Merge pull request #571 from alibaba-damo-academy/dev_lhn

hnluo

2023-05-31 4ce2e2d76c66ddecd6903f4ffce98eaba675c685

Merge pull request #571 from alibaba-damo-academy/dev_lhn

Dev lhn

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/README.md

New file
@@ -0,0 +1 @@
../../TEMPLATE/README.md

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/demo.py

@@ -1,39 +1,12 @@
import os
import logging
import torch
import soundfile

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
from modelscope.utils.logger import get_logger

logger = get_logger(log_level=logging.CRITICAL)
logger.setLevel(logging.CRITICAL)

os.environ["MODELSCOPE_CACHE"] = "./"
inference_pipeline = pipeline(
    task=Tasks.auto_speech_recognition,
    model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
    model_revision='v1.0.4'
    model_revision='v1.0.6',
    mode="paraformer_fake_streaming"
)

model_dir = os.path.join(os.environ["MODELSCOPE_CACHE"], "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online")
speech, sample_rate = soundfile.read(os.path.join(model_dir, "example/asr_example.wav"))
speech_length = speech.shape[0]

sample_offset = 0
chunk_size = [5, 10, 5] #[5, 10, 5] 600ms, [8, 8, 4] 480ms
stride_size =  chunk_size[1] * 960
param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size}
final_result = ""

for sample_offset in range(0, speech_length, min(stride_size, speech_length - sample_offset)):
    if sample_offset + stride_size >= speech_length - 1:
        stride_size = speech_length - sample_offset
        param_dict["is_final"] = True
    rec_result = inference_pipeline(audio_in=speech[sample_offset: sample_offset + stride_size],
                                    param_dict=param_dict)
    if len(rec_result) != 0:
        final_result += rec_result['text']
        print(rec_result)
print(final_result)
audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav'
rec_result = inference_pipeline(audio_in=audio_in)
print(rec_result)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/demo_online.py

New file
@@ -0,0 +1,40 @@
import os
import logging
import torch
import soundfile

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
from modelscope.utils.logger import get_logger

logger = get_logger(log_level=logging.CRITICAL)
logger.setLevel(logging.CRITICAL)

os.environ["MODELSCOPE_CACHE"] = "./"
inference_pipeline = pipeline(
    task=Tasks.auto_speech_recognition,
    model='damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online',
    model_revision='v1.0.6',
    mode="paraformer_streaming"
)

model_dir = os.path.join(os.environ["MODELSCOPE_CACHE"], "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online")
speech, sample_rate = soundfile.read(os.path.join(model_dir, "example/asr_example.wav"))
speech_length = speech.shape[0]

sample_offset = 0
chunk_size = [5, 10, 5] #[5, 10, 5] 600ms, [8, 8, 4] 480ms
stride_size =  chunk_size[1] * 960
param_dict = {"cache": dict(), "is_final": False, "chunk_size": chunk_size}
final_result = ""

for sample_offset in range(0, speech_length, min(stride_size, speech_length - sample_offset)):
    if sample_offset + stride_size >= speech_length - 1:
        stride_size = speech_length - sample_offset
        param_dict["is_final"] = True
    rec_result = inference_pipeline(audio_in=speech[sample_offset: sample_offset + stride_size],
                                    param_dict=param_dict)
    if len(rec_result) != 0:
        final_result += rec_result['text']
        print(rec_result)
print(final_result)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/finetune.py

New file
@@ -0,0 +1,37 @@
import os

from modelscope.metainfo import Trainers
from modelscope.trainers import build_trainer

from funasr.datasets.ms_dataset import MsDataset
from funasr.utils.modelscope_param import modelscope_args


def modelscope_finetune(params):
    if not os.path.exists(params.output_dir):
        os.makedirs(params.output_dir, exist_ok=True)
    # dataset split ["train", "validation"]
    ds_dict = MsDataset.load(params.data_path)
    kwargs = dict(
        model=params.model,
        model_revision='v1.0.6',
        data_dir=ds_dict,
        dataset_type=params.dataset_type,
        work_dir=params.output_dir,
        batch_bins=params.batch_bins,
        max_epoch=params.max_epoch,
        lr=params.lr)
    trainer = build_trainer(Trainers.speech_asr_trainer, default_args=kwargs)
    trainer.train()


if __name__ == '__main__':
    params = modelscope_args(model="damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online", data_path="./data")
    params.output_dir = "./checkpoint"              # m模型保存路径
    params.data_path = "./example_data/"            # 数据路径
    params.dataset_type = "small"                   # 小数据量设置small，若数据量大于1000小时，请使用large
    params.batch_bins = 1000                       # batch size，如果dataset_type="small"，batch_bins单位为fbank特征帧数，如果dataset_type="large"，batch_bins单位为毫秒，
    params.max_epoch = 20                           # 最大训练轮数
    params.lr = 0.00005                             # 设置学习率
    
    modelscope_finetune(params)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/infer.py

New file
@@ -0,0 +1,32 @@
import os
import shutil
import argparse
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

def modelscope_infer(args):
    os.environ['CUDA_VISIBLE_DEVICES'] = str(args.gpuid)
    inference_pipeline = pipeline(
        task=Tasks.auto_speech_recognition,
        model=args.model,
        output_dir=args.output_dir,
        batch_size=args.batch_size,
        model_revision='v1.0.6',
        mode="paraformer_fake_streaming",
        param_dict={"decoding_model": args.decoding_mode, "hotword": args.hotword_txt}
    )
    inference_pipeline(audio_in=args.audio_in)

