Spaces:
Running
Running
| # Initial installations handled separately (not in app.py) | |
| # Required imports | |
| import gradio as gr | |
| import fitz | |
| from transformers import BartTokenizer, BartForConditionalGeneration, pipeline | |
| import scipy.io.wavfile | |
| import numpy as np | |
| from IPython.display import Audio | |
| # Initialize tokenizers and models | |
| tokenizer = BartTokenizer.from_pretrained('facebook/bart-large-cnn') | |
| model = BartForConditionalGeneration.from_pretrained('facebook/bart-large-cnn') | |
| synthesiser = pipeline("text-to-speech", "suno/bark") | |
| # Function to extract abstract from PDF | |
| def extract_abstract(pdf_content): | |
| doc = fitz.open("pdf", pdf_content) | |
| 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." | |
| # Function to process text (summarize and convert to speech) | |
| def process_text(pdf_content): | |
| abstract_text = extract_abstract(pdf_content) | |
| # Generate summary | |
| inputs = tokenizer([abstract_text], max_length=1024, return_tensors='pt', truncation=True) | |
| summary_ids = model.generate(inputs['input_ids'], num_beams=4, max_length=40, min_length=10, length_penalty=2.0, early_stopping=True, no_repeat_ngram_size=2) | |
| summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True) | |
| # Convert summary to speech | |
| speech = synthesiser(summary, forward_params={"do_sample": True}) | |
| audio_data = speech["audio"].squeeze() | |
| normalized_audio_data = np.int16(audio_data / np.max(np.abs(audio_data)) * 32767) | |
| # Save audio to temporary file | |
| output_file = "temp_output.wav" | |
| scipy.io.wavfile.write(output_file, rate=speech["sampling_rate"], data=normalized_audio_data) | |
| return summary, output_file | |
| # Gradio Interface | |
| iface = gr.Interface( | |
| fn=process_text, | |
| inputs=gr.inputs.File(label="Upload PDF"), | |
| outputs=["text", "audio"], | |
| title="Summarization and Text-to-Speech", | |
| description="Upload a PDF to extract, summarize its abstract, and convert to speech." | |
| ) | |
| if __name__ == "__main__": | |
| iface.launch() |