def output_types(self):
if not self.calculate_loss and not self.training:
return {
"spec_pred_dec":
NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spec_pred_postnet":
NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"gate_pred":
NeuralType(('B', 'T'), LogitsType()),
"alignments":
NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
"pred_length":
NeuralType(('B'), LengthsType()),
}
return {
"spec_pred_dec":
NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spec_pred_postnet":
NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"gate_pred":
NeuralType(('B', 'T'), LogitsType()),
"spec_target":
NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spec_target_len":
NeuralType(('B'), LengthsType()),
"alignments":
NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
}
def input_types(self):
return {
"text": NeuralType(('B', 'T'), TokenIndex()),
"text_lengths": NeuralType(('B'), LengthsType()),
"spect": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spect_lengths": NeuralType(('B'), LengthsType()),
"speaker": NeuralType(('B'), IntType(), optional=True),
}
def input_types(self):
return {
"text": NeuralType(('B', 'T_text'), TokenIndex()),
"durs": NeuralType(('B', 'T_text'), TokenDurationType()),
"pitch": NeuralType(('B', 'T_audio'), RegressionValuesType()),
"speaker": NeuralType(('B'), Index(), optional=True),
"pace": NeuralType(optional=True),
"spec": NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType(), optional=True),
"attn_prior": NeuralType(('B', 'T_spec', 'T_text'), ProbsType(), optional=True),
"mel_lens": NeuralType(('B'), LengthsType(), optional=True),
"input_lens": NeuralType(('B'), LengthsType(), optional=True),
}
def input_types(self):
if self.training:
return {
"tokens": NeuralType(('B', 'T'), EmbeddedTextType()),
"token_len": NeuralType(('B'), LengthsType()),
"audio": NeuralType(('B', 'T'), AudioSignal()),
"audio_len": NeuralType(('B'), LengthsType()),
}
else:
return {
"tokens": NeuralType(('B', 'T'), EmbeddedTextType()),
"token_len": NeuralType(('B'), LengthsType()),
"audio": NeuralType(('B', 'T'), AudioSignal(), optional=True),
"audio_len": NeuralType(('B'), LengthsType(), optional=True),
}
def input_types(self):
return {
"x":
NeuralType(('B', 'T', 'D'), EncodedRepresentation()),
"x_len":
NeuralType(('B'), LengthsType()),
"dur_target":
NeuralType(('B', 'T'), TokenDurationType(), optional=True),
"pitch_target":
NeuralType(('B', 'T'), RegressionValuesType(), optional=True),
"energy_target":
NeuralType(('B', 'T'), RegressionValuesType(), optional=True),
"spec_len":
NeuralType(('B'), LengthsType(), optional=True),
}
def input_types(self):
return {
"x": NeuralType(('B', 'C', 'D', 'T'), SpectrogramType()),
"mag": NeuralType(('B', 'any', 'D', 'T'), SpectrogramType()),
"max_length": NeuralType(None, LengthsType()),
"repeat": NeuralType(None, IntType()),
}
def input_types(self):
return {
"spec_pred": NeuralType(('B', 'C', 'T'), MelSpectrogramType()),
"spec_target": NeuralType(('B', 'C', 'T'), MelSpectrogramType()),
"spec_target_len": NeuralType(('B'), LengthsType()),
"pad_value": NeuralType(),
}
def input_types(self):
return {
"log_durs_predicted": NeuralType(('B', 'T'),
TokenLogDurationType()),
"durs_tgt": NeuralType(('B', 'T'), TokenDurationType()),
"len": NeuralType(('B'), LengthsType()),
}
def input_types(self):
input_dict = {
"memory": NeuralType(('B', 'T', 'D'), EmbeddedTextType()),
"memory_lengths": NeuralType(('B'), LengthsType()),
}
if self.training:
input_dict["decoder_inputs"] = NeuralType(('B', 'D', 'T'), MelSpectrogramType())
return input_dict
def output_types(self):
return {
"out": NeuralType(('B', 'T', 'D'), EncodedRepresentation()),
"log_dur_preds": NeuralType(('B', 'T'), TokenDurationType()),
"pitch_preds": NeuralType(('B', 'T'), RegressionValuesType()),
"energy_preds": NeuralType(('B', 'T'), RegressionValuesType()),
"spec_len": NeuralType(('B'), LengthsType()),
}
def output_types(self):
output_dict = {
"mel_outputs": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"gate_outputs": NeuralType(('B', 'T'), LogitsType()),
"alignments": NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
}
if not self.training:
output_dict["mel_lengths"] = NeuralType(('B'), LengthsType())
return output_dict
def input_types(self):
return {
"log_durs_predicted": NeuralType(('B', 'T'),
TokenLogDurationType()),
"pitch_predicted": NeuralType(('B', 'T'), RegressionValuesType()),
"durs_tgt": NeuralType(('B', 'T'), TokenDurationType()),
"dur_lens": NeuralType(('B'), LengthsType()),
"pitch_tgt": NeuralType(('B', 'T'), RegressionValuesType()),
}
def input_types(self):
return {
"input":
NeuralType(('B', 'T', 'D'), EncodedRepresentation()),
"seq_lens":
NeuralType(('B'), LengthsType()),
"conditioning":
NeuralType(('B', 'T', 'D'), EncodedRepresentation(),
optional=True),
}
def input_types(self):
return {
"spec_pred_dec": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spec_pred_postnet": NeuralType(('B', 'D', 'T'),
MelSpectrogramType()),
"gate_pred": NeuralType(('B', 'T'), LogitsType()),
"spec_target": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spec_target_len": NeuralType(('B'), LengthsType()),
"pad_value": NeuralType(),
}
def input_types(self):
return {
"text":
NeuralType(('B', 'T'), TokenIndex()),
"text_lengths":
NeuralType(('B', ), LengthsType()),
"speaker_embeddings":
NeuralType(('B', 'D'),
AcousticEncodedRepresentation(),
optional=True),
}
def output_types(self):
return {
"z": NeuralType(('B', 'D', 'T'), NormalDistributionSamplesType()),
"y_m": NeuralType(('B', 'D', 'T'), NormalDistributionMeanType()),
"y_logs": NeuralType(('B', 'D', 'T'), NormalDistributionLogVarianceType()),
"logdet": NeuralType(('B'), LogDeterminantType()),
"log_durs_predicted": NeuralType(('B', 'T'), TokenLogDurationType()),
"log_durs_extracted": NeuralType(('B', 'T'), TokenLogDurationType()),
"spect_lengths": NeuralType(('B'), LengthsType()),
"attn": NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
}
def output_types(self):
if self.mode == OperationMode.training or self.mode == OperationMode.validation:
output_dict = {
"pred_normal_dist": NeuralType(('B', 'flowgroup', 'T'), NormalDistributionSamplesType()),
"logdet": NeuralType(elements_type=LogDeterminantType()),
"predicted_audio": NeuralType(('B', 'T'), AudioSignal()),
}
if self.mode == OperationMode.validation:
output_dict["spec"] = NeuralType(('B', 'T', 'D'), MelSpectrogramType())
output_dict["spec_len"] = NeuralType(('B'), LengthsType())
return output_dict
return {
"audio_pred": NeuralType(('B', 'T'), AudioSignal()),
}
def output_types(self):
if self.mode == OperationMode.training or self.mode == OperationMode.validation:
output_dict = {
"pred_normal_dist": NeuralType(('B', 'flowgroup', 'T'), NormalDistributionSamplesType()),
"log_s_list": NeuralType(('B', 'flowgroup', 'T'), VoidType()), # TODO: Figure out a good typing
"log_det_W_list": NeuralType(elements_type=VoidType()), # TODO: Figure out a good typing
}
if self.mode == OperationMode.validation:
output_dict["audio_pred"] = NeuralType(('B', 'T'), AudioSignal())
output_dict["spec"] = NeuralType(('B', 'T', 'D'), MelSpectrogramType())
output_dict["spec_len"] = NeuralType(('B'), LengthsType())
return output_dict
return {
"audio_pred": NeuralType(('B', 'T'), AudioSignal()),
}
class FastSpeech2HifiGanE2EModel(TextToWaveform):
"""An end-to-end speech synthesis model based on FastSpeech2 and HiFiGan that converts strings to audio without
using the intermediate mel spectrogram representation."""
def __init__(self, cfg: DictConfig, trainer: 'Trainer' = None):
if isinstance(cfg, dict):
cfg = OmegaConf.create(cfg)
super().__init__(cfg=cfg, trainer=trainer)
self.audio_to_melspec_precessor = instantiate(cfg.preprocessor)
self.encoder = instantiate(cfg.encoder)
self.variance_adapter = instantiate(cfg.variance_adaptor)
self.generator = instantiate(cfg.generator)
self.multiperioddisc = MultiPeriodDiscriminator()
self.multiscaledisc = MultiScaleDiscriminator()
self.melspec_fn = instantiate(cfg.preprocessor,
highfreq=None,
use_grads=True)
self.mel_val_loss = L1MelLoss()
self.durationloss = DurationLoss()
self.feat_matching_loss = FeatureMatchingLoss()
self.disc_loss = DiscriminatorLoss()
self.gen_loss = GeneratorLoss()
self.mseloss = torch.nn.MSELoss()
self.energy = cfg.add_energy_predictor
self.pitch = cfg.add_pitch_predictor
self.mel_loss_coeff = cfg.mel_loss_coeff
self.pitch_loss_coeff = cfg.pitch_loss_coeff
self.energy_loss_coeff = cfg.energy_loss_coeff
self.splice_length = cfg.splice_length
self.use_energy_pred = False
self.use_pitch_pred = False
self.log_train_images = False
self.logged_real_samples = False
self._tb_logger = None
self.sample_rate = cfg.sample_rate
self.hop_size = cfg.hop_size
# Parser and mappings are used for inference only.
self.parser = parsers.make_parser(name='en')
if 'mappings_filepath' in cfg:
mappings_filepath = cfg.get('mappings_filepath')
else:
logging.error(
"ERROR: You must specify a mappings.json file in the config file under model.mappings_filepath."
