T2I-Adapter-XL

src/diffusers/models/adapter.py

@@ -109,6 +109,8 @@ class T2IAdapter(ModelMixin, ConfigMixin):
 
         if adapter_type == "full_adapter":
             self.adapter = FullAdapter(in_channels, channels, num_res_blocks, downscale_factor)
+        elif adapter_type == "full_adapter_xl":
+            self.adapter = FullAdapter_XL(in_channels, channels, num_res_blocks, downscale_factor)
         elif adapter_type == "light_adapter":
             self.adapter = LightAdapter(in_channels, channels, num_res_blocks, downscale_factor)
         else:
@@ -165,6 +167,80 @@ class FullAdapter(nn.Module):
         return features
 
 
+class FullAdapter_XL(nn.Module):
+    def __init__(
+        self,
+        in_channels: int = 3,
+        channels: List[int] = [320, 640, 1280, 1280],
+        num_res_blocks: int = 2,
+        downscale_factor: int = 8,
+    ):
+        super().__init__()
+
+        in_channels = in_channels * downscale_factor**2
+        self.channels = channels
+        self.num_res_blocks = num_res_blocks
+
+        self.unshuffle = nn.PixelUnshuffle(downscale_factor)
+        self.conv_in = nn.Conv2d(in_channels, channels[0], kernel_size=3, padding=1)
+        self.body = []
+        for i in range(len(channels)):
+            for j in range(num_res_blocks):
+                if (i == 2) and (j == 0):
+                    self.body.append(
+                        AdapterResnetBlock_XL(channels[i - 1], channels[i], down=True))
+                elif (i == 1) and (j == 0):
+                    self.body.append(
+                        AdapterResnetBlock_XL(channels[i - 1], channels[i], down=False))
+                else:
+                    self.body.append(
+                        AdapterResnetBlock_XL(channels[i], channels[i], down=False))
+        self.body = nn.ModuleList(self.body)
+
+        self.total_downscale_factor = downscale_factor * 2 ** (len(channels) - 1)
+
+    def forward(self, x: torch.Tensor) -> List[torch.Tensor]:
+        x = self.unshuffle(x)
+        x = self.conv_in(x)
+
+        features = []
+
+        for i in range(len(self.channels)):
+            for j in range(self.num_res_blocks):
+                idx = i * self.num_res_blocks + j
+                x = self.body[idx](x)
+            features.append(x)
+
+        return features
+
+
+class AdapterResnetBlock_XL(nn.Module):
+    def __init__(self, channels_in, channels_out, down=False):
+        super().__init__()
+        if channels_in != channels_out:
+            self.in_conv = nn.Conv2d(channels_in, channels_out, 1)
+        else:
+            self.in_conv = None
+        self.block1 = nn.Conv2d(channels_out, channels_out, kernel_size=3, padding=1)
+        self.act = nn.ReLU()
+        self.block2 = nn.Conv2d(channels_out, channels_out, kernel_size=1)
+        self.downsample = None
+        if down:
+            self.downsample = Downsample2D(channels_in)
+
+    def forward(self, x):
+        if self.downsample is not None:
+            x = self.downsample(x)
+        if self.in_conv is not None:
+            x = self.in_conv(x)
+        h = x
+        h = self.block1(h)
+        h = self.act(h)
+        h = self.block2(h)
+
+        return h + x
+
+
 class AdapterBlock(nn.Module):
     def __init__(self, in_channels, out_channels, num_res_blocks, down=False):
         super().__init__()

src/diffusers/models/unet_2d_condition.py

@@ -965,6 +965,9 @@ class UNet2DConditionModel(ModelMixin, ConfigMixin, UNet2DConditionLoadersMixin)
                 cross_attention_kwargs=cross_attention_kwargs,
                 encoder_attention_mask=encoder_attention_mask,
             )
+            # To support T2I-Adapter-XL 
+            if is_adapter and len(down_block_additional_residuals) > 0:
+                sample += down_block_additional_residuals.pop(0)
 
         if is_controlnet:
             sample = sample + mid_block_additional_residual