pub fn attention(
x: &[f32],
seqlen: usize,
start_pos: usize,
w: &AttnWeights<'_>,
c: &AttnCfg,
fc: &[(f32, f32)],
cache: &mut WindowCache,
) -> AttnRunExpand description
2026-09-25: The ratio-0 attention forward. x: [seqlen][dim] bf16 values;
fc from freqs_cis over the maximum sequence length. Writes the new kv
rows into cache’s ring.