update

游雁

2024-07-17 a836eca98e30fa67d45167dac40f359ae42d42ec

 funasr/auto/auto_model.py

@@ -20,7 +20,7 @@
from funasr.download.file import download_from_url
from funasr.utils.timestamp_tools import timestamp_sentence
from funasr.utils.timestamp_tools import timestamp_sentence_en
from funasr.download.download_from_hub import download_model
from funasr.download.download_model_from_hub import download_model
from funasr.utils.vad_utils import slice_padding_audio_samples
from funasr.utils.vad_utils import merge_vad
from funasr.utils.load_utils import load_audio_text_image_video
@@ -111,11 +111,15 @@

    def __init__(self, **kwargs):

        try:
            from funasr.utils.version_checker import check_for_update

            check_for_update()
        except:
            pass

        log_level = getattr(logging, kwargs.get("log_level", "INFO").upper())
        logging.basicConfig(level=log_level)

        if not kwargs.get("disable_log", True):
            tables.print()

        model, kwargs = self.build_model(**kwargs)

@@ -164,7 +168,8 @@
        self.spk_kwargs = spk_kwargs
        self.model_path = kwargs.get("model_path")

    def build_model(self, **kwargs):
    @staticmethod
    def build_model(**kwargs):
        assert "model" in kwargs
        if "model_conf" not in kwargs:
            logging.info("download models from model hub: {}".format(kwargs.get("hub", "ms")))
@@ -210,6 +215,7 @@
        kwargs["frontend"] = frontend
        # build model
        model_class = tables.model_classes.get(kwargs["model"])
        assert model_class is not None, f'{kwargs["model"]} is not registered'
        model_conf = {}
        deep_update(model_conf, kwargs.get("model_conf", {}))
        deep_update(model_conf, kwargs)
@@ -237,6 +243,10 @@
        elif kwargs.get("bf16", False):
            model.to(torch.bfloat16)
        model.to(device)

        if not kwargs.get("disable_log", True):
            tables.print()

        return model, kwargs

    def __call__(self, *args, **cfg):
@@ -305,7 +315,7 @@
            speed_stats["rtf"] = f"{(time_escape) / batch_data_time:0.3f}"
            description = f"{speed_stats}, "
            if pbar:
                pbar.update(1)
                pbar.update(end_idx - beg_idx)
                pbar.set_description(description)
            time_speech_total += batch_data_time
            time_escape_total += time_escape
@@ -329,7 +339,9 @@
        #  FIX(gcf): concat the vad clips for sense vocie model for better aed
        if kwargs.get("merge_vad", False):
            for i in range(len(res)):
                res[i]["value"] = merge_vad(res[i]["value"], kwargs.get("merge_length", 15000))
                res[i]["value"] = merge_vad(
                    res[i]["value"], kwargs.get("merge_length_s", 15) * 1000
                )

        # step.2 compute asr model
        model = self.model
@@ -370,6 +382,9 @@
            if len(sorted_data) > 0 and len(sorted_data[0]) > 0:
                batch_size = max(batch_size, sorted_data[0][0][1] - sorted_data[0][0][0])

            if kwargs["device"] == "cpu":
                batch_size = 0

            beg_idx = 0
            beg_asr_total = time.time()
            time_speech_total_per_sample = speech_lengths / 16000