def main(model='Tacotron',input='training_data/train.txt',restore=True, summary_interval=100,base_dir='',hparams='',checkpoint_interval=500):
# parser = argparse.ArgumentParser()
# parser.add_argument('--base_dir', default='')
# parser.add_argument('--hparams', default='',
# help='Hyperparameter overrides as a comma-separated list of name=value pairs')
# parser.add_argument('--input', default='training_data/train.txt')
# parser.add_argument('--name', help='Name of logging directory.')
# parser.add_argument('--model', default='Tacotron')
# parser.add_argument('--restore', type=bool, default=True, help='Set this to False to do a fresh training')
# parser.add_argument('--summary_interval', type=int, default=100,
# help='Steps between running summary ops')
# parser.add_argument('--''', type=int, default=500,
# help='Steps between writing checkpoints')
# parser.add_argument('--tf_log_level', type=int, default=1, help='Tensorflow C++ log level.')
# args = parser.parse_args()
# accepted_models = ['Tacotron', 'Wavenet']
# if args.model not in accepted_models:
# raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
if model == 'Tacotron':
tacotron_train()
elif model == 'Wavenet':
raise NotImplementedError('Wavenet is still a work in progress, thank you for your patience!')
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir', default='')
parser.add_argument('--hparams', default='', help='Hyperparameter overrides as a comma-separated list of name=value pairs')
parser.add_argument('--model', default='Tacotron-2')
parser.add_argument('--mode', default='synthesis', help='mode for synthesis of tacotron after training')
parser.add_argument('--GTA', type=bool, default=True, help='Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode')
parser.add_argument('--restore', type=bool, default=True, help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval', type=int, default=250, help='Steps between running summary ops')
parser.add_argument('--checkpoint_interval', type=int, default=5000, help='Steps between writing checkpoints')
parser.add_argument('--eval_interval', type=int, default=10000, help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps', type=int, default=100000, help='total number of tacotron training steps')
parser.add_argument('--wavernn_train_epochs', type=int, default=300, help='total number of wavenet training epochs')
parser.add_argument('--text_list', default='', help='Text file contains list of texts to be synthesized. Valid if mode=eval')
parser.add_argument('--slack_url', default=None, help='slack webhook notification destination link')
args = parser.parse_args()
accepted_models = ['Tacotron', 'WaveRNN', 'Tacotron-2']
if args.model not in accepted_models:
raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveRNN':
wavernn_train(args, log_dir, hparams)
elif args.model == 'Tacotron-2':
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(args.model, accepted_models))
def train(args, log_dir, hparams):
state_file = os.path.join(log_dir, 'state_log')
# Get training states
(taco_state, GTA_state, wave_state), input_path = read_seq(state_file)
if not taco_state:
logging.debug(
'\n#############################################################\n'
)
logging.debug('Tacotron Train\n')
logging.debug(
'###########################################################\n')
checkpoint = tacotron_train(args, log_dir, hparams)
tf.reset_default_graph()
# Sleep 1/2 second to let previous graph close and avoid error messages while synthesis
sleep(0.5)
if checkpoint is None:
raise ('Error occured while training Tacotron, Exiting!')
taco_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state], input_path)
else:
checkpoint = os.path.join(log_dir, 'taco_pretrained/')
if not GTA_state:
logging.debug(
'\n#############################################################\n'
)
logging.debug('Tacotron GTA Synthesis\n')
logging.debug(
'###########################################################\n')
input_path = tacotron_synthesize(args, hparams, checkpoint)
tf.reset_default_graph()
# Sleep 1/2 second to let previous graph close and avoid error messages while Wavenet is training
sleep(0.5)
GTA_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state], input_path)
else:
input_path = os.path.join('tacotron_' + args.output_dir, 'gta',
'map.txt')
if input_path == '' or input_path is None:
raise RuntimeError(
'input_path has an unpleasant value -> {}'.format(input_path))
if not wave_state:
logging.debug(
'\n#############################################################\n'
)
logging.debug('Wavenet Train\n')
logging.debug(
'###########################################################\n')
checkpoint = wavenet_train(args, log_dir, hparams, input_path)
if checkpoint is None:
raise ('Error occured while training Wavenet, Exiting!')
