matrix_transform_desc_create# nvmath.bindings.cublasLt.matrix_transform_desc_create(int scale_type) → intptr_t[source]# See cublasLtMatrixTransformDescCreate.
