import utils.gen_utils as utils
(
inp_dir,
dataset,
lr,
batch_size,
epochs,
log_expdata,
embed,
layer,
mode,
embed_mode,
jobid,
) = utils.parse_args()
# embed_mode {mean, cls}
# mode {512_head, 512_tail, 256_head_tail}
network = "LR"
print("{} : {} : {} : {} : {}".format(dataset, embed, layer, mode, embed_mode))
n_classes = 2
seed = jobid
np.random.seed(seed)
tf.random.set_seed(seed)
start = time.time()
path = "explogs/"
def merge_features(embedding, other_features):
df = pd.merge(embedding, other_features, left_index=True, right_index=True)
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import tensorflow as tf
import numpy as np
import pickle
import time
import joblib
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import KFold
import utils.gen_utils as utils
inp_dir, dataset_type, network, lr, batch_size, epochs, seed, write_file, embed, layer, mode, embed_mode = utils.parse_args(
)
n_classes = 2
np.random.seed(seed)
tf.compat.v1.set_random_seed(seed)
start = time.time()
def classification(X_train, X_test, y_train, y_test, file_name):
model_name = 'LR-models/' + file_name + '-mean.joblib'
classifier = LogisticRegression(random_state=0)
classifier.fit(X_train, y_train)
# joblib.dump(classifier, model_name)
acc = classifier.score(X_test, y_test)
return acc
dataset, lr, batch_size, epochs, MODEL_INPUT, embed, layer, mode, embed_mode, jobid
pd.set_option('display.max_columns', None)
print(df.head(5))
# save the results of our experiment
if (log_expdata):
Path(path).mkdir(parents=True, exist_ok=True)
if (not os.path.exists(path + 'expdata.csv')):
df.to_csv(path + 'expdata.csv', mode='a', header=True)
else:
df.to_csv(path + 'expdata.csv', mode='a', header=False)
if __name__ == "__main__":
inp_dir, dataset, lr, batch_size, epochs, log_expdata, embed, layer, mode, embed_mode, jobid = utils.parse_args()
print('{} : {} : {} : {} : {}'.format(dataset, embed, layer, mode, embed_mode))
n_classes = 2
features_dim = 123
MODEL_INPUT = 'combined_features'
path = 'explogs/'
network = 'SVM'
np.random.seed(jobid)
tf.random.set_seed(jobid)
nrc, nrc_vad, readability, mairesse = [True, True, True, True]
feature_flags = [nrc, nrc_vad, readability, mairesse]
start = time.time()
if (re.search(r'base', embed)):