wave_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state], input_path)
if wave_state and GTA_state and taco_state:
logging.debug('TRAINING IS ALREADY COMPLETE!!')
def train(args, log_dir, hparams):
state_file = os.path.join(log_dir, 'state_log')
(taco_state, GTA_state, wave_state) = read_seq(state_file)
if not taco_state:
log('\n#############################################################\n'
)
log('Tacotron Train\n')
log('###########################################################\n')
checkpoint = tacotron_train(args, log_dir, hparams)
tf.reset_default_graph()
# Sleep 1/2 second to let previous graph close and avoid error messages while synthesis
sleep(0.5)
if checkpoint is None:
raise ('Error occured while training Tacotron, Exiting!')
taco_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state])
else:
checkpoint = os.path.join(log_dir, 'taco_pretrained')
if not GTA_state:
log('\n#############################################################\n'
)
log('Tacotron GTA Synthesis\n')
log('###########################################################\n')
args.mode = 'synthesis'
tacotron_synthesize(args, hparams, checkpoint)
args.mode = 'eval'
tacotron_synthesize(args, hparams, checkpoint, get_sentences(args))
tf.reset_default_graph()
# Sleep 1/2 second to let previous graph close and avoid error messages while Wavenet is training
sleep(0.5)
GTA_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state])
if not wave_state:
log('\n#############################################################\n'
)
log('WaveRNN Train\n')
log('###########################################################\n')
wavernn_preprocess(args, hparams)
wavernn_train(args, log_dir, hparams)
wave_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state])
if wave_state and GTA_state and taco_state:
log('TRAINING IS ALREADY COMPLETE!!')
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir', default='')
parser.add_argument('--hparams', default='',
help='Hyperparameter overrides as a comma-separated list of name=value pairs')
parser.add_argument('--tacotron_input', default='training_data/train.txt')
parser.add_argument('--wavenet_input', default='tacotron_output/gta/map.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='Tacotron-2')
parser.add_argument('--input_dir', default='training_data', help='folder to contain inputs sentences/targets')
parser.add_argument('--output_dir', default='output', help='folder to contain synthesized mel spectrograms')
parser.add_argument('--mode', default='synthesis', help='mode for synthesis of tacotron after training')
parser.add_argument('--GTA', default='True', help='Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode')
parser.add_argument('--restore', type=bool, default=True, help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval', type=int, default=2500,
help='Steps between running summary ops')
parser.add_argument('--embedding_interval', type=int, default=10000,
help='Steps between updating embeddings projection visualization')
parser.add_argument('--checkpoint_interval', type=int, default=5000,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval', type=int, default=10000,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps', type=int, default=150000, help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps', type=int, default=750000, help='total number of wavenet training steps')
parser.add_argument('--tf_log_level', type=int, default=1, help='Tensorflow C++ log level.')
parser.add_argument('--slack_url', default=None, help='slack webhook notification destination link')
args = parser.parse_args()
accepted_models = ['Tacotron', 'WaveNet', 'Tacotron-2']
if args.model not in accepted_models:
raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveNet':
wavenet_train(args, log_dir, hparams, args.wavenet_input)
elif args.model == 'Tacotron-2':
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(args.model, accepted_models))
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir', default='')
parser.add_argument('--hparams', default='',
help='Hyperparameter overrides as a comma-separated list of name=value pairs')
parser.add_argument('--tacotron_input', default='training_data/train.txt')
parser.add_argument('--wavenet_input', default='tacotron_output/gta/map.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='Tacotron-2')
parser.add_argument('--input_dir', default='training_data/', help='folder to contain inputs sentences/targets')
parser.add_argument('--output_dir', default='output/', help='folder to contain synthesized mel spectrograms')
parser.add_argument('--mode', default='synthesis', help='mode for synthesis of tacotron after training')
parser.add_argument('--GTA', default='True', help='Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode')
parser.add_argument('--restore', type=bool, default=True, help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval', type=int, default=250,
help='Steps between running summary ops')
parser.add_argument('--checkpoint_interval', type=int, default=5000,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval', type=int, default=10000,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps', type=int, default=160000, help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps', type=int, default=360000, help='total number of wavenet training steps')
parser.add_argument('--tf_log_level', type=int, default=1, help='Tensorflow C++ log level.')
