14a4d1a061872622d992ec39369bd24a78e292cc,models/tacotrongst.py,TacotronGST,forward,#TacotronGST#Any#Any#Any#Any#,44

Before Change

        encoder_outputs = self._add_speaker_embedding(encoder_outputs,
                                                      speaker_ids)
        gst_outputs = self.gst(mel_specs)
        gst_outputs = gst_outputs.expand(-1, encoder_outputs.size(1), -1)
        encoder_outputs = encoder_outputs + gst_outputs
        mel_outputs, alignments, stop_tokens = self.decoder(
            encoder_outputs, mel_specs, mask)
        mel_outputs = mel_outputs.view(B, -1, self.mel_dim)
        linear_outputs = self.postnet(mel_outputs)

After Change

        self._init_states()
        self.compute_speaker_embedding(speaker_ids)
        if self.num_speakers > 1:
            inputs = self._concat_speaker_embedding(inputs,
                                                self.speaker_embeddings)
        encoder_outputs = self.encoder(inputs)
        if self.num_speakers > 1:
            encoder_outputs = self._concat_speaker_embedding(encoder_outputs,
                                                        self.speaker_embeddings)
        gst_outputs = self.gst(mel_specs)
        encoder_outputs = self._concat_speaker_embedding(
            encoder_outputs, gst_outputs)
        mel_outputs, alignments, stop_tokens = self.decoder(

Instances