)
mappings_filepath = self.register_artifact('mappings_filepath',
mappings_filepath)
with open(mappings_filepath, 'r') as f:
mappings = json.load(f)
self.word2phones = mappings['word2phones']
self.phone2idx = mappings['phone2idx']
@property
def tb_logger(self):
if self._tb_logger is None:
if self.logger is None and self.logger.experiment is None:
return None
tb_logger = self.logger.experiment
if isinstance(self.logger, LoggerCollection):
for logger in self.logger:
if isinstance(logger, TensorBoardLogger):
tb_logger = logger.experiment
break
self._tb_logger = tb_logger
return self._tb_logger
def configure_optimizers(self):
gen_params = chain(
self.encoder.parameters(),
self.generator.parameters(),
self.variance_adapter.parameters(),
)
disc_params = chain(self.multiscaledisc.parameters(),
self.multiperioddisc.parameters())
opt1 = torch.optim.AdamW(disc_params, lr=self._cfg.lr)
opt2 = torch.optim.AdamW(gen_params, lr=self._cfg.lr)
num_procs = self._trainer.num_gpus * self._trainer.num_nodes
num_samples = len(self._train_dl.dataset)
batch_size = self._train_dl.batch_size
iter_per_epoch = np.ceil(num_samples / (num_procs * batch_size))
max_steps = iter_per_epoch * self._trainer.max_epochs
logging.info(f"MAX STEPS: {max_steps}")
sch1 = NoamAnnealing(opt1,
d_model=256,
warmup_steps=3000,
max_steps=max_steps,
min_lr=1e-5)
sch1_dict = {
'scheduler': sch1,
'interval': 'step',
}
sch2 = NoamAnnealing(opt2,
d_model=256,
warmup_steps=3000,
max_steps=max_steps,
min_lr=1e-5)
sch2_dict = {
'scheduler': sch2,
'interval': 'step',
}
return [opt1, opt2], [sch1_dict, sch2_dict]
@typecheck(
input_types={
"text":
NeuralType(('B', 'T'), TokenIndex()),
"text_length":
NeuralType(('B'), LengthsType()),
"splice":
NeuralType(optional=True),
"spec_len":
NeuralType(('B'), LengthsType(), optional=True),
"durations":
NeuralType(('B', 'T'), TokenDurationType(), optional=True),
"pitch":
NeuralType(('B', 'T'), RegressionValuesType(), optional=True),
"energies":
NeuralType(('B', 'T'), RegressionValuesType(), optional=True),
},
output_types={
"audio": NeuralType(('B', 'S', 'T'), MelSpectrogramType()),
"splices": NeuralType(),
"log_dur_preds": NeuralType(('B', 'T'), TokenLogDurationType()),
"pitch_preds": NeuralType(('B', 'T'), RegressionValuesType()),
"energy_preds": NeuralType(('B', 'T'), RegressionValuesType()),
"encoded_text_mask": NeuralType(('B', 'T', 'D'), MaskType()),
},
)
def forward(self,
*,
text,
text_length,
splice=True,
durations=None,
pitch=None,
energies=None,
spec_len=None):
encoded_text, encoded_text_mask = self.encoder(text=text,
text_length=text_length)
context, log_dur_preds, pitch_preds, energy_preds, spec_len = self.variance_adapter(
x=encoded_text,
x_len=text_length,
dur_target=durations,
pitch_target=pitch,
energy_target=energies,
spec_len=spec_len,
)
gen_in = context
splices = None
if splice:
# Splice generated spec
output = []
splices = []
for i, sample in enumerate(context):
start = np.random.randint(
low=0,
high=min(int(sample.size(0)), int(spec_len[i])) -
self.splice_length)
output.append(sample[start:start + self.splice_length, :])
splices.append(start)
gen_in = torch.stack(output)
output = self.generator(x=gen_in.transpose(1, 2))
return output, splices, log_dur_preds, pitch_preds, energy_preds, encoded_text_mask
def training_step(self, batch, batch_idx, optimizer_idx):
f, fl, t, tl, durations, pitch, energies = batch
spec, spec_len = self.audio_to_melspec_precessor(f, fl)
# train discriminator
if optimizer_idx == 0:
with torch.no_grad():
audio_pred, splices, _, _, _, _ = self(
spec=spec,
spec_len=spec_len,
text=t,
text_length=tl,
durations=durations,
pitch=pitch if not self.use_pitch_pred else None,
energies=energies if not self.use_energy_pred else None,
)
real_audio = []
for i, splice in enumerate(splices):
real_audio.append(
f[i, splice *
self.hop_size:(splice + self.splice_length) *
self.hop_size])
real_audio = torch.stack(real_audio).unsqueeze(1)
real_score_mp, gen_score_mp, _, _ = self.multiperioddisc(
real_audio, audio_pred)
real_score_ms, gen_score_ms, _, _ = self.multiscaledisc(
real_audio, audio_pred)
loss_mp, loss_mp_real, _ = self.disc_loss(real_score_mp,
gen_score_mp)
loss_ms, loss_ms_real, _ = self.disc_loss(real_score_ms,
gen_score_ms)
loss_mp /= len(loss_mp_real)
loss_ms /= len(loss_ms_real)
loss_disc = loss_mp + loss_ms
self.log("loss_discriminator", loss_disc, prog_bar=True)
self.log("loss_discriminator_ms", loss_ms)
self.log("loss_discriminator_mp", loss_mp)
return loss_disc
# train generator
elif optimizer_idx == 1:
audio_pred, splices, log_dur_preds, pitch_preds, energy_preds, encoded_text_mask = self(
spec=spec,
spec_len=spec_len,
text=t,
text_length=tl,
durations=durations,
pitch=pitch if not self.use_pitch_pred else None,
energies=energies if not self.use_energy_pred else None,
)
real_audio = []
for i, splice in enumerate(splices):
real_audio.append(
f[i, splice * self.hop_size:(splice + self.splice_length) *
self.hop_size])
real_audio = torch.stack(real_audio).unsqueeze(1)
# Do HiFiGAN generator loss
audio_length = torch.tensor([
self.splice_length * self.hop_size
for _ in range(real_audio.shape[0])
]).to(real_audio.device)
real_spliced_spec, _ = self.melspec_fn(real_audio.squeeze(),
seq_len=audio_length)
pred_spliced_spec, _ = self.melspec_fn(audio_pred.squeeze(),
seq_len=audio_length)
loss_mel = torch.nn.functional.l1_loss(real_spliced_spec,
pred_spliced_spec)
loss_mel *= self.mel_loss_coeff
_, gen_score_mp, real_feat_mp, gen_feat_mp = self.multiperioddisc(
real_audio, audio_pred)
_, gen_score_ms, real_feat_ms, gen_feat_ms = self.multiscaledisc(
real_audio, audio_pred)
loss_gen_mp, list_loss_gen_mp = self.gen_loss(gen_score_mp)
loss_gen_ms, list_loss_gen_ms = self.gen_loss(gen_score_ms)
loss_gen_mp /= len(list_loss_gen_mp)
loss_gen_ms /= len(list_loss_gen_ms)
total_loss = loss_gen_mp + loss_gen_ms + loss_mel
loss_feat_mp = self.feat_matching_loss(real_feat_mp, gen_feat_mp)
loss_feat_ms = self.feat_matching_loss(real_feat_ms, gen_feat_ms)
total_loss += loss_feat_mp + loss_feat_ms
self.log(name="loss_gen_disc_feat",
value=loss_feat_mp + loss_feat_ms)
self.log(name="loss_gen_disc_feat_ms", value=loss_feat_ms)
self.log(name="loss_gen_disc_feat_mp", value=loss_feat_mp)
self.log(name="loss_gen_mel", value=loss_mel)
self.log(name="loss_gen_disc", value=loss_gen_mp + loss_gen_ms)
self.log(name="loss_gen_disc_mp", value=loss_gen_mp)
self.log(name="loss_gen_disc_ms", value=loss_gen_ms)
dur_loss = self.durationloss(log_duration_pred=log_dur_preds,
duration_target=durations.float(),
mask=encoded_text_mask)
self.log(name="loss_gen_duration", value=dur_loss)
total_loss += dur_loss
if self.pitch:
pitch_loss = self.mseloss(
pitch_preds, pitch.float()) * self.pitch_loss_coeff
total_loss += pitch_loss
self.log(name="loss_gen_pitch", value=pitch_loss)
if self.energy:
energy_loss = self.mseloss(energy_preds,
energies) * self.energy_loss_coeff
total_loss += energy_loss
self.log(name="loss_gen_energy", value=energy_loss)
# Log images to tensorboard
if self.log_train_images:
self.log_train_images = False
if self.logger is not None and self.logger.experiment is not None:
self.tb_logger.add_image(
"train_mel_target",
plot_spectrogram_to_numpy(
real_spliced_spec[0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
spec_predict = pred_spliced_spec[0].data.cpu().numpy()
self.tb_logger.add_image(
"train_mel_predicted",
plot_spectrogram_to_numpy(spec_predict),
self.global_step,
dataformats="HWC",
)
self.log(name="loss_gen", prog_bar=True, value=total_loss)
return total_loss
def validation_step(self, batch, batch_idx):
f, fl, t, tl, _, _, _ = batch
spec, spec_len = self.audio_to_melspec_precessor(f, fl)
audio_pred, _, _, _, _, _ = self(spec=spec,
spec_len=spec_len,
text=t,
text_length=tl,
splice=False)
audio_pred.squeeze_()
pred_spec, _ = self.melspec_fn(audio_pred, seq_len=spec_len)
loss = self.mel_val_loss(spec_pred=pred_spec,
spec_target=spec,
spec_target_len=spec_len,
pad_value=-11.52)
return {
"val_loss": loss,
"audio_target": f.squeeze() if batch_idx == 0 else None,
"audio_pred": audio_pred if batch_idx == 0 else None,
}
def on_train_epoch_start(self):
# Switch to using energy predictions after 50% of training
if not self.use_energy_pred and self.current_epoch >= np.ceil(
0.5 * self._trainer.max_epochs):
logging.info(
f"Using energy predictions after epoch: {self.current_epoch}")
self.use_energy_pred = True
# Switch to using pitch predictions after 62.5% of training
if not self.use_pitch_pred and self.current_epoch >= np.ceil(
0.625 * self._trainer.max_epochs):
logging.info(
f"Using pitch predictions after epoch: {self.current_epoch}")
self.use_pitch_pred = True
def validation_epoch_end(self, outputs):
if self.tb_logger is not None:
_, audio_target, audio_predict = outputs[0].values()
if not self.logged_real_samples:
self.tb_logger.add_audio("val_target",
audio_target[0].data.cpu(),
self.global_step, self.sample_rate)
self.logged_real_samples = True
audio_predict = audio_predict[0].data.cpu()
self.tb_logger.add_audio("val_pred", audio_predict,
self.global_step, self.sample_rate)
avg_loss = torch.stack([
x['val_loss'] for x in outputs
]).mean() # This reduces across batches, not workers!
self.log('val_loss', avg_loss, sync_dist=True)
self.log_train_images = True
def __setup_dataloader_from_config(self,
cfg,
shuffle_should_be: bool = True,
name: str = "train"):
if "dataset" not in cfg or not isinstance(cfg.dataset, DictConfig):
raise ValueError(f"No dataset for {name}")
if "dataloader_params" not in cfg or not isinstance(
cfg.dataloader_params, DictConfig):
raise ValueError(f"No dataloder_params for {name}")
if shuffle_should_be:
if 'shuffle' not in cfg.dataloader_params:
logging.warning(
f"Shuffle should be set to True for {self}'s {name} dataloader but was not found in its "
"config. Manually setting to True")
with open_dict(cfg.dataloader_params):
cfg.dataloader_params.shuffle = True
elif not cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to False!!!"
)
elif not shuffle_should_be and cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to True!!!")
dataset = instantiate(cfg.dataset)
return torch.utils.data.DataLoader(dataset,
collate_fn=dataset.collate_fn,
**cfg.dataloader_params)
def setup_training_data(self, cfg):
self._train_dl = self.__setup_dataloader_from_config(cfg)
def setup_validation_data(self, cfg):
self._validation_dl = self.__setup_dataloader_from_config(
cfg, shuffle_should_be=False, name="validation")
def parse(self,
str_input: str,
additional_word2phones=None) -> torch.tensor:
"""
Parses text input and converts them to phoneme indices.
str_input (str): The input text to be converted.
additional_word2phones (dict): Optional dictionary mapping words to phonemes for updating the model's
word2phones. This will not overwrite the existing dictionary, just update it with OOV or new mappings.
Defaults to None, which will keep the existing mapping.
"""
# Update model's word2phones if applicable
if additional_word2phones is not None:
self.word2phones.update(additional_word2phones)
# Convert text -> normalized text -> list of phones per word -> indices
if str_input[-1] not in [".", "!", "?"]:
str_input = str_input + "."
norm_text = re.findall(r"""[\w']+|[.,!?;"]""",
self.parser._normalize(str_input))
try:
phones = [self.word2phones[t] for t in norm_text]
except KeyError as error:
logging.error(
f"ERROR: The following word in the input is not in the model's dictionary and could not be converted"
f" to phonemes: ({error}).\n"
f"You can pass in an `additional_word2phones` dictionary with a conversion for"
f" this word, e.g. {{'{error}': \['phone1', 'phone2', ...\]}} to update the model's mapping."
)
raise
tokens = []
for phone_list in phones:
inds = [self.phone2idx[p] for p in phone_list]
tokens += inds
x = torch.tensor(tokens).unsqueeze_(0).long().to(self.device)
return x
def convert_text_to_waveform(self, *, tokens):
"""
Accepts tokens returned from self.parse() and returns a list of tensors. Note: The tensors in the list can have
different lengths.
"""
self.eval()
token_len = torch.tensor([len(i) for i in tokens]).to(self.device)
audio, _, log_dur_pred, _, _, _ = self(text=tokens,
text_length=token_len,
splice=False)
audio = audio.squeeze(1)
durations = torch.sum(torch.exp(log_dur_pred) - 1, 1).to(torch.int)
audio_list = []
for i, sample in enumerate(audio):
audio_list.append(sample[:durations[i] * self.hop_size])
return audio_list
@classmethod
def list_available_models(cls) -> 'List[PretrainedModelInfo]':
"""
This method returns a list of pre-trained model which can be instantiated directly from NVIDIA's NGC cloud.
