Searched defs:max_seqlen_q (Results 1 – 7 of 7) sorted by relevance
| /aosp_15_r20/external/pytorch/aten/src/ATen/native/transformers/cuda/flash_attn/ |
| H A D | mask.h | 40 const int max_seqlen_q, const int warp_row_stride, in apply_mask_local()
77 const int max_seqlen_q, const int warp_row_stride) { in apply_mask_causal()
113 const int max_seqlen_k, max_seqlen_q; member
|
| H A D | flash_api.cpp | 298 const int num_heads, const int head_size, const int max_seqlen_k, const int max_seqlen_q, in set_params_splitkv()
552 int max_seqlen_q, in mha_varlen_fwd()
1030 const int max_seqlen_q, in mha_varlen_bwd()
|
| H A D | alibi.h | 20 const int max_seqlen_k, max_seqlen_q; member
|
| /aosp_15_r20/external/pytorch/torch/utils/ |
| H A D | flop_counter.py | 332 max_seqlen_q, argument
412 max_seqlen_q, argument
511 max_seqlen_q, argument
|
| /aosp_15_r20/external/pytorch/aten/src/ATen/native/transformers/cuda/ |
| H A D | attention_backward.cu | 262 int64_t max_seqlen_q, in _efficient_attention_backward()
817 int64_t max_seqlen_q = q_t.size(1); in _scaled_dot_product_efficient_attention_backward_cuda() local
|
| H A D | attention.cu | 995 int64_t max_seqlen_q = 0, max_seqlen_k = 0; in _efficient_attention_forward() local
|
| /aosp_15_r20/external/pytorch/test/cpp_extensions/ |
| H A D | open_registration_extension.cpp | 458 const int64_t max_seqlen_q = query.size(2); in custom_scaled_dot_product_fused_attention_overrideable() local
|