Fix processed_images dimen

Author: skirsten

src/diffusers/pipelines/stable_diffusion/pipeline_flax_stable_diffusion_img2img.py

@@ -121,7 +121,7 @@ def prepare_inputs(self, prompt: Union[str, List[str]], image: Union[Image.Image
         if isinstance(image, Image.Image):
             image = [image]
 
-        processed_images = jnp.array([preprocess(img, jnp.float32) for img in image])
+        processed_images = jnp.concatenate([preprocess(img, jnp.float32) for img in image])
 
         text_input = self.tokenizer(
             prompt,