ViperEkura
/
AstrAI
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

docs(transformer): 更新process_attention_mask函数文档

This commit is contained in:
ViperEkura 2025-11-05 23:41:11 +08:00
parent 7ccc4ab9ac
commit 805773c7fe
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
khaosz/model/transformer.py
View File

@ -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.
"""