[npu] fix where bug (#9573)

* [npu] fix where bug * fix
PaddlePaddle · Dec 9, 2024 · c565b77 · c565b77
1 parent 1917257
commit c565b77
Showing 1 changed file with 0 additions and 1 deletion.
diff --git a/paddlenlp/transformers/llama/modeling.py b/paddlenlp/transformers/llama/modeling.py
@@ -1593,7 +1593,6 @@ def _prepare_decoder_attention_mask(attention_mask, input_shape, past_key_values
         if get_env_device() in ["npu", "mlu", "intel_hpu"]:
             x = paddle.to_tensor(0.0, dtype="float32")
             y = paddle.to_tensor(paddle.finfo(dtype).min, dtype="float32")
-            expanded_attn_mask = expanded_attn_mask.astype("float32")
             expanded_attn_mask = paddle.where(expanded_attn_mask, x, y).astype(dtype)
         elif get_env_device() in ["xpu", "gcu"]:
             min_val = paddle.finfo(dtype).min if get_env_device() == "gcu" else -1e37  # mask value for xpu