app.py

#Initial installations
pip uninstall -y tensorflow
pip install tensorflow==2.14

pip install --upgrade pip
pip install --upgrade transformers scipy

pip install transformers
pip install pymupdf

## Summarization
import gradio as gr
import fitz  # PyMuPDF
from transformers import BartTokenizer, BartForConditionalGeneration, pipeline
import scipy.io.wavfile
import numpy as np

tokenizer = BartTokenizer.from_pretrained('facebook/bart-large-cnn')
model = BartForConditionalGeneration.from_pretrained('facebook/bart-large-cnn')

def extract_abstract(pdf_path):
    doc = fitz.open(pdf_path)
    first_page = doc[0].get_text()
    start_idx = first_page.lower().find("abstract")
    end_idx = first_page.lower().find("introduction")
    if start_idx != -1 and end_idx != -1:
        return first_page[start_idx:end_idx].strip()
    else:
        return "Abstract not found or '1 Introduction' not found in the first page."

# Specify the path to your PDF file
pdf_path = "/content/article11.pdf"  # Update the path

# Extract the abstract
abstract_text = extract_abstract(pdf_path)

# Print the extracted abstract
print("Extracted Abstract:")
print(abstract_text)


from IPython.core.display import display, HTML

# Function to display summary and reduction percentage aesthetically
def display_results(final_summary, original_text):
    reduction_percentage = 100 * (1 - len(final_summary) / len(original_text))
    html_content = f"""
    <div style='padding: 20px; background-color: #f3f3f3; border-radius: 10px;'>
        <h2 style='color: #2c3e50; text-align: center;'>Summary</h2>
        <p style='color: #34495e; font-size: 16px; text-align: justify;'>{final_summary}</p>
        <p style='color: #2c3e50;'><b>Reduction in Text:</b> {reduction_percentage:.2f}%</p>
    </div>
    """
    display(HTML(html_content))

# Summary generation and post-processing
inputs = tokenizer([abstract_text], max_length=1024, return_tensors='pt', truncation=True)
max_length_for_summary = 40
length_penalty_value = 2.0

summary_ids = model.generate(inputs['input_ids'],
                             num_beams=4,
                             max_length=max_length_for_summary,
                             min_length=10,
                             length_penalty=length_penalty_value,
                             early_stopping=True,
                             no_repeat_ngram_size=2)

summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
summary = ' '.join(summary.split())  # Remove extra spaces

# Handle truncated words and adjust periods
words = summary.split()
cleaned_summary = []
for i, word in enumerate(words):
    if '-' in word and i < len(words) - 1:
        word = word.replace('-', '') + words[i + 1]
        words[i + 1] = ""

    if '.' in word and i != len(words) - 1:
        word = word.replace('.', '')
        cleaned_summary.append(word + ' and')
    else:
        cleaned_summary.append(word)

# Capitalize first word and adjust following words
final_summary = ' '.join(cleaned_summary)
final_summary = final_summary[0].upper() + final_summary[1:]
final_summary = ' '.join(w[0].lower() + w[1:] if w.lower() != 'and' else w for w in final_summary.split())

# Displaying the results
display_results(final_summary, abstract_text)


##Text-to-Speech

# Initialize the Bark TTS pipeline
synthesiser = pipeline("text-to-speech", "suno/bark")

# Initialize the Bark TTS pipeline
synthesiser = pipeline("text-to-speech", "suno/bark")

# Convert the summarized text to speech
speech = synthesiser(final_summary, forward_params={"do_sample": True})

# Normalize the audio data
audio_data = speech["audio"].squeeze()
normalized_audio_data = np.int16(audio_data / np.max(np.abs(audio_data)) * 32767)

# Save the normalized audio data as a WAV file
output_file = "/content/bark_output.wav"
scipy.io.wavfile.write(output_file, rate=speech["sampling_rate"], data=normalized_audio_data)
print(f"Audio file saved as {output_file}")

# Display an audio player widget to play the generated speech
Audio(output_file)

# Gradio Interface
iface = gr.Interface(
    fn=process_text,
    inputs="text",
    outputs=["text", "audio"],
    title="Summarization and Text-to-Speech",
    description="Enter text to summarize and convert to speech."
)

if __name__ == "__main__":
    iface.launch()