Update app.py

app.py

@@ -70,35 +70,16 @@ tokenizer_3 = AutoTokenizer.from_pretrained(
 # Ensure model and scheduler are initialized in GPU-enabled function
 if torch.cuda.is_available():
     pipe = StableDiffusion3Pipeline.from_pretrained(
-        repo, 
+        repo,
+        vae=vae,
+        transformer=transformer,
         torch_dtype=torch.float16).to("cuda")
     pipe2 = StableDiffusion3Img2ImgPipeline.from_pretrained(
-        repo, 
+        repo,
+        vae=vae,
+        transformer=transformer,
         torch_dtype=torch.float16).to("cuda")
 
-torch.set_float32_matmul_precision("high")
-
-torch._inductor.config.conv_1x1_as_mm = True
-torch._inductor.config.coordinate_descent_tuning = True
-torch._inductor.config.epilogue_fusion = False
-torch._inductor.config.coordinate_descent_check_all_directions = True
-
-pipe.set_progress_bar_config(disable=True)
-
-pipe.transformer.to(memory_format=torch.channels_last)
-pipe.vae.to(memory_format=torch.channels_last)
-
-pipe.transformer = torch.compile(pipe.transformer, mode="max-autotune", fullgraph=True)
-pipe.vae.decode = torch.compile(pipe.vae.decode, mode="max-autotune", fullgraph=True)
-
-
-pipe2.set_progress_bar_config(disable=True)
-
-pipe2.transformer.to(memory_format=torch.channels_last)
-pipe2.vae.to(memory_format=torch.channels_last)
-
-pipe2.transformer = torch.compile(pipe2.transformer, mode="max-autotune", fullgraph=True)
-pipe2.vae.decode = torch.compile(pipe2.vae.decode, mode="max-autotune", fullgraph=True)
 
 
 pipe.scheduler = FlowMatchEulerDiscreteScheduler.from_config(pipe.scheduler.config)