phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-04 12:07:34 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Add rerank models to the dynamic model list; Fix integration tests

Browse source
This commit is contained in:
Jiayi 2025-09-28 14:45:16 -07:00
parent 3538477070
commit 816b68fdc7
8 changed files with 247 additions and 25 deletions
Download patch file Download diff file Expand all files Collapse all files

12
llama_stack/providers/utils/inference/openai_mixin.py
View file

@ -63,6 +63,10 @@ class OpenAIMixin(ModelsProtocolPrivate, NeedsRequestProviderData, ABC):
# Format: {"model_id": {"embedding_dimension": 1536, "context_length": 8192}}
embedding_model_metadata: dict[str, dict[str, int]] = {}
# List of rerank model IDs for this provider
# Can be set by subclasses or instances to provide rerank models
rerank_model_list: list[str] = []
# Cache of available models keyed by model ID
# This is set in list_models() and used in check_model_availability()
_model_cache: dict[str, Model] = {}
@ -400,6 +404,14 @@ class OpenAIMixin(ModelsProtocolPrivate, NeedsRequestProviderData, ABC):
model_type=ModelType.embedding,
metadata=metadata,
)
elif m.id in self.rerank_model_list:
# This is a rerank model
model = Model(
provider_id=self.__provider_id__, # type: ignore[attr-defined]
provider_resource_id=m.id,
identifier=m.id,
model_type=ModelType.rerank,
)
else:
# This is an LLM
model = Model(