Update README.md

README.md

@@ -42,6 +42,7 @@ pipe = Flux2KleinPipeline.from_pretrained(
     text_encoder=text_encoder_4bit,
     torch_dtype=dtype,
 )
+#pipe.to("cuda") # GPU < 32G
 pipe.enable_model_cpu_offload()  # save some VRAM by offloading the model to CPU,GPU < 32G
 
 prompt = "A cat holding a sign that says hello world"