funasr1.0 auto/ auto_model auto_frontend auto_tokenizer

游雁

2024-01-16 ce92fde1b754ae56aec7f62ff910c205a84bf159

 funasr/bin/inference.py

@@ -20,68 +20,8 @@
from funasr.utils.timestamp_tools import timestamp_sentence
from funasr.models.campplus.utils import sv_chunk, postprocess, distribute_spk
from funasr.models.campplus.cluster_backend import ClusterBackend
from funasr.auto.auto_model import AutoModel


def prepare_data_iterator(data_in, input_len=None, data_type=None, key=None):
    """
    
    :param input:
    :param input_len:
    :param data_type:
    :param frontend:
    :return:
    """
    data_list = []
    key_list = []
    filelist = [".scp", ".txt", ".json", ".jsonl"]
    
    chars = string.ascii_letters + string.digits
    if isinstance(data_in, str) and data_in.startswith('http'): # url
        data_in = download_from_url(data_in)
    if isinstance(data_in, str) and os.path.exists(data_in): # wav_path; filelist: wav.scp, file.jsonl;text.txt;
        _, file_extension = os.path.splitext(data_in)
        file_extension = file_extension.lower()
        if file_extension in filelist: #filelist: wav.scp, file.jsonl;text.txt;
            with open(data_in, encoding='utf-8') as fin:
                for line in fin:
                    key = "rand_key_" + ''.join(random.choice(chars) for _ in range(13))
                    if data_in.endswith(".jsonl"): #file.jsonl: json.dumps({"source": data})
                        lines = json.loads(line.strip())
                        data = lines["source"]
                        key = data["key"] if "key" in data else key
                    else: # filelist, wav.scp, text.txt: id \t data or data
                        lines = line.strip().split(maxsplit=1)
                        data = lines[1] if len(lines)>1 else lines[0]
                        key = lines[0] if len(lines)>1 else key
                    
                    data_list.append(data)
                    key_list.append(key)
        else:
            key = "rand_key_" + ''.join(random.choice(chars) for _ in range(13))
            data_list = [data_in]
            key_list = [key]
    elif isinstance(data_in, (list, tuple)):
        if data_type is not None and isinstance(data_type, (list, tuple)): # mutiple inputs
            data_list_tmp = []
            for data_in_i, data_type_i in zip(data_in, data_type):
                key_list, data_list_i = prepare_data_iterator(data_in=data_in_i, data_type=data_type_i)
                data_list_tmp.append(data_list_i)
            data_list = []
            for item in zip(*data_list_tmp):
                data_list.append(item)
        else:
            # [audio sample point, fbank, text]
            data_list = data_in
            key_list = ["rand_key_" + ''.join(random.choice(chars) for _ in range(13)) for _ in range(len(data_in))]
    else: # raw text; audio sample point, fbank; bytes
        if isinstance(data_in, bytes): # audio bytes
            data_in = load_bytes(data_in)
        if key is None:
            key = "rand_key_" + ''.join(random.choice(chars) for _ in range(13))
        data_list = [data_in]
        key_list = [key]
    
    return key_list, data_list

@hydra.main(config_name=None, version_base=None)
def main_hydra(cfg: DictConfig):
@@ -104,397 +44,6 @@
    res = model(input=kwargs["input"])
    print(res)

class AutoModel:
    
    def __init__(self, **kwargs):
        tables.print()
        
        model, kwargs = self.build_model(**kwargs)
        
        # if vad_model is not None, build vad model else None
        vad_model = kwargs.get("vad_model", None)
        vad_kwargs = kwargs.get("vad_model_revision", None)
        if vad_model is not None:
            logging.info("Building VAD model.")
            vad_kwargs = {"model": vad_model, "model_revision": vad_kwargs}
            vad_model, vad_kwargs = self.build_model(**vad_kwargs)

