Home
last modified time | relevance | path

Searched defs:num_head (Results 1 – 8 of 8) sorted by relevance

/external/pytorch/aten/src/ATen/native/cpu/
DNativeMultiheadAttnKernel.cpp25 int64_t num_head) { in cpu_transform_bias_rescale_qkv()
90 int64_t num_head) { in transform_bias_rescale_qkv_kernel_impl()
DFlashAttentionKernel.cpp151 int64_t num_head, in reshape_attn_mask_to_4d()
206 int64_t num_head = query.size(2); in cpu_flash_attention() local
449 int64_t num_head = query.size(2); in cpu_flash_attention_backward() local
/external/executorch/examples/mediatek/executor_runner/llama_runner/
DLlamaConfig.h25 size_t num_head = 32; member
/external/pytorch/aten/src/ATen/native/transformers/
Dattention.cpp233 const int64_t num_head) { in transform_bias_rescale_qkv_cpu()
266 const int64_t num_head, in native_multi_head_attention_cpu()
782 int64_t num_head = query.size(1); in _scaled_dot_product_flash_attention_cpu() local
857 const int64_t num_head, in triton_multi_head_attention()
Dsdp_utils_cpp.h289 auto num_head = params.query.sym_size(1); in check_attn_mask_shape() local
/external/openthread/third_party/tcplp/bsdtcp/
Dtcp_sack.c149 int num_head, num_saved, i; in tcp_update_sack_list() local
/external/pytorch/aten/src/ATen/native/transformers/cuda/
Dattention.cu379 const int64_t num_head) { in transform_bias_rescale_qkv_cuda()
484 const int64_t num_head, in native_multi_head_attention_cuda()
/external/executorch/extension/llm/custom_ops/
Dop_sdpa.cpp261 int64_t num_head = query.size(1); in cpu_flash_attention() local