Spaces:
No application file
No application file
File size: 7,565 Bytes
a85c9b8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 |
import numpy as np
import pytest
from embedchain.config.evaluation.base import AnswerRelevanceConfig
from embedchain.evaluation.metrics import AnswerRelevance
from embedchain.utils.evaluation import EvalData, EvalMetric
@pytest.fixture
def mock_data():
return [
EvalData(
contexts=[
"This is a test context 1.",
],
question="This is a test question 1.",
answer="This is a test answer 1.",
),
EvalData(
contexts=[
"This is a test context 2-1.",
"This is a test context 2-2.",
],
question="This is a test question 2.",
answer="This is a test answer 2.",
),
]
@pytest.fixture
def mock_answer_relevance_metric(monkeypatch):
monkeypatch.setenv("OPENAI_API_KEY", "test_api_key")
metric = AnswerRelevance()
return metric
def test_answer_relevance_init(monkeypatch):
monkeypatch.setenv("OPENAI_API_KEY", "test_api_key")
metric = AnswerRelevance()
assert metric.name == EvalMetric.ANSWER_RELEVANCY.value
assert metric.config.model == "gpt-4"
assert metric.config.embedder == "text-embedding-ada-002"
assert metric.config.api_key is None
assert metric.config.num_gen_questions == 1
monkeypatch.delenv("OPENAI_API_KEY")
def test_answer_relevance_init_with_config():
metric = AnswerRelevance(config=AnswerRelevanceConfig(api_key="test_api_key"))
assert metric.name == EvalMetric.ANSWER_RELEVANCY.value
assert metric.config.model == "gpt-4"
assert metric.config.embedder == "text-embedding-ada-002"
assert metric.config.api_key == "test_api_key"
assert metric.config.num_gen_questions == 1
def test_answer_relevance_init_without_api_key(monkeypatch):
monkeypatch.delenv("OPENAI_API_KEY", raising=False)
with pytest.raises(ValueError):
AnswerRelevance()
def test_generate_prompt(mock_answer_relevance_metric, mock_data):
prompt = mock_answer_relevance_metric._generate_prompt(mock_data[0])
assert "This is a test answer 1." in prompt
prompt = mock_answer_relevance_metric._generate_prompt(mock_data[1])
assert "This is a test answer 2." in prompt
def test_generate_questions(mock_answer_relevance_metric, mock_data, monkeypatch):
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type(
"obj",
(object,),
{"message": type("obj", (object,), {"content": "This is a test question response.\n"})},
)
]
},
)(),
)
prompt = mock_answer_relevance_metric._generate_prompt(mock_data[0])
questions = mock_answer_relevance_metric._generate_questions(prompt)
assert len(questions) == 1
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type("obj", (object,), {"message": type("obj", (object,), {"content": "question 1?\nquestion2?"})})
]
},
)(),
)
prompt = mock_answer_relevance_metric._generate_prompt(mock_data[1])
questions = mock_answer_relevance_metric._generate_questions(prompt)
assert len(questions) == 2
def test_generate_embedding(mock_answer_relevance_metric, mock_data, monkeypatch):
monkeypatch.setattr(
mock_answer_relevance_metric.client.embeddings,
"create",
lambda input, model: type("obj", (object,), {"data": [type("obj", (object,), {"embedding": [1, 2, 3]})]})(),
)
embedding = mock_answer_relevance_metric._generate_embedding("This is a test question.")
assert len(embedding) == 3
def test_compute_similarity(mock_answer_relevance_metric, mock_data):
original = np.array([1, 2, 3])
generated = np.array([[1, 2, 3], [1, 2, 3]])
similarity = mock_answer_relevance_metric._compute_similarity(original, generated)
assert len(similarity) == 2
assert similarity[0] == 1.0
assert similarity[1] == 1.0
def test_compute_score(mock_answer_relevance_metric, mock_data, monkeypatch):
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type(
"obj",
(object,),
{"message": type("obj", (object,), {"content": "This is a test question response.\n"})},
)
]
},
)(),
)
monkeypatch.setattr(
mock_answer_relevance_metric.client.embeddings,
"create",
lambda input, model: type("obj", (object,), {"data": [type("obj", (object,), {"embedding": [1, 2, 3]})]})(),
)
score = mock_answer_relevance_metric._compute_score(mock_data[0])
assert score == 1.0
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type("obj", (object,), {"message": type("obj", (object,), {"content": "question 1?\nquestion2?"})})
]
},
)(),
)
monkeypatch.setattr(
mock_answer_relevance_metric.client.embeddings,
"create",
lambda input, model: type("obj", (object,), {"data": [type("obj", (object,), {"embedding": [1, 2, 3]})]})(),
)
score = mock_answer_relevance_metric._compute_score(mock_data[1])
assert score == 1.0
def test_evaluate(mock_answer_relevance_metric, mock_data, monkeypatch):
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type(
"obj",
(object,),
{"message": type("obj", (object,), {"content": "This is a test question response.\n"})},
)
]
},
)(),
)
monkeypatch.setattr(
mock_answer_relevance_metric.client.embeddings,
"create",
lambda input, model: type("obj", (object,), {"data": [type("obj", (object,), {"embedding": [1, 2, 3]})]})(),
)
score = mock_answer_relevance_metric.evaluate(mock_data)
assert score == 1.0
monkeypatch.setattr(
mock_answer_relevance_metric.client.chat.completions,
"create",
lambda model, messages: type(
"obj",
(object,),
{
"choices": [
type("obj", (object,), {"message": type("obj", (object,), {"content": "question 1?\nquestion2?"})})
]
},
)(),
)
monkeypatch.setattr(
mock_answer_relevance_metric.client.embeddings,
"create",
lambda input, model: type("obj", (object,), {"data": [type("obj", (object,), {"embedding": [1, 2, 3]})]})(),
)
score = mock_answer_relevance_metric.evaluate(mock_data)
assert score == 1.0
|