Update app.py

app.py

@@ -24,6 +24,7 @@ saved_names = [
     hf_hub_download(repo_id, filename) for _, _, repo_id, _, filename, _ in sdxl_loras
 ]
 
+device = "cuda" #replace this to `mps` if on a MacOS Silicon
 
 def update_selection(selected_state: gr.SelectData):
     lora_repo = sdxl_loras[selected_state.index][2]
@@ -41,7 +42,7 @@ pipe = StableDiffusionXLPipeline.from_pretrained(
     torch_dtype=torch.float16,
 ).to("cpu")
 original_pipe = copy.deepcopy(pipe)
-pipe.to("cuda")
+pipe.to(device)
 
 last_lora = ""
 last_merged = False
@@ -58,7 +59,7 @@ def run_lora(prompt, negative, weight, selected_state):
     if last_lora != repo_name:
         if last_merged:
             pipe = copy.deepcopy(original_pipe)
-            pipe.to("cuda")
+            pipe.to(device)
         else:
             pipe.unload_lora_weights()
         is_compatible = sdxl_loras[selected_state.index][5]
@@ -85,6 +86,7 @@ def run_lora(prompt, negative, weight, selected_state):
                 #lora_model.merge_to(
                 #    pipe.text_encoder, pipe.unet, weights_sd, torch.float16, "cuda"
                 #)
+                pipe.to(device)
             last_merged = True
 
     image = pipe(