#!/usr/local/bin/python
from Model import Model
from Executor import Executor
if __name__ == '__main__':
model = Model()
model.assemble_graph()
silence_step = 0
skip_step = 20
exe = Executor(model, silence_step=silence_step, skip_step=skip_step)
exe.train_and_dev()
exe.restore_and_test()
if __name__ == '__main__':
seed = 12345
random.seed(seed)
np.random.seed(seed)
torch.manual_seed(seed)
n_gpu = torch.cuda.device_count()
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
if n_gpu > 0:
torch.cuda.manual_seed(seed)
silence_step = 0 # 从哪步开始可以进行evaluate
skip_step = 20 # 每隔多少步做evaluate
config = DicToObj(**config_model)
formatter = '%(asctime)s %(levelname)s %(message)s'
config = generate_name_path(config, path_parser)
train_logger = Logger(filename=config.saved_path + '/train.log',
fmt=formatter).logger
tb_logger = Tensorboard_Logger(config.saved_path)
exe = Executor(model_name="stocknet",
config=config,
silence_step=silence_step,
skip_step=skip_step,
train_logger=train_logger,
tb_logger=tb_logger)
exe.apply(xavier_init)
exe.train_and_dev(do_continue=False)
exe.restore_and_test()