Update app.py

app.py

@@ -11,7 +11,8 @@ model = AutoModelForCausalLM.from_pretrained(
     "AdnanRiaz107/CodePhi-3-Mini-1K4bit",
     torch_dtype="auto",  # Use auto for dtype selection
     trust_remote_code=True,
-    attn_implementation='eager',  # Keep this if you want to use 'eager'
+    attn_implementation='eager',
+    load_in_4bit= True,# Keep this if you want to use 'eager'
 )
 
 # Load the tokenizer