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument('--model', type=str, default="damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch")
    parser.add_argument('--audio_in', type=str, default="./data/test/wav.scp")
    parser.add_argument('--output_dir', type=str, default="./results/")
    parser.add_argument('--decoding_mode', type=str, default="normal")
    parser.add_argument('--model_revision', type=str, default=None)
    parser.add_argument('--mode', type=str, default=None)
    parser.add_argument('--hotword_txt', type=str, default=None)
    parser.add_argument('--batch_size', type=int, default=64)
    parser.add_argument('--gpuid', type=str, default="0")
    args = parser.parse_args()
    modelscope_infer(args)

 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/infer.sh

New file
@@ -0,0 +1,104 @@
#!/usr/bin/env bash

set -e
set -u
set -o pipefail

stage=1
stop_stage=2
model="damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online"
data_dir="./data/test"
output_dir="./results"
batch_size=32
gpu_inference=true    # whether to perform gpu decoding
gpuid_list="0,1"    # set gpus, e.g., gpuid_list="0,1"
njob=32    # the number of jobs for CPU decoding, if gpu_inference=false, use CPU decoding, please set njob
checkpoint_dir=
checkpoint_name="valid.cer_ctc.ave.pb"

. utils/parse_options.sh || exit 1;

if ${gpu_inference} == "true"; then
    nj=$(echo $gpuid_list | awk -F "," '{print NF}')
else
    nj=$njob
    batch_size=1
    gpuid_list=""
    for JOB in $(seq ${nj}); do
        gpuid_list=$gpuid_list"-1,"
    done
fi

mkdir -p $output_dir/split
split_scps=""
for JOB in $(seq ${nj}); do
    split_scps="$split_scps $output_dir/split/wav.$JOB.scp"
done
perl utils/split_scp.pl ${data_dir}/wav.scp ${split_scps}

if [ -n "${checkpoint_dir}" ]; then
  python utils/prepare_checkpoint.py ${model} ${checkpoint_dir} ${checkpoint_name}
  model=${checkpoint_dir}/${model}
fi

if [ $stage -le 1 ] && [ $stop_stage -ge 1 ];then
    echo "Decoding ..."
    gpuid_list_array=(${gpuid_list//,/ })
    for JOB in $(seq ${nj}); do
        {
        id=$((JOB-1))
        gpuid=${gpuid_list_array[$id]}
        mkdir -p ${output_dir}/output.$JOB
        python infer.py \
            --model ${model} \
            --audio_in ${output_dir}/split/wav.$JOB.scp \
            --output_dir ${output_dir}/output.$JOB \
            --batch_size ${batch_size} \
            --gpuid ${gpuid}
            --mode "paraformer_fake_streaming"
        }&
    done
    wait

    mkdir -p ${output_dir}/1best_recog
    for f in token score text; do
        if [ -f "${output_dir}/output.1/1best_recog/${f}" ]; then
          for i in $(seq "${nj}"); do
              cat "${output_dir}/output.${i}/1best_recog/${f}"
          done | sort -k1 >"${output_dir}/1best_recog/${f}"
        fi
    done
fi

if [ $stage -le 2 ] && [ $stop_stage -ge 2 ];then
    echo "Computing WER ..."
    cp ${output_dir}/1best_recog/text ${output_dir}/1best_recog/text.proc
    cp ${data_dir}/text ${output_dir}/1best_recog/text.ref
    python utils/compute_wer.py ${output_dir}/1best_recog/text.ref ${output_dir}/1best_recog/text.proc ${output_dir}/1best_recog/text.cer
    tail -n 3 ${output_dir}/1best_recog/text.cer
fi

if [ $stage -le 3 ] && [ $stop_stage -ge 3 ];then
    echo "SpeechIO TIOBE textnorm"
    echo "$0 --> Normalizing REF text ..."
    ./utils/textnorm_zh.py \
        --has_key --to_upper \
        ${data_dir}/text \
        ${output_dir}/1best_recog/ref.txt

    echo "$0 --> Normalizing HYP text ..."
    ./utils/textnorm_zh.py \
        --has_key --to_upper \
        ${output_dir}/1best_recog/text.proc \
        ${output_dir}/1best_recog/rec.txt
    grep -v $'\t$' ${output_dir}/1best_recog/rec.txt > ${output_dir}/1best_recog/rec_non_empty.txt

    echo "$0 --> computing WER/CER and alignment ..."
    ./utils/error_rate_zh \
        --tokenizer char \
        --ref ${output_dir}/1best_recog/ref.txt \
        --hyp ${output_dir}/1best_recog/rec_non_empty.txt \
        ${output_dir}/1best_recog/DETAILS.txt | tee ${output_dir}/1best_recog/RESULTS.txt
    rm -rf ${output_dir}/1best_recog/rec.txt ${output_dir}/1best_recog/rec_non_empty.txt
fi


 egs_modelscope/asr/paraformer/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/utils

New file
@@ -0,0 +1 @@
../../TEMPLATE/utils/

 funasr/bin/asr_infer.py

@@ -305,6 +305,7 @@
            nbest: int = 1,
            frontend_conf: dict = None,
            hotword_list_or_file: str = None,
            decoding_ind: int = 0,
            **kwargs,
    ):
        assert check_argument_types()
@@ -415,6 +416,7 @@
        self.nbest = nbest
        self.frontend = frontend
        self.encoder_downsampling_factor = 1
        self.decoding_ind = decoding_ind
        if asr_train_args.encoder == "data2vec_encoder" or asr_train_args.encoder_conf["input_layer"] == "conv2d":
            self.encoder_downsampling_factor = 4

@@ -452,7 +454,7 @@
        batch = to_device(batch, device=self.device)

