phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-12 13:57:57 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

featu: support passing "extra body" throught to providers

Browse source 
# What does this PR do?
Allows passing through extra_body parameters to inference providers.


closes #2720

## Test Plan
CI and added new test
This commit is contained in:
Eric Huang 2025-10-10 15:42:51 -07:00
parent cb7fb0705b
commit 0ed59497bc
35 changed files with 1893 additions and 200 deletions
Download patch file Download diff file Expand all files Collapse all files

10
llama_stack/core/routers/inference.py
View file

@ -32,13 +32,13 @@ from llama_stack.apis.inference import (
OpenAIAssistantMessageParam,
OpenAIChatCompletion,
OpenAIChatCompletionChunk,
OpenAIChatCompletionRequest,
OpenAIChatCompletionRequestWithExtraBody,
OpenAIChatCompletionToolCall,
OpenAIChatCompletionToolCallFunction,
OpenAIChoice,
OpenAIChoiceLogprobs,
OpenAICompletion,
OpenAICompletionRequest,
OpenAICompletionRequestWithExtraBody,
OpenAICompletionWithInputMessages,
OpenAIEmbeddingsResponse,
OpenAIMessageParam,
@ -183,7 +183,7 @@ class InferenceRouter(Inference):
async def openai_completion(
self,
params: Annotated[OpenAICompletionRequest, Body(...)],
params: Annotated[OpenAICompletionRequestWithExtraBody, Body(...)],
) -> OpenAICompletion:
logger.debug(
f"InferenceRouter.openai_completion: model={params.model}, stream={params.stream}, prompt={params.prompt}",
@ -218,7 +218,7 @@ class InferenceRouter(Inference):
async def openai_chat_completion(
self,
params: Annotated[OpenAIChatCompletionRequest, Body(...)],
params: Annotated[OpenAIChatCompletionRequestWithExtraBody, Body(...)],
) -> OpenAIChatCompletion | AsyncIterator[OpenAIChatCompletionChunk]:
logger.debug(
f"InferenceRouter.openai_chat_completion: model={params.model}, stream={params.stream}, messages={params.messages}",
@ -317,7 +317,7 @@ class InferenceRouter(Inference):
raise NotImplementedError("Get chat completion is not supported: inference store is not configured.")
async def _nonstream_openai_chat_completion(
self, provider: Inference, params: OpenAIChatCompletionRequest
self, provider: Inference, params: OpenAIChatCompletionRequestWithExtraBody
) -> OpenAIChatCompletion:
response = await provider.openai_chat_completion(params)
for choice in response.choices: