return
#=======================================
##########################################################################
train_samples = mm.load_train_data(params['TRAIN_FILE'])
char_dict, ord_dict = mm.build_vocab(train_samples)
NUM_EMB = len(char_dict)
DATA_LENGTH = max(map(len, train_samples))
MAX_LENGTH = params["MAX_LENGTH"]
to_use = [
sample for sample in train_samples
if mm.verified_and_below(sample, MAX_LENGTH)
]
positive_samples = [
mm.encode(sample, MAX_LENGTH, char_dict) for sample in to_use
]
POSITIVE_NUM = len(positive_samples)
print('Starting ObjectiveGAN for {:7s}'.format(PREFIX))
print('Data points in train_file {:7d}'.format(len(train_samples)))
print('Max data length is {:7d}'.format(DATA_LENGTH))
print('Max length to use is {:7d}'.format(MAX_LENGTH))
print('Avg length to use is {:7f}'.format(
np.mean([len(s) for s in to_use])))
print('Num valid data points is {:7d}'.format(POSITIVE_NUM))
print('Size of alphabet is {:7d}'.format(NUM_EMB))
mm.print_params(params)
##########################################################################
mean_r, std_r = np.mean(rewards), np.std(rewards)
min_r, max_r = np.min(rewards), np.max(rewards)
print('Mean: {:.3f} , Std: {:.3f}'.format(mean_r, std_r),end='')
print(', Min: {:.3f} , Max: {:.3f}\n'.format(min_r, max_r))
np.set_printoptions(precision=8, suppress=False)
return
#=======================================
##########################################################################
train_samples = mm.load_train_data(params['TRAIN_FILE'])
char_dict, ord_dict = mm.build_vocab(train_samples)
NUM_EMB = len(char_dict)
DATA_LENGTH = max(map(len, train_samples))
MAX_LENGTH = params["MAX_LENGTH"]
to_use = [sample for sample in train_samples if mm.verified_and_below(
sample, MAX_LENGTH)]
positive_samples = [mm.encode(sample, MAX_LENGTH, char_dict) for sample in to_use]
POSITIVE_NUM = len(positive_samples)
print('Starting ObjectiveGAN for {:7s}'.format(PREFIX))
print('Data points in train_file {:7d}'.format(len(train_samples)))
print('Max data length is {:7d}'.format(DATA_LENGTH))
print('Max length to use is {:7d}'.format(MAX_LENGTH))
print('Avg length to use is {:7f}'.format(
np.mean([len(s) for s in to_use])))
print('Num valid data points is {:7d}'.format(POSITIVE_NUM))
print('Size of alphabet is {:7d}'.format(NUM_EMB))
mm.print_params(params)
##########################################################################