        # b. Forward Encoder
        enc, enc_len = self.asr_model.encode(**batch)
        enc, enc_len = self.asr_model.encode(**batch, ind=self.decoding_ind)
        if isinstance(enc, tuple):
            enc = enc[0]
        # assert len(enc) == 1, len(enc)

 funasr/bin/asr_inference_launch.py

@@ -1638,6 +1638,8 @@
        return inference_uniasr(**kwargs)
    elif mode == "paraformer":
        return inference_paraformer(**kwargs)
    elif mode == "paraformer_fake_streaming":
        return inference_paraformer(**kwargs)
    elif mode == "paraformer_streaming":
        return inference_paraformer_online(**kwargs)
    elif mode.startswith("paraformer_vad"):
@@ -1920,4 +1922,4 @@


if __name__ == "__main__":
    main()
    main()

 funasr/bin/build_trainer.py

@@ -23,6 +23,8 @@
        from funasr.tasks.asr import ASRTask as ASRTask
    elif mode == "paraformer":
        from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    elif mode == "paraformer_streaming":
        from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    elif mode == "paraformer_vad_punc":
        from funasr.tasks.asr import ASRTaskParaformer as ASRTask
    elif mode == "uniasr":

 funasr/build_utils/build_asr_model.py

@@ -23,7 +23,7 @@
from funasr.models.joint_net.joint_network import JointNetwork
from funasr.models.e2e_asr import ASRModel
from funasr.models.e2e_asr_mfcca import MFCCA
from funasr.models.e2e_asr_paraformer import Paraformer, ParaformerBert, BiCifParaformer, ContextualParaformer
from funasr.models.e2e_asr_paraformer import Paraformer, ParaformerOnline, ParaformerBert, BiCifParaformer, ContextualParaformer
from funasr.models.e2e_tp import TimestampPredictor
from funasr.models.e2e_uni_asr import UniASR
from funasr.models.e2e_asr_transducer import TransducerModel, UnifiedTransducerModel
@@ -82,6 +82,7 @@
        asr=ASRModel,
        uniasr=UniASR,
        paraformer=Paraformer,
        paraformer_online=ParaformerOnline,
        paraformer_bert=ParaformerBert,
        bicif_paraformer=BiCifParaformer,
        contextual_paraformer=ContextualParaformer,
@@ -293,7 +294,7 @@
            token_list=token_list,
            **args.model_conf,
        )
    elif args.model in ["paraformer", "paraformer_bert", "bicif_paraformer", "contextual_paraformer"]:
    elif args.model in ["paraformer", "paraformer_online", "paraformer_bert", "bicif_paraformer", "contextual_paraformer"]:
        # predictor
        predictor_class = predictor_choices.get_class(args.predictor)
        predictor = predictor_class(**args.predictor_conf)

 funasr/models/decoder/sanm_decoder.py

@@ -935,6 +935,7 @@
        hlens: torch.Tensor,
        ys_in_pad: torch.Tensor,
        ys_in_lens: torch.Tensor,
        chunk_mask: torch.Tensor = None,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """Forward decoder.

@@ -955,9 +956,13 @@
        """
        tgt = ys_in_pad
        tgt_mask = myutils.sequence_mask(ys_in_lens, device=tgt.device)[:, :, None]

        
        memory = hs_pad
        memory_mask = myutils.sequence_mask(hlens, device=memory.device)[:, None, :]
        if chunk_mask is not None:
            memory_mask = memory_mask * chunk_mask
            if tgt_mask.size(1) != memory_mask.size(1):
                memory_mask = torch.cat((memory_mask, memory_mask[:, -2:-1, :]), dim=1)

        x = tgt
        x, tgt_mask, memory, memory_mask, _ = self.decoders(

 funasr/models/e2e_asr_paraformer.py

@@ -153,6 +153,7 @@
            speech_lengths: torch.Tensor,
            text: torch.Tensor,
            text_lengths: torch.Tensor,
            decoding_ind: int = None,
    ) -> Tuple[torch.Tensor, Dict[str, torch.Tensor], torch.Tensor]:
        """Frontend + Encoder + Decoder + Calc loss
        Args:
@@ -160,6 +161,7 @@
                speech_lengths: (Batch, )
                text: (Batch, Length)
                text_lengths: (Batch,)
                decoding_ind: int
        """
        assert text_lengths.dim() == 1, text_lengths.shape
        # Check that batch_size is unified
@@ -176,7 +178,11 @@
        speech = speech[:, :speech_lengths.max()]

        # 1. Encoder
        encoder_out, encoder_out_lens = self.encode(speech, speech_lengths)
        if hasattr(self.encoder, "overlap_chunk_cls"):
            ind = self.encoder.overlap_chunk_cls.random_choice(self.training, decoding_ind)
            encoder_out, encoder_out_lens = self.encode(speech, speech_lengths, ind=ind)
        else:
            encoder_out, encoder_out_lens = self.encode(speech, speech_lengths)
        intermediate_outs = None
        if isinstance(encoder_out, tuple):
            intermediate_outs = encoder_out[1]
@@ -272,12 +278,13 @@
        return {"feats": feats, "feats_lengths": feats_lengths}