args = parser.parse_args()
accepted_models = ['Tacotron', 'WaveNet', 'Both', 'Tacotron-2']
if args.model not in accepted_models:
raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveNet':
wavenet_train(args, log_dir, hparams, args.wavenet_input)
elif args.model in ('Both', 'Tacotron-2'):
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(args.model, accepted_models))
def train(args, log_dir, hparams):
log('\n#############################################################\n')
log('Tacotron Train\n')
log('###########################################################\n')
checkpoint = tacotron_train(args, log_dir, hparams)
tf.reset_default_graph()
if checkpoint is None:
raise('Error occured while training Tacotron, Exiting!')
log('\n#############################################################\n')
log('Tacotron GTA Synthesis\n')
log('###########################################################\n')
input_path = tacotron_synthesize(args, hparams, checkpoint)
log('\n#############################################################\n')
log('Wavenet Train\n')
log('###########################################################\n')
wavenet_train(args, log_dir, hparams, input_path)
def train(args, log_dir, hparams):
log('\n#############################################################\n')
log('Tacotron Train\n')
log('###########################################################\n')
checkpoint = tacotron_train(args, log_dir, hparams)
tf.reset_default_graph()
if checkpoint is None:
raise('Error occured while training Tacotron, Exiting!')
log('\n#############################################################\n')
log('Tacotron GTA Synthesis\n')
log('###########################################################\n')
input_path = tacotron_synthesize(args, hparams, checkpoint)
log('\n#############################################################\n')
log('Wavenet Train\n')
log('###########################################################\n')
wavenet_train(args, log_dir, hparams, input_path)
def train(args, log_dir, hparams):
state_file = os.path.join(log_dir, 'state_log')
#Get training states
(taco_state, GTA_state, wave_state), input_path = read_seq(state_file)
print('taco_state, GTA_state, wave_state:', taco_state, GTA_state,
wave_state)
if not taco_state:
log('\n#############################################################\n'
)
log('Tacotron Train\n')
log('###########################################################\n')
checkpoint = tacotron_train(args, log_dir, hparams)
tf.reset_default_graph()
#Sleep 1/2 second to let previous graph close and avoid error messages while synthesis
sleep(0.5)
if checkpoint is None:
raise ('Error occured while training Tacotron, Exiting!')
taco_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state], input_path)
else:
checkpoint = os.path.join(log_dir, 'taco_pretrained/')
# print('checkpoint:',checkpoint)
log('tacotron_train done!!')
if not GTA_state:
log('\n#############################################################\n'
)
log('Tacotron GTA Synthesis\n')
log('###########################################################\n')
input_path = tacotron_synthesize(args, hparams, checkpoint)
tf.reset_default_graph()
#Sleep 1/2 second to let previous graph close and avoid error messages while Wavenet is training
sleep(0.5)
GTA_state = 1
save_seq(state_file, [taco_state, GTA_state, wave_state], input_path)
else:
input_path = os.path.join(log_dir, 'tacotron_' + args.output_dir,
'gta', 'map.txt')
# input_path = './tacotron_output/gta/map.txt'
log('Tacotron GTA Synthesis done')
def train(args, log_dir, hparams):
log('\n#############################################################\n')
log('Tacotron Train\n')
log('###########################################################\n')
tacotron_train(args, log_dir, hparams)
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir', default='')
parser.add_argument(
'--hparams',
default='',
help=
'Hyperparameter overrides as a comma-separated list of name=value pairs'
)
parser.add_argument('--tacotron_input', default='training_data/train.txt')
parser.add_argument('--wavenet_input',
default='tacotron_output/gta/map.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='Tacotron')
parser.add_argument('--input_dir',
default='training_data',
help='folder to contain inputs sentences/targets')
parser.add_argument('--output_dir',
default='output',
help='folder to contain synthesized mel spectrograms')
parser.add_argument('--mode',
default='synthesis',
help='mode for synthesis of tacotron after training')
parser.add_argument(
'--GTA',
default='True',
help=
'Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode'
)
parser.add_argument('--restore',
type=bool,
default=True,
help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval',
type=int,
default=250,
help='Steps between running summary ops')
parser.add_argument('--checkpoint_interval',
type=int,
default=1000,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval',
type=int,
default=1000,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps',
type=int,
default=150000,
help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps',
type=int,
default=1300000,
help='total number of wavenet training steps')
parser.add_argument('--tf_log_level',
type=int,
default=1,
help='Tensorflow C++ log level.')
