Fix PerceiverMLP and test (#16405)

Co-authored-by: Jaesun Park <jaesun.park1@navercorp.com>

src/transformers/models/perceiver/modeling_perceiver.py

@@ -420,7 +420,7 @@ class PerceiverMLP(nn.Module):
             self.intermediate_act_fn = ACT2FN[config.hidden_act]
         else:
             self.intermediate_act_fn = config.hidden_act
-        self.dense2 = nn.Linear(input_size, input_size)
+        self.dense2 = nn.Linear(widening_factor * input_size, input_size)
 
     def forward(self, hidden_states):
         hidden_states = self.dense1(hidden_states)

tests/perceiver/test_modeling_perceiver.py

@@ -82,6 +82,8 @@ class PerceiverModelTester:
         num_self_attends_per_block=2,
         num_self_attention_heads=1,
         num_cross_attention_heads=1,
+        self_attention_widening_factor=4,
+        cross_attention_widening_factor=4,
         is_training=True,
         use_input_mask=True,
         use_labels=True,
@@ -109,6 +111,8 @@ class PerceiverModelTester:
         self.num_self_attends_per_block = num_self_attends_per_block
         self.num_self_attention_heads = num_self_attention_heads
         self.num_cross_attention_heads = num_cross_attention_heads
+        self.self_attention_widening_factor = self_attention_widening_factor
+        self.cross_attention_widening_factor = cross_attention_widening_factor
         self.is_training = is_training
         self.use_input_mask = use_input_mask
         self.use_labels = use_labels
@@ -174,10 +178,14 @@ class PerceiverModelTester:
         return PerceiverConfig(
             num_latents=self.num_latents,
             d_latents=self.d_latents,
+            qk_channels=self.d_latents,
+            v_channels=self.d_latents,
             num_blocks=self.num_blocks,
             num_self_attends_per_block=self.num_self_attends_per_block,
             num_self_attention_heads=self.num_self_attention_heads,
             num_cross_attention_heads=self.num_cross_attention_heads,
+            self_attention_widening_factor=self.self_attention_widening_factor,
+            cross_attention_widening_factor=self.cross_attention_widening_factor,
             vocab_size=self.vocab_size,
             hidden_act=self.hidden_act,
             attention_probs_dropout_prob=self.attention_probs_dropout_prob,