def test_different_qa(self): ref = "this is a string" self.assertDictClose( metrics.qa([[ref, ref], [ref, ref]], ["", ""]), { "em": 0, "f1": 0 })
def test_qa_bytes(self): ref = b"this is a string" self.assertDictClose( metrics.qa([["", ref], [ref, ref]], [ref, ref]), { "em": 100, "f1": 100 })
def test_same_qa(self): ref = "this is a string" self.assertDictClose( metrics.qa([["", ref], [ref, ref]], [ref, ref]), { "em": 100, "f1": 100, })
def test_qa_small(self): self.assertDictClose( metrics.qa([["abc abd", "$$$$"]], ["abd"]), { "f1": 100 * 2.0 / 3.0, "em": 0. }, )
def test_qa_big(self): self.assertDictClose( metrics.qa( [ ["big moose", "hippo"], ["correct1"], ["correct2.1", "correct2.2"], ["a", "b"], ], [ "a big Moose!", "wrong", "correct2.2", "c", ], ), {"em": 50., "f1": 50.}, )