Update app.py

app.py

@@ -12,7 +12,7 @@ model = AutoModelForCausalLM.from_pretrained(
     torch_dtype="auto",  # Use auto for dtype selection
     trust_remote_code=True,
     attn_implementation='eager',
-    load_in_4bit= True,# Keep this if you want to use 'eager'
+    load_in_8bit= True,# Keep this if you want to use 'eager'
 )
 
 # Load the tokenizer