49 lines
1.5 KiB
Python
49 lines
1.5 KiB
Python
import torch
|
|
from torch.utils.data import DataLoader
|
|
|
|
import commons
|
|
import utils
|
|
from data_utils import TextAudioSpeakerLoader, TextAudioSpeakerCollate
|
|
from tqdm import tqdm
|
|
from multiprocessing import Pool
|
|
|
|
from text import cleaned_text_to_sequence, get_bert
|
|
|
|
config_path = 'configs/config.json'
|
|
hps = utils.get_hparams_from_file(config_path)
|
|
|
|
def process_line(line):
|
|
_id, spk, language_str, text, phones, tone, word2ph = line.strip().split("|")
|
|
phone = phones.split(" ")
|
|
tone = [int(i) for i in tone.split(" ")]
|
|
word2ph = [int(i) for i in word2ph.split(" ")]
|
|
# print(text, word2ph,phone, tone, language_str)
|
|
w2pho = [i for i in word2ph]
|
|
word2ph = [i for i in word2ph]
|
|
phone, tone, language = cleaned_text_to_sequence(phone, tone, language_str)
|
|
|
|
if hps.data.add_blank:
|
|
phone = commons.intersperse(phone, 0)
|
|
tone = commons.intersperse(tone, 0)
|
|
language = commons.intersperse(language, 0)
|
|
for i in range(len(word2ph)):
|
|
word2ph[i] = word2ph[i] * 2
|
|
word2ph[0] += 1
|
|
wav_path = f'{_id}'
|
|
|
|
bert_path = wav_path.replace(".wav", ".bert.pt")
|
|
try:
|
|
bert = torch.load(bert_path)
|
|
assert bert.shape[-1] == len(phone)
|
|
except:
|
|
bert = get_bert(text, word2ph, language_str)
|
|
assert bert.shape[-1] == len(phone)
|
|
torch.save(bert, bert_path)
|
|
|
|
with open(hps.data.training_files) as f:
|
|
lines = f.readlines()
|
|
|
|
with Pool() as pool:
|
|
for _ in tqdm(pool.imap_unordered(process_line, lines)):
|
|
pass
|