[LoRA] pop the LoRA scale so that it doesn't get propagated to the weeds (#7338)

* pop scale from the top-level unet instead of getting it.

* improve readability.

* Apply suggestions from code review

Co-authored-by: YiYi Xu <[email protected]>

* fix a little bit.

---------

Co-authored-by: YiYi Xu <[email protected]>

Author: sayakpaul

src/diffusers/models/unets/unet_2d_condition.py

@@ -1081,25 +1081,15 @@ def forward(
                 A tuple of tensors that if specified are added to the residuals of down unet blocks.
             mid_block_additional_residual: (`torch.Tensor`, *optional*):
                 A tensor that if specified is added to the residual of the middle unet block.
+            down_intrablock_additional_residuals (`tuple` of `torch.Tensor`, *optional*):
+                additional residuals to be added within UNet down blocks, for example from T2I-Adapter side model(s)
             encoder_attention_mask (`torch.Tensor`):
                 A cross-attention mask of shape `(batch, sequence_length)` is applied to `encoder_hidden_states`. If
                 `True` the mask is kept, otherwise if `False` it is discarded. Mask will be converted into a bias,
                 which adds large negative values to the attention scores corresponding to "discard" tokens.
             return_dict (`bool`, *optional*, defaults to `True`):
                 Whether or not to return a [`~models.unets.unet_2d_condition.UNet2DConditionOutput`] instead of a plain
                 tuple.
-            cross_attention_kwargs (`dict`, *optional*):
-                A kwargs dictionary that if specified is passed along to the [`AttnProcessor`].
-            added_cond_kwargs: (`dict`, *optional*):
-                A kwargs dictionary containin additional embeddings that if specified are added to the embeddings that
-                are passed along to the UNet blocks.
-            down_block_additional_residuals (`tuple` of `torch.Tensor`, *optional*):
-                additional residuals to be added to UNet long skip connections from down blocks to up blocks for
-                example from ControlNet side model(s)
-            mid_block_additional_residual (`torch.Tensor`, *optional*):
-                additional residual to be added to UNet mid block output, for example from ControlNet side model
-            down_intrablock_additional_residuals (`tuple` of `torch.Tensor`, *optional*):
-                additional residuals to be added within UNet down blocks, for example from T2I-Adapter side model(s)
 
         Returns:
             [`~models.unets.unet_2d_condition.UNet2DConditionOutput`] or `tuple`:
@@ -1185,7 +1175,13 @@ def forward(
             cross_attention_kwargs["gligen"] = {"objs": self.position_net(**gligen_args)}
 
         # 3. down
-        lora_scale = cross_attention_kwargs.get("scale", 1.0) if cross_attention_kwargs is not None else 1.0
+        # we're popping the `scale` instead of getting it because otherwise `scale` will be propagated
+        # to the internal blocks and will raise deprecation warnings. this will be confusing for our users.
+        if cross_attention_kwargs is not None:
+            lora_scale = cross_attention_kwargs.pop("scale", 1.0)
+        else:
+            lora_scale = 1.0
+
         if USE_PEFT_BACKEND:
             # weight the lora layers by setting `lora_scale` for each PEFT layer
             scale_lora_layers(self, lora_scale)