out_data[:, col] /= max_values[col] - min_values[col]
np.place(out_data[:, col], max_mask, 1.0)
np.place(out_data[:, col], min_mask, 0.0)
return out_data
if __name__ == "__main__":
epochs = 5000
batch_size = 128
input_size = 34
latent_size = 8
model = DAE(input_size, latent_size)
model.to('cuda')
torch.backends.cudnn.benchmark = True
loss_fn = nn.MSELoss()
optimizer = torch.optim.SGD(model.parameters(), lr=0.1, momentum=0.9)
DECAY = 0.95
scheduler = LambdaLR(optimizer, lr_lambda=lambda t: DECAY**t)
data = pd.read_csv("model/training_test_data.csv")
data.sample(frac=1, random_state=200)
data = data.to_numpy()
size = data.shape[0]
training_data = data[:int(0.7 * size)]
validation_data = data[int(0.7 * size):int(0.9 * size)]
test_data = data[int(0.9 * size):]
type=str,
help="Where to save raw acoustic output")
parser = add_decoder_args(parser)
parser.add_argument('--save-output',
action="store_true",
help="Saves output of model from test")
args = parser.parse_args()
if __name__ == '__main__':
torch.set_grad_enabled(False)
device = torch.device("cuda" if args.cuda else "cpu")
model = load_model(device, args.model_path, args.cuda)
denoiser = DAE()
denoiser.load_state_dict(
torch.load('./models/denoiser_deepspeech_final.pth'))
denoiser = denoiser.to(device)
denoiser.eval()
if args.decoder == "beam":
from decoder import BeamCTCDecoder
decoder = BeamCTCDecoder(model.labels,
lm_path=args.lm_path,
alpha=args.alpha,
beta=args.beta,
cutoff_top_n=args.cutoff_top_n,
cutoff_prob=args.cutoff_prob,
beam_width=args.beam_width,
num_processes=args.lm_workers)
elif args.decoder == "greedy":
decoder = GreedyDecoder(model.labels,