UserWarning: indexing with dtype torch.uint8 is now deprecated, please use a dtype torch.bool instead.

内部实则是bool。
原：

    idxs = torch.Tensor(idxs_0.shape[0]).byte()
    idxs[:] = 1
    idxs[idxs_0 < 0] = 0
    idxs[idxs_end >= int(N / args.gps_step)] = 0
    delta_p_gt = delta_p_gt[idxs]
    idxs_0_bias = idxs_0[idxs]
    idxs_end_bis = idxs_end[idxs]

可以看出，实则idxs内部为bool性质。

改后：

    # 加一个.bool()
    idxs = torch.Tensor(idxs_0.shape[0]).byte().bool()
    idxs[:] = 1
    idxs[idxs_0 < 0] = 0
    idxs[idxs_end >= int(N / args.gps_step)] = 0
    delta_p_gt = delta_p_gt[idxs]
    idxs_0_bias = idxs_0[idxs]
    idxs_end_bis = idxs_end[idxs]