Mila 0.13.48
Deep Neural Network Library
Loading...
Searching...
No Matches
Compute.CudaMultiHeadAttentionOp Module Reference

Classes

class  Mila::Dnn::Compute::Cuda::MultiHeadAttention::CudaMultiHeadAttentionOp< TPrecision >
 CUDA implementation of Multi-Head Attention using column-major cuBLASLt optimization. More...
class  Mila::Dnn::Compute::Cuda::MultiHeadAttention::CudaMultiHeadAttentionOpRegistrar

Files

file  /__w/Mila/Mila/Mila/Src/Dnn/Compute/Devices/Cuda/Operations/Attention/MHA/CudaMhaOp.ixx
file  /__w/Mila/Mila/Mila/Src/Dnn/Compute/Devices/Cuda/Operations/Attention/MHA/CudaMhaOp.Dispatch.ixx
file  /__w/Mila/Mila/Mila/Src/Dnn/Compute/Devices/Cuda/Operations/Attention/MHA/CudaMhaOp.Plans.ixx