train

游雁

2023-05-16 3d9f094e9652d4b84894c6fd4eae39a4a753b0f0

 funasr/models/e2e_vad.py

@@ -311,7 +311,7 @@
                                0.000001))

    def ComputeScores(self, feats: torch.Tensor, in_cache: Dict[str, torch.Tensor]) -> None:
        scores = self.encoder(feats, in_cache)  # return B * T * D
        scores = self.encoder(feats, in_cache).to('cpu')  # return B * T * D
        assert scores.shape[1] == feats.shape[1], "The shape between feats and scores does not match"
        self.vad_opts.nn_eval_block_size = scores.shape[1]
        self.frm_cnt += scores.shape[1]  # count total frames
@@ -469,7 +469,7 @@
                        - 1)) / self.vad_opts.noise_frame_num_used_for_snr

        return frame_state
     

    def forward(self, feats: torch.Tensor, waveform: torch.tensor, in_cache: Dict[str, torch.Tensor] = dict(),
                is_final: bool = False
                ) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
@@ -499,11 +499,11 @@
        return segments, in_cache

    def forward_online(self, feats: torch.Tensor, waveform: torch.tensor, in_cache: Dict[str, torch.Tensor] = dict(),
                is_final: bool = False, max_end_sil: int = 800
                ) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
                       is_final: bool = False, max_end_sil: int = 800
                       ) -> Tuple[List[List[List[int]]], Dict[str, torch.Tensor]]:
        self.max_end_sil_frame_cnt_thresh = max_end_sil - self.vad_opts.speech_to_sil_time_thres
        self.waveform = waveform  # compute decibel for each frame
        

        self.ComputeScores(feats, in_cache)
        self.ComputeDecibel()
        if not is_final: