opt_kwargs = {'lr': args.lr, 'lr_decay': args.lr_decay, 'weight_decay': args.weight_decay, 'dampening': args.dampening,
'momentum': args.momentum}
l2_dist = nn.CosineSimilarity(dim=1, eps=1e-12) if args.cos_sim else nn.PairwiseDistance(p=2)
if args.acoustic_feature == 'fbank':
transform = transforms.Compose([
totensor()
])
transform_T = transforms.Compose([
concateinputfromMFB(num_frames=c.NUM_FRAMES_SPECT, input_per_file=args.test_input_per_file,
remove_vad=args.remove_vad),
])
transform_V = transforms.Compose([
ConcateVarInput(remove_vad=args.remove_vad),
# varLengthFeat(remove_vad=args.remove_vad),
# concateinputfromMFB(num_frames=c.NUM_FRAMES_SPECT, input_per_file=args.test_input_per_file,
# remove_vad=args.remove_vad),
])
else:
transform = transforms.Compose([
truncatedinput(),
toMFB(),
totensor(),
# tonormal()
])
file_loader = read_audio
if args.log_scale:
ConcateOrgInput(remove_vad=args.remove_vad),
])
transform_T = transforms.Compose([
ConcateOrgInput(remove_vad=args.remove_vad),
>>>>>>> Server/Server:TrainAndTest/test_egs.py
])
elif args.input_length == 'fix':
transform = transforms.Compose([
<<<<<<< HEAD:TrainAndTest/test_vox1.py
concateinputfromMFB(remove_vad=args.remove_vad),
])
transform_T = transforms.Compose([
concateinputfromMFB(input_per_file=args.test_input_per_file, remove_vad=args.remove_vad),
=======
ConcateVarInput(frame_shift=args.frame_shift, remove_vad=args.remove_vad),
])
transform_T = transforms.Compose([
ConcateVarInput(frame_shift=args.frame_shift, remove_vad=args.remove_vad),
>>>>>>> Server/Server:TrainAndTest/test_egs.py
])
else:
raise ValueError('input length must be var or fix.')
if args.mvnorm:
transform.transforms.append(mvnormal())
transform_T.transforms.append(mvnormal())
# pdb.set_trace()
if args.feat_format == 'kaldi':
file_loader = read_mat
opt_kwargs = {'lr': args.lr,
'lr_decay': args.lr_decay,
'weight_decay': args.weight_decay,
'dampening': args.dampening,
'momentum': args.momentum}
l2_dist = nn.CosineSimilarity(dim=1, eps=1e-6) if args.cos_sim else PairwiseDistance(2)
if args.acoustic_feature == 'fbank':
transform = transforms.Compose([
concateinputfromMFB(num_frames=c.NUM_FRAMES_SPECT, remove_vad=args.remove_vad),
# varLengthFeat(),
to2tensor()
])
transform_T = transforms.Compose([
ConcateVarInput(num_frames=c.NUM_FRAMES_SPECT, remove_vad=args.remove_vad),
# to2tensor()
])
transform_V = transforms.Compose([
varLengthFeat(remove_vad=args.remove_vad),
to2tensor()
])
else:
transform = transforms.Compose([
truncatedinput(),
toMFB(),
totensor(),
# tonormal()
])
file_loader = read_audio
np.random.seed(args.seed)
torch.manual_seed(args.seed)
if args.cuda:
cudnn.benchmark = True
# create logger
# Define visulaize SummaryWriter instance
kwargs = {'num_workers': 12, 'pin_memory': True} if args.cuda else {}
l2_dist = nn.CosineSimilarity(
dim=1, eps=1e-6) if args.cos_sim else PairwiseDistance(2)
if args.acoustic_feature == 'fbank':
transform = transforms.Compose([
ConcateVarInput(),
])
transform_T = transforms.Compose([
ConcateVarInput(),
])
file_loader = read_mat
else:
transform = transforms.Compose([
truncatedinput(),
toMFB(),
totensor(),
# tonormal()
])
file_loader = read_audio
# pdb.set_trace()
transform = transforms.Compose([
truncatedinput(),
toMFB(),
totensor(),
# tonormal()
])
if args.test_input == 'var':
transform_V = transforms.Compose([
ConcateOrgInput(remove_vad=args.remove_vad,
feat_type=args.feat_format),
])
elif args.test_input == 'fix':
transform_V = transforms.Compose([
ConcateVarInput(remove_vad=args.remove_vad,
num_frames=args.chunk_size,
frame_shift=args.chunk_size,
feat_type=args.feat_format),
])
if args.log_scale:
transform.transforms.append(tolog())
transform_V.transforms.append(tolog())
# pdb.set_trace()
if args.feat_format == 'kaldi':
file_loader = read_mat
elif args.feat_format == 'npy':
file_loader = np.load
elif args.feat_format == 'wav':
file_loader = load_mat