phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-18 16:59:47 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat(api): introduce /rerank

Browse source 
# What does this PR do?


## Test Plan
This commit is contained in:
Eric Huang 2025-08-14 16:14:38 -07:00
parent e69acbafbf
commit 306625025a
8 changed files with 336 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

14
llama_stack/providers/remote/inference/llama_openai_compat/llama.py
View file

@ -5,6 +5,11 @@
# the root directory of this source tree.
import logging
from llama_stack.apis.inference import (
OpenAIChatCompletionContentPartImageParam,
OpenAIChatCompletionContentPartTextParam,
RerankResponse,
)
from llama_stack.providers.remote.inference.llama_openai_compat.config import LlamaCompatConfig
from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
from llama_stack.providers.utils.inference.openai_mixin import OpenAIMixin
@ -55,3 +60,12 @@ class LlamaCompatInferenceAdapter(OpenAIMixin, LiteLLMOpenAIMixin):
async def shutdown(self):
await super().shutdown()
async def rerank(
self,
model: str,
query: str | OpenAIChatCompletionContentPartTextParam | OpenAIChatCompletionContentPartImageParam,
items: list[str | OpenAIChatCompletionContentPartTextParam | OpenAIChatCompletionContentPartImageParam],
max_num_results: int | None = None,
) -> RerankResponse:
raise NotImplementedError("Reranking is not supported for Llama OpenAI Compat")