|
onnx2versal
|
Vector implementation for MK*KN, streams input, outputs, weights, stores bias, requires (2*K + N)*4 <= 24576 and N % 8 == 0 GemmReluMKKNStream<7,36,10> total = 2774 (slower than GemmReluMKKNStream for larger N)
#include <gemm.h>
Public Member Functions | |
| GemmReluMKKNTwoAccsStream (float(&b)[N]) | |
| void | filter (input_stream< float > *in, input_stream< float > *weight, output_window< float > *out) |
Static Public Member Functions | |
| static void | registerKernelClass () |