phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-11 13:44:38 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat(tests): make inference_recorder into api_recorder (include tool_invoke) (#3403)

Browse source 
Renames `inference_recorder.py` to `api_recorder.py` and extends it to
support recording/replaying tool invocations in addition to inference
calls.

This allows us to record web-search, etc. tool calls and thereafter
apply recordings for `tests/integration/responses`

## Test Plan

```
export OPENAI_API_KEY=...
export TAVILY_SEARCH_API_KEY=...

./scripts/integration-tests.sh --stack-config ci-tests \
   --suite responses --inference-mode record-if-missing
```
This commit is contained in:
Ashwin Bharambe 2025-10-09 14:27:51 -07:00 • committed by GitHub
parent 26fd5dbd34
commit f50ce11a3b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
284 changed files with 296191 additions and 631 deletions
Download patch file Download diff file Expand all files Collapse all files

18
tests/integration/responses/test_tool_responses.py
View file

@ -46,11 +46,13 @@ def test_response_non_streaming_web_search(compat_client, text_model_id, case):
@pytest.mark.parametrize("case", file_search_test_cases)
def test_response_non_streaming_file_search(compat_client, text_model_id, tmp_path, case):
def test_response_non_streaming_file_search(
compat_client, text_model_id, embedding_model_id, embedding_dimension, tmp_path, case
):
if isinstance(compat_client, LlamaStackAsLibraryClient):
pytest.skip("Responses API file search is not yet supported in library client.")
vector_store = new_vector_store(compat_client, "test_vector_store")
vector_store = new_vector_store(compat_client, "test_vector_store", embedding_model_id, embedding_dimension)
if case.file_content:
file_name = "test_response_non_streaming_file_search.txt"
@ -101,11 +103,13 @@ def test_response_non_streaming_file_search(compat_client, text_model_id, tmp_pa
assert case.expected.lower() in response.output_text.lower().strip()
def test_response_non_streaming_file_search_empty_vector_store(compat_client, text_model_id):
def test_response_non_streaming_file_search_empty_vector_store(
compat_client, text_model_id, embedding_model_id, embedding_dimension
):
if isinstance(compat_client, LlamaStackAsLibraryClient):
pytest.skip("Responses API file search is not yet supported in library client.")
vector_store = new_vector_store(compat_client, "test_vector_store")
vector_store = new_vector_store(compat_client, "test_vector_store", embedding_model_id, embedding_dimension)
# Create the response request, which should query our vector store
response = compat_client.responses.create(
@ -127,12 +131,14 @@ def test_response_non_streaming_file_search_empty_vector_store(compat_client, te
assert response.output_text
def test_response_sequential_file_search(compat_client, text_model_id, tmp_path):
def test_response_sequential_file_search(
compat_client, text_model_id, embedding_model_id, embedding_dimension, tmp_path
):
"""Test file search with sequential responses using previous_response_id."""
if isinstance(compat_client, LlamaStackAsLibraryClient):
pytest.skip("Responses API file search is not yet supported in library client.")
vector_store = new_vector_store(compat_client, "test_vector_store")
vector_store = new_vector_store(compat_client, "test_vector_store", embedding_model_id, embedding_dimension)
# Create a test file with content
file_content = "The Llama 4 Maverick model has 128 experts in its mixture of experts architecture."