speech_asr
2023-04-19 124d49c6ee7983df0d0efe1ff6652061b179ce03
funasr/bin/tp_inference.py
@@ -116,8 +116,8 @@
            enc = enc[0]
        # c. Forward Predictor
        _, _, us_alphas, us_cif_peak = self.tp_model.calc_predictor_timestamp(enc, enc_len, text_lengths.to(self.device)+1)
        return us_alphas, us_cif_peak
        _, _, us_alphas, us_peaks = self.tp_model.calc_predictor_timestamp(enc, enc_len, text_lengths.to(self.device)+1)
        return us_alphas, us_peaks
def inference(