Update app.py

app.py

@@ -1,15 +1,21 @@
 from flask import Flask, request, render_template, jsonify
 import torch
 from nltk.tokenize import word_tokenize
-from transformers import pipeline, AutoModelForSeq2SeqLM, AutoTokenizer, PegasusTokenizerFast, T5Tokenizer, T5ForConditionalGeneration, MBartForConditionalGeneration, MBart50TokenizerFast
+from transformers import pipeline, AutoModelForSeq2SeqLM, AutoTokenizer, T5Tokenizer, T5ForConditionalGeneration, MBartForConditionalGeneration, MBart50TokenizerFast
 from LDict import find_legal_terms, legal_terms_lower
 import nltk
-import re
-import logging
+import re,os, logging
+
+# Set environment variables for writable directories
+os.environ["TRANSFORMERS_CACHE"] = "/tmp/transformers_cache"
+nltk.data.path.append("/tmp/nltk_data")
+
 logging.basicConfig(level=logging.ERROR)
 
-nltk.download('punkt')
-nltk.download('punkt_tab')
+# Download necessary NLTK data
+nltk.download('punkt', download_dir="/tmp/nltk_data")
+nltk.download('punkt_tab', download_dir="/tmp/nltk_data")
+
 
 app = Flask(__name__)
 
@@ -54,22 +60,17 @@ def summarize_text(text, method):
         inputs_legal = port_tokenizer(text, max_length=1024, truncation=True, return_tensors="pt")
         summary_ids_legal = model_port.generate(inputs_legal["input_ids"], max_length=250, num_beams=4, early_stopping=True)
         Summarized_method2 = port_tokenizer.decode(summary_ids_legal[0], skip_special_tokens=True)
-        print("\n\n\n Summarized MEthod2",Summarized_method2, "\n\n\n\n")
         cleaned_summary2 = remove_parentheses(Summarized_method2)
-        print("\n\n\n Cleaned Summarized MEthod2",cleaned_summary2, "\n\n\n\n")        
         #Paraphrase
         p_inputs = tokenizer_t5.encode(cleaned_summary2, return_tensors="pt", max_length=512, truncation=True)
         p_summary_ids = model_t5.generate(p_inputs, max_length=150, min_length=50, length_penalty=2.0, num_beams=4, early_stopping=True)
         method2 = tokenizer_t5.decode(p_summary_ids[0], skip_special_tokens=True)
-        print("\n\n\n Summarized Paraphrased MEthod2",method2, "\n\n\n\n")
         return method2
 
     elif method == "method1":
         summarization_pipeline = pipeline('summarization', model=model_pegasus, tokenizer=tokenizer_pegasus, device=0 if device == "cuda" else -1)
         method1 = summarization_pipeline(text, max_length=100, min_length=30, truncation=True)[0]['summary_text']
-        print("\n\n\n Summarized MEthod1",method1, "\n\n\n\n")                 
         cleaned_summary1 = remove_parentheses(method1)
-        print("\n\n\n Summarized Cleaned MEthod1",cleaned_summary1, "\n\n\n\n")                
         return cleaned_summary1
 
 
@@ -86,17 +87,13 @@ def index():
     if request.method == 'POST':
         try:
             input_text = request.form['input_text']
-            logging.info(f"Received data for translation: {input_text}")  # Log incoming data
             method = request.form['method']
             
             simplified_text = simplify_text(input_text)
-            logging.info(f"Received data for translation: {simplified_text}")
             summarized_text = summarize_text(simplified_text, method)
-            logging.info(f"Received data for translation: {summarized_text}")
 
             return jsonify({
-                "summarized_text": summarized_text,
-            })
+                "summarized_text": summarized_text, })
         except Exception as e:
             logging.error(f"Error occurred: {e}", exc_info=True)
             return jsonify({"error": str(e)}), 500
@@ -106,18 +103,15 @@ def index():
 def translate():
     try:
         data = request.get_json()
-        logging.info(f"Received data for translation: {data}")  # Log incoming data
         text = data['text']
         translated_text = translate_to_hindi(text)
 
         return jsonify({
-            "translated_text": translated_text
-        })
+            "translated_text": translated_text})
     except Exception as e:
         logging.error(f"Error occurred during translation: {e}", exc_info=True)
         return jsonify({"error": str(e)}), 500
 
 
 if __name__ == '__main__':
-    app.run(port=5003)
-    
+    app.run(port=5003)