Add Flux model support for InstantX style controlnet residuals

comfy/ldm/flux/controlnet_xlabs.py

@@ -82,7 +82,7 @@ def forward_orig(
             block_res_sample = controlnet_block(block_res_sample)
             controlnet_block_res_samples = controlnet_block_res_samples + (block_res_sample,)
 
-        return {"output": (controlnet_block_res_samples * 10)[:19]}
+        return {"input": (controlnet_block_res_samples * 10)[:19]}
 
     def forward(self, x, timesteps, context, y, guidance=None, hint=None, **kwargs):
         hint = hint * 2.0 - 1.0

comfy/ldm/flux/model.py

@@ -114,19 +114,28 @@ def forward_orig(
         ids = torch.cat((txt_ids, img_ids), dim=1)
         pe = self.pe_embedder(ids)
 
-        for i in range(len(self.double_blocks)):
-            img, txt = self.double_blocks[i](img=img, txt=txt, vec=vec, pe=pe)
+        for i, block in enumerate(self.double_blocks):
+           img, txt = block(img=img, txt=txt, vec=vec, pe=pe)
 
-            if control is not None: #Controlnet
+           if control is not None: # Controlnet
+               control_i = control.get("input")
+               if i < len(control_i):
+                   add = control_i[i]
+                   if add is not None:
+                       img += add
+
+        img = torch.cat((txt, img), 1)
+
+        for i, block in enumerate(self.single_blocks):
+            img = block(img, vec=vec, pe=pe)
+
+            if control is not None: # Controlnet
                 control_o = control.get("output")
                 if i < len(control_o):
                     add = control_o[i]
                     if add is not None:
-                        img += add
+                        img[:, txt.shape[1] :, ...] += add
 
-        img = torch.cat((txt, img), 1)
-        for block in self.single_blocks:
-            img = block(img, vec=vec, pe=pe)
         img = img[:, txt.shape[1] :, ...]
 
         img = self.final_layer(img, vec)  # (N, T, patch_size ** 2 * out_channels)