minor changes from github issues

        # Build samples mapping

        verbose = torch.distributed.get_rank() == 0

        start_time = time.time()

        print_rank_0(' > building sapmles index mapping for {} ...'.format(

        print_rank_0(' > building samples index mapping for {} ...'.format(

            name))

        # First compile and then import.

        from megatron.data import helpers

            seed,

            verbose,

            2 if binary_head else 1)

        print_rank_0(' > done building sapmles index maping')

        print_rank_0(' > done building samples index maping')

        np.save(indexmap_filename, samples_mapping, allow_pickle=True)

        print_rank_0(' > saved the index mapping in {}'.format(

            indexmap_filename))

    if iteration % args.log_interval == 0:

        elapsed_time = timers('interval-time').elapsed()

        elapsed_time_per_iteration = elapsed_time / total_iterations

        if writer and torch.distributed.get_rank() == 0:

        if writer:

            if args.log_timers_to_tensorboard:

                writer.add_scalar('iteration-time',

                                  elapsed_time_per_iteration, iteration)

        string += '{} value: {:.6E} | '.format(key, total_loss_dict[key].item())

        ppl = math.exp(min(20, total_loss_dict[key].item()))

        string += '{} PPL: {:.6E} | '.format(key, ppl)

        if writer and is_last_rank():

        if writer:

            writer.add_scalar('{} validation'.format(key),

                              total_loss_dict[key].item(),

                              iteration)