GroupNorm patcher (#593)

* add gn wrapper and corresponding patcher

* add gn wrapper and corresponding patcher

ldm_patched/ldm/modules/diffusionmodules/openaimodel.py

@@ -30,7 +30,7 @@ class TimestepBlock(nn.Module):
     """
 
     @abstractmethod
-    def forward(self, x, emb):
+    def forward(self, x, emb, transformer_options={}):
         """
         Apply the module to `x` given `emb` timestep embeddings.
         """
@@ -46,7 +46,7 @@ def forward_timestep_embed(ts, x, emb, context=None, transformer_options={}, out
         if isinstance(layer, VideoResBlock):
             x = layer(x, emb, num_video_frames, image_only_indicator)
         elif isinstance(layer, TimestepBlock):
-            x = layer(x, emb)
+            x = layer(x, emb, transformer_options)
         elif isinstance(layer, SpatialVideoTransformer):
             x = layer(x, context, time_context, num_video_frames, image_only_indicator, transformer_options)
             if "transformer_index" in transformer_options:
@@ -234,7 +234,7 @@ class ResBlock(TimestepBlock):
         else:
             self.skip_connection = operations.conv_nd(dims, channels, self.out_channels, 1, dtype=dtype, device=device)
 
-    def forward(self, x, emb):
+    def forward(self, x, emb, transformer_options={}):
         """
         Apply the block to a Tensor, conditioned on a timestep embedding.
         :param x: an [N x C x ...] Tensor of features.
@@ -242,19 +242,29 @@ class ResBlock(TimestepBlock):
         :return: an [N x C x ...] Tensor of outputs.
         """
         return checkpoint(
-            self._forward, (x, emb), self.parameters(), self.use_checkpoint
+            self._forward, (x, emb, transformer_options), self.parameters(), self.use_checkpoint
         )
 
 
-    def _forward(self, x, emb):
+    def _forward(self, x, emb, transformer_options={}):
         if self.updown:
             in_rest, in_conv = self.in_layers[:-1], self.in_layers[-1]
-            h = in_rest(x)
+            if "groupnorm_wrapper" in transformer_options:
+                in_norm, in_rest = in_rest[0], in_rest[1:]
+                h = transformer_options["groupnorm_wrapper"](in_norm, x, transformer_options)
+                h = in_rest(h)
+            else:
+                h = in_rest(x)
             h = self.h_upd(h)
             x = self.x_upd(x)
             h = in_conv(h)
         else:
-            h = self.in_layers(x)
+            if "groupnorm_wrapper" in transformer_options:
+                in_norm = self.in_layers[0]
+                h = transformer_options["groupnorm_wrapper"](in_norm, x, transformer_options)
+                h = self.in_layers[1:](h)
+            else:
+                h = self.in_layers(x)
 
         emb_out = None
         if not self.skip_t_emb:
@@ -263,7 +273,10 @@ class ResBlock(TimestepBlock):
                 emb_out = emb_out[..., None]
         if self.use_scale_shift_norm:
             out_norm, out_rest = self.out_layers[0], self.out_layers[1:]
-            h = out_norm(h)
+            if "groupnorm_wrapper" in transformer_options:
+                h = transformer_options["groupnorm_wrapper"](out_norm, h, transformer_options)
+            else:
+                h = out_norm(h)
             if emb_out is not None:
                 scale, shift = th.chunk(emb_out, 2, dim=1)
                 h *= (1 + scale)
@@ -274,7 +287,11 @@ class ResBlock(TimestepBlock):
                 if self.exchange_temb_dims:
                     emb_out = rearrange(emb_out, "b t c ... -> b c t ...")
                 h = h + emb_out
-            h = self.out_layers(h)
+            if "groupnorm_wrapper" in transformer_options:
+                h = transformer_options["groupnorm_wrapper"](self.out_layers[0], h, transformer_options)
+                h = self.out_layers[1:](h)
+            else:
+                h = self.out_layers(h)
         return self.skip_connection(x) + h
 
 
@@ -924,6 +941,10 @@ class UNetModel(nn.Module):
 
         if self.predict_codebook_ids:
             h = self.id_predictor(h)
+        elif "groupnorm_wrapper" in transformer_options:
+            out_norm, out_rest = self.out[0], self.out[1:]
+            h = transformer_options["groupnorm_wrapper"](out_norm, h, transformer_options)
+            h = out_rest(h)
         else:
             h = self.out(h)
 

modules_forge/unet_patcher.py

@@ -159,6 +159,10 @@ class UnetPatcher(ModelPatcher):
         self.append_transformer_option('controlnet_conditioning_modifiers', modifier, ensure_uniqueness)
         return
 
+    def set_groupnorm_wrapper(self, wrapper):
+        self.set_transformer_option('groupnorm_wrapper', wrapper)
+        return
+
     def set_controlnet_model_function_wrapper(self, wrapper):
         self.set_transformer_option('controlnet_model_function_wrapper', wrapper)
         return