def full_weight()

in tt_embeddings_ops.py [0:0]


    def full_weight(self) -> torch.Tensor:
        assert (
            self.num_tables == 1
        ), "full_weight() only supported for num_tables == 1 for now"
        return tt_matrix_to_full(
            self.tt_p_shapes,
            self.tt_q_shapes,
            self.tt_ranks,
            self.tt_cores,
            [1, 0, 2, 3],
        )