From 805773c7fe4055f944b061b073370b084c60f598 Mon Sep 17 00:00:00 2001
From: ViperEkura <3081035982@qq.com>
Date: Wed, 5 Nov 2025 23:41:11 +0800
Subject: [PATCH] =?UTF-8?q?docs(transformer):=20=E6=9B=B4=E6=96=B0process?=
 =?UTF-8?q?=5Fattention=5Fmask=E5=87=BD=E6=95=B0=E6=96=87=E6=A1=A3?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 khaosz/model/transformer.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/khaosz/model/transformer.py b/khaosz/model/transformer.py
index 63904bc..e6af718 100644
--- a/khaosz/model/transformer.py
+++ b/khaosz/model/transformer.py
@@ -18,10 +18,10 @@ def process_attention_mask(
     Create attention mask for GQA
     Args:
         seq_mask (Tensor): A tensor indicating whether each position is valid or not.
+        input_tensor (Tensor): The input tensor.
         start_pos (int): The starting position of the sequence.
         seq_len (int): The length of the sequence.
         is_causal (bool): Whether the attention is causal or not.
-        device (torch.device): The device to use.
     Returns:
         Tensor: The attention mask tensor.
     """