ggml_mul_mat_set_prec method
change the precision of a matrix multiplication set to GGML_PREC_F32 for higher precision (useful for phi-2)
Implementation
void ggml_mul_mat_set_prec(
ffi.Pointer<ggml_tensor> a,
int prec,
) {
return _ggml_mul_mat_set_prec(
a,
prec,
);
}