Fix bug in ResnetBlock2D.forward where LoRA Scale gets Overwritten (#6736)

Fix bug in ResnetBlock2D.forward when not USE_PEFT_BACKEND and using scale_shift for time emb where the lora scale  gets overwritten.

Co-authored-by: Sayak Paul <[email protected]>

Author: dg845

src/diffusers/models/resnet.py

@@ -384,9 +384,9 @@ def forward(
                 raise ValueError(
                     f" `temb` should not be None when `time_embedding_norm` is {self.time_embedding_norm}"
                 )
-            scale, shift = torch.chunk(temb, 2, dim=1)
+            time_scale, time_shift = torch.chunk(temb, 2, dim=1)
             hidden_states = self.norm2(hidden_states)
-            hidden_states = hidden_states * (1 + scale) + shift
+            hidden_states = hidden_states * (1 + time_scale) + time_shift
         else:
             hidden_states = self.norm2(hidden_states)