infer.py

from models import SynthesizerTrn
from scipy.io.wavfile import write
from khmer_phonemizer import phonemize_single
import utils
import commons
import torch
import sys

_pad        = '_'
_punctuation = '. '
_letters_ipa = 'acefhijklmnoprstuwzĕŋŏŭɑɓɔɗəɛɡɨɲʋʔʰː'

# Export all symbols:
symbols = [_pad] + list(_punctuation) + list(_letters_ipa)


# Special symbol ids
SPACE_ID = symbols.index(" ")

_symbol_to_id = {s: i for i, s in enumerate(symbols)}

def text_to_sequence(text):
    sequence = []
    for symbol in text:
        symbol_id = _symbol_to_id[symbol]
        sequence += [symbol_id]
    return sequence


def get_text(text, hps):
    text_norm = text_to_sequence(text)
    if hps.data.add_blank:
        text_norm = commons.intersperse(text_norm, 0)
    text_norm = torch.LongTensor(text_norm)
    return text_norm


hps = utils.get_hparams_from_file("config.json")
net_g = SynthesizerTrn(
    len(symbols),
    hps.data.filter_length // 2 + 1,
    hps.train.segment_size // hps.data.hop_length,
    **hps.model
)

_ = net_g.eval()
_ = utils.load_checkpoint("G_60000.pth", net_g, None)

text = " ".join(phonemize_single(sys.argv[1]) + ["."])
stn_tst = get_text(text, hps)

with torch.no_grad():
    x_tst = stn_tst.unsqueeze(0)
    x_tst_lengths = torch.LongTensor([stn_tst.size(0)])
    audio = (
        net_g.infer(
            x_tst, x_tst_lengths, noise_scale=0.667, noise_scale_w=0.8, length_scale=1
        )[0][0, 0]
        .data.cpu()
        .float()
        .numpy()
    )
    write("audio.wav", rate=hps.data.sampling_rate, data=audio)
    print("saved audio.wav")