phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-25 02:34:29 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

test_basic_openai_responses_api_streaming

Browse source
This commit is contained in:
Ishaan Jaff 2025-04-19 13:11:33 -07:00
parent b61ae8dc86
commit a508da4c8d
1 changed files with 10 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

11
tests/llm_responses_api_testing/base_responses_api.py
View file

@ -133,7 +133,7 @@ class BaseResponsesAPITest(ABC):
validate_responses_api_response(response, final_chunk=True)
@pytest.mark.parametrize("sync_mode", [True])
@pytest.mark.parametrize("sync_mode", [True, False])
@pytest.mark.asyncio
async def test_basic_openai_responses_api_streaming(self, sync_mode):
litellm._turn_on_debug()
@ -178,6 +178,15 @@ class BaseResponsesAPITest(ABC):
# assert the response completed event includes the usage
assert response_completed_event.response.usage is not None
# basic test assert the usage seems reasonable
print("response_completed_event.response.usage=", response_completed_event.response.usage)
assert response_completed_event.response.usage.input_tokens > 0 and response_completed_event.response.usage.input_tokens < 100
assert response_completed_event.response.usage.output_tokens > 0 and response_completed_event.response.usage.output_tokens < 100
assert response_completed_event.response.usage.total_tokens > 0 and response_completed_event.response.usage.total_tokens < 1000
# total tokens should be the sum of input and output tokens
assert response_completed_event.response.usage.total_tokens == response_completed_event.response.usage.input_tokens + response_completed_event.response.usage.output_tokens