FunASR java http client

VirtuosoQ

2024-04-26 e9d2cfc3a134b00f4e98271fbee3838d1ccecbcc

 funasr/frontends/whisper_frontend.py

@@ -1,8 +1,8 @@
from typing import Tuple
import torch
import torch.nn as nn
import whisper
from whisper.audio import HOP_LENGTH, N_FFT, N_SAMPLES


from funasr.register import tables
from torch.nn.utils.rnn import pad_sequence

@@ -17,30 +17,40 @@
    def __init__(
            self,
            fs: int = 16000,
            whisper_model: str = "large-v3",
            whisper_model: str = None,
            do_pad_trim: bool = True,
            n_mels: int = 80,
            permute: bool = False,
            **kwargs,
    ):
        super().__init__()
        assert fs == 16000
        self.fs = fs

        import whisper
        from whisper.audio import HOP_LENGTH, N_FFT, N_SAMPLES
        self.n_fft = N_FFT
        self.win_length = N_FFT
        self.hop_length = HOP_LENGTH
        self.pad_samples = N_SAMPLES
        self.frame_shift = self.hop_length
        self.lfr_n = 1
        self.n_mels = n_mels
        if whisper_model == "large-v3" or whisper_model == "large":
            self.n_mels = 128
        else:
            self.n_mels = 80

        self.mel_filters = whisper.audio.mel_filters
        filters_path = kwargs.get("filters_path", None)
        self.filters_path = filters_path
        if filters_path is not None:
            from funasr.models.sense_voice.whisper_lib.audio import mel_filters
            self.mel_filters = mel_filters
        else:
            self.mel_filters = whisper.audio.mel_filters
        self.do_pad_trim = do_pad_trim
        if do_pad_trim:
            self.pad_or_trim = whisper.pad_or_trim
        self.permute = permute

        assert whisper_model in whisper.available_models()
        # assert whisper_model in whisper.available_models()

    def output_size(self) -> int:
        return self.n_mels
@@ -57,8 +67,10 @@

        # whisper deletes the last frame by default (Shih-Lun)
        magnitudes = stft[..., :-1].abs() ** 2

        filters = self.mel_filters(audio.device, self.n_mels)
        if self.filters_path is not None:
            filters = self.mel_filters(audio.device, self.n_mels, self.filters_path)
        else:
            filters = self.mel_filters(audio.device, self.n_mels)
        mel_spec = filters @ magnitudes

        log_spec = torch.clamp(mel_spec, min=1e-10).log10()
@@ -77,11 +89,12 @@
        return log_spec, olens

    def forward(
            self, input: torch.Tensor, input_lengths: torch.Tensor
            self, input: torch.Tensor, input_lengths: torch.Tensor, **kwargs,
    ) -> Tuple[torch.Tensor, torch.Tensor]:
        batch_size = input.size(0)
        feats = []
        feats_lens = []
        input = input.to(torch.float32)
        for i in range(batch_size):
            if self.do_pad_trim:
                feat = self.pad_or_trim(input[i], self.pad_samples)
@@ -98,5 +111,6 @@
            feats_pad = pad_sequence(feats,
                                     batch_first=True,
                                     padding_value=0.0)

        if self.permute:
            feats_pad = feats_pad.permute(0, 2, 1)
        return feats_pad, feats_lens