import os
import sys
import time
import logging
import numpy as np
sys.path.append('../')
logging.getLogger('tensorflow').disabled = True
import tensorflow as tf
from utils import checkmate as cm
from utils import data_helpers as dh
from utils import param_parser as parser
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score
args = parser.parameter_parser()
MODEL = dh.get_model_name()
logger = dh.logger_fn("tflog", "logs/Test-{0}.log".format(time.asctime()))
CPT_DIR = 'runs/' + MODEL + '/checkpoints/'
BEST_CPT_DIR = 'runs/' + MODEL + '/bestcheckpoints/'
SAVE_DIR = 'output/' + MODEL
def create_input_data(data: dict):
return zip(data['f_pad_seqs'], data['b_pad_seqs'], data['onehot_labels'])
def test_cnn():
"""Test CNN model."""
# Print parameters used for the model
# Load word2vec model
print("Loading data...")
word2idx, embedding_matrix = dh.load_word2vec_matrix(args.word2vec_file)
# Load sentences, labels, and training parameters
print("Data processing...")
train_data = TextData(args, args.train_file, word2idx, embedding_matrix)
test_data = TextData(args, args.test_file, word2idx, embedding_matrix)
train_loader = torch.utils.data.DataLoader(train_data,
args.batch_size,
shuffle=True,
num_workers=1)
test_loader = torch.utils.data.DataLoader(test_data,
args.batch_size,
shuffle=False,
num_workers=1)
model = CNN(args).to(device)
#print(model)
for epoch in range(1, args.epochs + 1):
train(args, model, train_loader, device, epoch)
test(model, device, test_loader)
#torch.save(model.state_dict(), "../data/TextCNN.pt")
if __name__ == '__main__':
args = parser.parameter_parser() # add parser by using argparse module
train_CNN(args)
x = print("Press any key to continue...")
