Skip to main content

flash_attention2_backward_dq

vision_rs::models::yolo::kernels::attention::flash_attn2

Function flash_attention2_backward_dq 

Source 
pub fn flash_attention2_backward_dq<T: Triton, D: Float, const HEAD_DIM: i32>(
    q_ptr: T::Pointer<D>,
    k_ptr: T::Pointer<D>,
    v_ptr: T::Pointer<D>,
    o_ptr: T::Pointer<D>,
    do_ptr: T::Pointer<D>,
    l_ptr: T::Pointer<D>,
    dq_ptr: T::Pointer<D>,
    n_ctx_q: i32,
    n_ctx_k: i32,
    softmax_scale: f32,
)
where
    T::I32Tensor: Tensor<i32, 1> + Comparison<i32, BoolTensor = T::BoolTensor>,
    T::Pointer<D>: AddOffsets<i32, 1, T::I32Tensor, Output = T::Tensor<T::Pointer<D>>>,
Expand description

Flash Attention 2 backward: computes dQ.

Grid: (N_CTX_Q, BH, 1) — same grid shape as the forward pass.