fix diffusers

requirements.txt

@@ -2,12 +2,12 @@
 torch
 torchvision
 torchaudio
-accelerate==0.25.0
-torchmetrics==1.2.1
-tqdm==4.66.1
-transformers==4.36.2
-diffusers==0.25.0
-einops==0.7.0
-bitsandbytes==0.39.0
-scipy==1.11.1
+accelerate==0.30.0
+torchmetrics==1.4.0
+tqdm==4.66.4
+transformers==4.40.2
+diffusers==0.27.2
+einops==0.8.0
+bitsandbytes==0.43.1
+scipy==1.13.0
 opencv-python

src/idm_vton/unet_hacked_garmnet.py

@@ -37,7 +37,7 @@ from diffusers.models.embeddings import (
     ImageHintTimeEmbedding,
     ImageProjection,
     ImageTimeEmbedding,
-    PositionNet,
+    GLIGENTextBoundingboxProjection,
     TextImageProjection,
     TextImageTimeEmbedding,
     TextTimeEmbedding,
@@ -695,7 +695,7 @@ 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
             )
 

src/idm_vton/unet_hacked_tryon.py

@@ -37,7 +37,7 @@ from diffusers.models.embeddings import (
     ImageHintTimeEmbedding,
     ImageProjection,
     ImageTimeEmbedding,
-    PositionNet,
+    GLIGENTextBoundingboxProjection,
     TextImageProjection,
     TextImageTimeEmbedding,
     TextTimeEmbedding,
@@ -764,7 +764,7 @@ 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
             )
 
@@ -841,9 +841,9 @@ class UNet2DConditionModel(ModelMixin, ConfigMixin, UNet2DConditionLoadersMixin)
         def fn_recursive_attn_processor(name: str, module: torch.nn.Module, processor):
             if hasattr(module, "set_processor"):
                 if not isinstance(processor, dict):
-                    module.set_processor(processor, _remove_lora=_remove_lora)
+                    module.set_processor(processor)
                 else:
-                    module.set_processor(processor.pop(f"{name}.processor"), _remove_lora=_remove_lora)
+                    module.set_processor(processor.pop(f"{name}.processor"))
 
             for sub_name, child in module.named_children():
                 fn_recursive_attn_processor(f"{name}.{sub_name}", child, processor)