Returns:
List of available pre-trained models.
"""
list_of_models = []
model = PretrainedModelInfo(
pretrained_model_name="tts_en_e2e_fastspeech2hifigan",
location=
"https://api.ngc.nvidia.com/v2/models/nvidia/nemo/tts_en_e2e_fastspeech2hifigan/versions/1.0.0/files/tts_en_e2e_fastspeech2hifigan.nemo",
description=
"This model is trained on LJSpeech sampled at 22050Hz with and can be used to generate female English voices with an American accent.",
class_=cls,
)
list_of_models.append(model)
return list_of_models
class MixerTTSModel(SpectrogramGenerator, Exportable):
"""MixerTTS pipeline."""
def __init__(self, cfg: DictConfig, trainer: 'Trainer' = None):
super().__init__(cfg=cfg, trainer=trainer)
cfg = self._cfg
if "text_normalizer" in cfg.train_ds.dataset:
self.normalizer = instantiate(cfg.train_ds.dataset.text_normalizer)
self.text_normalizer_call = self.normalizer.normalize
self.text_normalizer_call_args = {}
if cfg.train_ds.dataset.get("text_normalizer_call_args",
None) is not None:
self.text_normalizer_call_args = cfg.train_ds.dataset.text_normalizer_call_args
self.tokenizer = instantiate(cfg.train_ds.dataset.text_tokenizer)
num_tokens = len(self.tokenizer.tokens)
self.tokenizer_pad = self.tokenizer.pad
self.tokenizer_unk = self.tokenizer.oov
self.pitch_loss_scale = cfg.pitch_loss_scale
self.durs_loss_scale = cfg.durs_loss_scale
self.mel_loss_scale = cfg.mel_loss_scale
self.aligner = instantiate(cfg.alignment_module)
self.forward_sum_loss = ForwardSumLoss()
self.bin_loss = BinLoss()
self.add_bin_loss = False
self.bin_loss_scale = 0.0
self.bin_loss_start_ratio = cfg.bin_loss_start_ratio
self.bin_loss_warmup_epochs = cfg.bin_loss_warmup_epochs
self.cond_on_lm_embeddings = cfg.get("cond_on_lm_embeddings", False)
if self.cond_on_lm_embeddings:
self.lm_padding_value = (self._train_dl.dataset.lm_padding_value
if self._train_dl is not None else
self._get_lm_padding_value(
cfg.train_ds.dataset.lm_model))
self.lm_embeddings = self._get_lm_embeddings(
cfg.train_ds.dataset.lm_model)
self.lm_embeddings.weight.requires_grad = False
self.self_attention_module = instantiate(
cfg.self_attention_module,
n_lm_tokens_channels=self.lm_embeddings.weight.shape[1])
self.encoder = instantiate(cfg.encoder,
num_tokens=num_tokens,
padding_idx=self.tokenizer_pad)
self.symbol_emb = self.encoder.to_embed
self.duration_predictor = instantiate(cfg.duration_predictor)
self.pitch_mean, self.pitch_std = float(cfg.pitch_mean), float(
cfg.pitch_std)
self.pitch_predictor = instantiate(cfg.pitch_predictor)
self.pitch_emb = instantiate(cfg.pitch_emb)
self.preprocessor = instantiate(cfg.preprocessor)
self.decoder = instantiate(cfg.decoder)
self.proj = nn.Linear(self.decoder.d_model, cfg.n_mel_channels)
def _get_lm_model_tokenizer(self, lm_model="albert"):
if getattr(self, "_lm_model_tokenizer", None) is not None:
return self._lm_model_tokenizer
if self._train_dl is not None and self._train_dl.dataset is not None:
self._lm_model_tokenizer = self._train_dl.dataset.lm_model_tokenizer
if lm_model == "albert":
self._lm_model_tokenizer = AlbertTokenizer.from_pretrained(
'albert-base-v2')
else:
raise NotImplementedError(
f"{lm_model} lm model is not supported. Only albert is supported at this moment."
)
return self._lm_model_tokenizer
def _get_lm_embeddings(self, lm_model="albert"):
if lm_model == "albert":
return transformers.AlbertModel.from_pretrained(
'albert-base-v2').embeddings.word_embeddings
else:
raise NotImplementedError(
f"{lm_model} lm model is not supported. Only albert is supported at this moment."
)
def _get_lm_padding_value(self, lm_model="albert"):
if lm_model == "albert":
return transformers.AlbertTokenizer.from_pretrained(
'albert-base-v2')._convert_token_to_id('<pad>')
else:
raise NotImplementedError(
f"{lm_model} lm model is not supported. Only albert is supported at this moment."
)
def _metrics(
self,
true_durs,
true_text_len,
pred_durs,
true_pitch,
pred_pitch,
true_spect=None,
pred_spect=None,
true_spect_len=None,
attn_logprob=None,
attn_soft=None,
attn_hard=None,
attn_hard_dur=None,
):
text_mask = get_mask_from_lengths(true_text_len)
mel_mask = get_mask_from_lengths(true_spect_len)
loss = 0.0
# dur loss and metrics
durs_loss = F.mse_loss(pred_durs, (true_durs + 1).float().log(),
reduction='none')
durs_loss = durs_loss * text_mask.float()
durs_loss = durs_loss.sum() / text_mask.sum()
durs_pred = pred_durs.exp() - 1
durs_pred = torch.clamp_min(durs_pred, min=0)
durs_pred = durs_pred.round().long()
acc = ((true_durs == durs_pred) *
text_mask).sum().float() / text_mask.sum() * 100
acc_dist_1 = (((true_durs - durs_pred).abs() <= 1) *
text_mask).sum().float() / text_mask.sum() * 100
acc_dist_3 = (((true_durs - durs_pred).abs() <= 3) *
text_mask).sum().float() / text_mask.sum() * 100
pred_spect = pred_spect.transpose(1, 2)
# mel loss
mel_loss = F.mse_loss(pred_spect, true_spect,
reduction='none').mean(dim=-2)
mel_loss = mel_loss * mel_mask.float()
mel_loss = mel_loss.sum() / mel_mask.sum()
loss = loss + self.durs_loss_scale * durs_loss + self.mel_loss_scale * mel_loss
# aligner loss
bin_loss, ctc_loss = None, None
ctc_loss = self.forward_sum_loss(attn_logprob=attn_logprob,
in_lens=true_text_len,
out_lens=true_spect_len)
loss = loss + ctc_loss
if self.add_bin_loss:
bin_loss = self.bin_loss(hard_attention=attn_hard,
soft_attention=attn_soft)
loss = loss + self.bin_loss_scale * bin_loss
true_avg_pitch = average_pitch(true_pitch.unsqueeze(1),
attn_hard_dur).squeeze(1)
# pitch loss
pitch_loss = F.mse_loss(pred_pitch, true_avg_pitch,
reduction='none') # noqa
pitch_loss = (pitch_loss * text_mask).sum() / text_mask.sum()
loss = loss + self.pitch_loss_scale * pitch_loss
return loss, durs_loss, acc, acc_dist_1, acc_dist_3, pitch_loss, mel_loss, ctc_loss, bin_loss
@torch.jit.unused
def run_aligner(self, text, text_len, text_mask, spect, spect_len,
attn_prior):
text_emb = self.symbol_emb(text)
attn_soft, attn_logprob = self.aligner(
spect,
text_emb.permute(0, 2, 1),
mask=text_mask == 0,
attn_prior=attn_prior,
)
attn_hard = binarize_attention_parallel(attn_soft, text_len, spect_len)
attn_hard_dur = attn_hard.sum(2)[:, 0, :]
assert torch.all(torch.eq(attn_hard_dur.sum(dim=1), spect_len))
return attn_soft, attn_logprob, attn_hard, attn_hard_dur
@typecheck(
input_types={
"text":
NeuralType(('B', 'T_text'), TokenIndex()),
"text_len":
NeuralType(('B', ), LengthsType()),
"pitch":
NeuralType(('B', 'T_audio'), RegressionValuesType(),
optional=True),
"spect":
NeuralType(('B', 'D', 'T_spec'),
MelSpectrogramType(),
optional=True),
"spect_len":
NeuralType(('B', ), LengthsType(), optional=True),
"attn_prior":
NeuralType(('B', 'T_spec', 'T_text'), ProbsType(), optional=True),
"lm_tokens":
NeuralType(('B', 'T_lm_tokens'), TokenIndex(), optional=True),
},
output_types={
"pred_spect":
NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType()),
"durs_predicted":
NeuralType(('B', 'T_text'), TokenDurationType()),
"log_durs_predicted":
NeuralType(('B', 'T_text'), TokenLogDurationType()),
"pitch_predicted":
NeuralType(('B', 'T_text'), RegressionValuesType()),
"attn_soft":
NeuralType(('B', 'S', 'T_spec', 'T_text'), ProbsType()),
"attn_logprob":
NeuralType(('B', 'S', 'T_spec', 'T_text'), LogprobsType()),
"attn_hard":
NeuralType(('B', 'S', 'T_spec', 'T_text'), ProbsType()),
"attn_hard_dur":
NeuralType(('B', 'T_text'), TokenDurationType()),
},
)
def forward(self,
text,
text_len,
pitch=None,
spect=None,
spect_len=None,
attn_prior=None,
lm_tokens=None):
if self.training:
assert pitch is not None
text_mask = get_mask_from_lengths(text_len).unsqueeze(2)
enc_out, enc_mask = self.encoder(text, text_mask)
# aligner
attn_soft, attn_logprob, attn_hard, attn_hard_dur = None, None, None, None
if spect is not None:
attn_soft, attn_logprob, attn_hard, attn_hard_dur = self.run_aligner(
text, text_len, text_mask, spect, spect_len, attn_prior)
if self.cond_on_lm_embeddings:
lm_emb = self.lm_embeddings(lm_tokens)
lm_features = self.self_attention_module(
enc_out,
lm_emb,
lm_emb,
q_mask=enc_mask.squeeze(2),
kv_mask=lm_tokens != self.lm_padding_value)
# duration predictor
log_durs_predicted = self.duration_predictor(enc_out, enc_mask)
durs_predicted = torch.clamp(log_durs_predicted.exp() - 1, 0)
# pitch predictor
pitch_predicted = self.pitch_predictor(enc_out, enc_mask)
# avg pitch, add pitch_emb
if not self.training:
if pitch is not None:
pitch = average_pitch(pitch.unsqueeze(1),
attn_hard_dur).squeeze(1)
pitch_emb = self.pitch_emb(pitch.unsqueeze(1))
else:
pitch_emb = self.pitch_emb(pitch_predicted.unsqueeze(1))
else:
pitch = average_pitch(pitch.unsqueeze(1), attn_hard_dur).squeeze(1)
pitch_emb = self.pitch_emb(pitch.unsqueeze(1))
enc_out = enc_out + pitch_emb.transpose(1, 2)
if self.cond_on_lm_embeddings:
enc_out = enc_out + lm_features
# regulate length
len_regulated_enc_out, dec_lens = regulate_len(attn_hard_dur, enc_out)
dec_out, dec_lens = self.decoder(
len_regulated_enc_out,
get_mask_from_lengths(dec_lens).unsqueeze(2))
pred_spect = self.proj(dec_out)
return (
pred_spect,
durs_predicted,
log_durs_predicted,
pitch_predicted,
attn_soft,
attn_logprob,
attn_hard,
attn_hard_dur,
)
def infer(
self,
text,
text_len=None,
text_mask=None,
spect=None,
spect_len=None,
attn_prior=None,
use_gt_durs=False,
lm_tokens=None,
pitch=None,
):
if text_mask is None:
text_mask = get_mask_from_lengths(text_len).unsqueeze(2)
enc_out, enc_mask = self.encoder(text, text_mask)
# aligner
attn_hard_dur = None
if use_gt_durs:
attn_soft, attn_logprob, attn_hard, attn_hard_dur = self.run_aligner(
text, text_len, text_mask, spect, spect_len, attn_prior)