    def encode(
            self, speech: torch.Tensor, speech_lengths: torch.Tensor
            self, speech: torch.Tensor, speech_lengths: torch.Tensor, ind: int = 0,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """Frontend + Encoder. Note that this method is used by asr_inference.py
        Args:
                speech: (Batch, Length, ...)
                speech_lengths: (Batch, )
                ind: int
        """
        with autocast(False):
            # 1. Extract feats
@@ -299,11 +306,25 @@
        # feats: (Batch, Length, Dim)
        # -> encoder_out: (Batch, Length2, Dim2)
        if self.encoder.interctc_use_conditioning:
            encoder_out, encoder_out_lens, _ = self.encoder(
                feats, feats_lengths, ctc=self.ctc
            )
            if hasattr(self.encoder, "overlap_chunk_cls"):
                encoder_out, encoder_out_lens, _ = self.encoder(
                    feats, feats_lengths, ctc=self.ctc, ind=ind
                )
                encoder_out, encoder_out_lens = self.encoder.overlap_chunk_cls.remove_chunk(encoder_out,
                                                                                            encoder_out_lens,
                                                                                            chunk_outs=None)
            else:
                encoder_out, encoder_out_lens, _ = self.encoder(
                    feats, feats_lengths, ctc=self.ctc
                )
        else:
            encoder_out, encoder_out_lens, _ = self.encoder(feats, feats_lengths)
            if hasattr(self.encoder, "overlap_chunk_cls"):
                encoder_out, encoder_out_lens, _ = self.encoder(feats, feats_lengths, ind=ind)
                encoder_out, encoder_out_lens = self.encoder.overlap_chunk_cls.remove_chunk(encoder_out,
                                                                                            encoder_out_lens,
                                                                                            chunk_outs=None)
            else:
                encoder_out, encoder_out_lens, _ = self.encoder(feats, feats_lengths)
        intermediate_outs = None
        if isinstance(encoder_out, tuple):
            intermediate_outs = encoder_out[1]
@@ -592,9 +613,137 @@
    """

    def __init__(
            self, *args, **kwargs,
            self,
            vocab_size: int,
            token_list: Union[Tuple[str, ...], List[str]],
            frontend: Optional[AbsFrontend],
            specaug: Optional[AbsSpecAug],
            normalize: Optional[AbsNormalize],
            encoder: AbsEncoder,
            decoder: AbsDecoder,
            ctc: CTC,
            ctc_weight: float = 0.5,
            interctc_weight: float = 0.0,
            ignore_id: int = -1,
            blank_id: int = 0,
            sos: int = 1,
            eos: int = 2,
            lsm_weight: float = 0.0,
            length_normalized_loss: bool = False,
            report_cer: bool = True,
            report_wer: bool = True,
            sym_space: str = "<space>",
            sym_blank: str = "<blank>",
            extract_feats_in_collect_stats: bool = True,
            predictor=None,
            predictor_weight: float = 0.0,
            predictor_bias: int = 0,
            sampling_ratio: float = 0.2,
            decoder_attention_chunk_type: str = 'chunk',
            share_embedding: bool = False,
            preencoder: Optional[AbsPreEncoder] = None,
            postencoder: Optional[AbsPostEncoder] = None,
            use_1st_decoder_loss: bool = False,
    ):
        super().__init__(*args, **kwargs)
        assert check_argument_types()
        assert 0.0 <= ctc_weight <= 1.0, ctc_weight
        assert 0.0 <= interctc_weight < 1.0, interctc_weight

        super().__init__(
            vocab_size=vocab_size,
            token_list=token_list,
            frontend=frontend,
            specaug=specaug,
            normalize=normalize,
            preencoder=preencoder,
            encoder=encoder,
            postencoder=postencoder,
            decoder=decoder,
            ctc=ctc,
            ctc_weight=ctc_weight,
            interctc_weight=interctc_weight,
            ignore_id=ignore_id,
            blank_id=blank_id,
            sos=sos,
            eos=eos,
            lsm_weight=lsm_weight,
            length_normalized_loss=length_normalized_loss,
            report_cer=report_cer,
            report_wer=report_wer,
            sym_space=sym_space,
            sym_blank=sym_blank,
            extract_feats_in_collect_stats=extract_feats_in_collect_stats,
            predictor=predictor,
            predictor_weight=predictor_weight,
            predictor_bias=predictor_bias,
            sampling_ratio=sampling_ratio,
        )
        # note that eos is the same as sos (equivalent ID)
        self.blank_id = blank_id
        self.sos = vocab_size - 1 if sos is None else sos
        self.eos = vocab_size - 1 if eos is None else eos
        self.vocab_size = vocab_size
        self.ignore_id = ignore_id
        self.ctc_weight = ctc_weight
        self.interctc_weight = interctc_weight
        self.token_list = token_list.copy()

        self.frontend = frontend
        self.specaug = specaug
        self.normalize = normalize
        self.preencoder = preencoder
        self.postencoder = postencoder
        self.encoder = encoder

        if not hasattr(self.encoder, "interctc_use_conditioning"):
            self.encoder.interctc_use_conditioning = False
        if self.encoder.interctc_use_conditioning:
            self.encoder.conditioning_layer = torch.nn.Linear(
                vocab_size, self.encoder.output_size()
            )

        self.error_calculator = None

        if ctc_weight == 1.0:
            self.decoder = None
        else:
            self.decoder = decoder

        self.criterion_att = LabelSmoothingLoss(
            size=vocab_size,
            padding_idx=ignore_id,
            smoothing=lsm_weight,
            normalize_length=length_normalized_loss,
        )

        if report_cer or report_wer:
            self.error_calculator = ErrorCalculator(
                token_list, sym_space, sym_blank, report_cer, report_wer
            )

        if ctc_weight == 0.0:
            self.ctc = None
        else:
            self.ctc = ctc

        self.extract_feats_in_collect_stats = extract_feats_in_collect_stats
        self.predictor = predictor
        self.predictor_weight = predictor_weight
        self.predictor_bias = predictor_bias
        self.sampling_ratio = sampling_ratio
        self.criterion_pre = mae_loss(normalize_length=length_normalized_loss)
        self.step_cur = 0
        self.scama_mask = None
        if hasattr(self.encoder, "overlap_chunk_cls") and self.encoder.overlap_chunk_cls is not None:
            from funasr.modules.streaming_utils.chunk_utilis import build_scama_mask_for_cross_attention_decoder
            self.build_scama_mask_for_cross_attention_decoder_fn = build_scama_mask_for_cross_attention_decoder
            self.decoder_attention_chunk_type = decoder_attention_chunk_type

        self.share_embedding = share_embedding
        if self.share_embedding:
            self.decoder.embed = None

        self.use_1st_decoder_loss = use_1st_decoder_loss

    def forward(
            self,
@@ -602,6 +751,7 @@
            speech_lengths: torch.Tensor,
            text: torch.Tensor,
            text_lengths: torch.Tensor,
            decoding_ind: int = None,
    ) -> Tuple[torch.Tensor, Dict[str, torch.Tensor], torch.Tensor]:
        """Frontend + Encoder + Decoder + Calc loss
        Args:
@@ -609,6 +759,7 @@
                speech_lengths: (Batch, )
                text: (Batch, Length)
                text_lengths: (Batch,)
                decoding_ind: int
        """
        assert text_lengths.dim() == 1, text_lengths.shape
        # Check that batch_size is unified
@@ -625,7 +776,11 @@
        speech = speech[:, :speech_lengths.max()]

