update

paddlenlp/transformers/llama/modeling.py

@@ -118,8 +118,7 @@ def _get_interleave_power_of_2(n):
 def build_alibi_tensor(
     bool_attention_mask: Tensor, num_heads: int, dtype: paddle.dtype, tensor_parallel_degree=1
 ) -> Tensor:
-    attention_mask = bool_attention_mask.astype("float32")
-    batch_size, seq_length = attention_mask.shape[0], attention_mask.shape[-1]
+    batch_size, seq_length = bool_attention_mask.shape[0], bool_attention_mask.shape[-1]
     slopes = paddle.to_tensor(_get_interleave(num_heads), dtype="float32")
     alibi = slopes.unsqueeze(axis=[1, 2]) * paddle.arange(seq_length, dtype="float32").unsqueeze(axis=[0, 1]).expand(
         [num_heads, -1, -1]
@@ -1533,15 +1532,21 @@ def forward(
         if position_ids is None:
             position_ids = paddle.arange(seq_length, dtype="int64").expand((batch_size, seq_length))
 
-        attention_mask = self._prepare_decoder_attention_mask(
-            attention_mask, (batch_size, seq_length), cache_length, inputs_embeds.dtype
-        )  # [bs, 1, seq_len, seq_len]
+        is_casual_mask = (
+            True if hasattr(self.config, "use_casual_mask") and self.config.use_casual_mask is True else False
+        )
+        if is_casual_mask:
+            attention_mask = None
+        else:
+            attention_mask = self._prepare_decoder_attention_mask(
+                attention_mask, (batch_size, seq_length), cache_length, inputs_embeds.dtype
+            )  # [bs, 1, seq_len, seq_len]
 
         is_casual = False
 
         if self.config.use_flash_attention and get_env_device() != "gcu":
-            if hasattr(self.config, "use_casual_mask"):
-                is_casual = self.config.use_casual_mask
+            if is_casual_mask:
+                is_casual = True
             else:
                 is_casual = is_casual_mask(attention_mask)
             if get_env_device() != "npu":