speech_asr
2023-04-20 eac9f111b502e4581b14dc718731bf7dc1c7d5f6
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(