if self.cond_on_lm_embeddings:
lm_emb = self.lm_embeddings(lm_tokens)
lm_features = self.self_attention_module(
enc_out,
lm_emb,
lm_emb,
q_mask=enc_mask.squeeze(2),
kv_mask=lm_tokens != self.lm_padding_value)
# duration predictor
log_durs_predicted = self.duration_predictor(enc_out, enc_mask)
durs_predicted = torch.clamp(log_durs_predicted.exp() - 1, 0)
# avg pitch, pitch predictor
if use_gt_durs and pitch is not None:
pitch = average_pitch(pitch.unsqueeze(1), attn_hard_dur).squeeze(1)
pitch_emb = self.pitch_emb(pitch.unsqueeze(1))
else:
pitch_predicted = self.pitch_predictor(enc_out, enc_mask)
pitch_emb = self.pitch_emb(pitch_predicted.unsqueeze(1))
# add pitch emb
enc_out = enc_out + pitch_emb.transpose(1, 2)
if self.cond_on_lm_embeddings:
enc_out = enc_out + lm_features
if use_gt_durs:
if attn_hard_dur is not None:
len_regulated_enc_out, dec_lens = regulate_len(
attn_hard_dur, enc_out)
else:
raise NotImplementedError
else:
len_regulated_enc_out, dec_lens = regulate_len(
durs_predicted, enc_out)
dec_out, _ = self.decoder(len_regulated_enc_out,
get_mask_from_lengths(dec_lens).unsqueeze(2))
pred_spect = self.proj(dec_out)
return pred_spect
def on_train_epoch_start(self):
bin_loss_start_epoch = np.ceil(self.bin_loss_start_ratio *
self._trainer.max_epochs)
# Add bin loss when current_epoch >= bin_start_epoch
if not self.add_bin_loss and self.current_epoch >= bin_loss_start_epoch:
logging.info(
f"Using hard attentions after epoch: {self.current_epoch}")
self.add_bin_loss = True
if self.add_bin_loss:
self.bin_loss_scale = min(
(self.current_epoch - bin_loss_start_epoch) /
self.bin_loss_warmup_epochs, 1.0)
def training_step(self, batch, batch_idx):
attn_prior, lm_tokens = None, None
if self.cond_on_lm_embeddings:
audio, audio_len, text, text_len, attn_prior, pitch, _, lm_tokens = batch
else:
audio, audio_len, text, text_len, attn_prior, pitch, _ = batch
spect, spect_len = self.preprocessor(input_signal=audio,
length=audio_len)
# pitch normalization
zero_pitch_idx = pitch == 0
pitch = (pitch - self.pitch_mean) / self.pitch_std
pitch[zero_pitch_idx] = 0.0
(
pred_spect,
_,
pred_log_durs,
pred_pitch,
attn_soft,
attn_logprob,
attn_hard,
attn_hard_dur,
) = self(
text=text,
text_len=text_len,
pitch=pitch,
spect=spect,
spect_len=spect_len,
attn_prior=attn_prior,
lm_tokens=lm_tokens,
)
(
loss,
durs_loss,
acc,
acc_dist_1,
acc_dist_3,
pitch_loss,
mel_loss,
ctc_loss,
bin_loss,
) = self._metrics(
pred_durs=pred_log_durs,
pred_pitch=pred_pitch,
true_durs=attn_hard_dur,
true_text_len=text_len,
true_pitch=pitch,
true_spect=spect,
pred_spect=pred_spect,
true_spect_len=spect_len,
attn_logprob=attn_logprob,
attn_soft=attn_soft,
attn_hard=attn_hard,
attn_hard_dur=attn_hard_dur,
)
train_log = {
'train_loss':
loss,
'train_durs_loss':
durs_loss,
'train_pitch_loss':
torch.tensor(1.0).to(durs_loss.device)
if pitch_loss is None else pitch_loss,
'train_mel_loss':
mel_loss,
'train_durs_acc':
acc,
'train_durs_acc_dist_3':
acc_dist_3,
'train_ctc_loss':
torch.tensor(1.0).to(durs_loss.device)
if ctc_loss is None else ctc_loss,
'train_bin_loss':
torch.tensor(1.0).to(durs_loss.device)
if bin_loss is None else bin_loss,
}
return {'loss': loss, 'progress_bar': train_log, 'log': train_log}
def validation_step(self, batch, batch_idx):
attn_prior, lm_tokens = None, None
if self.cond_on_lm_embeddings:
audio, audio_len, text, text_len, attn_prior, pitch, _, lm_tokens = batch
else:
audio, audio_len, text, text_len, attn_prior, pitch, _ = batch
spect, spect_len = self.preprocessor(input_signal=audio,
length=audio_len)
# pitch normalization
zero_pitch_idx = pitch == 0
pitch = (pitch - self.pitch_mean) / self.pitch_std
pitch[zero_pitch_idx] = 0.0
(
pred_spect,
_,
pred_log_durs,
pred_pitch,
attn_soft,
attn_logprob,
attn_hard,
attn_hard_dur,
) = self(
text=text,
text_len=text_len,
pitch=pitch,
spect=spect,
spect_len=spect_len,
attn_prior=attn_prior,
lm_tokens=lm_tokens,
)
(
loss,
durs_loss,
acc,
acc_dist_1,
acc_dist_3,
pitch_loss,
mel_loss,
ctc_loss,
bin_loss,
) = self._metrics(
pred_durs=pred_log_durs,
pred_pitch=pred_pitch,
true_durs=attn_hard_dur,
true_text_len=text_len,
true_pitch=pitch,
true_spect=spect,
pred_spect=pred_spect,
true_spect_len=spect_len,
attn_logprob=attn_logprob,
attn_soft=attn_soft,
attn_hard=attn_hard,
attn_hard_dur=attn_hard_dur,
)
# without ground truth internal features except for durations
pred_spect, _, pred_log_durs, pred_pitch, attn_soft, attn_logprob, attn_hard, attn_hard_dur = self(
text=text,
text_len=text_len,
pitch=None,
spect=spect,
spect_len=spect_len,
attn_prior=attn_prior,
lm_tokens=lm_tokens,
)
*_, with_pred_features_mel_loss, _, _ = self._metrics(
pred_durs=pred_log_durs,
pred_pitch=pred_pitch,
true_durs=attn_hard_dur,
true_text_len=text_len,
true_pitch=pitch,
true_spect=spect,
pred_spect=pred_spect,
true_spect_len=spect_len,
attn_logprob=attn_logprob,
attn_soft=attn_soft,
attn_hard=attn_hard,
attn_hard_dur=attn_hard_dur,
)
val_log = {
'val_loss':
loss,
'val_durs_loss':
durs_loss,
'val_pitch_loss':
torch.tensor(1.0).to(durs_loss.device)
if pitch_loss is None else pitch_loss,
'val_mel_loss':
mel_loss,
'val_with_pred_features_mel_loss':
with_pred_features_mel_loss,
'val_durs_acc':
acc,
'val_durs_acc_dist_3':
acc_dist_3,
'val_ctc_loss':
torch.tensor(1.0).to(durs_loss.device)
if ctc_loss is None else ctc_loss,
'val_bin_loss':
torch.tensor(1.0).to(durs_loss.device)
if bin_loss is None else bin_loss,
}
self.log_dict(val_log,
prog_bar=False,
on_epoch=True,
logger=True,
sync_dist=True)
if batch_idx == 0 and self.current_epoch % 5 == 0 and isinstance(
self.logger, WandbLogger):
specs = []
pitches = []
for i in range(min(3, spect.shape[0])):
specs += [
wandb.Image(
plot_spectrogram_to_numpy(
spect[i, :, :spect_len[i]].data.cpu().numpy()),
caption=f"gt mel {i}",
),
wandb.Image(
plot_spectrogram_to_numpy(
pred_spect.transpose(
1, 2)[i, :, :spect_len[i]].data.cpu().numpy()),
caption=f"pred mel {i}",
),
]
pitches += [
wandb.Image(
plot_pitch_to_numpy(
average_pitch(pitch.unsqueeze(1),
attn_hard_dur).squeeze(1)
[i, :text_len[i]].data.cpu().numpy(),
ylim_range=[-2.5, 2.5],
),
caption=f"gt pitch {i}",
),
]
pitches += [
wandb.Image(
plot_pitch_to_numpy(
pred_pitch[i, :text_len[i]].data.cpu().numpy(),
ylim_range=[-2.5, 2.5]),
caption=f"pred pitch {i}",
),
]
self.logger.experiment.log({"specs": specs, "pitches": pitches})
@typecheck(
input_types={
"tokens":
NeuralType(('B', 'T_text'), TokenIndex(), optional=True),
"tokens_len":
NeuralType(('B'), LengthsType(), optional=True),
"lm_tokens":
NeuralType(('B', 'T_lm_tokens'), TokenIndex(), optional=True),
"raw_texts": [NeuralType(optional=True)],
"lm_model":
NeuralType(optional=True),
},
output_types={
"spect": NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType()),
},
)
def generate_spectrogram(
self,
tokens: Optional[torch.Tensor] = None,
tokens_len: Optional[torch.Tensor] = None,
lm_tokens: Optional[torch.Tensor] = None,
raw_texts: Optional[List[str]] = None,
lm_model: str = "albert",
):
if tokens is not None:
if tokens_len is None:
# it is assumed that padding is consecutive and only at the end
tokens_len = (tokens != self.tokenizer.pad).sum(dim=-1)
else:
if raw_texts is None:
logging.error("raw_texts must be specified if tokens is None")
t_seqs = [self.tokenizer(t) for t in raw_texts]
tokens = torch.nn.utils.rnn.pad_sequence(
sequences=[
torch.tensor(t, dtype=torch.long, device=self.device)
for t in t_seqs
],
batch_first=True,
padding_value=self.tokenizer.pad,
)
tokens_len = torch.tensor([len(t) for t in t_seqs],
dtype=torch.long,
device=tokens.device)
if self.cond_on_lm_embeddings and lm_tokens is None:
if raw_texts is None:
logging.error(
"raw_texts must be specified if lm_tokens is None")
lm_model_tokenizer = self._get_lm_model_tokenizer(lm_model)
lm_padding_value = lm_model_tokenizer._convert_token_to_id('<pad>')
lm_space_value = lm_model_tokenizer._convert_token_to_id('▁')
assert isinstance(self.tokenizer,
EnglishCharsTokenizer) or isinstance(
self.tokenizer, EnglishPhonemesTokenizer)
preprocess_texts_as_tts_input = [
self.tokenizer.text_preprocessing_func(t) for t in raw_texts
]
lm_tokens_as_ids_list = [
lm_model_tokenizer.encode(t, add_special_tokens=False)
for t in preprocess_texts_as_tts_input
]
if self.tokenizer.pad_with_space:
lm_tokens_as_ids_list = [[lm_space_value] + t +
[lm_space_value]
for t in lm_tokens_as_ids_list]
lm_tokens = torch.full(
(len(lm_tokens_as_ids_list),
max([len(t) for t in lm_tokens_as_ids_list])),
fill_value=lm_padding_value,
device=tokens.device,
)
for i, lm_tokens_i in enumerate(lm_tokens_as_ids_list):
lm_tokens[i, :len(lm_tokens_i)] = torch.tensor(
lm_tokens_i, device=tokens.device)
pred_spect = self.infer(tokens, tokens_len,
lm_tokens=lm_tokens).transpose(1, 2)
return pred_spect
def parse(self, text: str, normalize=True) -> torch.Tensor:
if normalize and getattr(self, "text_normalizer_call",
None) is not None:
text = self.text_normalizer_call(text,
**self.text_normalizer_call_args)
return torch.tensor(
self.tokenizer.encode(text)).long().unsqueeze(0).to(self.device)
@staticmethod
def _loader(cfg):
try:
_ = cfg.dataset.manifest_filepath
except omegaconf.errors.MissingMandatoryValue:
logging.warning(
"manifest_filepath was skipped. No dataset for this model.")
return None
dataset = instantiate(cfg.dataset)
return torch.utils.data.DataLoader( # noqa
dataset=dataset,
collate_fn=dataset.collate_fn,
**cfg.dataloader_params,
)
def setup_training_data(self, cfg):
self._train_dl = self._loader(cfg)
def setup_validation_data(self, cfg):
self._validation_dl = self._loader(cfg)
def setup_test_data(self, cfg):
"""Omitted."""