        # if punc_model is not None, build punc model else None
        punc_model = kwargs.get("punc_model", None)
        punc_kwargs = kwargs.get("punc_model_revision", None)
        if punc_model is not None:
            logging.info("Building punc model.")
            punc_kwargs = {"model": punc_model, "model_revision": punc_kwargs}
            punc_model, punc_kwargs = self.build_model(**punc_kwargs)

        # if spk_model is not None, build spk model else None
        spk_model = kwargs.get("spk_model", None)
        spk_kwargs = kwargs.get("spk_model_revision", None)
        if spk_model is not None:
            logging.info("Building SPK model.")
            spk_kwargs = {"model": spk_model, "model_revision": spk_kwargs}
            spk_model, spk_kwargs = self.build_model(**spk_kwargs)
            self.cb_model = ClusterBackend()
            spk_mode = kwargs.get("spk_mode", 'punc_segment')
            if spk_mode not in ["default", "vad_segment", "punc_segment"]:
                logging.error("spk_mode should be one of default, vad_segment and punc_segment.")
            self.spk_mode = spk_mode
            self.preset_spk_num = kwargs.get("preset_spk_num", None)
            if self.preset_spk_num:
                logging.warning("Using preset speaker number: {}".format(self.preset_spk_num))
            logging.warning("Many to print when using speaker model...")
            
        self.kwargs = kwargs
        self.model = model
        self.vad_model = vad_model
        self.vad_kwargs = vad_kwargs
        self.punc_model = punc_model
        self.punc_kwargs = punc_kwargs
        self.spk_model = spk_model
        self.spk_kwargs = spk_kwargs
        self.model_path = kwargs["model_path"]
  
        
    def build_model(self, **kwargs):
        assert "model" in kwargs
        if "model_conf" not in kwargs:
            logging.info("download models from model hub: {}".format(kwargs.get("model_hub", "ms")))
            kwargs = download_model(**kwargs)
        
        set_all_random_seed(kwargs.get("seed", 0))
        
        device = kwargs.get("device", "cuda")
        if not torch.cuda.is_available() or kwargs.get("ngpu", 0):
            device = "cpu"
            # kwargs["batch_size"] = 1
        kwargs["device"] = device
        
        if kwargs.get("ncpu", None):
            torch.set_num_threads(kwargs.get("ncpu"))
        
        # build tokenizer
        tokenizer = kwargs.get("tokenizer", None)
        if tokenizer is not None:
            tokenizer_class = tables.tokenizer_classes.get(tokenizer)
            tokenizer = tokenizer_class(**kwargs["tokenizer_conf"])
            kwargs["tokenizer"] = tokenizer
            kwargs["token_list"] = tokenizer.token_list
            vocab_size = len(tokenizer.token_list)
        else:
            vocab_size = -1
        
        # build frontend
        frontend = kwargs.get("frontend", None)
        if frontend is not None:
            frontend_class = tables.frontend_classes.get(frontend)
            frontend = frontend_class(**kwargs["frontend_conf"])
            kwargs["frontend"] = frontend
            kwargs["input_size"] = frontend.output_size()
        
        # build model
        model_class = tables.model_classes.get(kwargs["model"])
        model = model_class(**kwargs, **kwargs["model_conf"], vocab_size=vocab_size)
        model.eval()
        model.to(device)
        
        # init_param
        init_param = kwargs.get("init_param", None)
        if init_param is not None:
            logging.info(f"Loading pretrained params from {init_param}")
            load_pretrained_model(
                model=model,
                init_param=init_param,
                ignore_init_mismatch=kwargs.get("ignore_init_mismatch", False),
                oss_bucket=kwargs.get("oss_bucket", None),
            )
        
        return model, kwargs
    
    def __call__(self, input, input_len=None, **cfg):
        if self.vad_model is None:
            return self.generate(input, input_len=input_len, **cfg)
            
        else:
            return self.generate_with_vad(input, input_len=input_len, **cfg)
        
    def generate(self, input, input_len=None, model=None, kwargs=None, key=None, **cfg):
        kwargs = self.kwargs if kwargs is None else kwargs
        kwargs.update(cfg)
        model = self.model if model is None else model

