test_matching.py•2.11 kB
"""
test mathcing
"""
import sys
sys.path.append('../..')
import unittest
import os
import appbuilder
class TestMatching(unittest.TestCase):
def test_example(self):
# 初始化所需要的组件
embedding = appbuilder.Embedding()
matching = appbuilder.Matching(embedding)
# 定义输入query和文本列表
query = appbuilder.Message("你好")
contexts = appbuilder.Message(["世界", "你好"])
# 根据query，对文本列表做相似度排序
contexts_matched = matching(query, contexts)
self.assertListEqual(contexts_matched.content, ['你好', '世界'])
def test_run(self):
embedding = appbuilder.Embedding()
matching = appbuilder.Matching(embedding)
query = appbuilder.Message("你好")
contexts = appbuilder.Message(["世界", "你好"])
contexts_matched = matching(query, contexts)
self.assertListEqual(
contexts_matched.content,
['你好', '世界']
)
def test_return_score(self):
embedding = appbuilder.Embedding()
matching = appbuilder.Matching(embedding)
query = appbuilder.Message("你好")
contexts = appbuilder.Message(["世界", "你好"])
contexts_matched = matching(query, contexts, return_score=True)
scores = [i[0] for i in contexts_matched.content]
contexts = [i[1] for i in contexts_matched.content]
self.assertListEqual(
contexts,
['你好', '世界']
)
self.assertGreater(scores[0], scores[1])
def test_semantics(self):
embedding = appbuilder.Embedding()
matching = appbuilder.Matching(embedding)
query = appbuilder.Message("你好")
query_embedding = embedding(query)
contexts = appbuilder.Message(["世界", "你好"])
context_embedding = embedding.batch(contexts)
semantics = matching.semantics(query_embedding, context_embedding)
self.assertEqual(len(semantics.content), 2)
if __name__ == '__main__':
unittest.main()