File size: 4,436 Bytes
60dc3af |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 |
import os
from trainer import Trainer, TrainerArgs
from TTS.config.shared_configs import BaseAudioConfig
from TTS.tts.configs.shared_configs import BaseDatasetConfig
from TTS.tts.configs.vits_config import VitsConfig
from TTS.tts.datasets import load_tts_samples
from TTS.tts.models.vits import Vits, VitsArgs
from TTS.tts.utils.speakers import SpeakerManager
from TTS.tts.utils.text.tokenizer import TTSTokenizer
from TTS.utils.audio import AudioProcessor
# to read tsv files from common voice
import pandas as pd
# output_path = '/media/popos/Barracuda/Models/TTS_new/trained_common_voice'
# dataset_path = "/media/popos/Barracuda/Datasets/CommonVoiceMozillaIta/it_29-03-2021/cv-corpus-6.1-2020-12-11/it"
output_path = '/run/media/opensuse/Barracuda/Models/TTS_new/trained_common_voice'
dataset_path = "/run/media/opensuse/Barracuda/Datasets/CommonVoiceMozillaIta/cv-corpus-9.0-2022-04-27/it"
pretrained_path = '/run/media/opensuse/Barracuda/Models/TTS_new/trained_common_voice/vits_vctk-June-05-2022_03+45PM-0cf3265a/'
dataset_config = BaseDatasetConfig(
name="vctk", meta_file_train="", language="it-it", path=dataset_path
)
# custom formatter implementation
def commonvoice_formatter(root_path, manifest_file, **kwargs):
# from root path we have train.tsv, test.tsv and val.tsv or use validated.tsv that contains all
txt_file = os.path.join(root_path, 'train.tsv')
df = pd.read_csv(txt_file, sep='\t')
items = []
for i, data in df.iterrows():
items.append({
"text": data['sentence'],
"audio_file": os.path.join(root_path, 'clips', data['path']),
"speaker_name": data['client_id']
})
return items
audio_config = BaseAudioConfig(
sample_rate=22050,
win_length=1024,
hop_length=256,
num_mels=80,
preemphasis=0.0,
ref_level_db=20,
log_func="np.log",
do_trim_silence=True,
trim_db=23.0,
mel_fmin=0,
mel_fmax=None,
spec_gain=1.0,
signal_norm=False,
do_amp_to_db_linear=False,
resample=True,
)
vitsArgs = VitsArgs(
use_speaker_embedding=True,
)
config = VitsConfig(
model_args=vitsArgs,
audio=audio_config,
run_name="vits_vctk",
batch_size=32,
eval_batch_size=16,
batch_group_size=5,
num_loader_workers=4,
num_eval_loader_workers=4,
run_eval=True,
test_delay_epochs=-1,
epochs=1000,
text_cleaner="english_cleaners",
use_phonemes=False,
phoneme_cache_path=os.path.join(output_path, "phoneme_cache"),
compute_input_seq_cache=True,
print_step=25,
print_eval=False,
mixed_precision=True,
max_text_len=325, # change this if you have a larger VRAM than 16GB
output_path=output_path,
datasets=[dataset_config],
)
# INITIALIZE THE AUDIO PROCESSOR
# Audio processor is used for feature extraction and audio I/O.
# It mainly serves to the dataloader and the training loggers.
ap = AudioProcessor.init_from_config(config)
# INITIALIZE THE TOKENIZER
# Tokenizer is used to convert text to sequences of token IDs.
# config is updated with the default characters if not defined in the config.
tokenizer, config = TTSTokenizer.init_from_config(config)
# LOAD DATA SAMPLES
# Each sample is a list of ```[text, audio_file_path, speaker_name]```
# You can define your custom sample loader returning the list of samples.
# Or define your custom formatter and pass it to the `load_tts_samples`.
# Check `TTS.tts.datasets.load_tts_samples` for more details.
train_samples, eval_samples = load_tts_samples(
dataset_config, eval_split=True, formatter=commonvoice_formatter)
# init speaker manager for multi-speaker training
# it maps speaker-id to speaker-name in the model and data-loader
speaker_manager = SpeakerManager()
speaker_manager.set_speaker_ids_from_data(train_samples + eval_samples)
config.model_args.num_speakers = speaker_manager.num_speakers
# init model
model = Vits(config, ap, tokenizer, speaker_manager)
# init the trainer and 🚀
if pretrained_path:
trainer = Trainer(
TrainerArgs(
continue_path=pretrained_path,
),
config,
output_path,
model=model,
train_samples=train_samples,
eval_samples=eval_samples,
)
else:
trainer = Trainer(
TrainerArgs(),
config,
output_path,
model=model,
train_samples=train_samples,
eval_samples=eval_samples,
)
trainer.fit()
|