def train(net: SiameseEMG):
gesture_list = list(range(8))
train_set = CapgTriplet(gesture_list,
sequence_len=20,
frame_x=False,
train=True)
net.dataset = train_set
net.fit_with_dataset()
return net
def main(train_args):
# 1. 设置好optimizer
# 2. 定义好model
args = {**train_args, **hyperparameters}
all_gestures = list(range(8))
model = SiameseLSTM(args['input_size'], args['hidden_size'],
len(all_gestures), args['layer'], args['dropout'])
name = args['name']
sub_folder = args['sub_folder']
# from emg.utils import config_tensorboard
# tensorboard_cb = config_tensorboard(name, sub_folder, model, (1, 10, 128))
#
# from emg.utils.lr_scheduler import DecayLR
# lr_callback = DecayLR(start_lr=args['lr'], gamma=0.5, step_size=args['lr_step'])
net = SiameseEMG(module=model,
model_name=name,
sub_folder=sub_folder,
hyperparamters=args,
optimizer=torch.optim.Adam,
gesture_list=[],
callbacks=[])
net = train(net)
def main(train_args, TEST_MODE=False):
args = train_args
all_gestures = list(range(20))
model = SiameseCNN(len(all_gestures))
name = args['name']
sub_folder = args['sub_folder']
# from emg.utils import config_tensorboard
# tensorboard_cb = config_tensorboard(name, sub_folder)
#
# from emg.utils.lr_scheduler import DecayLR
# lr_callback = DecayLR(start_lr=args['lr'], gamma=0.5, step_size=args['lr_step'])
net = SiameseEMG(module=model,
model_name=name,
sub_folder=sub_folder,
hyperparamters=args,
optimizer=torch.optim.Adam,
gesture_list=[],
callbacks=[])
net = train(net)