def main(args):
new_model = args.new_model
rnn = RNN()
if not new_model:
try:
rnn.set_weights(config.rnn_weight)
except:
print("Either set --new_model or ensure {} exists".format(
config.rnn_weight))
raise
rnn_input = []
rnn_output = []
for i in range(130):
# print('Building {}th...'.format(i))
input = np.load('./rnn_data/rnn_input_' + str(i) + '.npy')
output = np.load('./rnn_data/rnn_output_' + str(i) + '.npy')
# sequence pre-processing, for training LSTM the rnn_input must be (samples/episodes, time steps, features)
input = pad_sequences(input,
maxlen=40,
dtype='float32',
padding='post',
truncating='post')
output = pad_sequences(output,
maxlen=40,
dtype='float32',
padding='post',
truncating='post')
rnn_input.append(input)
rnn_output.append(output)
input = rnn_input[0]
output = rnn_output[0]
for i in range(len(rnn_input) - 1):
input = np.concatenate((input, rnn_input[i + 1]), axis=0)
output = np.concatenate((output, rnn_output[i + 1]), axis=0)
print(input.shape)
print(output.shape)
rnn.train(input, output)
rnn.plot_loss()
import matplotlib.pyplot as plt
import sys
sys.path.append("..")
import config
dir_name = config.dir_name
episode = config.episode
frame = config.frame
np.set_printoptions(precision=4, edgeitems=6, linewidth=100, suppress=True)
vae = VAE()
vae.set_weights(config.vae_weight)
rnn = RNN()
rnn.set_weights(config.rnn_weight)
rnn_input = np.load(dir_name + '/rnn_input_0.npy')
rnn_output = np.load(dir_name + '/rnn_output_0.npy')
rnn_input = pad_sequences(rnn_input,
maxlen=30,
dtype='float32',
padding='pre',
truncating='pre')
rnn_output = pad_sequences(rnn_output,
maxlen=30,
dtype='float32',
padding='pre',
truncating='pre')
# print(rnn_input.shape)
