phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-31 08:40:05 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

mark tests

Browse source
This commit is contained in:
Xi Yan 2025-03-18 21:56:11 -07:00
parent 583de3d80c
commit aaa5974dce
2 changed files with 24 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

18
tests/integration/eval/test_eval.py
View file

@ -16,12 +16,15 @@ from ..datasets.test_datasets import data_url_from_file
@pytest.mark.parametrize("scoring_fn_id", ["basic::equality"])
@pytest.mark.skip(reason="TODO(xiyan): fix this")
def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
dataset = llama_stack_client.datasets.register(
purpose="eval/messages-answer",
source={
"type": "uri",
"uri": data_url_from_file(Path(__file__).parent.parent / "datasets" / "test_dataset.csv"),
"uri": data_url_from_file(
Path(__file__).parent.parent / "datasets" / "test_dataset.csv"
),
},
)
response = llama_stack_client.datasets.list()
@ -65,12 +68,15 @@ def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
@pytest.mark.parametrize("scoring_fn_id", ["basic::subset_of"])
@pytest.mark.skip(reason="TODO(xiyan): fix this")
def test_evaluate_benchmark(llama_stack_client, text_model_id, scoring_fn_id):
dataset = llama_stack_client.datasets.register(
purpose="eval/messages-answer",
source={
"type": "uri",
"uri": data_url_from_file(Path(__file__).parent.parent / "datasets" / "test_dataset.csv"),
"uri": data_url_from_file(
Path(__file__).parent.parent / "datasets" / "test_dataset.csv"
),
},
)
benchmark_id = str(uuid.uuid4())
@ -93,10 +99,14 @@ def test_evaluate_benchmark(llama_stack_client, text_model_id, scoring_fn_id):
},
)
assert response.job_id == "0"
job_status = llama_stack_client.eval.jobs.status(job_id=response.job_id, benchmark_id=benchmark_id)
job_status = llama_stack_client.eval.jobs.status(
job_id=response.job_id, benchmark_id=benchmark_id
)
assert job_status and job_status == "completed"
eval_response = llama_stack_client.eval.jobs.retrieve(job_id=response.job_id, benchmark_id=benchmark_id)
eval_response = llama_stack_client.eval.jobs.retrieve(
job_id=response.job_id, benchmark_id=benchmark_id
)
assert eval_response is not None
assert len(eval_response.generations) == 5
assert scoring_fn_id in eval_response.scores