        batch_size = kwargs.get("batch_size", 1)
        # if kwargs.get("device", "cpu") == "cpu":
        #     batch_size = 1
        
        key_list, data_list = prepare_data_iterator(input, input_len=input_len, data_type=kwargs.get("data_type", None), key=key)
        
        speed_stats = {}
        asr_result_list = []
        num_samples = len(data_list)
        pbar = tqdm(colour="blue", total=num_samples+1, dynamic_ncols=True)
        time_speech_total = 0.0
        time_escape_total = 0.0
        for beg_idx in range(0, num_samples, batch_size):
            end_idx = min(num_samples, beg_idx + batch_size)
            data_batch = data_list[beg_idx:end_idx]
            key_batch = key_list[beg_idx:end_idx]
            batch = {"data_in": data_batch, "key": key_batch}
            if (end_idx - beg_idx) == 1 and isinstance(data_batch[0], torch.Tensor): # fbank
                batch["data_in"] = data_batch[0]
                batch["data_lengths"] = input_len
        
            time1 = time.perf_counter()
            with torch.no_grad():
                results, meta_data = model.inference(**batch, **kwargs)
            time2 = time.perf_counter()
            
            asr_result_list.extend(results)
            pbar.update(1)
            
            # batch_data_time = time_per_frame_s * data_batch_i["speech_lengths"].sum().item()
            batch_data_time = meta_data.get("batch_data_time", -1)
            time_escape = time2 - time1
            speed_stats["load_data"] = meta_data.get("load_data", 0.0)
            speed_stats["extract_feat"] = meta_data.get("extract_feat", 0.0)
            speed_stats["forward"] = f"{time_escape:0.3f}"
            speed_stats["batch_size"] = f"{len(results)}"
            speed_stats["rtf"] = f"{(time_escape) / batch_data_time:0.3f}"
            description = (
                f"{speed_stats}, "
            )
            pbar.set_description(description)
            time_speech_total += batch_data_time
            time_escape_total += time_escape
            
        pbar.update(1)
        pbar.set_description(f"rtf_avg: {time_escape_total/time_speech_total:0.3f}")
        torch.cuda.empty_cache()
        return asr_result_list
    
    def generate_with_vad(self, input, input_len=None, **cfg):
        
        # step.1: compute the vad model
        self.vad_kwargs.update(cfg)
        beg_vad = time.time()
        res = self.generate(input, input_len=input_len, model=self.vad_model, kwargs=self.vad_kwargs, **cfg)
        end_vad = time.time()
        print(f"time cost vad: {end_vad - beg_vad:0.3f}")


        # step.2 compute asr model
        model = self.model
        kwargs = self.kwargs
        kwargs.update(cfg)
        batch_size = int(kwargs.get("batch_size_s", 300))*1000
        batch_size_threshold_ms = int(kwargs.get("batch_size_threshold_s", 60))*1000
        kwargs["batch_size"] = batch_size
        
        key_list, data_list = prepare_data_iterator(input, input_len=input_len, data_type=kwargs.get("data_type", None))
        results_ret_list = []
        time_speech_total_all_samples = 0.0

        beg_total = time.time()
        pbar_total = tqdm(colour="red", total=len(res) + 1, dynamic_ncols=True)
        for i in range(len(res)):
            key = res[i]["key"]
            vadsegments = res[i]["value"]
            input_i = data_list[i]
            speech = load_audio_text_image_video(input_i, fs=kwargs["frontend"].fs, audio_fs=kwargs.get("fs", 16000))
            speech_lengths = len(speech)
            n = len(vadsegments)
            data_with_index = [(vadsegments[i], i) for i in range(n)]
            sorted_data = sorted(data_with_index, key=lambda x: x[0][1] - x[0][0])
            results_sorted = []
            
            if not len(sorted_data):
                logging.info("decoding, utt: {}, empty speech".format(key))
                continue

