Spaces:
Sleeping
Sleeping
File size: 2,131 Bytes
522a147 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
import os
from flask import Flask, render_template, request, jsonify
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
app = Flask(__name__)
MODEL_NAME = "facebook/bart-large-cnn"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME)
@app.route('/')
def index():
return render_template('index.html')
@app.route('/summarize', methods=['POST'])
def summarize():
try:
data = request.get_json()
text = data.get('text', '').strip()
if not text or len(text.split()) < 15:
return jsonify({'error': 'Text too short.'}), 400
inputs = tokenizer(text, return_tensors="pt", max_length=1024, truncation=True)
summary_ids = model.generate(
inputs["input_ids"],
max_length=300,
min_length=150,
length_penalty=2.5,
num_beams=5,
no_repeat_ngram_size=3,
early_stopping=False
)
summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
# --- SAFE CLEANING LOGIC ---
# 1. Fix punctuation spacing
summary = summary.replace(" .", ".").replace(" ,", ",").replace(" !", "!").replace(" ?", "?")
# 2. Capitalize ONLY the first letter of each sentence
# This prevents "James Webb" from becoming "James webb"
parts = summary.split(". ")
cleaned_sentences = []
for p in parts:
p = p.strip()
if p:
# Upcase only index 0, keep everything else exactly as the AI wrote it
sentence = p[0].upper() + p[1:]
cleaned_sentences.append(sentence)
summary = ". ".join(cleaned_sentences)
if summary and not summary.endswith('.'):
summary += '.'
return jsonify({'summary': summary})
except Exception as e:
return jsonify({'error': str(e)}), 500
if __name__ == '__main__':
app.run(host='0.0.0.0', port=7860) |