pass
@classmethod
def list_available_models(cls):
"""Empty."""
pass
@property
def input_types(self):
return {
"text":
NeuralType(('B', 'T_text'), TokenIndex()),
"lm_tokens":
NeuralType(('B', 'T_lm_tokens'), TokenIndex(), optional=True),
}
@property
def output_types(self):
return {
"spect": NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType()),
}
def forward_for_export(self, text, lm_tokens=None):
text_mask = (text != self.tokenizer_pad).unsqueeze(2)
spect = self.infer(text=text, text_mask=text_mask,
lm_tokens=lm_tokens).transpose(1, 2)
return spect.to(torch.float)
class FastPitchModel(SpectrogramGenerator, Exportable):
"""FastPitch model (https://arxiv.org/abs/2006.06873) that is used to generate mel spectrogram from text."""
def __init__(self, cfg: DictConfig, trainer: Trainer = None):
# Convert to Hydra 1.0 compatible DictConfig
cfg = model_utils.convert_model_config_to_dict_config(cfg)
cfg = model_utils.maybe_update_config_version(cfg)
# Setup normalizer
self.normalizer = None
self.text_normalizer_call = None
self.text_normalizer_call_kwargs = {}
self._setup_normalizer(cfg)
self.learn_alignment = cfg.get("learn_alignment", False)
# Setup vocabulary (=tokenizer) and input_fft_kwargs (supported only with self.learn_alignment=True)
input_fft_kwargs = {}
if self.learn_alignment:
self.vocab = None
self.ds_class_name = cfg.train_ds.dataset._target_.split(".")[-1]
if self.ds_class_name == "TTSDataset":
self._setup_tokenizer(cfg)
assert self.vocab is not None
input_fft_kwargs["n_embed"] = len(self.vocab.tokens)
input_fft_kwargs["padding_idx"] = self.vocab.pad
elif self.ds_class_name == "AudioToCharWithPriorAndPitchDataset":
logging.warning(
"AudioToCharWithPriorAndPitchDataset class has been deprecated. No support for"
" training or finetuning. Only inference is supported.")
tokenizer_conf = self._get_default_text_tokenizer_conf()
self._setup_tokenizer(tokenizer_conf)
assert self.vocab is not None
input_fft_kwargs["n_embed"] = len(self.vocab.tokens)
input_fft_kwargs["padding_idx"] = self.vocab.pad
else:
raise ValueError(
f"Unknown dataset class: {self.ds_class_name}")
self._parser = None
self._tb_logger = None
super().__init__(cfg=cfg, trainer=trainer)
self.bin_loss_warmup_epochs = cfg.get("bin_loss_warmup_epochs", 100)
self.log_train_images = False
loss_scale = 0.1 if self.learn_alignment else 1.0
dur_loss_scale = loss_scale
pitch_loss_scale = loss_scale
if "dur_loss_scale" in cfg:
dur_loss_scale = cfg.dur_loss_scale
if "pitch_loss_scale" in cfg:
pitch_loss_scale = cfg.pitch_loss_scale
self.mel_loss = MelLoss()
self.pitch_loss = PitchLoss(loss_scale=pitch_loss_scale)
self.duration_loss = DurationLoss(loss_scale=dur_loss_scale)
self.aligner = None
if self.learn_alignment:
self.aligner = instantiate(self._cfg.alignment_module)
self.forward_sum_loss = ForwardSumLoss()
self.bin_loss = BinLoss()
self.preprocessor = instantiate(self._cfg.preprocessor)
input_fft = instantiate(self._cfg.input_fft, **input_fft_kwargs)
output_fft = instantiate(self._cfg.output_fft)
duration_predictor = instantiate(self._cfg.duration_predictor)
pitch_predictor = instantiate(self._cfg.pitch_predictor)
self.fastpitch = FastPitchModule(
input_fft,
output_fft,
duration_predictor,
pitch_predictor,
self.aligner,
cfg.n_speakers,
cfg.symbols_embedding_dim,
cfg.pitch_embedding_kernel_size,
cfg.n_mel_channels,
)
self._input_types = self._output_types = None
def _get_default_text_tokenizer_conf(self):
text_tokenizer: TextTokenizerConfig = TextTokenizerConfig()
return OmegaConf.create(OmegaConf.to_yaml(text_tokenizer))
def _setup_normalizer(self, cfg):
if "text_normalizer" in cfg:
normalizer_kwargs = {}
if "whitelist" in cfg.text_normalizer:
normalizer_kwargs["whitelist"] = self.register_artifact(
'text_normalizer.whitelist', cfg.text_normalizer.whitelist)
self.normalizer = instantiate(cfg.text_normalizer,
**normalizer_kwargs)
self.text_normalizer_call = self.normalizer.normalize
if "text_normalizer_call_kwargs" in cfg:
self.text_normalizer_call_kwargs = cfg.text_normalizer_call_kwargs
def _setup_tokenizer(self, cfg):
text_tokenizer_kwargs = {}
if "g2p" in cfg.text_tokenizer:
g2p_kwargs = {}
if "phoneme_dict" in cfg.text_tokenizer.g2p:
g2p_kwargs["phoneme_dict"] = self.register_artifact(
'text_tokenizer.g2p.phoneme_dict',
cfg.text_tokenizer.g2p.phoneme_dict,
)
if "heteronyms" in cfg.text_tokenizer.g2p:
g2p_kwargs["heteronyms"] = self.register_artifact(
'text_tokenizer.g2p.heteronyms',
cfg.text_tokenizer.g2p.heteronyms,
)
text_tokenizer_kwargs["g2p"] = instantiate(cfg.text_tokenizer.g2p,
**g2p_kwargs)
self.vocab = instantiate(cfg.text_tokenizer, **text_tokenizer_kwargs)
@property
def tb_logger(self):
if self._tb_logger is None:
if self.logger is None and self.logger.experiment is None:
return None
tb_logger = self.logger.experiment
if isinstance(self.logger, LoggerCollection):
for logger in self.logger:
if isinstance(logger, TensorBoardLogger):
tb_logger = logger.experiment
break
self._tb_logger = tb_logger
return self._tb_logger
@property
def parser(self):
if self._parser is not None:
return self._parser
if self.learn_alignment:
ds_class_name = self._cfg.train_ds.dataset._target_.split(".")[-1]
if ds_class_name == "TTSDataset":
self._parser = self.vocab.encode
elif ds_class_name == "AudioToCharWithPriorAndPitchDataset":
if self.vocab is None:
tokenizer_conf = self._get_default_text_tokenizer_conf()
self._setup_tokenizer(tokenizer_conf)
self._parser = self.vocab.encode
else:
raise ValueError(f"Unknown dataset class: {ds_class_name}")
else:
self._parser = parsers.make_parser(
labels=self._cfg.labels,
name='en',
unk_id=-1,
blank_id=-1,
do_normalize=True,
abbreviation_version="fastpitch",
make_table=False,
)
return self._parser
def parse(self, str_input: str, normalize=True) -> torch.tensor:
if self.training:
logging.warning("parse() is meant to be called in eval mode.")
if normalize and self.text_normalizer_call is not None:
str_input = self.text_normalizer_call(
str_input, **self.text_normalizer_call_kwargs)
if self.learn_alignment:
eval_phon_mode = contextlib.nullcontext()
if hasattr(self.vocab, "set_phone_prob"):
eval_phon_mode = self.vocab.set_phone_prob(prob=1.0)
# Disable mixed g2p representation if necessary
with eval_phon_mode:
tokens = self.parser(str_input)
else:
tokens = self.parser(str_input)
x = torch.tensor(tokens).unsqueeze_(0).long().to(self.device)
return x
@typecheck(
input_types={
"text":
NeuralType(('B', 'T_text'), TokenIndex()),
"durs":
NeuralType(('B', 'T_text'), TokenDurationType()),
"pitch":
NeuralType(('B', 'T_audio'), RegressionValuesType()),
"speaker":
NeuralType(('B'), Index(), optional=True),
"pace":
NeuralType(optional=True),
"spec":
NeuralType(('B', 'D', 'T_spec'),
MelSpectrogramType(),
optional=True),
"attn_prior":
NeuralType(('B', 'T_spec', 'T_text'), ProbsType(), optional=True),
"mel_lens":
NeuralType(('B'), LengthsType(), optional=True),
"input_lens":
NeuralType(('B'), LengthsType(), optional=True),
})
def forward(
self,
*,
text,
durs=None,
pitch=None,
speaker=None,
pace=1.0,
spec=None,
attn_prior=None,
mel_lens=None,
input_lens=None,
):
return self.fastpitch(
text=text,
durs=durs,
pitch=pitch,
speaker=speaker,
pace=pace,
spec=spec,
attn_prior=attn_prior,
mel_lens=mel_lens,
input_lens=input_lens,
)
@typecheck(output_types={
"spect": NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType())
})
def generate_spectrogram(self,
tokens: 'torch.tensor',
speaker: Optional[int] = None,
pace: float = 1.0) -> torch.tensor:
if self.training:
logging.warning(
"generate_spectrogram() is meant to be called in eval mode.")
