class TestIhrmLogin(unittest.TestCase):
def setUp(self):
self.login_api = TestLoginApi()
def tearDown(self):
pass
@parameterized.expand(test_data('./data/test_login.json'))
def test01_login_success(self, data, httpcode, success, code, message):
jsonData = data
response = self.login_api.login(jsonData)
logging.info('登录的结果为:', response.json())
assesr_common(httpcode, success, code, message, response, self)
'''
def main (verbose = False):
path_train = '../../data/restaurants_train.tsv'
path_test = '../../data/restaurants_gold.tsv'
prompt = 'please input path to SemEval 2014 CoNLL formatted training data: '
input_path_train = input(prompt)
if input_path_train == '':
pass
else:
path_train = input_path_train
prompt = 'please input path to SemEval 2014 CoNLL formatted test data: '
input_path_test = input(prompt)
if input_path_test == '':
pass
else:
path_test = input_path_test
df_train, df_test = load_test_train(path_train, path_test)
embed_prompt = "please enter path to 300-dimensional Word2vec Google News vectors: "
w2v_path = input (embed_prompt)
xu_embed_prompt = "please enter path to Xu et al.'s 100-d restaurant domain vectors: "
xuv_path = input (xu_embed_prompt)
print("Thank you. Please wait while embeddings load")
#load embeddings
w2v, xuv = load_embeddings(w2v_path, xuv_path)
X_train, X_valid, y_train, y_valid = training_data (w2v, xuv, df_train['word'], df_train['label'])
X_test, label_index = test_data (w2v, xuv, df_test['word'], df_test['label'])
# input("Ready to run?")
# run DE_CNN NN
batch_size = 128
embedding_dims = X_train.shape[2]
kernel_size = 5
epochs = 200
model = run_cnn(X_train, y_train, X_valid, y_valid, embedding_dims = embedding_dims,
batch_size = batch_size, kernel_size = kernel_size, epochs = 200,
verbose = verbose)
#predict
predict(model, X_test, df_test, label_index)
module_name='partition_thrift')
return make_client(partition_thrift.Partition, '127.0.0.1', 6000)
def file_info(filename):
with open(filename, 'rb') as file:
file_content = file.read()
return {filename: file_content}
if __name__ == '__main__':
# get time threshold
threshold = float(input('Please input latency threshold: '))
# get test data
dataiter = test_data()
images, labels = dataiter.next()
start = time.time()
# get partition point and exit point
ep, pp = Optimize(threshold)
print('Branch is %d, and partition point is %d' % (ep, pp))
# infer left part
out = infer(CLIENT, ep, pp, images)
print('Left part of model inference complete.')
# save intermediate for RPC process
intermediate = out.detach().numpy()
import os
from gensim.models import word2vec
import numpy as np
from utils import train_data, test_data
DIM = 256
if __name__ == "__main__":
print("Loading train data ...", flush=True)
train_x1, _ = train_data("data/training_label.txt", True)
train_x0 = train_data("data/training_nolabel.txt", False)
print("Loading test data ...", flush=True)
test_x = test_data("data/testing_data.txt")
#print(np.percentile([len(i) for i in train_x1], 75))
#print(np.percentile([len(i) for i in train_x0], 75))
#print(np.percentile([len(i) for i in test_x], 75))
#exit()
print("Word2Vec ...", flush=True)
model = word2vec.Word2Vec(train_x1 + train_x0 + test_x, size=256, window=5, min_count=5, workers=12, iter=10, sg=1)
print("Saving model ...", flush=True)
model.save("w2v_model/w2v.model")
labels=tf.ones_like(real_predict)))
dis_loss = fake_loss + real_loss
var_list = tf.trainable_variables()
g_var_list = [x for x in var_list if 'g_' in x.name]
d_var_list = [x for x in var_list if 'd_' in x.name]
with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
g_opt = tf.train.AdamOptimizer(0.0002, 0.5, 0.9)
d_opt = tf.train.AdamOptimizer(0.0002, 0.5, 0.9)
optim_gen = g_opt.compute_gradients(gen_loss, var_list=g_var_list)
optim_g = g_opt.apply_gradients(optim_gen)
optim_dis = d_opt.compute_gradients(dis_loss, var_list=d_var_list)
optim_d = d_opt.apply_gradients(optim_dis)
# the optim operation
sample_gen = generator(x_place, reuse=True)
#test files
#restore from the checkpoint dir
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
saver = tf.train.Saver()
# print('22')
ckpt = tf.train.get_checkpoint_state(check_dir)
# restore from the check point
saver.restore(sess, ckpt.model_checkpoint_path)
for i in range(batch_idx):
_, x = test_data(i)
feed_dict = {x_place: x}
sample_imgs = sess.run(sample_gen, feed_dict=feed_dict)
save_batch_imgs(sample_imgs, i, fig_dir)