if args.embedding == 'random':
embedding = 'random'
else:
embedding = args.embedding
model_name = args.model
print(model_name)
x = import_module('models.' + model_name)
config = Config(dataset, outputdir, embedding)
# reset config
config.model_name = args.model
config.save_path = os.path.join(outputdir, args.model + '.ckpt')
config.log_path = os.path.join(outputdir, args.model + '.log')
config.dropout = float(args.dropout)
config.require_improvement = int(args.require_improvement)
config.num_epochs = int(args.num_epochs)
config.batch_size = int(args.batch_size)
config.max_length = int(args.max_length)
config.learning_rate = float(args.learning_rate)
config.embed = int(args.embed_dim)
config.bucket = int(args.bucket)
config.wordNgrams = int(args.wordNgrams)
config.lr_decay_rate = float(args.lr_decay_rate)
start_time = time.time()
print("Loading data...")
vocab, train_data, dev_data, test_data = build_dataset(
config, args.use_word_segment, min_freq=int(args.min_freq))
time_dif = get_time_dif(start_time)