if isinstance(speaker, int):
speaker = torch.tensor([speaker]).to(self.device)
spect, *_ = self(text=tokens,
durs=None,
pitch=None,
speaker=speaker,
pace=pace)
return spect
def training_step(self, batch, batch_idx):
attn_prior, durs, speaker = None, None, None
if self.learn_alignment:
if self.ds_class_name == "TTSDataset":
if SpeakerID in self._train_dl.dataset.sup_data_types_set:
audio, audio_lens, text, text_lens, attn_prior, pitch, _, speaker = batch
else:
audio, audio_lens, text, text_lens, attn_prior, pitch, _ = batch
else:
raise ValueError(
f"Unknown vocab class: {self.vocab.__class__.__name__}")
else:
audio, audio_lens, text, text_lens, durs, pitch, speaker = batch
mels, spec_len = self.preprocessor(input_signal=audio,
length=audio_lens)
mels_pred, _, _, log_durs_pred, pitch_pred, attn_soft, attn_logprob, attn_hard, attn_hard_dur, pitch = self(
text=text,
durs=durs,
pitch=pitch,
speaker=speaker,
pace=1.0,
spec=mels if self.learn_alignment else None,
attn_prior=attn_prior,
mel_lens=spec_len,
input_lens=text_lens,
)
if durs is None:
durs = attn_hard_dur
mel_loss = self.mel_loss(spect_predicted=mels_pred, spect_tgt=mels)
dur_loss = self.duration_loss(log_durs_predicted=log_durs_pred,
durs_tgt=durs,
len=text_lens)
loss = mel_loss + dur_loss
if self.learn_alignment:
ctc_loss = self.forward_sum_loss(attn_logprob=attn_logprob,
in_lens=text_lens,
out_lens=spec_len)
bin_loss_weight = min(
self.current_epoch / self.bin_loss_warmup_epochs, 1.0) * 1.0
bin_loss = self.bin_loss(
hard_attention=attn_hard,
soft_attention=attn_soft) * bin_loss_weight
loss += ctc_loss + bin_loss
pitch_loss = self.pitch_loss(pitch_predicted=pitch_pred,
pitch_tgt=pitch,
len=text_lens)
loss += pitch_loss
self.log("t_loss", loss)
self.log("t_mel_loss", mel_loss)
self.log("t_dur_loss", dur_loss)
self.log("t_pitch_loss", pitch_loss)
if self.learn_alignment:
self.log("t_ctc_loss", ctc_loss)
self.log("t_bin_loss", bin_loss)
# Log images to tensorboard
if self.log_train_images and isinstance(self.logger,
TensorBoardLogger):
self.log_train_images = False
self.tb_logger.add_image(
"train_mel_target",
plot_spectrogram_to_numpy(mels[0].data.cpu().float().numpy()),
self.global_step,
dataformats="HWC",
)
spec_predict = mels_pred[0].data.cpu().float().numpy()
self.tb_logger.add_image(
"train_mel_predicted",
plot_spectrogram_to_numpy(spec_predict),
self.global_step,
dataformats="HWC",
)
if self.learn_alignment:
attn = attn_hard[0].data.cpu().float().numpy().squeeze()
self.tb_logger.add_image(
"train_attn",
plot_alignment_to_numpy(attn.T),
self.global_step,
dataformats="HWC",
)
soft_attn = attn_soft[0].data.cpu().float().numpy().squeeze()
self.tb_logger.add_image(
"train_soft_attn",
plot_alignment_to_numpy(soft_attn.T),
self.global_step,
dataformats="HWC",
)
return loss
def validation_step(self, batch, batch_idx):
attn_prior, durs, speaker = None, None, None
if self.learn_alignment:
if self.ds_class_name == "TTSDataset":
if SpeakerID in self._train_dl.dataset.sup_data_types_set:
audio, audio_lens, text, text_lens, attn_prior, pitch, _, speaker = batch
else:
audio, audio_lens, text, text_lens, attn_prior, pitch, _ = batch
else:
raise ValueError(
f"Unknown vocab class: {self.vocab.__class__.__name__}")
else:
audio, audio_lens, text, text_lens, durs, pitch, speaker = batch
mels, mel_lens = self.preprocessor(input_signal=audio,
length=audio_lens)
# Calculate val loss on ground truth durations to better align L2 loss in time
mels_pred, _, _, log_durs_pred, pitch_pred, _, _, _, attn_hard_dur, pitch = self(
text=text,
durs=durs,
pitch=pitch,
speaker=speaker,
pace=1.0,
spec=mels if self.learn_alignment else None,
attn_prior=attn_prior,
mel_lens=mel_lens,
input_lens=text_lens,
)
if durs is None:
durs = attn_hard_dur
mel_loss = self.mel_loss(spect_predicted=mels_pred, spect_tgt=mels)
dur_loss = self.duration_loss(log_durs_predicted=log_durs_pred,
durs_tgt=durs,
len=text_lens)
pitch_loss = self.pitch_loss(pitch_predicted=pitch_pred,
pitch_tgt=pitch,
len=text_lens)
loss = mel_loss + dur_loss + pitch_loss
return {
"val_loss": loss,
"mel_loss": mel_loss,
"dur_loss": dur_loss,
"pitch_loss": pitch_loss,
"mel_target": mels if batch_idx == 0 else None,
"mel_pred": mels_pred if batch_idx == 0 else None,
}
def validation_epoch_end(self, outputs):
collect = lambda key: torch.stack([x[key] for x in outputs]).mean()
val_loss = collect("val_loss")
mel_loss = collect("mel_loss")
dur_loss = collect("dur_loss")
pitch_loss = collect("pitch_loss")
self.log("v_loss", val_loss)
self.log("v_mel_loss", mel_loss)
self.log("v_dur_loss", dur_loss)
self.log("v_pitch_loss", pitch_loss)
_, _, _, _, spec_target, spec_predict = outputs[0].values()
if isinstance(self.logger, TensorBoardLogger):
self.tb_logger.add_image(
"val_mel_target",
plot_spectrogram_to_numpy(
spec_target[0].data.cpu().float().numpy()),
self.global_step,
dataformats="HWC",
)
spec_predict = spec_predict[0].data.cpu().float().numpy()
self.tb_logger.add_image(
"val_mel_predicted",
plot_spectrogram_to_numpy(spec_predict),
self.global_step,
dataformats="HWC",
)
self.log_train_images = True
def __setup_dataloader_from_config(self,
cfg,
shuffle_should_be: bool = True,
name: str = "train"):
if "dataset" not in cfg or not isinstance(cfg.dataset, DictConfig):
raise ValueError(f"No dataset for {name}")
if "dataloader_params" not in cfg or not isinstance(
cfg.dataloader_params, DictConfig):
raise ValueError(f"No dataloder_params for {name}")
if shuffle_should_be:
if 'shuffle' not in cfg.dataloader_params:
logging.warning(
f"Shuffle should be set to True for {self}'s {name} dataloader but was not found in its "
"config. Manually setting to True")
with open_dict(cfg.dataloader_params):
cfg.dataloader_params.shuffle = True
elif not cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to False!!!"
)
elif not shuffle_should_be and cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to True!!!")
if cfg.dataset._target_ == "nemo.collections.tts.torch.data.TTSDataset":
phon_mode = contextlib.nullcontext()
if hasattr(self.vocab, "set_phone_prob"):
phon_mode = self.vocab.set_phone_prob(
prob=None if name ==
"val" else self.vocab.phoneme_probability)
with phon_mode:
dataset = instantiate(
cfg.dataset,
text_normalizer=self.normalizer,
text_normalizer_call_kwargs=self.
text_normalizer_call_kwargs,
text_tokenizer=self.vocab,
)
else:
dataset = instantiate(cfg.dataset)
return torch.utils.data.DataLoader(dataset,
collate_fn=dataset.collate_fn,
**cfg.dataloader_params)
def setup_training_data(self, cfg):
self._train_dl = self.__setup_dataloader_from_config(cfg)
def setup_validation_data(self, cfg):
self._validation_dl = self.__setup_dataloader_from_config(
cfg, shuffle_should_be=False, name="val")
def setup_test_data(self, cfg):
"""Omitted."""
pass
@classmethod
def list_available_models(cls) -> 'List[PretrainedModelInfo]':
"""
This method returns a list of pre-trained model which can be instantiated directly from NVIDIA's NGC cloud.
Returns:
List of available pre-trained models.
"""
list_of_models = []
model = PretrainedModelInfo(
pretrained_model_name="tts_en_fastpitch",
location=
"https://api.ngc.nvidia.com/v2/models/nvidia/nemo/tts_en_fastpitch/versions/1.8.1/files/tts_en_fastpitch_align.nemo",
description=
"This model is trained on LJSpeech sampled at 22050Hz with and can be used to generate female English voices with an American accent.",
class_=cls,
)
list_of_models.append(model)
return list_of_models
# Methods for model exportability
def _prepare_for_export(self, **kwargs):
super()._prepare_for_export(**kwargs)
# Define input_types and output_types as required by export()
self._input_types = {
"text": NeuralType(('B', 'T_text'), TokenIndex()),
"pitch": NeuralType(('B', 'T_text'), RegressionValuesType()),
"pace": NeuralType(('B', 'T_text'), optional=True),
"volume": NeuralType(('B', 'T_text')),
"speaker": NeuralType(('B'), Index()),
}
self._output_types = {
"spect":
NeuralType(('B', 'D', 'T_spec'), MelSpectrogramType()),
"num_frames":
NeuralType(('B'), TokenDurationType()),
"durs_predicted":
NeuralType(('B', 'T_text'), TokenDurationType()),
"log_durs_predicted":
NeuralType(('B', 'T_text'), TokenLogDurationType()),
"pitch_predicted":
NeuralType(('B', 'T_text'), RegressionValuesType()),
"volume_aligned":
NeuralType(('B', 'T_spec'), RegressionValuesType()),
}
def _export_teardown(self):
self._input_types = self._output_types = None
@property
def disabled_deployment_input_names(self):
"""Implement this method to return a set of input names disabled for export"""
disabled_inputs = set()
if self.fastpitch.speaker_emb is None:
disabled_inputs.add("speaker")
return disabled_inputs
@property
def input_types(self):
return self._input_types
@property
def output_types(self):
return self._output_types
def input_example(self, max_batch=1, max_dim=44):
"""
Generates input examples for tracing etc.
Returns:
A tuple of input examples.
"""
par = next(self.fastpitch.parameters())
sz = (max_batch, max_dim)
inp = torch.randint(0,
self.fastpitch.encoder.word_emb.num_embeddings,
sz,
device=par.device,
dtype=torch.int64)
pitch = torch.randn(sz, device=par.device, dtype=torch.float32) * 0.5
pace = torch.clamp(
(torch.randn(sz, device=par.device, dtype=torch.float32) + 1) *
0.1,
min=0.01)
volume = torch.clamp(
(torch.randn(sz, device=par.device, dtype=torch.float32) + 1) *
0.1,
min=0.01)
inputs = {'text': inp, 'pitch': pitch, 'pace': pace, 'volume': volume}
if self.fastpitch.speaker_emb is not None:
inputs['speaker'] = torch.randint(
0,
self.fastpitch.speaker_emb.num_embeddings, (max_batch, ),
device=par.device,
dtype=torch.int64)
return (inputs, )
def forward_for_export(self, text, pitch, pace, volume, speaker=None):
return self.fastpitch.infer(text=text,
pitch=pitch,
pace=pace,
volume=volume,
speaker=speaker)
def input_types(self):
return {
'text': NeuralType(('B', 'T'), EmbeddedTextType()),
'text_len': NeuralType(('B',), LengthsType()),
}
def output_types(self):
return {
'durs': NeuralType(('B', 'T'), LengthsType()),
}
def input_types(self):
return {
'durs_true': NeuralType(('B', 'T'), LengthsType()),
'len_true': NeuralType(('B',), LengthsType()),
'durs_pred': NeuralType(('B', 'T'), LengthsType()),
}
def input_types(self):
return {
'mel_true': NeuralType(('B', 'D', 'T'), SpectrogramType()),
'len_true': NeuralType(('B',), LengthsType()),
'mel_pred': NeuralType(('B', 'D', 'T'), SpectrogramType()),
}
class FastPitchModel(SpectrogramGenerator):
"""FastPitch Model that is used to generate mel spectrograms from text"""
def __init__(self, cfg: DictConfig, trainer: Trainer = None):
if isinstance(cfg, dict):
cfg = OmegaConf.create(cfg)
self.learn_alignment = False
if "learn_alignment" in cfg:
self.learn_alignment = cfg.learn_alignment
self._parser = None
self._tb_logger = None
super().__init__(cfg=cfg, trainer=trainer)
schema = OmegaConf.structured(FastPitchConfig)
# ModelPT ensures that cfg is a DictConfig, but do this second check in case ModelPT changes
if isinstance(cfg, dict):
cfg = OmegaConf.create(cfg)
elif not isinstance(cfg, DictConfig):
raise ValueError(
f"cfg was type: {type(cfg)}. Expected either a dict or a DictConfig"
)
# Ensure passed cfg is compliant with schema
OmegaConf.merge(cfg, schema)
self.bin_loss_warmup_epochs = 100
self.aligner = None
self.log_train_images = False
self.mel_loss = MelLoss()
loss_scale = 0.1 if self.learn_alignment else 1.0
self.pitch_loss = PitchLoss(loss_scale=loss_scale)
self.duration_loss = DurationLoss(loss_scale=loss_scale)
input_fft_kwargs = {}
if self.learn_alignment:
self.aligner = instantiate(self._cfg.alignment_module)
self.forward_sum_loss = ForwardSumLoss()
self.bin_loss = BinLoss()
self.vocab = AudioToCharWithDursF0Dataset.make_vocab(
**self._cfg.train_ds.dataset.vocab)
input_fft_kwargs["n_embed"] = len(self.vocab.labels)
input_fft_kwargs["padding_idx"] = self.vocab.pad
self.preprocessor = instantiate(self._cfg.preprocessor)
input_fft = instantiate(self._cfg.input_fft, **input_fft_kwargs)
output_fft = instantiate(self._cfg.output_fft)
duration_predictor = instantiate(self._cfg.duration_predictor)
pitch_predictor = instantiate(self._cfg.pitch_predictor)
self.fastpitch = FastPitchModule(
input_fft,
output_fft,
duration_predictor,
pitch_predictor,
self.aligner,
cfg.n_speakers,
cfg.symbols_embedding_dim,
cfg.pitch_embedding_kernel_size,
cfg.n_mel_channels,
)
@property
def tb_logger(self):
if self._tb_logger is None:
if self.logger is None and self.logger.experiment is None:
return None
tb_logger = self.logger.experiment
if isinstance(self.logger, LoggerCollection):
for logger in self.logger:
if isinstance(logger, TensorBoardLogger):
tb_logger = logger.experiment
break
self._tb_logger = tb_logger
return self._tb_logger
@property
def parser(self):
if self._parser is not None:
return self._parser
if self.learn_alignment:
vocab = AudioToCharWithDursF0Dataset.make_vocab(
**self._cfg.train_ds.dataset.vocab)
self._parser = vocab.encode
else:
self._parser = parsers.make_parser(
labels=self._cfg.labels,
name='en',
unk_id=-1,
blank_id=-1,
do_normalize=True,
abbreviation_version="fastpitch",
make_table=False,
)
return self._parser
def parse(self, str_input: str) -> torch.tensor:
if str_input[-1] not in [".", "!", "?"]:
str_input = str_input + "."