        # 1. Encoder
        encoder_out, encoder_out_lens = self.encode(speech, speech_lengths)
        if hasattr(self.encoder, "overlap_chunk_cls"):
            ind = self.encoder.overlap_chunk_cls.random_choice(self.training, decoding_ind)
            encoder_out, encoder_out_lens = self.encode(speech, speech_lengths, ind=ind)
        else:
            encoder_out, encoder_out_lens = self.encode(speech, speech_lengths)
        intermediate_outs = None
        if isinstance(encoder_out, tuple):
            intermediate_outs = encoder_out[1]
@@ -638,8 +793,12 @@

        # 1. CTC branch
        if self.ctc_weight != 0.0:
            if hasattr(self.encoder, "overlap_chunk_cls"):
                encoder_out_ctc, encoder_out_lens_ctc = self.encoder.overlap_chunk_cls.remove_chunk(encoder_out,
                                                                                                    encoder_out_lens,
                                                                                                    chunk_outs=None)
            loss_ctc, cer_ctc = self._calc_ctc_loss(
                encoder_out, encoder_out_lens, text, text_lengths
                encoder_out_ctc, encoder_out_lens_ctc, text, text_lengths
            )

            # Collect CTC branch stats
@@ -652,8 +811,14 @@
            for layer_idx, intermediate_out in intermediate_outs:
                # we assume intermediate_out has the same length & padding
                # as those of encoder_out
                if hasattr(self.encoder, "overlap_chunk_cls"):
                    encoder_out_ctc, encoder_out_lens_ctc = \
                        self.encoder.overlap_chunk_cls.remove_chunk(
                            intermediate_out,
                            encoder_out_lens,
                            chunk_outs=None)
                loss_ic, cer_ic = self._calc_ctc_loss(
                    intermediate_out, encoder_out_lens, text, text_lengths
                    encoder_out_ctc, encoder_out_lens_ctc, text, text_lengths
                )
                loss_interctc = loss_interctc + loss_ic

@@ -672,7 +837,7 @@

        # 2b. Attention decoder branch
        if self.ctc_weight != 1.0:
            loss_att, acc_att, cer_att, wer_att, loss_pre = self._calc_att_loss(
            loss_att, acc_att, cer_att, wer_att, loss_pre, pre_loss_att = self._calc_att_predictor_loss(
                encoder_out, encoder_out_lens, text, text_lengths
            )

@@ -684,8 +849,12 @@
        else:
            loss = self.ctc_weight * loss_ctc + (1 - self.ctc_weight) * loss_att + loss_pre * self.predictor_weight

        if self.use_1st_decoder_loss and pre_loss_att is not None:
            loss = loss + pre_loss_att

        # Collect Attn branch stats
        stats["loss_att"] = loss_att.detach() if loss_att is not None else None
        stats["pre_loss_att"] = pre_loss_att.detach() if pre_loss_att is not None else None
        stats["acc"] = acc_att
        stats["cer"] = cer_att
        stats["wer"] = wer_att
@@ -697,14 +866,67 @@
        loss, stats, weight = force_gatherable((loss, stats, batch_size), loss.device)
        return loss, stats, weight

    def encode(
        self, speech: torch.Tensor, speech_lengths: torch.Tensor, ind: int = 0,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """Frontend + Encoder. Note that this method is used by asr_inference.py
        Args:
                        speech: (Batch, Length, ...)
                        speech_lengths: (Batch, )
        """
        with autocast(False):
            # 1. Extract feats
            feats, feats_lengths = self._extract_feats(speech, speech_lengths)

            # 2. Data augmentation
            if self.specaug is not None and self.training:
                feats, feats_lengths = self.specaug(feats, feats_lengths)

            # 3. Normalization for feature: e.g. Global-CMVN, Utterance-CMVN
            if self.normalize is not None:
                feats, feats_lengths = self.normalize(feats, feats_lengths)
        # Pre-encoder, e.g. used for raw input data
        if self.preencoder is not None:
            feats, feats_lengths = self.preencoder(feats, feats_lengths)
        
        # 4. Forward encoder
        # feats: (Batch, Length, Dim)
        # -> encoder_out: (Batch, Length2, Dim2)
        if self.encoder.interctc_use_conditioning:
            encoder_out, encoder_out_lens, _ = self.encoder(
                feats, feats_lengths, ctc=self.ctc, ind=ind
            )
        else:
            encoder_out, encoder_out_lens, _ = self.encoder(feats, feats_lengths, ind=ind)
        intermediate_outs = None
        if isinstance(encoder_out, tuple):
            intermediate_outs = encoder_out[1]
            encoder_out = encoder_out[0]