parser.add_argument('--slack_url',
default=None,
help='slack webhook notification destination link')
args = parser.parse_args()
log_dir, hparams = prepare_run(args)
tacotron_train(args, log_dir, hparams)
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir',
default='dataset/MultiSpeaker',
required=True)
parser.add_argument('--hparams',
default='',
help='Hyperparameter overrides from a json file')
parser.add_argument('--tacotron_input', default='training_data/train.txt')
parser.add_argument('--wavenet_input',
default='tacotron_output/gta/map.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='MultiSpeaker', required=True)
parser.add_argument('--input_dir',
default='training_data',
help='folder to contain inputs sentences/targets')
parser.add_argument('--output_dir',
default='output/',
help='folder to contain synthesized mel spectrograms')
parser.add_argument('--mode',
default='synthesis',
help='mode for synthesis of tacotron after training')
parser.add_argument(
'--GTA',
default='True',
help=
'Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode'
)
parser.add_argument('--restore',
type=bool,
default=True,
help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval',
type=int,
default=200,
help='Steps between running summary ops')
parser.add_argument('--checkpoint_interval',
type=int,
default=500,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval',
type=int,
default=200,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps',
type=int,
default=500000,
help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps',
type=int,
default=360000,
help='total number of wavenet training steps')
parser.add_argument('--tf_log_level',
type=int,
default=2,
help='Tensorflow C++ log level.')
args = parser.parse_args()
accepted_models = [
'Tacotron', 'WaveNet', 'Both', 'Tacotron-2', 'MultiSpeaker'
]
if args.model not in accepted_models:
raise ValueError(
'please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveNet':
wavenet_train(args, log_dir, hparams, args.wavenet_input)
elif args.model == 'MultiSpeaker':
multispeaker_train(args, log_dir, hparams)
elif args.model in ('Both', 'Tacotron-2'):
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(
args.model, accepted_models))
parser.add_argument('--checkpoint_interval', type=int, default=500,
>>>>>>> f33090dba9ba4bc52db8367abdc48841d13c48f8
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval', type=int, default=1000,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps', type=int, default=1000000, help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps', type=int, default=750000, help='total number of wavenet training steps')
parser.add_argument('--tf_log_level', type=int, default=1, help='Tensorflow C++ log level.')
parser.add_argument('--slack_url', default=None, help='slack webhook notification destination link')
args = parser.parse_args()
accepted_models = ['Tacotron', 'WaveNet', 'Tacotron-2']
if args.model not in accepted_models:
raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveNet':
wavenet_train(args, log_dir, hparams, args.wavenet_input)
elif args.model == 'Tacotron-2':
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(args.model, accepted_models))
if __name__ == '__main__':
main()
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--base_dir', default='')
parser.add_argument('--hparams', default='',
help='Hyperparameter overrides as a comma-separated list of name=value pairs')
parser.add_argument('--tacotron_input', default='../data/train_emt4_jessa.txt')
parser.add_argument('--wavenet_input', default='tacotron_output/gta/map.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='Tacotron-2')
parser.add_argument('--input_dir', default='../data', help='folder to contain inputs sentences/targets')
parser.add_argument('--output_dir', default='output', help='folder to contain synthesized mel spectrograms')
parser.add_argument('--mode', default='synthesis', help='mode for synthesis of tacotron after training')
parser.add_argument('--GTA', default='True', help='Ground truth aligned synthesis, defaults to True, only considered in Tacotron synthesis mode')
parser.add_argument('--restore', action='store_true', default=False, help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval', type=int, default=1000000,
help='Steps between running summary ops')
parser.add_argument('--embedding_interval', type=int, default=1000000,
help='Steps between updating embeddings projection visualization')
parser.add_argument('--checkpoint_interval', type=int, default=250,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval', type=int, default=20,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps', type=int, default=1000000, help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps', type=int, default=500000, help='total number of wavenet training steps')
parser.add_argument('--tf_log_level', type=int, default=1, help='Tensorflow C++ log level.')
