelerator()

# Print a description of the current configuration
print("Accelerator State:", accelerator.state)

# Define the path to your custom model or use a pre-trained Hugging Face model
model_path = (
    "../Models/models--sshleifer--distilbart-cnn-12-6/snapshots"
    "/a4f8f3ea906ed274767e9906dbaede7531d660ff"
)

# Initialize the text summarization pipeline for CPU usage
try:
    text_summary = pipeline(
        "summarization",
        model=model_path,  # Use the custom model path
        device=-1  # Force usage of CPU
    )
except Exception as e:
    print(f"Error initializing the summarization pipeline: {e}")
    print("Switching to the default model from Hugging Face.")
    # Fallback to default model from Hugging Face if the custom model fails
    text_summary = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6", device=-1)

# Define the Gradio interface function
def summary(input_text):
    try:
        # Generate summary using the text_summary pipeline
        output = text_summary(input_text)
        return output[0]['summary_text']
    except Exception as e:
        return f"An error occurred while summarizing: {e}"

# Close any existing Gradio interfaces
gr.close_all()

# Define the Gradio interface
demo = gr.Interface(
    fn=summary,
    inputs=gr.Textbox(label="Input Text to Summarize", lines=6, placeholder="Enter text here..."),
    outputs=gr.Textbox(label="Summarized Text", lines=4),
    title="GenAIProject01: Text Summarizer",
    description="THIS APPLICATION WILL BE USED TO SUMMARIZE INPUT TEXT USING A PRE-TRAINED MODEL."
)

# Launch the Gradio app
demo.launch(share=True)