phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-08 03:00:56 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

chore(misc): update datasets, benchmarks to use alpha, beta prefixes (#3891)

Browse source 
This will be landed together with
https://github.com/llamastack/llama-stack-client-python/pull/282 (hence
CI will be red on this one.)

I have verified locally that tests pass with the updated version of the
client-sdk.
This commit is contained in:
Ashwin Bharambe 2025-10-22 15:26:35 -07:00 • committed by GitHub
parent 7918188f1e
commit 0e57233a0a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 15 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama_stack/core/server/server.py
View file

@ -174,7 +174,9 @@ class StackApp(FastAPI):
@asynccontextmanager @asynccontextmanager
async def lifespan(app: StackApp): async def lifespan(app: StackApp):
logger.info("Starting up") server_version = parse_version("llama-stack")
logger.info(f"Starting up Llama Stack server (version: {server_version})")
assert app.stack is not None assert app.stack is not None
app.stack.create_registry_refresh_task() app.stack.create_registry_refresh_task()
yield yield

10
tests/integration/datasets/test_datasets.py
View file

@ -78,18 +78,18 @@ def data_url_from_file(file_path: str) -> str:
], ],
) )
def test_register_and_iterrows(llama_stack_client, purpose, source, provider_id, limit): def test_register_and_iterrows(llama_stack_client, purpose, source, provider_id, limit):
dataset = llama_stack_client.datasets.register( dataset = llama_stack_client.beta.datasets.register(
purpose=purpose, purpose=purpose,
source=source, source=source,
) )
assert dataset.identifier is not None assert dataset.identifier is not None
assert dataset.provider_id == provider_id assert dataset.provider_id == provider_id
iterrow_response = llama_stack_client.datasets.iterrows(dataset.identifier, limit=limit) iterrow_response = llama_stack_client.beta.datasets.iterrows(dataset.identifier, limit=limit)
assert len(iterrow_response.data) == limit assert len(iterrow_response.data) == limit
dataset_list = llama_stack_client.datasets.list() dataset_list = llama_stack_client.beta.datasets.list()
assert dataset.identifier in [d.identifier for d in dataset_list] assert dataset.identifier in [d.identifier for d in dataset_list]
llama_stack_client.datasets.unregister(dataset.identifier) llama_stack_client.beta.datasets.unregister(dataset.identifier)
dataset_list = llama_stack_client.datasets.list() dataset_list = llama_stack_client.beta.datasets.list()
assert dataset.identifier not in [d.identifier for d in dataset_list] assert dataset.identifier not in [d.identifier for d in dataset_list]

14
tests/integration/eval/test_eval.py
View file

@ -17,17 +17,17 @@ from ..datasets.test_datasets import data_url_from_file
@pytest.mark.parametrize("scoring_fn_id", ["basic::equality"]) @pytest.mark.parametrize("scoring_fn_id", ["basic::equality"])
def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id): def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
dataset = llama_stack_client.datasets.register( dataset = llama_stack_client.beta.datasets.register(
purpose="eval/messages-answer", purpose="eval/messages-answer",
source={ source={
"type": "uri", "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() response = llama_stack_client.beta.datasets.list()
assert any(x.identifier == dataset.identifier for x in response) assert any(x.identifier == dataset.identifier for x in response)
rows = llama_stack_client.datasets.iterrows( rows = llama_stack_client.beta.datasets.iterrows(
dataset_id=dataset.identifier, dataset_id=dataset.identifier,
limit=3, limit=3,
) )
@ -37,12 +37,12 @@ def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
scoring_fn_id, scoring_fn_id,
] ]
benchmark_id = str(uuid.uuid4()) benchmark_id = str(uuid.uuid4())
llama_stack_client.benchmarks.register( llama_stack_client.alpha.benchmarks.register(
benchmark_id=benchmark_id, benchmark_id=benchmark_id,
dataset_id=dataset.identifier, dataset_id=dataset.identifier,
scoring_functions=scoring_functions, scoring_functions=scoring_functions,
) )
list_benchmarks = llama_stack_client.benchmarks.list() list_benchmarks = llama_stack_client.alpha.benchmarks.list()
assert any(x.identifier == benchmark_id for x in list_benchmarks) assert any(x.identifier == benchmark_id for x in list_benchmarks)
response = llama_stack_client.alpha.eval.evaluate_rows( response = llama_stack_client.alpha.eval.evaluate_rows(
@ -66,7 +66,7 @@ def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
@pytest.mark.parametrize("scoring_fn_id", ["basic::subset_of"]) @pytest.mark.parametrize("scoring_fn_id", ["basic::subset_of"])
def test_evaluate_benchmark(llama_stack_client, text_model_id, scoring_fn_id): def test_evaluate_benchmark(llama_stack_client, text_model_id, scoring_fn_id):
dataset = llama_stack_client.datasets.register( dataset = llama_stack_client.beta.datasets.register(
purpose="eval/messages-answer", purpose="eval/messages-answer",
source={ source={
"type": "uri", "type": "uri",
@ -74,7 +74,7 @@ def test_evaluate_benchmark(llama_stack_client, text_model_id, scoring_fn_id):
}, },
) )
benchmark_id = str(uuid.uuid4()) benchmark_id = str(uuid.uuid4())
llama_stack_client.benchmarks.register( llama_stack_client.alpha.benchmarks.register(
benchmark_id=benchmark_id, benchmark_id=benchmark_id,
dataset_id=dataset.identifier, dataset_id=dataset.identifier,
scoring_functions=[scoring_fn_id], scoring_functions=[scoring_fn_id],