parser.add_argument('--slack_url', default=None, help='slack webhook notification destination link')
parser.add_argument('--emt_disc', action='store_true', default=False, help='whether to use emotion discriminator as part of loss')
parser.add_argument('--spk_disc', action='store_true', default=False, help='whether to use speaker discriminator as part of loss')
parser.add_argument('--intercross', action='store_true', default=True, help='whether to use intercross training')
parser.add_argument('--synth_style_type', default=None, help='vary the emotion, speaker id, or neither')
parser.add_argument('--tacotron_test_steps', type=int, default=3, help='Num batches to process when running evaluation')
parser.add_argument('--remove_long_samps', action='store_true', default=False, help='Will remove out the longest samples from EMT4/VCTK')
parser.add_argument('--test_max_len', action='store_true', default=False,help='Will create batches with the longest samples first to test max batch size')
parser.add_argument('--unpaired', action='store_true', default=False,help='Will create batches with the longest samples first to test max batch size')
parser.add_argument('--TEST', action='store_true', default=False,help='Uses small groups of batches to make testing faster')
parser.add_argument('--TEST_INPUTS', action='store_true', default=False,help='Fixes all input data to be the same for testing')
parser.add_argument('--max_to_keep', type=int, default=50, help='how many checkpoints to save')
parser.add_argument('--recon_emb_loss', action='store_true', default=False, help='Adds loss for reconstructing embeddings')
parser.add_argument('--intercross_both', action='store_true', default=False, help='does intercross for emotion and spk for both datasets')
parser.add_argument('--intercross_spk_only', action='store_true', default=False,help='does intercross for emotion and spk for both datasets')
parser.add_argument('--unpaired_loss_derate', type=float, default=1, help='how much to derate the unpaired mel out emb disc loss')
parser.add_argument('--unpaired_emt_loss_derate', type=float, default=1,help='how much to derate the unpaired mel out emb disc loss')
parser.add_argument('--lock_ref_enc', action='store_true', default=False, help='does not allow retraining of reference encoders')
parser.add_argument('--lock_gst', action='store_true', default=False, help='does not allow retraining of global style tokens')
parser.add_argument('--nat_gan', action='store_true', default=False, help='whether to use the naturalness gan')
parser.add_argument('--restart_nat_gan_d', action='store_true', default=False, help='whether to use the naturalness gan')
parser.add_argument('--nat_gan_derate', type=float, default=.1, help='how much to derate the unpaired mel out emb disc loss')
parser.add_argument('--restore_nat_gan_d_sep', action='store_true', default=False,help='whether to use the naturalness gan')
parser.add_argument('--save_output_vars', action='store_true', default=False, help='saves csvs of output vars')
parser.add_argument('--opt_ref_no_mo', action='store_true', default=False, help='dont train encoders based on synthesized samples style embeddings')
parser.add_argument('--restart_optimizer_r', action='store_true', default=False, help='retrains the reference encoder optimizer')
parser.add_argument('--pretrained_emb_disc', action='store_true', default=False, help='whether to use pretrained emt disc')
parser.add_argument('--pretrained_emb_disc_all', action='store_true', default=False,help='use pretrained emb disc on references and unpaired')
parser.add_argument('--no_general', action='store_true', default=False, help='mel output loss is not being classified as general')
parser.add_argument('--restore_std', action='store_true', default=False,help='allows the restoring of a model without optimzer_r to a new model with optimizer_r')
parser.add_argument('--emt_attn', action='store_true', default=False,help='allows the restoring of a model without optimzer_r to a new model with optimizer_r')
parser.add_argument('--emt_ref_gru', default='none', help='whether to use the a gru at the end of the reference embedding cnn')
parser.add_argument('--emt_only', action='store_true', default=False,help='does only one condition - emotion')
parser.add_argument('--attn', default=None, help='what type of attention to use')
parser.add_argument('--up_ref_match_p', action='store_true', default=False, help='feeds in the same references as paired for unpaired')