        # Post-encoder, e.g. NLU
        if self.postencoder is not None:
            encoder_out, encoder_out_lens = self.postencoder(
                encoder_out, encoder_out_lens
            )

        assert encoder_out.size(0) == speech.size(0), (
            encoder_out.size(),
            speech.size(0),
        )
        assert encoder_out.size(1) <= encoder_out_lens.max(), (
            encoder_out.size(),
            encoder_out_lens.max(),
        )

        if intermediate_outs is not None:
            return (encoder_out, intermediate_outs), encoder_out_lens

        return encoder_out, encoder_out_lens

    def encode_chunk(
            self, speech: torch.Tensor, speech_lengths: torch.Tensor, cache: dict = None
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        """Frontend + Encoder. Note that this method is used by asr_inference.py
<<<<<<< HEAD
=======

>>>>>>> 4cd79db451786548d8a100f25c3b03da0eb30f4b
        Args:
                speech: (Batch, Length, ...)
                speech_lengths: (Batch, )
@@ -750,11 +972,240 @@

        return encoder_out, torch.tensor([encoder_out.size(1)])

    def _calc_att_predictor_loss(
        self,
        encoder_out: torch.Tensor,
        encoder_out_lens: torch.Tensor,
        ys_pad: torch.Tensor,
        ys_pad_lens: torch.Tensor,
    ):
        encoder_out_mask = (~make_pad_mask(encoder_out_lens, maxlen=encoder_out.size(1))[:, None, :]).to(
            encoder_out.device)
        if self.predictor_bias == 1:
            _, ys_pad = add_sos_eos(ys_pad, self.sos, self.eos, self.ignore_id)
            ys_pad_lens = ys_pad_lens + self.predictor_bias
        mask_chunk_predictor = None
        if self.encoder.overlap_chunk_cls is not None:
            mask_chunk_predictor = self.encoder.overlap_chunk_cls.get_mask_chunk_predictor(None,
                                                                                           device=encoder_out.device,
                                                                                           batch_size=encoder_out.size(
                                                                                               0))
            mask_shfit_chunk = self.encoder.overlap_chunk_cls.get_mask_shfit_chunk(None, device=encoder_out.device,
                                                                                   batch_size=encoder_out.size(0))
            encoder_out = encoder_out * mask_shfit_chunk
        pre_acoustic_embeds, pre_token_length, pre_alphas, _ = self.predictor(encoder_out,
                                                                              ys_pad,
                                                                              encoder_out_mask,
                                                                              ignore_id=self.ignore_id,
                                                                              mask_chunk_predictor=mask_chunk_predictor,
                                                                              target_label_length=ys_pad_lens,
                                                                              )
        predictor_alignments, predictor_alignments_len = self.predictor.gen_frame_alignments(pre_alphas,
                                                                                             encoder_out_lens)

        scama_mask = None
        if self.encoder.overlap_chunk_cls is not None and self.decoder_attention_chunk_type == 'chunk':
            encoder_chunk_size = self.encoder.overlap_chunk_cls.chunk_size_pad_shift_cur
            attention_chunk_center_bias = 0
            attention_chunk_size = encoder_chunk_size
            decoder_att_look_back_factor = self.encoder.overlap_chunk_cls.decoder_att_look_back_factor_cur
            mask_shift_att_chunk_decoder = self.encoder.overlap_chunk_cls.\
                get_mask_shift_att_chunk_decoder(None,
                                                 device=encoder_out.device,
                                                 batch_size=encoder_out.size(0)
                                                 )
            scama_mask = self.build_scama_mask_for_cross_attention_decoder_fn(
                predictor_alignments=predictor_alignments,
                encoder_sequence_length=encoder_out_lens,
                chunk_size=1,
                encoder_chunk_size=encoder_chunk_size,
                attention_chunk_center_bias=attention_chunk_center_bias,
                attention_chunk_size=attention_chunk_size,
                attention_chunk_type=self.decoder_attention_chunk_type,
                step=None,
                predictor_mask_chunk_hopping=mask_chunk_predictor,
                decoder_att_look_back_factor=decoder_att_look_back_factor,
                mask_shift_att_chunk_decoder=mask_shift_att_chunk_decoder,
                target_length=ys_pad_lens,
                is_training=self.training,
            )
        elif self.encoder.overlap_chunk_cls is not None:
            encoder_out, encoder_out_lens = self.encoder.overlap_chunk_cls.remove_chunk(encoder_out,
                                                                                        encoder_out_lens,
                                                                                        chunk_outs=None)
        # 0. sampler
        decoder_out_1st = None
        pre_loss_att = None
        if self.sampling_ratio > 0.0:
            if self.step_cur < 2:
                logging.info("enable sampler in paraformer, sampling_ratio: {}".format(self.sampling_ratio))
            if self.use_1st_decoder_loss:
                sematic_embeds, decoder_out_1st, pre_loss_att = \
                    self.sampler_with_grad(encoder_out, encoder_out_lens, ys_pad,
                                           ys_pad_lens, pre_acoustic_embeds, scama_mask)
            else:
                sematic_embeds, decoder_out_1st = \
                    self.sampler(encoder_out, encoder_out_lens, ys_pad,
                                 ys_pad_lens, pre_acoustic_embeds, scama_mask)
        else:
            if self.step_cur < 2:
                logging.info("disable sampler in paraformer, sampling_ratio: {}".format(self.sampling_ratio))
            sematic_embeds = pre_acoustic_embeds

        # 1. Forward decoder
        decoder_outs = self.decoder(
            encoder_out, encoder_out_lens, sematic_embeds, ys_pad_lens, scama_mask
        )
        decoder_out, _ = decoder_outs[0], decoder_outs[1]

        if decoder_out_1st is None:
            decoder_out_1st = decoder_out
        # 2. Compute attention loss
        loss_att = self.criterion_att(decoder_out, ys_pad)
        acc_att = th_accuracy(
            decoder_out_1st.view(-1, self.vocab_size),
            ys_pad,
            ignore_label=self.ignore_id,
        )
        loss_pre = self.criterion_pre(ys_pad_lens.type_as(pre_token_length), pre_token_length)

