Upload modeling_opt.py

modeling_opt.py

@@ -167,8 +167,6 @@ class OPTAttention(nn.Module):
     def __init__(
         self,
         config: OPTConfig,
-        embed_dim: int,
-        num_heads: int,
         dropout: float = 0.0,
         is_decoder: bool = False,
         bias: bool = True,
@@ -188,10 +186,13 @@ class OPTAttention(nn.Module):
         attn_softmax='vanilla',
     ):
         super().__init__()
-        self.embed_dim = embed_dim
-        self.num_heads = num_heads
-        self.dropout = dropout
+        self.embed_dim = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.dropout = config.attention_dropout
+        self.enable_bias = config.enable_bias
         self.head_dim = embed_dim // num_heads
+        self.is_causal = True
+
 
         if (self.head_dim * num_heads) != self.embed_dim:
             raise ValueError(