            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])
            
            batch_size_ms_cum = 0
            beg_idx = 0
            beg_asr_total = time.time()
            time_speech_total_per_sample = speech_lengths/16000
            time_speech_total_all_samples += time_speech_total_per_sample

            for j, _ in enumerate(range(0, n)):
                batch_size_ms_cum += (sorted_data[j][0][1] - sorted_data[j][0][0])
                if j < n - 1 and (
                    batch_size_ms_cum + sorted_data[j + 1][0][1] - sorted_data[j + 1][0][0]) < batch_size and (
                    sorted_data[j + 1][0][1] - sorted_data[j + 1][0][0]) < batch_size_threshold_ms:
                    continue
                batch_size_ms_cum = 0
                end_idx = j + 1
                speech_j, speech_lengths_j = slice_padding_audio_samples(speech, speech_lengths, sorted_data[beg_idx:end_idx])       
                results = self.generate(speech_j, input_len=None, model=model, kwargs=kwargs, **cfg)
                if self.spk_model is not None:
                    all_segments = []
                    # compose vad segments: [[start_time_sec, end_time_sec, speech], [...]]
                    for _b in range(len(speech_j)):
                        vad_segments = [[sorted_data[beg_idx:end_idx][_b][0][0]/1000.0, \
                                        sorted_data[beg_idx:end_idx][_b][0][1]/1000.0, \
                                        speech_j[_b]]]
                        segments = sv_chunk(vad_segments)
                        all_segments.extend(segments)
                        speech_b = [i[2] for i in segments]
                        spk_res = self.generate(speech_b, input_len=None, model=self.spk_model, kwargs=kwargs, **cfg)
                        results[_b]['spk_embedding'] = spk_res[0]['spk_embedding']
                beg_idx = end_idx
                if len(results) < 1:
                    continue
                results_sorted.extend(results)


            pbar_total.update(1)
            end_asr_total = time.time()
            time_escape_total_per_sample = end_asr_total - beg_asr_total
            pbar_total.set_description(f"rtf_avg_per_sample: {time_escape_total_per_sample / time_speech_total_per_sample:0.3f}, "
                                 f"time_speech_total_per_sample: {time_speech_total_per_sample: 0.3f}, "
                                 f"time_escape_total_per_sample: {time_escape_total_per_sample:0.3f}")

            restored_data = [0] * n
            for j in range(n):
                index = sorted_data[j][1]
                restored_data[index] = results_sorted[j]
            result = {}
            
            # results combine for texts, timestamps, speaker embeddings and others
            # TODO: rewrite for clean code
            for j in range(n):
                for k, v in restored_data[j].items():
                    if k.startswith("timestamp"):
                        if k not in result:
                            result[k] = []
                        for t in restored_data[j][k]:
                            t[0] += vadsegments[j][0]
                            t[1] += vadsegments[j][0]
                        result[k].extend(restored_data[j][k])
                    elif k == 'spk_embedding':
                        if k not in result:
                            result[k] = restored_data[j][k]
                        else:
                            result[k] = torch.cat([result[k], restored_data[j][k]], dim=0)
                    elif k == 'text':
                        if k not in result:
                            result[k] = restored_data[j][k]
                        else:
                            result[k] += " " + restored_data[j][k]
                    else:
                        if k not in result:
                            result[k] = restored_data[j][k]
                        else:
                            result[k] += restored_data[j][k]
                            
            # step.3 compute punc model
            if self.punc_model is not None:
                self.punc_kwargs.update(cfg)
                punc_res = self.generate(result["text"], model=self.punc_model, kwargs=self.punc_kwargs, **cfg)
                result["text_with_punc"] = punc_res[0]["text"]
                     