        # Compute cer/wer using attention-decoder
        if self.training or self.error_calculator is None:
            cer_att, wer_att = None, None
        else:
            ys_hat = decoder_out_1st.argmax(dim=-1)
            cer_att, wer_att = self.error_calculator(ys_hat.cpu(), ys_pad.cpu())

        return loss_att, acc_att, cer_att, wer_att, loss_pre, pre_loss_att

    def sampler(self, encoder_out, encoder_out_lens, ys_pad, ys_pad_lens, pre_acoustic_embeds, chunk_mask=None):

        tgt_mask = (~make_pad_mask(ys_pad_lens, maxlen=ys_pad_lens.max())[:, :, None]).to(ys_pad.device)
        ys_pad_masked = ys_pad * tgt_mask[:, :, 0]
        if self.share_embedding:
            ys_pad_embed = self.decoder.output_layer.weight[ys_pad_masked]
        else:
            ys_pad_embed = self.decoder.embed(ys_pad_masked)
        with torch.no_grad():
            decoder_outs = self.decoder(
                encoder_out, encoder_out_lens, pre_acoustic_embeds, ys_pad_lens, chunk_mask
            )
            decoder_out, _ = decoder_outs[0], decoder_outs[1]
            pred_tokens = decoder_out.argmax(-1)
            nonpad_positions = ys_pad.ne(self.ignore_id)
            seq_lens = (nonpad_positions).sum(1)
            same_num = ((pred_tokens == ys_pad) & nonpad_positions).sum(1)
            input_mask = torch.ones_like(nonpad_positions)
            bsz, seq_len = ys_pad.size()
            for li in range(bsz):
                target_num = (((seq_lens[li] - same_num[li].sum()).float()) * self.sampling_ratio).long()
                if target_num > 0:
                    input_mask[li].scatter_(dim=0, index=torch.randperm(seq_lens[li])[:target_num].cuda(), value=0)
            input_mask = input_mask.eq(1)
            input_mask = input_mask.masked_fill(~nonpad_positions, False)
            input_mask_expand_dim = input_mask.unsqueeze(2).to(pre_acoustic_embeds.device)

        sematic_embeds = pre_acoustic_embeds.masked_fill(~input_mask_expand_dim, 0) + ys_pad_embed.masked_fill(
            input_mask_expand_dim, 0)
        return sematic_embeds * tgt_mask, decoder_out * tgt_mask

    def sampler_with_grad(self, encoder_out, encoder_out_lens, ys_pad, ys_pad_lens, pre_acoustic_embeds, chunk_mask=None):
        tgt_mask = (~make_pad_mask(ys_pad_lens, maxlen=ys_pad_lens.max())[:, :, None]).to(ys_pad.device)
        ys_pad_masked = ys_pad * tgt_mask[:, :, 0]
        if self.share_embedding:
            ys_pad_embed = self.decoder.output_layer.weight[ys_pad_masked]
        else:
            ys_pad_embed = self.decoder.embed(ys_pad_masked)
        decoder_outs = self.decoder(
            encoder_out, encoder_out_lens, pre_acoustic_embeds, ys_pad_lens, chunk_mask
        )
        pre_loss_att = self.criterion_att(decoder_outs[0], ys_pad)
        decoder_out, _ = decoder_outs[0], decoder_outs[1]
        pred_tokens = decoder_out.argmax(-1)
        nonpad_positions = ys_pad.ne(self.ignore_id)
        seq_lens = (nonpad_positions).sum(1)
        same_num = ((pred_tokens == ys_pad) & nonpad_positions).sum(1)
        input_mask = torch.ones_like(nonpad_positions)
        bsz, seq_len = ys_pad.size()
        for li in range(bsz):
            target_num = (((seq_lens[li] - same_num[li].sum()).float()) * self.sampling_ratio).long()
            if target_num > 0:
                input_mask[li].scatter_(dim=0, index=torch.randperm(seq_lens[li])[:target_num].cuda(), value=0)
        input_mask = input_mask.eq(1)
        input_mask = input_mask.masked_fill(~nonpad_positions, False)
        input_mask_expand_dim = input_mask.unsqueeze(2).to(pre_acoustic_embeds.device)

        sematic_embeds = pre_acoustic_embeds.masked_fill(~input_mask_expand_dim, 0) + ys_pad_embed.masked_fill(
            input_mask_expand_dim, 0)

        return sematic_embeds * tgt_mask, decoder_out * tgt_mask, pre_loss_att

    def calc_predictor(self, encoder_out, encoder_out_lens):

        encoder_out_mask = (~make_pad_mask(encoder_out_lens, maxlen=encoder_out.size(1))[:, None, :]).to(
            encoder_out.device)
        mask_chunk_predictor = None
        if self.encoder.overlap_chunk_cls is not None:
            mask_chunk_predictor = self.encoder.overlap_chunk_cls.get_mask_chunk_predictor(None,
                                                                                           device=encoder_out.device,
                                                                                           batch_size=encoder_out.size(
                                                                                               0))
            mask_shfit_chunk = self.encoder.overlap_chunk_cls.get_mask_shfit_chunk(None, device=encoder_out.device,
                                                                                   batch_size=encoder_out.size(0))
            encoder_out = encoder_out * mask_shfit_chunk
        pre_acoustic_embeds, pre_token_length, pre_alphas, pre_peak_index = self.predictor(encoder_out,
                                                                                           None,
                                                                                           encoder_out_mask,
                                                                                           ignore_id=self.ignore_id,
                                                                                           mask_chunk_predictor=mask_chunk_predictor,
                                                                                           target_label_length=None,
                                                                                           )
        predictor_alignments, predictor_alignments_len = self.predictor.gen_frame_alignments(pre_alphas,
                                                                                             encoder_out_lens+1 if self.predictor.tail_threshold > 0.0 else encoder_out_lens)

