pub async fn compute_flash_attention_async(
query: Float32Array,
keys: Vec<Float32Array>,
values: Vec<Float32Array>,
dim: u32,
block_size: u32,
) -> Result<Float32Array>Expand description
Async flash attention computation
pub async fn compute_flash_attention_async(
query: Float32Array,
keys: Vec<Float32Array>,
values: Vec<Float32Array>,
dim: u32,
block_size: u32,
) -> Result<Float32Array>Async flash attention computation