Update app.py

app.py

@@ -23,7 +23,7 @@ def load_model():
     
     return Llama(
         model_path=model_path,
-        n_ctx=1024,       # Match model's training context length
+        n_ctx=1024,
         n_threads=4,
         verbose=False
     )
@@ -31,9 +31,9 @@ def load_model():
 # Load model at startup
 llm = load_model()
 
-# Generation function with validation
+# Generation function with anti-repetition
 def generate_text(prompt, max_tokens=MAX_TOKENS, temp=0.7, top_p=0.95):
-    """Generate text with repetition prevention and input validation"""
+    """Generate text with improved repetition handling"""
     if not prompt.strip():
         return "⚠️ Please enter a valid prompt."
         
@@ -47,8 +47,8 @@ def generate_text(prompt, max_tokens=MAX_TOKENS, temp=0.7, top_p=0.95):
             temperature=temp,
             top_p=top_p,
             echo=False,
-            repeat_penalty=1.2,
-            no_repeat_ngram_size=3
+            # Only use supported parameters
+            repeat_penalty=1.5  # Increased from 1.2
         )
         return output["choices"][0]["text"].strip()
     except Exception as e: