Update app.py

app.py

@@ -135,7 +135,7 @@ def build_pipeline(lora_name: str) -> ICLoraPipeline:
         gemma_root=gemma_root,
         loras=[lora],
         #loras=[],
-        # quantization=QuantizationPolicy.fp8_cast(),
+        quantization=QuantizationPolicy.fp8_cast(),
     )
     return pipe