tokens = self.parser(str_input)
x = torch.tensor(tokens).unsqueeze_(0).long().to(self.device)
return x
@typecheck(
input_types={
"text":
NeuralType(('B', 'T'), TokenIndex()),
"durs":
NeuralType(('B', 'T'), TokenDurationType()),
"pitch":
NeuralType(('B', 'T'), RegressionValuesType()),
"speaker":
NeuralType(('B'), Index()),
"pace":
NeuralType(optional=True),
"spec":
NeuralType(('B', 'D', 'T'), MelSpectrogramType(), optional=True),
"attn_prior":
NeuralType(('B', 'T', 'T'), ProbsType(), optional=True),
"mel_lens":
NeuralType(('B'), LengthsType(), optional=True),
"input_lens":
NeuralType(('B'), LengthsType(), optional=True),
})
def forward(
self,
*,
text,
durs=None,
pitch=None,
speaker=0,
pace=1.0,
spec=None,
attn_prior=None,
mel_lens=None,
input_lens=None,
):
return self.fastpitch(
text=text,
durs=durs,
pitch=pitch,
speaker=speaker,
pace=pace,
spec=spec,
attn_prior=attn_prior,
mel_lens=mel_lens,
input_lens=input_lens,
)
@typecheck(output_types={
"spect": NeuralType(('B', 'C', 'T'), MelSpectrogramType())
})
def generate_spectrogram(self,
tokens: 'torch.tensor',
speaker: int = 0,
pace: float = 1.0) -> torch.tensor:
self.eval()
spect, *_ = self(text=tokens,
durs=None,
pitch=None,
speaker=speaker,
pace=pace)
return spect.transpose(1, 2)
def training_step(self, batch, batch_idx):
attn_prior, durs, speakers = None, None, None
if self.learn_alignment:
audio, audio_lens, text, text_lens, attn_prior, pitch = batch
else:
audio, audio_lens, text, text_lens, durs, pitch, speakers = batch
mels, spec_len = self.preprocessor(input_signal=audio,
length=audio_lens)
mels_pred, _, log_durs_pred, pitch_pred, attn_soft, attn_logprob, attn_hard, attn_hard_dur, pitch = self(
text=text,
durs=durs,
pitch=pitch,
speaker=speakers,
pace=1.0,
spec=mels if self.learn_alignment else None,
attn_prior=attn_prior,
mel_lens=spec_len,
input_lens=text_lens,
)
if durs is None:
durs = attn_hard_dur
mel_loss = self.mel_loss(spect_predicted=mels_pred, spect_tgt=mels)
dur_loss = self.duration_loss(log_durs_predicted=log_durs_pred,
durs_tgt=durs,
len=text_lens)
loss = mel_loss + dur_loss
if self.learn_alignment:
ctc_loss = self.forward_sum_loss(attn_logprob=attn_logprob,
in_lens=text_lens,
out_lens=spec_len)
bin_loss_weight = min(
self.current_epoch / self.bin_loss_warmup_epochs, 1.0) * 1.0
bin_loss = self.bin_loss(
hard_attention=attn_hard,
soft_attention=attn_soft) * bin_loss_weight
loss += ctc_loss + bin_loss
pitch_loss = self.pitch_loss(pitch_predicted=pitch_pred,
pitch_tgt=pitch,
len=text_lens)
loss += pitch_loss
self.log("t_loss", loss)
self.log("t_mel_loss", mel_loss)
self.log("t_dur_loss", dur_loss)
self.log("t_pitch_loss", pitch_loss)
if self.learn_alignment:
self.log("t_ctc_loss", ctc_loss)
self.log("t_bin_loss", bin_loss)
# Log images to tensorboard
if self.log_train_images:
self.log_train_images = False
self.tb_logger.add_image(
"train_mel_target",
plot_spectrogram_to_numpy(mels[0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
spec_predict = mels_pred[0].data.cpu().numpy().T
self.tb_logger.add_image(
"train_mel_predicted",
plot_spectrogram_to_numpy(spec_predict),
self.global_step,
dataformats="HWC",
)
if self.learn_alignment:
attn = attn_hard[0].data.cpu().numpy().squeeze()
self.tb_logger.add_image(
"train_attn",
plot_alignment_to_numpy(attn.T),
self.global_step,
dataformats="HWC",
)
soft_attn = attn_soft[0].data.cpu().numpy().squeeze()
self.tb_logger.add_image(
"train_soft_attn",
plot_alignment_to_numpy(soft_attn.T),
self.global_step,
dataformats="HWC",
)
return loss
def validation_step(self, batch, batch_idx):
attn_prior, durs, speakers = None, None, None
if self.learn_alignment:
audio, audio_lens, text, text_lens, attn_prior, pitch = batch
else:
audio, audio_lens, text, text_lens, durs, pitch, speakers = batch
mels, mel_lens = self.preprocessor(input_signal=audio,
length=audio_lens)
# Calculate val loss on ground truth durations to better align L2 loss in time
mels_pred, _, log_durs_pred, pitch_pred, _, _, _, attn_hard_dur, pitch = self(
text=text,
durs=durs,
pitch=pitch,
speaker=speakers,
pace=1.0,
spec=mels if self.learn_alignment else None,
attn_prior=attn_prior,
mel_lens=mel_lens,
input_lens=text_lens,
)
if durs is None:
durs = attn_hard_dur
mel_loss = self.mel_loss(spect_predicted=mels_pred, spect_tgt=mels)
dur_loss = self.duration_loss(log_durs_predicted=log_durs_pred,
durs_tgt=durs,
len=text_lens)
pitch_loss = self.pitch_loss(pitch_predicted=pitch_pred,
pitch_tgt=pitch,
len=text_lens)
loss = mel_loss + dur_loss + pitch_loss
return {
"val_loss": loss,
"mel_loss": mel_loss,
"dur_loss": dur_loss,
"pitch_loss": pitch_loss,
"mel_target": mels if batch_idx == 0 else None,
"mel_pred": mels_pred if batch_idx == 0 else None,
}
def validation_epoch_end(self, outputs):
collect = lambda key: torch.stack([x[key] for x in outputs]).mean()
val_loss = collect("val_loss")
mel_loss = collect("mel_loss")
dur_loss = collect("dur_loss")
pitch_loss = collect("pitch_loss")
self.log("v_loss", val_loss)
self.log("v_mel_loss", mel_loss)
self.log("v_dur_loss", dur_loss)
self.log("v_pitch_loss", pitch_loss)
_, _, _, _, spec_target, spec_predict = outputs[0].values()
self.tb_logger.add_image(
"val_mel_target",
plot_spectrogram_to_numpy(spec_target[0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
spec_predict = spec_predict[0].data.cpu().numpy()
self.tb_logger.add_image(
"val_mel_predicted",
plot_spectrogram_to_numpy(spec_predict.T),
self.global_step,
dataformats="HWC",
)
self.log_train_images = True
def __setup_dataloader_from_config(self,
cfg,
shuffle_should_be: bool = True,
name: str = "train"):
if "dataset" not in cfg or not isinstance(cfg.dataset, DictConfig):
raise ValueError(f"No dataset for {name}")
if "dataloader_params" not in cfg or not isinstance(
cfg.dataloader_params, DictConfig):
raise ValueError(f"No dataloder_params for {name}")
if shuffle_should_be:
if 'shuffle' not in cfg.dataloader_params:
logging.warning(
f"Shuffle should be set to True for {self}'s {name} dataloader but was not found in its "
"config. Manually setting to True")
with open_dict(cfg.dataloader_params):
cfg.dataloader_params.shuffle = True
elif not cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to False!!!"
)
elif not shuffle_should_be and cfg.dataloader_params.shuffle:
logging.error(
f"The {name} dataloader for {self} has shuffle set to True!!!")
kwargs_dict = {}
if cfg.dataset._target_ == "nemo.collections.asr.data.audio_to_text.FastPitchDataset":
kwargs_dict["parser"] = self.parser
dataset = instantiate(cfg.dataset, **kwargs_dict)
return torch.utils.data.DataLoader(dataset,
collate_fn=dataset.collate_fn,
**cfg.dataloader_params)
def setup_training_data(self, cfg):
self._train_dl = self.__setup_dataloader_from_config(cfg)
def setup_validation_data(self, cfg):
self._validation_dl = self.__setup_dataloader_from_config(
cfg, shuffle_should_be=False, name="val")
def setup_test_data(self, cfg):
"""Omitted."""
pass
@classmethod
def list_available_models(cls) -> 'List[PretrainedModelInfo]':
"""
This method returns a list of pre-trained model which can be instantiated directly from NVIDIA's NGC cloud.
Returns:
List of available pre-trained models.
"""
list_of_models = []
model = PretrainedModelInfo(
pretrained_model_name="tts_en_fastpitch",
location=
"https://api.ngc.nvidia.com/v2/models/nvidia/nemo/tts_en_fastpitch/versions/1.0.0/files/tts_en_fastpitch.nemo",
description=
"This model is trained on LJSpeech sampled at 22050Hz with and can be used to generate female English voices with an American accent.",
class_=cls,
)
list_of_models.append(model)
return list_of_models
class GlowTTSModule(NeuralModule):
def __init__(
self, encoder_module: NeuralModule, decoder_module: NeuralModule, n_speakers: int = 1, gin_channels: int = 0
):
"""
Main GlowTTS module. Contains the encoder and decoder.