        scama_mask = None
        if self.encoder.overlap_chunk_cls is not None and self.decoder_attention_chunk_type == 'chunk':
            encoder_chunk_size = self.encoder.overlap_chunk_cls.chunk_size_pad_shift_cur
            attention_chunk_center_bias = 0
            attention_chunk_size = encoder_chunk_size
            decoder_att_look_back_factor = self.encoder.overlap_chunk_cls.decoder_att_look_back_factor_cur
            mask_shift_att_chunk_decoder = self.encoder.overlap_chunk_cls.\
                get_mask_shift_att_chunk_decoder(None,
                                                 device=encoder_out.device,
                                                 batch_size=encoder_out.size(0)
                                                 )
            scama_mask = self.build_scama_mask_for_cross_attention_decoder_fn(
                predictor_alignments=predictor_alignments,
                encoder_sequence_length=encoder_out_lens,
                chunk_size=1,
                encoder_chunk_size=encoder_chunk_size,
                attention_chunk_center_bias=attention_chunk_center_bias,
                attention_chunk_size=attention_chunk_size,
                attention_chunk_type=self.decoder_attention_chunk_type,
                step=None,
                predictor_mask_chunk_hopping=mask_chunk_predictor,
                decoder_att_look_back_factor=decoder_att_look_back_factor,
                mask_shift_att_chunk_decoder=mask_shift_att_chunk_decoder,
                target_length=None,
                is_training=self.training,
            )
        self.scama_mask = scama_mask

        return pre_acoustic_embeds, pre_token_length, pre_alphas, pre_peak_index

    def calc_predictor_chunk(self, encoder_out, cache=None):

        pre_acoustic_embeds, pre_token_length = \
            self.predictor.forward_chunk(encoder_out, cache["encoder"])
        return pre_acoustic_embeds, pre_token_length

    def cal_decoder_with_predictor(self, encoder_out, encoder_out_lens, sematic_embeds, ys_pad_lens):
        decoder_outs = self.decoder(
            encoder_out, encoder_out_lens, sematic_embeds, ys_pad_lens, self.scama_mask
        )
        decoder_out = decoder_outs[0]
        decoder_out = torch.log_softmax(decoder_out, dim=-1)
        return decoder_out, ys_pad_lens

    def cal_decoder_with_predictor_chunk(self, encoder_out, sematic_embeds, cache=None):
        decoder_outs = self.decoder.forward_chunk(
@@ -1800,4 +2251,4 @@
                    "torch tensor: {}, {}, loading from tf tensor: {}, {}".format(name, data_tf.size(), name_tf,
                                                                                  var_dict_tf[name_tf].shape))

        return var_dict_torch_update
        return var_dict_torch_update

 funasr/models/encoder/sanm_encoder.py

@@ -633,6 +633,8 @@
                self.embed = torch.nn.Linear(input_size, output_size)
        elif input_layer == "pe":
            self.embed = SinusoidalPositionEncoder()
        elif input_layer == "pe_online":
            self.embed = StreamSinusoidalPositionEncoder()
        else:
            raise ValueError("unknown input_layer: " + input_layer)
        self.normalize_before = normalize_before
@@ -818,6 +820,59 @@
            return (xs_pad, intermediate_outs), olens, None
        return xs_pad, olens, None

    def _add_overlap_chunk(self, feats: np.ndarray, cache: dict = {}):
        if len(cache) == 0:
            return feats
        cache["feats"] = to_device(cache["feats"], device=feats.device)
        overlap_feats = torch.cat((cache["feats"], feats), dim=1)
        cache["feats"] = overlap_feats[:, -(cache["chunk_size"][0] + cache["chunk_size"][2]):, :]
        return overlap_feats

    def forward_chunk(self,
                      xs_pad: torch.Tensor,
                      ilens: torch.Tensor,
                      cache: dict = None,
                      ctc: CTC = None,
                      ):
        xs_pad *= self.output_size() ** 0.5
        if self.embed is None:
            xs_pad = xs_pad
        else:
            xs_pad = self.embed(xs_pad, cache)
        if cache["tail_chunk"]:
            xs_pad = to_device(cache["feats"], device=xs_pad.device)
        else:
            xs_pad = self._add_overlap_chunk(xs_pad, cache)
        encoder_outs = self.encoders0(xs_pad, None, None, None, None)
        xs_pad, masks = encoder_outs[0], encoder_outs[1]
        intermediate_outs = []
        if len(self.interctc_layer_idx) == 0:
            encoder_outs = self.encoders(xs_pad, None, None, None, None)
            xs_pad, masks = encoder_outs[0], encoder_outs[1]
        else:
            for layer_idx, encoder_layer in enumerate(self.encoders):
                encoder_outs = encoder_layer(xs_pad, None, None, None, None)
                xs_pad, masks = encoder_outs[0], encoder_outs[1]
                if layer_idx + 1 in self.interctc_layer_idx:
                    encoder_out = xs_pad

                    # intermediate outputs are also normalized
                    if self.normalize_before:
                        encoder_out = self.after_norm(encoder_out)

                    intermediate_outs.append((layer_idx + 1, encoder_out))

                    if self.interctc_use_conditioning:
                        ctc_out = ctc.softmax(encoder_out)
                        xs_pad = xs_pad + self.conditioning_layer(ctc_out)

        if self.normalize_before:
            xs_pad = self.after_norm(xs_pad)

        if len(intermediate_outs) > 0:
            return (xs_pad, intermediate_outs), None, None
        return xs_pad, ilens, None

    def gen_tf2torch_map_dict(self):
        tensor_name_prefix_torch = self.tf2torch_tensor_name_prefix_torch
        tensor_name_prefix_tf = self.tf2torch_tensor_name_prefix_tf