phoenix-oss/llama-stack
1
0
Fork 0
forked from phoenix-oss/llama-stack-mirror
Code Pull requests Releases Packages 2 Activity Actions

feat: OpenAI Responses API (#1989)

Browse source 
# What does this PR do?

This provides an initial [OpenAI Responses
API](https://platform.openai.com/docs/api-reference/responses)
implementation. The API is not yet complete, and this is more a
proof-of-concept to show how we can store responses in our key-value
stores and use them to support the Responses API concepts like
`previous_response_id`.

## Test Plan

I've added a new
`tests/integration/openai_responses/test_openai_responses.py` as part of
a test-driven development for this new API. I'm only testing this
locally with the remote-vllm provider for now, but it should work with
any of our inference providers since the only API it requires out of the
inference provider is the `openai_chat_completion` endpoint.

```
VLLM_URL="http://localhost:8000/v1" \
INFERENCE_MODEL="meta-llama/Llama-3.2-3B-Instruct" \
llama stack build --template remote-vllm --image-type venv --run
```

```
LLAMA_STACK_CONFIG="http://localhost:8321" \
python -m pytest -v \
  tests/integration/openai_responses/test_openai_responses.py \
  --text-model "meta-llama/Llama-3.2-3B-Instruct"
 ```

---------

Signed-off-by: Ben Browning <bbrownin@redhat.com>
Co-authored-by: Ashwin Bharambe <ashwin.bharambe@gmail.com>
This commit is contained in:
Ben Browning 2025-04-28 17:06:00 -04:00 • committed by GitHub
parent 79851d93aa
commit 8dfce2f596
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
21 changed files with 1766 additions and 59 deletions
Download patch file Download diff file Expand all files Collapse all files

43
llama_stack/apis/agents/agents.py
View file

@ -38,6 +38,13 @@ from llama_stack.apis.safety import SafetyViolation
from llama_stack.apis.tools import ToolDef
from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
from .openai_responses import (
OpenAIResponseInputMessage,
OpenAIResponseInputTool,
OpenAIResponseObject,
OpenAIResponseObjectStream,
)
class Attachment(BaseModel):
"""An attachment to an agent turn.
@ -593,3 +600,39 @@ class Agents(Protocol):
:returns: A ListAgentSessionsResponse.
"""
...
# We situate the OpenAI Responses API in the Agents API just like we did things
# for Inference. The Responses API, in its intent, serves the same purpose as
# the Agents API above -- it is essentially a lightweight "agentic loop" with
# integrated tool calling.
#
# Both of these APIs are inherently stateful.
@webmethod(route="/openai/v1/responses/{id}", method="GET")
async def get_openai_response(
self,
id: str,
) -> OpenAIResponseObject:
"""Retrieve an OpenAI response by its ID.
:param id: The ID of the OpenAI response to retrieve.
:returns: An OpenAIResponseObject.
"""
...
@webmethod(route="/openai/v1/responses", method="POST")
async def create_openai_response(
self,
input: Union[str, List[OpenAIResponseInputMessage]],
model: str,
previous_response_id: Optional[str] = None,
store: Optional[bool] = True,
stream: Optional[bool] = False,
tools: Optional[List[OpenAIResponseInputTool]] = None,
) -> Union[OpenAIResponseObject, AsyncIterator[OpenAIResponseObjectStream]]:
"""Create a new OpenAI response.
:param input: Input message(s) to create the response.
:param model: The underlying LLM used for completions.
:param previous_response_id: (Optional) if specified, the new response will be a continuation of the previous response. This can be used to easily fork-off new responses from existing responses.
"""