phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-12 20:12:33 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat(api)!: support passing extra_body to embeddings and vector_stores APIs

Browse source 
Applies the same pattern from #3777 to embeddings and vector_stores.create() endpoints.

Breaking change: Method signatures now accept a single params object with Pydantic extra="allow" instead of individual parameters. Provider-specific params can be passed via extra_body and accessed through params.model_extra.

Updated APIs: openai_embeddings(), openai_create_vector_store(), openai_create_vector_store_file_batch()
This commit is contained in:
Ashwin Bharambe 2025-10-11 15:27:47 -07:00
parent cfd2e303db
commit 74e2976c1e
20 changed files with 364 additions and 297 deletions
Download patch file Download diff file Expand all files Collapse all files

31
llama_stack/apis/inference/inference.py
View file

@ -1140,6 +1140,25 @@ class OpenAIChatCompletionRequestWithExtraBody(BaseModel, extra="allow"):
user: str | None = None
# extra_body can be accessed via .model_extra
@json_schema_type
class OpenAIEmbeddingsRequestWithExtraBody(BaseModel, extra="allow"):
"""Request parameters for OpenAI-compatible embeddings endpoint.
:param model: The identifier of the model to use. The model must be an embedding model registered with Llama Stack and available via the /models endpoint.
:param input: Input text to embed, encoded as a string or array of strings. To embed multiple inputs in a single request, pass an array of strings.
:param encoding_format: (Optional) The format to return the embeddings in. Can be either "float" or "base64". Defaults to "float".
:param dimensions: (Optional) The number of dimensions the resulting output embeddings should have. Only supported in text-embedding-3 and later models.
:param user: (Optional) A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
"""
model: str
input: str | list[str]
encoding_format: str | None = "float"
dimensions: int | None = None
user: str | None = None
@runtime_checkable
@trace_protocol
class InferenceProvider(Protocol):
@ -1200,21 +1219,11 @@ class InferenceProvider(Protocol):
@webmethod(route="/embeddings", method="POST", level=LLAMA_STACK_API_V1)
async def openai_embeddings(
self,
model: str,
input: str | list[str],
encoding_format: str | None = "float",
dimensions: int | None = None,
user: str | None = None,
params: Annotated[OpenAIEmbeddingsRequestWithExtraBody, Body(...)],
) -> OpenAIEmbeddingsResponse:
"""Create embeddings.
Generate OpenAI-compatible embeddings for the given input using the specified model.
:param model: The identifier of the model to use. The model must be an embedding model registered with Llama Stack and available via the /models endpoint.
:param input: Input text to embed, encoded as a string or array of strings. To embed multiple inputs in a single request, pass an array of strings.
:param encoding_format: (Optional) The format to return the embeddings in. Can be either "float" or "base64". Defaults to "float".
:param dimensions: (Optional) The number of dimensions the resulting output embeddings should have. Only supported in text-embedding-3 and later models.
:param user: (Optional) A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.
:returns: An OpenAIEmbeddingsResponse containing the embeddings.
"""
...