Skip to content

[npu] fix where bug #9573

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Dec 9, 2024
Merged

[npu] fix where bug #9573

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c58bec4
[npu] fix where bug
tianhaodongbd Dec 5, 2024
f38ed4c
fix
tianhaodongbd Dec 5, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion paddlenlp/transformers/llama/modeling.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -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
Expand Down