phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-23 00:27:26 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

chore: Updating how default embedding model is set in stack (#3818)

Browse source 
# What does this PR do?

Refactor setting default vector store provider and embedding model to
use an optional `vector_stores` config in the `StackRunConfig` and clean
up code to do so (had to add back in some pieces of VectorDB). Also
added remote Qdrant and Weaviate to starter distro (based on other PR
where inference providers were added for UX).

New config is simply (default for Starter distro):

```yaml
vector_stores:
  default_provider_id: faiss
  default_embedding_model:
    provider_id: sentence-transformers
    model_id: nomic-ai/nomic-embed-text-v1.5
```

## Test Plan
CI and Unit tests.

---------

Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
Co-authored-by: Ashwin Bharambe <ashwin.bharambe@gmail.com>
This commit is contained in:
Francisco Arceo 2025-10-20 17:22:45 -04:00 • committed by GitHub
parent 2c43285e22
commit 48581bf651
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
48 changed files with 973 additions and 818 deletions
Download patch file Download diff file Expand all files Collapse all files

126
tests/unit/core/test_stack_validation.py
View file

@ -4,90 +4,64 @@
# This source code is licensed under the terms described in the LICENSE file in
# the root directory of this source tree.
"""
Unit tests for Stack validation functions.
"""
"""Unit tests for Stack validation functions."""
from unittest.mock import AsyncMock
import pytest
from llama_stack.apis.models import Model, ModelType
from llama_stack.core.stack import validate_default_embedding_model
from llama_stack.apis.models import ListModelsResponse, Model, ModelType
from llama_stack.core.datatypes import QualifiedModel, StackRunConfig, StorageConfig, VectorStoresConfig
from llama_stack.core.stack import validate_vector_stores_config
from llama_stack.providers.datatypes import Api
class TestStackValidation:
"""Test Stack validation functions."""
class TestVectorStoresValidation:
async def test_validate_missing_model(self):
"""Test validation fails when model not found."""
run_config = StackRunConfig(
image_name="test",
providers={},
storage=StorageConfig(backends={}, stores={}),
vector_stores=VectorStoresConfig(
default_provider_id="faiss",
default_embedding_model=QualifiedModel(
provider_id="p",
model_id="missing",
),
),
)
mock_models = AsyncMock()
mock_models.list_models.return_value = ListModelsResponse(data=[])
@pytest.mark.parametrize(
"models,should_raise",
[
([], False), # No models
(
[
Model(
identifier="emb1",
model_type=ModelType.embedding,
metadata={"default_configured": True},
provider_id="p",
provider_resource_id="emb1",
)
],
False,
), # Single default
(
[
Model(
identifier="emb1",
model_type=ModelType.embedding,
metadata={"default_configured": True},
provider_id="p",
provider_resource_id="emb1",
),
Model(
identifier="emb2",
model_type=ModelType.embedding,
metadata={"default_configured": True},
provider_id="p",
provider_resource_id="emb2",
),
],
True,
), # Multiple defaults
(
[
Model(
identifier="emb1",
model_type=ModelType.embedding,
metadata={"default_configured": True},
provider_id="p",
provider_resource_id="emb1",
),
Model(
identifier="llm1",
model_type=ModelType.llm,
metadata={"default_configured": True},
provider_id="p",
provider_resource_id="llm1",
),
],
False,
), # Ignores non-embedding
],
)
async def test_validate_default_embedding_model(self, models, should_raise):
"""Test validation with various model configurations."""
mock_models_impl = AsyncMock()
mock_models_impl.list_models.return_value = models
impls = {Api.models: mock_models_impl}
with pytest.raises(ValueError, match="not found"):
await validate_vector_stores_config(run_config.vector_stores, {Api.models: mock_models})
if should_raise:
with pytest.raises(ValueError, match="Multiple embedding models marked as default_configured=True"):
await validate_default_embedding_model(impls)
else:
await validate_default_embedding_model(impls)
async def test_validate_success(self):
"""Test validation passes with valid model."""
run_config = StackRunConfig(
image_name="test",
providers={},
storage=StorageConfig(backends={}, stores={}),
vector_stores=VectorStoresConfig(
default_provider_id="faiss",
default_embedding_model=QualifiedModel(
provider_id="p",
model_id="valid",
),
),
)
mock_models = AsyncMock()
mock_models.list_models.return_value = ListModelsResponse(
data=[
Model(
identifier="p/valid", # Must match provider_id/model_id format
model_type=ModelType.embedding,
metadata={"embedding_dimension": 768},
provider_id="p",
provider_resource_id="valid",
)
]
)
async def test_validate_default_embedding_model_no_models_api(self):
"""Test validation when models API is not available."""
await validate_default_embedding_model({})
await validate_vector_stores_config(run_config.vector_stores, {Api.models: mock_models})