small update

app.py

@@ -107,7 +107,7 @@ It's quantized using the BitsAndBytes library to 4-bit using the [bnb-my-repo](h
     
     # Append original README content if available
     if original_readme and not original_readme.isspace():
-        model_card += "\n\n# Original Model Information\n" + original_readme
+        model_card += "\n\n# 📄 Original Model Information\n\n" + original_readme
     
     return model_card
 
@@ -147,7 +147,7 @@ def quantize_model(
         quantization_config=quantization_config,
         device_map="cpu",
         use_auth_token=auth_token.token,
-        torch_dtype=torch.bfloat16,
+        torch_dtype="auto",
     )
     progress(0.33, desc="Quantizing")