funasr/bin/build_trainer.py
@@ -627,8 +627,6 @@ torch.backends.cudnn.deterministic = args.cudnn_deterministic # ddp init os.environ['CUDA_VISIBLE_DEVICES'] = str(args.gpu_id) args.distributed = args.ngpu > 1 or args.dist_world_size > 1 distributed_option = build_distributed(args) # for logging
