diff --git a/modules/default_pipeline.py b/modules/default_pipeline.py
index 525e6ec..ccc3f8e 100644
--- a/modules/default_pipeline.py
+++ b/modules/default_pipeline.py
@@ -9,6 +9,8 @@ xl_base_filename = os.path.join(modelfile_path, 'sd_xl_base_1.0.safetensors')
 xl_refiner_filename = os.path.join(modelfile_path, 'sd_xl_refiner_1.0.safetensors')
 
 xl_base = core.load_model(xl_base_filename)
+xl_refiner = core.load_model(xl_refiner_filename)
+del xl_base.vae
 
 
 @torch.no_grad()
@@ -16,16 +18,28 @@ def process(positive_prompt, negative_prompt, width=1024, height=1024, batch_siz
     positive_conditions = core.encode_prompt_condition(clip=xl_base.clip, prompt=positive_prompt)
     negative_conditions = core.encode_prompt_condition(clip=xl_base.clip, prompt=negative_prompt)
 
+    positive_conditions_refiner = core.encode_prompt_condition(clip=xl_refiner.clip, prompt=positive_prompt)
+    negative_conditions_refiner = core.encode_prompt_condition(clip=xl_refiner.clip, prompt=negative_prompt)
+
     empty_latent = core.generate_empty_latent(width=width, height=height, batch_size=batch_size)
 
     sampled_latent = core.ksample(
         unet=xl_base.unet,
         positive_condition=positive_conditions,
         negative_condition=negative_conditions,
-        latent_image=empty_latent
+        latent_image=empty_latent,
+        steps=30, start_at_step=0, end_at_step=20, return_with_leftover_noise=True, add_noise=True
     )
 
-    decoded_latent = core.decode_vae(vae=xl_base.vae, latent_image=sampled_latent)
+    sampled_latent = core.ksample(
+        unet=xl_refiner.unet,
+        positive_condition=positive_conditions_refiner,
+        negative_condition=negative_conditions_refiner,
+        latent_image=sampled_latent,
+        steps=30, start_at_step=20, end_at_step=30, return_with_leftover_noise=False, add_noise=False
+    )
+
+    decoded_latent = core.decode_vae(vae=xl_refiner.vae, latent_image=sampled_latent)
 
     images = core.image_to_numpy(decoded_latent)
     return images