pub fn attention(
x: &[f32],
seqlen: usize,
start_pos: usize,
w: &AttnWeights<'_>,
c: &AttnCfg,
fc: &[(f32, f32)],
cache: &mut WindowCache,
) -> AttnRunExpand description
Attention.forward for a ratio-0 layer. x: [seqlen][dim] bf16 values; fc from
freqs_cis over max_seq_len. Mutates cache exactly as the reference’s ring buffer.