pub fn causal_mask(
seq_len: usize,
past_len: usize,
sliding_window: Option<usize>,
dtype: DType,
device: &Device,
) -> Result<Tensor>Expand description
Build a causal additive attention mask (1, 1, seq_len, total_len) with -1e9 on masked positions.
ยงErrors
Returns an error if tensor creation fails.