def main():
random_state = 42
np.random.seed(random_state)
data_dir_path = './data'
model_dir_path = './models'
report_dir_path = './reports'
url_data = load_url_data(data_dir_path)
text_model = extract_text_model(url_data['text'])
batch_size = 64
epochs = 50
classifier = CnnLstmPredictor()
history = classifier.fit(text_model=text_model,
model_dir_path=model_dir_path,
url_data=url_data,
batch_size=batch_size,
epochs=epochs)
plot_and_save_history(
history, CnnLstmPredictor.model_name,
report_dir_path + '/' + CnnLstmPredictor.model_name + '-history.png')
def main():
data_dir_path = './data'
model_dir_path = './models'
predictor = LstmEmbedPredictor()
predictor.load_model(model_dir_path)
url_data = load_url_data(data_dir_path)
count = 0
for url, label in zip(url_data['text'], url_data['label']):
predicted_label = predictor.predict(url)
print('predicted: ' + str(predicted_label) + ' actual: ' + str(label))
count += 1
if count > 20:
break
