Update app.py

app.py

@@ -10,6 +10,8 @@ model = AutoModelForCausalLM.from_pretrained(model_name).to("cpu")  # Force CPU
 
 import tempfile
 
+import tempfile
+
 def review_code(code_snippet):
     print("✅ Received Code:", code_snippet)  # Debugging log
     
@@ -30,7 +32,7 @@ Do NOT generate new functions or extra text—only return the fixed version of t
     inputs = tokenizer(prompt, return_tensors="pt").to("cpu")  # Move to CPU
     outputs = model.generate(
         **inputs,
-        max_length=60,  # ✅ Keeps response concise & correct
+        max_new_tokens=50,  # ✅ Generate only 50 new tokens (fixes warning)
         do_sample=False,
         num_beams=4,  # ✅ Ensures better correction quality
         repetition_penalty=2.5  # ✅ Prevents repeated/unnecessary output