Merge branch 'main' into main

Author: piEsposito

src/diffusers/models/attention.py

@@ -244,7 +244,9 @@ def __init__(
         self.to_k = nn.Linear(context_dim, inner_dim, bias=False)
         self.to_v = nn.Linear(context_dim, inner_dim, bias=False)
 
-        self.to_out = nn.Sequential(nn.Linear(inner_dim, query_dim), nn.Dropout(dropout))
+        self.to_out = nn.ModuleList([])
+        self.to_out.append(nn.Linear(inner_dim, query_dim))
+        self.to_out.append(nn.Dropout(dropout))
 
     def reshape_heads_to_batch_dim(self, tensor):
         batch_size, seq_len, dim = tensor.shape
@@ -283,7 +285,11 @@ def forward(self, hidden_states, context=None, mask=None):
         else:
             hidden_states = self._sliced_attention(query, key, value, sequence_length, dim)
 
-        return self.to_out(hidden_states)
+        # linear proj
+        hidden_states = self.to_out[0](hidden_states)
+        # dropout
+        hidden_states = self.to_out[1](hidden_states)
+        return hidden_states
 
     def _attention(self, query, key, value):
         # TODO: use baddbmm for better performance
@@ -354,12 +360,19 @@ def __init__(
         super().__init__()
         inner_dim = int(dim * mult)
         dim_out = dim_out if dim_out is not None else dim
-        project_in = GEGLU(dim, inner_dim)
+        self.net = nn.ModuleList([])
 
-        self.net = nn.Sequential(project_in, nn.Dropout(dropout), nn.Linear(inner_dim, dim_out))
+        # project in
+        self.net.append(GEGLU(dim, inner_dim))
+        # project dropout
+        self.net.append(nn.Dropout(dropout))
+        # project out
+        self.net.append(nn.Linear(inner_dim, dim_out))
 
     def forward(self, hidden_states):
-        return self.net(hidden_states)
+        for module in self.net:
+            hidden_states = module(hidden_states)
+        return hidden_states
 
 
 # feedforward