src/summarizer.py

from transformers import pipeline
import torch
import logging

class TextSummarizer:
    def __init__(self, model_name="facebook/bart-large-cnn"):
        """
        Initialize summarization pipeline
        
        Args:
            model_name (str): Hugging Face model for summarization
        """
        try:
            # Configure device
            device = 0 if torch.cuda.is_available() else -1
            logging.info(f"Using device: {'cuda' if device == 0 else 'cpu'}")
            
            # Initialize pipeline with explicit device mapping and lower precision
            self.summarizer = pipeline(
                "summarization", 
                model=model_name,
                device=device,
                torch_dtype=torch.float32
            )
            logging.info("Summarization pipeline initialized successfully")
            
        except Exception as e:
            logging.error(f"Failed to load summarization model: {str(e)}")
            raise RuntimeError(f"Failed to load summarization model: {str(e)}")
    
    def generate_summary(self, text, max_length=400, min_length=100):
        """
        Generate summary for given text
        
        Args:
            text (str): Input text to summarize
            max_length (int): Maximum length of summary
            min_length (int): Minimum length of summary
            
        Returns:
            str: Generated summary
        """
        try:
            # Validate input text
            if not text or len(text.strip()) == 0:
                return "No text provided for summarization."
            
            # Ensure min_length is less than max_length
            min_length = min(min_length, max_length)
            
            # Generate summary with chunking for long texts
            max_chunk_length = 1024  # BART's max input length
            chunks = [text[i:i + max_chunk_length] for i in range(0, len(text), max_chunk_length)]
            summaries = []
            
            for chunk in chunks:
                if chunk.strip():
                    summary = self.summarizer(
                        chunk,
                        max_length=max_length // len(chunks),  # Distribute length across chunks
                        min_length=min_length // len(chunks),
                        do_sample=False
                    )[0]['summary_text']
                    summaries.append(summary)
            
            return " ".join(summaries)
            
        except Exception as e:
            logging.error(f"Error during summarization: {str(e)}")
            return f"Error during summarization: {str(e)}"