mlflow--mlflow
wehub-resource-sync
73 行
1.8 KiB
Python
73 行
1.8 KiB
Python
import re
|
|
|
|
from mlflow.metrics.genai import EvaluationExample
|
|
|
|
|
|
def test_evaluation_example_str():
|
|
example1 = str(
|
|
EvaluationExample(
|
|
input="This is an input",
|
|
output="This is an output",
|
|
score=5,
|
|
justification="This is a justification",
|
|
grading_context={"foo": "bar"},
|
|
)
|
|
)
|
|
example1_expected = """
|
|
Example Input:
|
|
This is an input
|
|
|
|
Example Output:
|
|
This is an output
|
|
|
|
Additional information used by the model:
|
|
key: foo
|
|
value:
|
|
bar
|
|
|
|
Example score: 5
|
|
Example justification: This is a justification
|
|
"""
|
|
assert re.sub(r"\s+", "", example1_expected) == re.sub(r"\s+", "", example1)
|
|
|
|
example2 = str(
|
|
EvaluationExample(
|
|
input="This is an input", output="This is an output", score=5, justification="It works"
|
|
)
|
|
)
|
|
example2_expected = """
|
|
Example Input:
|
|
This is an input
|
|
|
|
Example Output:
|
|
This is an output
|
|
|
|
Example score: 5
|
|
Example justification: It works
|
|
"""
|
|
assert re.sub(r"\s+", "", example2_expected) == re.sub(r"\s+", "", example2)
|
|
|
|
example3 = str(
|
|
EvaluationExample(
|
|
input="This is an input",
|
|
output="This is an output",
|
|
score=5,
|
|
justification="This is a justification",
|
|
grading_context="Baz baz",
|
|
)
|
|
)
|
|
example3_expected = """
|
|
Example Input:
|
|
This is an input
|
|
|
|
Example Output:
|
|
This is an output
|
|
|
|
Additional information used by the model:
|
|
Baz baz
|
|
|
|
Example score: 5
|
|
Example justification: This is a justification
|
|
"""
|
|
assert re.sub(r"\s+", "", example3_expected) == re.sub(r"\s+", "", example3)
|