Merge pull request #456 from alibaba-damo-academy/dev_zc

zhifu gao

2023-05-05 06b4aab43bea7cd5be30f163acc1b68d5527d7db

 funasr/bin/build_trainer.py

@@ -27,6 +27,10 @@
        from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    elif mode == "uniasr":
        from funasr.tasks.asr import ASRTaskUniASR as ASRTask
    elif mode == "mfcca":
        from funasr.tasks.asr import ASRTaskMFCCA as ASRTask
    elif mode == "tp":
        from funasr.tasks.asr import ASRTaskAligner as ASRTask
    else:
        raise ValueError("Unknown mode: {}".format(mode))
    parser = ASRTask.get_parser()
@@ -49,7 +53,8 @@
                  scheduler_conf=None,
                  specaug=None,
                  specaug_conf=None,
                  param_dict=None):
                  param_dict=None,
                  **kwargs):
    mode = modelscope_dict['mode']
    args, ASRTask = parse_args(mode=mode)
    # ddp related
@@ -78,7 +83,8 @@
        finetune_configs = yaml.safe_load(f)
        # set data_types
        if dataset_type == "large":
            finetune_configs["dataset_conf"]["data_types"] = "sound,text"
            if 'data_types' not in finetune_configs['dataset_conf']:
                finetune_configs["dataset_conf"]["data_types"] = "sound,text"
    finetune_configs = update_dct(configs, finetune_configs)
    for key, value in finetune_configs.items():
        if hasattr(args, key):
@@ -126,7 +132,8 @@
        if args.dataset_type == "small":
            args.batch_bins = batch_bins
        elif args.dataset_type == "large":
            args.dataset_conf["batch_conf"]["batch_size"] = batch_bins
            if "batch_size" not in args.dataset_conf["batch_conf"]:
                args.dataset_conf["batch_conf"]["batch_size"] = batch_bins
        else:
            raise ValueError(f"Not supported dataset_type={args.dataset_type}")
    if args.normalize in ["null", "none", "None"]: