phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-03 18:00:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat!: standardize base_url for inference (#4177)

Browse source 
# What does this PR do?

Completes #3732 by removing runtime URL transformations and requiring
users to provide full URLs in configuration. All providers now use
'base_url' consistently and respect the exact URL provided without
appending paths like /v1 or /openai/v1 at runtime.

BREAKING CHANGE: Users must update configs to include full URL paths
(e.g., http://localhost:11434/v1 instead of http://localhost:11434).

Closes #3732 

## Test Plan

Existing tests should pass even with the URL changes, due to default
URLs being altered.

Add unit test to enforce URL standardization across remote inference
providers (verifies all use 'base_url' field with HttpUrl | None type)

Signed-off-by: Charlie Doern <cdoern@redhat.com>
This commit is contained in:
Charlie Doern 2025-11-19 11:44:28 -05:00 • committed by GitHub
parent 91f1b352b4
commit d5cd0eea14
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
67 changed files with 282 additions and 227 deletions
Download patch file Download diff file Expand all files Collapse all files

4
tests/unit/providers/inference/test_inference_client_caching.py
View file

@ -120,7 +120,7 @@ from llama_stack.providers.remote.inference.watsonx.watsonx import WatsonXInfere
VLLMInferenceAdapter,
"llama_stack.providers.remote.inference.vllm.VLLMProviderDataValidator",
{
"url": "http://fake",
"base_url": "http://fake",
},
),
],
@ -153,7 +153,7 @@ def test_litellm_provider_data_used(config_cls, adapter_cls, provider_data_valid
"""Validate data for LiteLLM-based providers. Similar to test_openai_provider_data_used, but without the
assumption that there is an OpenAI-compatible client object."""
inference_adapter = adapter_cls(config=config_cls())
inference_adapter = adapter_cls(config=config_cls(base_url="http://fake"))
inference_adapter.__provider_spec__ = MagicMock()
inference_adapter.__provider_spec__.provider_data_validator = provider_data_validator

6
tests/unit/providers/inference/test_remote_vllm.py
View file

@ -40,7 +40,7 @@ from llama_stack_api import (
@pytest.fixture(scope="function")
async def vllm_inference_adapter():
config = VLLMInferenceAdapterConfig(url="http://mocked.localhost:12345")
config = VLLMInferenceAdapterConfig(base_url="http://mocked.localhost:12345")
inference_adapter = VLLMInferenceAdapter(config=config)
inference_adapter.model_store = AsyncMock()
await inference_adapter.initialize()
@ -204,7 +204,7 @@ async def test_vllm_completion_extra_body():
via extra_body to the underlying OpenAI client through the InferenceRouter.
"""
# Set up the vLLM adapter
config = VLLMInferenceAdapterConfig(url="http://mocked.localhost:12345")
config = VLLMInferenceAdapterConfig(base_url="http://mocked.localhost:12345")
vllm_adapter = VLLMInferenceAdapter(config=config)
vllm_adapter.__provider_id__ = "vllm"
await vllm_adapter.initialize()
@ -277,7 +277,7 @@ async def test_vllm_chat_completion_extra_body():
via extra_body to the underlying OpenAI client through the InferenceRouter for chat completion.
"""
# Set up the vLLM adapter
config = VLLMInferenceAdapterConfig(url="http://mocked.localhost:12345")
config = VLLMInferenceAdapterConfig(base_url="http://mocked.localhost:12345")
vllm_adapter = VLLMInferenceAdapter(config=config)
vllm_adapter.__provider_id__ = "vllm"
await vllm_adapter.initialize()