parser.add_argument('--tfr_up_only', action='store_true', default=False,help='feeds in the same references as paired for unpaired')
parser.add_argument('--no_mo_style_loss', action='store_true', default=False,help='feeds in the same references as paired for unpaired')
parser.add_argument('--l2_spk_emb', action='store_true', default=False,help='feeds in the same references as paired for unpaired')
parser.add_argument('--flip_spk_emt', action='store_true', default=False,help='pass in emt as spk ref and vice versa - used for testing reversing the attention')
parser.add_argument('--adain', action='store_true', default=False,help='use adaptive image normalization on references')
parser.add_argument('--synth_constraint', action='store_true', default=False, help='use adaptive image normalization on references')
parser.add_argument('--adv_emb_disc', action='store_true', default=False,help='use adversarial training on style embeddings')
args = parser.parse_args()
accepted_models = ['Tacotron', 'WaveNet', 'Tacotron-2']
if args.model not in accepted_models:
raise ValueError('please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
synth_metadata_filename = r"synth_emt4_jessa.txt"
args.synth_metadata_filename = os.path.join(r"../data", synth_metadata_filename)
import socket
if socket.gethostname() in ['A3907623','MININT-39T168F']:
hparams.tacotron_num_gpus = 1
hparams.tacotron_batch_size = 32
elif socket.gethostname() == 'area51.cs.washington.edu':
hparams.tacotron_num_gpus = 2
hparams.tacotron_batch_size = 64
os.environ["CUDA_VISIBLE_DEVICES"]="1,2"
args.input_dir = '/data/tts_emotion'
args.synth_metadata_filename = os.path.join(r"/data/tts_emotion",synth_metadata_filename)
print("over-riding input directory for running on Area51")
if hparams.tacotron_fine_tuning and not(args.restore):
raise ValueError('fine_tuning set to true but not restoring the model!')
#need to use the intercross both method with the zo/jessa datasets
if args.tacotron_input.endswith('jessa.txt'):
assert(args.intercross_both)
if args.emt_attn and args.attn==None:
raise ValueError("can't have emotion attention and no attention type")
if args.flip_spk_emt:
assert(not (args.unpaired))
if args.model == 'Tacotron':
tacotron_train(args, log_dir, hparams)
elif args.model == 'WaveNet':
wavenet_train(args, log_dir, hparams, args.wavenet_input)
elif args.model == 'Tacotron-2':
train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(args.model, accepted_models))
def main():
parser = argparse.ArgumentParser()
parser.add_argument(
'--hparams',
default='',
help=
'Hyperparameter overrides as a comma-separated list of name=value pairs'
)
parser.add_argument('--output-model-path',
dest='output_model_path',
required=True,
type=str,
default=os.path.dirname(os.path.realpath(__file__)),
help='Philly model output path.')
parser.add_argument('--log-dir',
dest='log_dir',
required=True,
type=str,
default=os.path.dirname(os.path.realpath(__file__)),
help='Philly log dir.')
parser.add_argument('--tacotron_input',
default='clean100/training_data/train.txt')
parser.add_argument('--name', help='Name of logging directory.')
parser.add_argument('--model', default='Tacotron')
parser.add_argument('--restore',
type=bool,
default=True,
help='Set this to False to do a fresh training')
parser.add_argument('--summary_interval',
type=int,
default=100,
help='Steps between running summary ops')
parser.add_argument(
'--embedding_interval',
type=int,
default=10000,
help='Steps between updating embeddings projection visualization')
parser.add_argument('--checkpoint_interval',
type=int,
default=5000,
help='Steps between writing checkpoints')
parser.add_argument('--eval_interval',
type=int,
default=1000,
help='Steps between eval on test data')
parser.add_argument('--tacotron_train_steps',
type=int,
default=150000,
help='total number of tacotron training steps')
parser.add_argument('--wavenet_train_steps',
type=int,
default=750000,
help='total number of wavenet training steps')
parser.add_argument('--tf_log_level',
type=int,
default=1,
help='Tensorflow C++ log level.')
parser.add_argument('--slack_url',
default=None,
help='slack webhook notification destination link')
args = parser.parse_args()
accepted_models = ['Tacotron']
if args.model not in accepted_models:
raise ValueError(
'please enter a valid model to train: {}'.format(accepted_models))
log_dir, hparams = prepare_run(args)
if args.model == 'Tacotron':
hparams.tacotron_batch_size = 16
tacotron_train(args, log_dir, hparams)
else:
raise ValueError('Model provided {} unknown! {}'.format(
args.model, accepted_models))
