funasr/utils/prepare_data.py
@@ -162,6 +162,10 @@ if args.dataset_type == "large" and args.train_data_file is not None: return distributed = distributed_option.distributed if not hasattr(args, "train_set"): args.train_set = "train" if not hasattr(args, "dev_set"): args.dev_set = "validation" if not distributed or distributed_option.dist_rank == 0: filter_wav_text(args.data_dir, args.train_set) filter_wav_text(args.data_dir, args.dev_set)