Args:
encoder_module (NeuralModule): Text encoder for predicting latent distribution statistics
decoder_module (NeuralModule): Invertible spectrogram decoder
n_speakers (int): Number of speakers
gin_channels (int): Channels in speaker embeddings
"""
super().__init__()
self.encoder = encoder_module
self.decoder = decoder_module
if n_speakers > 1:
self.emb_g = nn.Embedding(n_speakers, gin_channels)
nn.init.uniform_(self.emb_g.weight, -0.1, 0.1)
@property
def input_types(self):
return {
"text": NeuralType(('B', 'T'), TokenIndex()),
"text_lengths": NeuralType(('B'), LengthsType()),
"spect": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"spect_lengths": NeuralType(('B'), LengthsType()),
"speaker": NeuralType(('B'), IntType(), optional=True),
}
@property
def output_types(self):
return {
"z": NeuralType(('B', 'D', 'T'), NormalDistributionSamplesType()),
"y_m": NeuralType(('B', 'D', 'T'), NormalDistributionMeanType()),
"y_logs": NeuralType(('B', 'D', 'T'), NormalDistributionLogVarianceType()),
"logdet": NeuralType(('B'), LogDeterminantType()),
"log_durs_predicted": NeuralType(('B', 'T'), TokenLogDurationType()),
"log_durs_extracted": NeuralType(('B', 'T'), TokenLogDurationType()),
"spect_lengths": NeuralType(('B'), LengthsType()),
"attn": NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
}
@typecheck()
def forward(self, *, text, text_lengths, spect, spect_lengths, speaker=None):
if speaker is not None:
speaker = F.normalize(self.emb_g(speaker)).unsqueeze(-1) # [b, h]
x_m, x_logs, log_durs_predicted, x_mask = self.encoder(
text=text, text_lengths=text_lengths, speaker_embeddings=speaker
)
y_max_length = spect.size(2)
y_max_length = (y_max_length // self.decoder.n_sqz) * self.decoder.n_sqz
spect = spect[:, :, :y_max_length]
spect_lengths = (spect_lengths // self.decoder.n_sqz) * self.decoder.n_sqz
y_mask = torch.unsqueeze(glow_tts_submodules.sequence_mask(spect_lengths, y_max_length), 1).to(x_mask.dtype)
attn_mask = torch.unsqueeze(x_mask, -1) * torch.unsqueeze(y_mask, 2)
z, logdet = self.decoder(spect=spect, spect_mask=y_mask, speaker_embeddings=speaker, reverse=False)
with torch.no_grad():
x_s_sq_r = torch.exp(-2 * x_logs)
logp1 = torch.sum(-0.5 * math.log(2 * math.pi) - x_logs, [1]).unsqueeze(-1) # [b, t, 1]
logp2 = torch.matmul(x_s_sq_r.transpose(1, 2), -0.5 * (z ** 2)) # [b, t, d] x [b, d, t'] = [b, t, t']
logp3 = torch.matmul((x_m * x_s_sq_r).transpose(1, 2), z) # [b, t, d] x [b, d, t'] = [b, t, t']
logp4 = torch.sum(-0.5 * (x_m ** 2) * x_s_sq_r, [1]).unsqueeze(-1) # [b, t, 1]
logp = logp1 + logp2 + logp3 + logp4 # [b, t, t']
attn = (glow_tts_submodules.maximum_path(logp, attn_mask.squeeze(1)).unsqueeze(1).detach()).squeeze(1)
y_m = torch.matmul(x_m, attn)
y_logs = torch.matmul(x_logs, attn)
log_durs_extracted = torch.log(1e-8 + torch.sum(attn, -1)) * x_mask.squeeze()
return z, y_m, y_logs, logdet, log_durs_predicted, log_durs_extracted, spect_lengths, attn
@typecheck(
input_types={
"text": NeuralType(('B', 'T'), TokenIndex()),
"text_lengths": NeuralType(('B',), LengthsType()),
"speaker": NeuralType(('B'), IntType(), optional=True),
"noise_scale": NeuralType(optional=True),
"length_scale": NeuralType(optional=True),
},
output_types={
"y": NeuralType(('B', 'D', 'T'), MelSpectrogramType()),
"attn": NeuralType(('B', 'T', 'T'), SequenceToSequenceAlignmentType()),
},
)
def generate_spect(self, *, text, text_lengths, speaker=None, noise_scale=0.3, length_scale=1.0):
if speaker is not None:
speaker = F.normalize(self.emb_g(speaker)).unsqueeze(-1) # [b, h]
x_m, x_logs, log_durs_predicted, x_mask = self.encoder(
text=text, text_lengths=text_lengths, speaker_embeddings=speaker
)
w = torch.exp(log_durs_predicted) * x_mask.squeeze() * length_scale
w_ceil = torch.ceil(w)
spect_lengths = torch.clamp_min(torch.sum(w_ceil, [1]), 1).long()
y_max_length = None
spect_lengths = (spect_lengths // self.decoder.n_sqz) * self.decoder.n_sqz
y_mask = torch.unsqueeze(glow_tts_submodules.sequence_mask(spect_lengths, y_max_length), 1).to(x_mask.dtype)
attn_mask = torch.unsqueeze(x_mask, -1) * torch.unsqueeze(y_mask, 2)
attn = glow_tts_submodules.generate_path(w_ceil.squeeze(1), attn_mask.squeeze(1))
y_m = torch.matmul(x_m, attn)
y_logs = torch.matmul(x_logs, attn)
z = (y_m + torch.exp(y_logs) * torch.randn_like(y_m) * noise_scale) * y_mask
y, _ = self.decoder(spect=z, spect_mask=y_mask, speaker_embeddings=speaker, reverse=True)
return y, attn
def save_to(self, save_path: str):
"""TODO: Implement"""
@classmethod
def restore_from(cls, restore_path: str):
"""TODO: Implement"""
def input_types(self):
return {
"audio": NeuralType(('B', 'T'), AudioSignal()),
"audio_len": NeuralType(('B'), LengthsType()),
"run_inverse": NeuralType(optional=True),
}
def input_types(self):
return {
"audio": NeuralType(('B', 'T'), AudioSignal()),
"audio_len": NeuralType(('B'), LengthsType()),
}
class GlowTTSModel(SpectrogramGenerator):
"""
GlowTTS model used to generate spectrograms from text
Consists of a text encoder and an invertible spectrogram decoder
"""
def __init__(self, cfg: DictConfig, trainer: Trainer = None):
if isinstance(cfg, dict):
cfg = OmegaConf.create(cfg)
self.parser = instantiate(cfg.parser)
super().__init__(cfg=cfg, trainer=trainer)
schema = OmegaConf.structured(GlowTTSConfig)
# ModelPT ensures that cfg is a DictConfig, but do this second check in case ModelPT changes
if isinstance(cfg, dict):
cfg = OmegaConf.create(cfg)
elif not isinstance(cfg, DictConfig):
raise ValueError(
f"cfg was type: {type(cfg)}. Expected either a dict or a DictConfig"
)
# Ensure passed cfg is compliant with schema
OmegaConf.merge(cfg, schema)
self.preprocessor = instantiate(self._cfg.preprocessor)
encoder = instantiate(self._cfg.encoder)
decoder = instantiate(self._cfg.decoder)
self.glow_tts = GlowTTSModule(encoder,
decoder,
n_speakers=cfg.n_speakers,
gin_channels=cfg.gin_channels)
self.loss = GlowTTSLoss()
def parse(self, str_input: str) -> torch.tensor:
if str_input[-1] not in [".", "!", "?"]:
str_input = str_input + "."
tokens = self.parser(str_input)
x = torch.tensor(tokens).unsqueeze_(0).long().to(self.device)
return x
@typecheck(
input_types={
"x": NeuralType(('B', 'T'), TokenIndex()),
"x_lengths": NeuralType(('B'), LengthsType()),
"y": NeuralType(('B', 'D', 'T'),
MelSpectrogramType(),
optional=True),
"y_lengths": NeuralType(('B'), LengthsType(), optional=True),
"gen": NeuralType(optional=True),
"noise_scale": NeuralType(optional=True),
"length_scale": NeuralType(optional=True),
})
def forward(self,
*,
x,
x_lengths,
y=None,
y_lengths=None,
gen=False,
noise_scale=0.3,
length_scale=1.0):
if gen:
return self.glow_tts.generate_spect(text=x,
text_lengths=x_lengths,
noise_scale=noise_scale,
length_scale=length_scale)
else:
return self.glow_tts(text=x,
text_lengths=x_lengths,
spect=y,
spect_lengths=y_lengths)
def step(self, y, y_lengths, x, x_lengths):
z, y_m, y_logs, logdet, logw, logw_, y_lengths, attn = self(
x=x, x_lengths=x_lengths, y=y, y_lengths=y_lengths, gen=False)
l_mle, l_length, logdet = self.loss(
z=z,
y_m=y_m,
y_logs=y_logs,
logdet=logdet,
logw=logw,
logw_=logw_,
x_lengths=x_lengths,
y_lengths=y_lengths,
)
loss = sum([l_mle, l_length])
return l_mle, l_length, logdet, loss, attn
def training_step(self, batch, batch_idx):
y, y_lengths, x, x_lengths = batch
y, y_lengths = self.preprocessor(input_signal=y, length=y_lengths)
l_mle, l_length, logdet, loss, _ = self.step(y, y_lengths, x,
x_lengths)
output = {
"loss": loss, # required
"progress_bar": {
"l_mle": l_mle,
"l_length": l_length,
"logdet": logdet
},
"log": {
"loss": loss,
"l_mle": l_mle,
"l_length": l_length,
"logdet": logdet
},
}
return output
def validation_step(self, batch, batch_idx):
y, y_lengths, x, x_lengths = batch
y, y_lengths = self.preprocessor(input_signal=y, length=y_lengths)
l_mle, l_length, logdet, loss, attn = self.step(
y, y_lengths, x, x_lengths)
y_gen, attn_gen = self(x=x, x_lengths=x_lengths, gen=True)
return {
"loss": loss,
"l_mle": l_mle,
"l_length": l_length,
"logdet": logdet,
"y": y,
"y_gen": y_gen,
"x": x,
"attn": attn,
"attn_gen": attn_gen,
"progress_bar": {
"l_mle": l_mle,
"l_length": l_length,
"logdet": logdet
},
}
def validation_epoch_end(self, outputs):
avg_loss = torch.stack([x['loss'] for x in outputs]).mean()
avg_mle = torch.stack([x['l_mle'] for x in outputs]).mean()
avg_length_loss = torch.stack([x['l_length'] for x in outputs]).mean()
avg_logdet = torch.stack([x['logdet'] for x in outputs]).mean()
tensorboard_logs = {
'val_loss': avg_loss,
'val_mle': avg_mle,
'val_length_loss': avg_length_loss,
'val_logdet': avg_logdet,
}
if self.logger is not None and self.logger.experiment is not None:
tb_logger = self.logger.experiment
if isinstance(self.logger, LoggerCollection):
for logger in self.logger:
if isinstance(logger, TensorBoardLogger):
tb_logger = logger.experiment
break
separated_phonemes = "|".join(
[self.parser.symbols[c] for c in outputs[0]['x'][0]])
tb_logger.add_text("separated phonemes", separated_phonemes,
self.global_step)
tb_logger.add_image(
"real_spectrogram",
plot_spectrogram_to_numpy(
outputs[0]['y'][0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
tb_logger.add_image(
"generated_spectrogram",
plot_spectrogram_to_numpy(
outputs[0]['y_gen'][0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
tb_logger.add_image(
"alignment_for_real_sp",
plot_alignment_to_numpy(
outputs[0]['attn'][0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
tb_logger.add_image(
"alignment_for_generated_sp",
plot_alignment_to_numpy(
outputs[0]['attn_gen'][0].data.cpu().numpy()),
self.global_step,
dataformats="HWC",
)
log_audio_to_tb(tb_logger, outputs[0]['y'][0], "true_audio_gf",
self.global_step)
log_audio_to_tb(tb_logger, outputs[0]['y_gen'][0],
"generated_audio_gf", self.global_step)
return {'val_loss': avg_loss, 'log': tensorboard_logs}
def _setup_dataloader_from_config(self, cfg: DictConfig):
if 'manifest_filepath' in cfg and cfg['manifest_filepath'] is None:
logging.warning(
f"Could not load dataset as `manifest_filepath` was None. Provided config : {cfg}"
)
return None
if 'augmentor' in cfg:
augmentor = process_augmentations(cfg['augmentor'])
else:
augmentor = None
dataset = _AudioTextDataset(
manifest_filepath=cfg['manifest_filepath'],
parser=self.parser,
sample_rate=cfg['sample_rate'],
int_values=cfg.get('int_values', False),
augmentor=augmentor,
max_duration=cfg.get('max_duration', None),
min_duration=cfg.get('min_duration', None),
max_utts=cfg.get('max_utts', 0),
trim=cfg.get('trim_silence', True),
load_audio=cfg.get('load_audio', True),
add_misc=cfg.get('add_misc', False),
)
return torch.utils.data.DataLoader(
dataset=dataset,
batch_size=cfg['batch_size'],
collate_fn=dataset.collate_fn,
drop_last=cfg.get('drop_last', False),
shuffle=cfg['shuffle'],
num_workers=cfg.get('num_workers', 0),
)
def setup_training_data(self, train_data_config: Optional[DictConfig]):
self._train_dl = self._setup_dataloader_from_config(
cfg=train_data_config)
def setup_validation_data(self, val_data_config: Optional[DictConfig]):
self._validation_dl = self._setup_dataloader_from_config(
cfg=val_data_config)
def setup_test_data(self, test_data_config: Optional[DictConfig]):
self._test_dl = self._setup_dataloader_from_config(
cfg=test_data_config)
def generate_spectrogram(self,
tokens: 'torch.tensor',
noise_scale: float = 0.0,
length_scale: float = 1.0) -> torch.tensor:
self.eval()
token_len = torch.tensor([tokens.shape[1]]).to(self.device)
spect, _ = self(x=tokens,
x_lengths=token_len,
gen=True,
noise_scale=noise_scale,
length_scale=length_scale)
return spect
@classmethod
def list_available_models(cls) -> 'List[PretrainedModelInfo]':
"""
This method returns a list of pre-trained model which can be instantiated directly from NVIDIA's NGC cloud.
Returns:
List of available pre-trained models.
"""
list_of_models = []
model = PretrainedModelInfo(
pretrained_model_name="tts_en_glowtts",
location=
"https://api.ngc.nvidia.com/v2/models/nvidia/nemo/tts_en_glowtts/versions/1.0.0rc1/files/tts_en_glowtts.nemo",
description=
"This model is trained on LJSpeech sampled at 22050Hz, and can be used to generate female English voices with an American accent.",
class_=cls,
)
list_of_models.append(model)
return list_of_models
