Fix error

Hi

@Kvikontent

,

Merging this Pull Request will make your space work. You won't have an error at the execution anymore.

_Successfully tested on my duplicated space:_ https://huggingface.co/spaces/Fabrice-TIERCELIN/SDXL-Turbo-Realtime

app.py

@@ -8,11 +8,7 @@ pipe = DiffusionPipeline.from_pretrained("stabilityai/sdxl-turbo")
 def generate_image(prompt):
     # Run the diffusion model to generate an image
     output = pipe(prompt, num_inference_steps=50, guidance_scale=7.5)
-    image_tensor = output.images[0]
-    image_array = image_tensor.cpu().numpy().transpose(1, 2, 0)
-    image_array = (image_array * 255).astype(np.uint8)
-    image = Image.fromarray(image_array)
-    return image
+    return output.images[0]
 
 gr_interface = gr.Interface(
     fn=generate_image,