1 less line

src/pipeline.py

@@ -47,13 +47,11 @@ def load_pipeline() -> Pipeline:
         vae=vae,
         local_files_only=True,
         torch_dtype=torch.bfloat16,
-    )
+    ).to("cuda")
 
-    pipeline.transformer.to(memory_format=torch.channels_last) 
-    pipeline.vae.to(memory_format=torch.channels_last)
+    pipeline.to(memory_format=torch.channels_last)
     quantize_(pipeline.vae, int8_weight_only()) 
     pipeline.vae = torch.compile(pipeline.vae, mode="max-autotune", fullgraph=True)
-    pipeline.to("cuda")
     
     for _ in range(2):
         pipeline("cat", num_inference_steps=4)