Searched defs:k_padded (Results 1 – 2 of 2) sorted by relevance
| /external/pytorch/aten/src/ATen/native/transformers/cuda/flash_attn/ |
| D | flash_api.cpp | 421 at::Tensor q_padded, k_padded, v_padded; in mha_fwd() local
660 at::Tensor q_padded, k_padded, v_padded; in mha_varlen_fwd() local
1409 at::Tensor k, v, k_padded, v_padded; in mha_fwd_kvcache() local
|
| /external/pytorch/aten/src/ATen/native/transformers/cuda/ |
| D | attention.cu | 878 Tensor output, q_padded, k_padded, v_padded, logsumexp, output_shape, in _flash_attention_forward() local
|