phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-28 02:11:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into feat/litellm_sambanova_usage

Browse source
This commit is contained in:
jhpiedrahitao 2025-04-01 07:57:21 -05:00
commit 9c9f9577e2
173 changed files with 3073 additions and 3118 deletions
Download patch file Download diff file Expand all files Collapse all files

29
tests/integration/inference/test_text_inference.py
View file

@ -117,6 +117,33 @@ def test_text_completion_streaming(client_with_models, text_model_id, test_case)
assert len(content_str) > 10
@pytest.mark.parametrize(
"test_case",
[
"inference:completion:stop_sequence",
],
)
def test_text_completion_stop_sequence(client_with_models, text_model_id, inference_provider_type, test_case):
skip_if_model_doesnt_support_completion(client_with_models, text_model_id)
# This is only supported/tested for remote vLLM: https://github.com/meta-llama/llama-stack/issues/1771
if inference_provider_type != "remote::vllm":
pytest.xfail(f"{inference_provider_type} doesn't support 'stop' parameter yet")
tc = TestCase(test_case)
response = client_with_models.inference.completion(
content=tc["content"],
stream=True,
model_id=text_model_id,
sampling_params={
"max_tokens": 50,
"stop": ["1963"],
},
)
streamed_content = [chunk.delta for chunk in response]
content_str = "".join(streamed_content).lower().strip()
assert "1963" not in content_str
@pytest.mark.parametrize(
"test_case",
[
@ -266,6 +293,7 @@ def test_text_chat_completion_first_token_profiling(client_with_models, text_mod
model_id=text_model_id,
messages=messages,
stream=False,
timeout=120, # Increase timeout to 2 minutes for large conversation history
)
message_content = response.completion_message.content.lower().strip()
assert len(message_content) > 0
@ -292,6 +320,7 @@ def test_text_chat_completion_streaming(client_with_models, text_model_id, test_
model_id=text_model_id,
messages=[{"role": "user", "content": question}],
stream=True,
timeout=120, # Increase timeout to 2 minutes for large conversation history
)
streamed_content = [str(chunk.event.delta.text.lower().strip()) for chunk in response]
assert len(streamed_content) > 0