elif args.model == 'lcnn29_mfcc':
model = model_.lcnn_29layers_v2(n_z=args.latent_size,
proj_size=len(train_dataset.speakers_list)
if args.softmax != 'none' else 0,
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'TDNN':
model = model_.TDNN(n_z=args.latent_size,
proj_size=len(train_dataset.speakers_list)
if args.softmax != 'none' else 0,
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(n_z=args.latent_size,
proj_size=len(train_dataset.speakers_list)
if args.softmax != 'none' else 0,
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'FTDNN':
model = model_.FTDNN(n_z=args.latent_size,
proj_size=len(train_dataset.speakers_list)
if args.softmax != 'none' else 0,
ncoef=args.ncoef,
sm_type=args.softmax)
if args.pretrained_path is not None:
ckpt = torch.load(args.pretrained_path,
map_location=lambda storage, loc: storage)
try:
model.load_state_dict(ckpt['model_state'], strict=True)
if args.softmax != 'none' or args.pretrain else 0,
ncoef=args.ncoef,
sm_type=args.softmax,
delta=args.delta)
elif args.model == 'TDNN_mod':
model = model_.TDNN_mod(n_z=args.latent_size,
proj_size=train_dataset.n_speakers
if args.softmax != 'none' or args.pretrain else 0,
ncoef=args.ncoef,
sm_type=args.softmax,
delta=args.delta)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(
n_z=args.latent_size,
proj_size=train_dataset.n_speakers
if args.softmax != 'none' or args.pretrain else 0,
ncoef=args.ncoef,
sm_type=args.softmax,
delta=args.delta)
elif args.model == 'transformer':
model = model_.transformer_enc(
n_z=args.latent_size,
proj_size=train_dataset.n_speakers
if args.softmax != 'none' or args.pretrain else 0,
ncoef=args.ncoef,
sm_type=args.softmax,
delta=args.delta)
if args.cuda:
model = model.to(device)
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'lcnn29_mfcc':
model = model_.lcnn_29layers_v2(n_z=args.latent_size,
proj_size=len(list(
labels_dict.keys())),
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'TDNN':
model = model_.TDNN(n_z=args.latent_size,
proj_size=len(list(labels_dict.keys())),
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(n_z=args.latent_size,
proj_size=len(list(labels_dict.keys())),
ncoef=args.ncoef,
sm_type=args.softmax)
elif args.model == 'FTDNN':
model = model_.FTDNN(n_z=args.latent_size,
proj_size=len(list(labels_dict.keys())),
ncoef=args.ncoef,
sm_type=args.softmax)
if args.cp_path_2 is not None:
model_2 = type(model)(n_z=args.latent_size,
proj_size=len(list(labels_dict.keys())),
ncoef=args.ncoef,
sm_type=args.softmax)
ckpt_2 = torch.load(args.cp_path_2,
map_location=lambda storage, loc: storage)
model_2.load_state_dict(ckpt_2['model_state'], strict=True)
ncoef=args.ncoef)
elif args.model == 'lcnn9_mfcc':
model = model_.lcnn_9layers(n_z=args.latent_size,
proj_size=None,
ncoef=args.ncoef)
elif args.model == 'lcnn29_mfcc':
model = model_.lcnn_29layers_v2(n_z=args.latent_size,
proj_size=None,
ncoef=args.ncoef)
elif args.model == 'TDNN':
model = model_.TDNN(n_z=args.latent_size,
proj_size=None,
ncoef=args.ncoef)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(n_z=args.latent_size,
proj_size=None,
ncoef=args.ncoef)
elif args.model == 'FTDNN':
model = model_.FTDNN(n_z=args.latent_size,
proj_size=None,
ncoef=args.ncoef)
ckpt = torch.load(args.cp_path, map_location=lambda storage, loc: storage)
model.load_state_dict(ckpt['model_state'], strict=False)
model.eval()
if args.cuda:
model = model.cuda()
scp_list = glob.glob(args.path_to_data + '*.scp')
proj_size=0,
ncoef=args.ncoef,
delta=args.delta)
elif args.model == 'TDNN_aspp':
model = model_.TDNN_aspp(n_z=args.latent_size,
proj_size=0,
ncoef=args.ncoef,
delta=args.delta)
elif args.model == 'TDNN_mod':
model = model_.TDNN_mod(n_z=args.latent_size,
proj_size=0,
ncoef=args.ncoef,
delta=args.delta)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(n_z=args.latent_size,
proj_size=0,
ncoef=args.ncoef,
delta=args.delta)
elif args.model == 'transformer':
model = model_.transformer_enc(n_z=args.latent_size,
proj_size=0,
ncoef=args.ncoef,
delta=args.delta)
ckpt = torch.load(args.cp_path, map_location=lambda storage, loc: storage)
model.load_state_dict(ckpt['model_state'], strict=False)
model.eval()
if args.cuda:
model = model.to(device)
def train(lr, l2, max_gnorm, momentum, margin, lambda_, swap, latent_size, n_frames, model, ncoef, epochs, batch_size, valid_batch_size, n_workers, cuda, train_hdf_file, valid_hdf_file, cp_path, softmax, delta, logdir):
if cuda:
device=get_freer_gpu()
if args.model == 'resnet_qrnn':
import cupy
cupy.cuda.Device(int(str(device).split(':')[-1])).use()
cp_name = get_file_name(cp_path)
if args.logdir:
from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter(log_dir=logdir+cp_name, comment=args.model, purge_step=True)