            # speaker embedding cluster after resorted
            if self.spk_model is not None:
                all_segments = sorted(all_segments, key=lambda x: x[0])
                spk_embedding = result['spk_embedding']
                labels = self.cb_model(spk_embedding, oracle_num=self.preset_spk_num)
                del result['spk_embedding']
                sv_output = postprocess(all_segments, None, labels, spk_embedding.cpu())
                if self.spk_mode == 'vad_segment':
                    sentence_list = []
                    for res, vadsegment in zip(restored_data, vadsegments):
                        sentence_list.append({"start": vadsegment[0],\
                                                "end": vadsegment[1],
                                                "sentence": res['text'],
                                                "timestamp": res['timestamp']})
                else: # punc_segment
                    sentence_list = timestamp_sentence(punc_res[0]['punc_array'], \
                                                        result['timestamp'], \
                                                        result['text'])
                distribute_spk(sentence_list, sv_output)
                result['sentence_info'] = sentence_list
                    
            result["key"] = key
            results_ret_list.append(result)
            pbar_total.update(1)
            
        pbar_total.update(1)
        end_total = time.time()
        time_escape_total_all_samples = end_total - beg_total
        pbar_total.set_description(f"rtf_avg_all_samples: {time_escape_total_all_samples / time_speech_total_all_samples:0.3f}, "
                             f"time_speech_total_all_samples: {time_speech_total_all_samples: 0.3f}, "
                             f"time_escape_total_all_samples: {time_escape_total_all_samples:0.3f}")
        return results_ret_list


class AutoFrontend:
    def __init__(self, **kwargs):
        assert "model" in kwargs
        if "model_conf" not in kwargs:
            logging.info("download models from model hub: {}".format(kwargs.get("model_hub", "ms")))
            kwargs = download_model(**kwargs)
        
        # build frontend
        frontend = kwargs.get("frontend", None)
        if frontend is not None:
            frontend_class = tables.frontend_classes.get(frontend)
            frontend = frontend_class(**kwargs["frontend_conf"])

        self.frontend = frontend
        if "frontend" in kwargs:
            del kwargs["frontend"]
        self.kwargs = kwargs

    
    def __call__(self, input, input_len=None, kwargs=None, **cfg):
        
        kwargs = self.kwargs if kwargs is None else kwargs
        kwargs.update(cfg)


        key_list, data_list = prepare_data_iterator(input, input_len=input_len)
        batch_size = kwargs.get("batch_size", 1)
        device = kwargs.get("device", "cpu")
        if device == "cpu":
            batch_size = 1
        
        meta_data = {}
        
        result_list = []
        num_samples = len(data_list)
        pbar = tqdm(colour="blue", total=num_samples + 1, dynamic_ncols=True)
        
        time0 = time.perf_counter()
        for beg_idx in range(0, num_samples, batch_size):
            end_idx = min(num_samples, beg_idx + batch_size)
            data_batch = data_list[beg_idx:end_idx]
            key_batch = key_list[beg_idx:end_idx]

            # extract fbank feats
            time1 = time.perf_counter()
            audio_sample_list = load_audio_text_image_video(data_batch, fs=self.frontend.fs, audio_fs=kwargs.get("fs", 16000))
            time2 = time.perf_counter()
            meta_data["load_data"] = f"{time2 - time1:0.3f}"
            speech, speech_lengths = extract_fbank(audio_sample_list, data_type=kwargs.get("data_type", "sound"),
                                                   frontend=self.frontend, **kwargs)
            time3 = time.perf_counter()
            meta_data["extract_feat"] = f"{time3 - time2:0.3f}"
            meta_data["batch_data_time"] = speech_lengths.sum().item() * self.frontend.frame_shift * self.frontend.lfr_n / 1000
            
            speech.to(device=device), speech_lengths.to(device=device)
            batch = {"input": speech, "input_len": speech_lengths, "key": key_batch}
            result_list.append(batch)
            
            pbar.update(1)
            description = (
                f"{meta_data}, "
            )
            pbar.set_description(description)
        
        time_end = time.perf_counter()
        pbar.set_description(f"time escaped total: {time_end - time0:0.3f}")
        
        return result_list


if __name__ == '__main__':