speech_asr
2023-04-13 66fc9f0a9b387495511623d2efbd5e7a10accce7
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(