Why TTS generated audio always sounds different ?

[arthurwu4work]

Hi ,

I use English multispeaker model to generate TTS audio
But the generated audio always sounds different

My script

import sys
import soundfile as sf
import torch
import nemo.collections.tts as nemo_tts

# Load mel spectrogram generator
spec_generator = nemo_tts.models.FastPitchModel.restore_from(restore_path="models/nemo/tts_en_fastpitch_multispeaker.nemo")
# Load vocoder
vocoder = nemo_tts.models.HifiGanModel.restore_from(restore_path="models/nemo/tts_en_hifitts_hifigan_ft_fastpitch.nemo")
# Generate audio
with torch.no_grad():
    parsed = spec_generator.parse("You can type your sentence here to get nemo to produce speech.")
    spectrogram = spec_generator.generate_spectrogram(tokens=parsed, speaker=15)
    audio = vocoder.convert_spectrogram_to_audio(spec=spectrogram)
    if isinstance(audio, torch.Tensor):
        audio = audio.to('cpu').detach().numpy()

# Save the audio to disk in a file called speech.wav
sf.write("speech2.wav", audio.T, 44100, format='WAV')

My environment details

OS: ubuntu 20.04
NeMo : brench 1.20.0
python : 3.8
pytorch : 2.0.1

this is the audio file
TTS.zip