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

Classes

class  Mila::Dnn::Compute::CudaMultiHeadAttentionOp< TInput, TOutput >
 CUDA implementation of the Multi-Head Attention operation for transformer models. More...
 
class  Mila::Dnn::Compute::CudaMultiHeadAttentionOpRegistrar
 Class responsible for registering the CudaMultiHeadAttentionOp operation. More...
 

Files

file  /home/runner/work/Mila/Mila/Mila/Src/Dnn/Compute/Operations/Cuda/CudaMultiHeadAttentionOp.ixx
 Implementation of the CUDA-based Multi-Head Attention operation for transformer models.