Update app.py

app.py

@@ -50,7 +50,7 @@ def predict():
     formatted_text = f"<s>[INST] {input_text} [/INST]"
 
     # Tokenize the input
-    inputs = tokenizer(formatted_text, return_tensors='pt', padding=True, truncation=True)
+    inputs = tokenizer(formatted_text, return_tensors='pt', padding=True, truncation=True, max_length=1024)
 
     # Generate the output
     outputs = model.generate(