diff --git a/models/unet_2d_condition.py b/models/unet_2d_condition.py
index e18f69f..804a164 100644
--- a/models/unet_2d_condition.py
+++ b/models/unet_2d_condition.py
@@ -38,7 +38,7 @@ from diffusers.models.embeddings import (
     ImageHintTimeEmbedding,
     ImageProjection,
     ImageTimeEmbedding,
-    #PositionNet,
+    GLIGENTextBoundingboxProjection,
     TextImageProjection,
     TextImageTimeEmbedding,
     TextTimeEmbedding,
@@ -619,7 +619,7 @@ class UNet2DConditionModel(ModelMixin, ConfigMixin, UNet2DConditionLoadersMixin)
                 positive_len = cross_attention_dim[0]
 
             feature_type = "text-only" if attention_type == "gated" else "text-image"
-            self.position_net = PositionNet(
+            self.position_net = GLIGENTextBoundingboxProjection(
                 positive_len=positive_len, out_dim=cross_attention_dim, feature_type=feature_type
             )