Update app.py

app.py

@@ -14,12 +14,30 @@ def load_models():
     # Text embedding model
     embed_model = SentenceTransformer('sentence-transformers/all-mpnet-base-v2')
     
-    # IBM Granite models
-    summary_tokenizer = AutoTokenizer.from_pretrained("ibm/granite-13b-instruct-v2")
-    summary_model = AutoModelForCausalLM.from_pretrained("ibm/granite-13b-instruct-v2")
+    TOKEN = "TOKEN"  
+
+    # IBM Granite models with proper token and trust_remote_code settings
+    summary_tokenizer = AutoTokenizer.from_pretrained(
+        "ibm/granite-13b-instruct-v2",
+        token=TOKEN,
+        trust_remote_code=True
+    )
+    summary_model = AutoModelForCausalLM.from_pretrained(
+        "ibm/granite-13b-instruct-v2",
+        token=TOKEN,
+        trust_remote_code=True
+    )
     
-    qa_tokenizer = AutoTokenizer.from_pretrained("ibm/granite-13b-instruct-v2")
-    qa_model = AutoModelForCausalLM.from_pretrained("ibm/granite-13b-instruct-v2")
+    qa_tokenizer = AutoTokenizer.from_pretrained(
+        "ibm/granite-13b-instruct-v2",
+        token=TOKEN,
+        trust_remote_code=True
+    )
+    qa_model = AutoModelForCausalLM.from_pretrained(
+        "ibm/granite-13b-instruct-v2",
+        token=TOKEN,
+        trust_remote_code=True
+    )
     
     return embed_model, summary_model, summary_tokenizer, qa_model, qa_tokenizer
 
@@ -30,16 +48,13 @@ def process_file(uploaded_file):
     if file_type == 'pdf':
         pdf_reader = PdfReader(uploaded_file)
         for page in pdf_reader.pages:
-            text += page.extract_text()
-            
+            text += page.extract_text() or ""
     elif file_type == 'txt':
         text = uploaded_file.read().decode('utf-8')
-        
     elif file_type == 'docx':
         doc = Document(uploaded_file)
         for para in doc.paragraphs:
             text += para.text + "\n"
-            
     return clean_text(text)
 
 def clean_text(text):
@@ -111,4 +126,4 @@ def main():
             st.info(f"Answer: {answer}")
 
 if __name__ == "__main__":
-    main()
+    main()