fix: dont pass hidden_size twice to NeobertMLP

model.py

@@ -139,7 +139,7 @@ class EncoderBlock(nn.Module):
         if XFORMERS_AVAILABLE:
             self.ffn = SwiGLU(config.hidden_size, intermediate_size, config.hidden_size, bias=False)
         else:
-            self.ffn = NeobertMLP(config.hidden_size, intermediate_size, config.hidden_size, bias=False)
+            self.ffn = NeobertMLP(config.hidden_size, intermediate_size, bias=False)
 
         # Layer norms
         self.attention_norm = nn.RMSNorm(config.hidden_size, config.norm_eps)