Combine Input 運算子

std::tuple<at::Tensor, at::Tensor, at::Tensor> tbe_input_combine_cpu(const std::vector<at::Tensor> &indices_list, const std::vector<at::Tensor> &offsets_list, const std::vector<at::Tensor> &per_sample_weights, const at::Tensor &include_last_offsets)

std::tuple<at::Tensor, at::Tensor, at::Tensor> padding_fused_tbe_input_combine_cpu(const std::vector<at::Tensor> &indices_list, const std::vector<at::Tensor> &offsets_list, const std::vector<at::Tensor> &per_sample_weights, const at::Tensor &include_last_offsets, int64_t batch_size)