else:
writer = None
train_dataset = Loader(hdf5_name = train_hdf_file, max_nb_frames = int(n_frames), delta = delta)
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=n.workers, worker_init_fn=set_np_randomseed)
valid_dataset = Loader_valid(hdf5_name = valid_hdf_file, max_nb_frames = int(n_frames), delta = delta)
valid_loader = torch.utils.data.DataLoader(valid_dataset, batch_size=valid_batch_size, shuffle=True, num_workers=n_workers, worker_init_fn=set_np_randomseed)
if model == 'resnet_mfcc':
model=model_.ResNet_mfcc(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=ncoef, sm_type=softmax, delta=delta)
elif model == 'resnet_34':
model=model_.ResNet_34(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=ncoef, sm_type=softmax, delta=delta)
elif model == 'resnet_lstm':
model=model_.ResNet_lstm(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=ncoef, sm_type=softmax, delta=delta)
elif model == 'resnet_qrnn':
model=model_.ResNet_qrnn(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=ncoef, sm_type=softmax, delta=delta)
elif model == 'resnet_stats':
model=model_.ResNet_stats(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=ncoef, sm_type=softmax, delta=delta)
elif args.model == 'resnet_large':
model = model_.ResNet_large(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'resnet_small':
model = model_.ResNet_small(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'resnet_2d':
model = model_.ResNet_2d(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN':
model = model_.TDNN(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_att':
model = model_.TDNN_att(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_multihead':
model = model_.TDNN_multihead(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_lstm':
model = model_.TDNN_lstm(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_aspp':
model = model_.TDNN_aspp(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_mod':
model = model_.TDNN_mod(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'TDNN_multipool':
model = model_.TDNN_multipool(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
elif args.model == 'transformer':
model = model_.transformer_enc(n_z=int(latent_size), proj_size=train_dataset.n_speakers, ncoef=args.ncoef, sm_type=softmax, delta=delta)
if cuda:
model=model.to(device)
else:
device=None
optimizer=optim.SGD(model.parameters(), lr=lr, momentum=momentum, weight_decay=l2)
trainer=TrainLoop(model, optimizer, train_loader, valid_loader, max_gnorm=max_gnorm, margin=margin, lambda_=lambda_, verbose=-1, device=device, cp_name=cp_name, save_cp=True, checkpoint_path=cp_path, swap=swap, softmax=True, pretrain=False, mining=True, cuda=cuda, logger=writer)
return trainer.train(n_epochs=epochs)
if args.model == 'TDNN_lstm' or args.model == 'all':
batch = torch.rand(3, 3 if args.delta else 1, args.ncoef, 200)
model = model_.TDNN_lstm(n_z=args.latent_size, ncoef=args.ncoef, delta=args.delta, proj_size=10, sm_type='softmax')
mu, emb = model.forward(batch)
out = model.out_proj(mu, torch.ones(mu.size(0)))
print('TDNN_lstm', mu.size(), emb.size(), out.size())
if args.model == 'TDNN_aspp' or args.model == 'all':
batch = torch.rand(3, 3 if args.delta else 1, args.ncoef, 200)
model = model_.TDNN_aspp(n_z=args.latent_size, ncoef=args.ncoef, delta=args.delta, proj_size=10, sm_type='softmax')
mu, emb = model.forward(batch)
out = model.out_proj(mu, torch.ones(mu.size(0)))
print('TDNN_aspp', mu.size(), emb.size(), out.size())
if args.model == 'TDNN_mod' or args.model == 'all':
batch = torch.rand(3, 3 if args.delta else 1, args.ncoef, 200)
model = model_.TDNN_mod(n_z=args.latent_size, ncoef=args.ncoef, delta=args.delta, proj_size=10, sm_type='softmax')
mu, emb = model.forward(batch)
out = model.out_proj(mu, torch.ones(mu.size(0)))
print('TDNN_mod', mu.size(), emb.size(), out.size())
if args.model == 'TDNN_multipool' or args.model == 'all':
batch = torch.rand(3, 3 if args.delta else 1, args.ncoef, 200)
model = model_.TDNN_multipool(n_z=args.latent_size, ncoef=args.ncoef, delta=args.delta, proj_size=10, sm_type='softmax')
mu, emb = model.forward(batch)
out = model.out_proj(mu, torch.ones(mu.size(0)))
print('TDNN_multipool', mu.size(), emb.size(), out.size())
if args.model == 'transformer' or args.model == 'all':
batch = torch.rand(3, 3 if args.delta else 1, args.ncoef, 200)
model = model_.transformer_enc(n_z=args.latent_size, ncoef=args.ncoef, delta=args.delta, proj_size=10, sm_type='softmax')
mu, emb = model.forward(batch)
out = model.out_proj(mu, torch.ones(mu.size(0)))
print('transformer', mu.size(), emb.size(), out.size())
