phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-07-27 06:28:50 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

make refreshing happen for all routing tables, naming changes, ollama fixes

Browse source
This commit is contained in:
Ashwin Bharambe 2025-07-24 10:24:10 -07:00
parent 487e073378
commit 0fe110d94a
6 changed files with 67 additions and 63 deletions
Download patch file Download diff file Expand all files Collapse all files

14
llama_stack/providers/remote/inference/ollama/ollama.py
View file

@ -5,6 +5,7 @@
# the root directory of this source tree.
import asyncio
import base64
import uuid
from collections.abc import AsyncGenerator, AsyncIterator
@ -97,14 +98,16 @@ class OllamaInferenceAdapter(
def __init__(self, config: OllamaImplConfig) -> None:
self.register_helper = ModelRegistryHelper(MODEL_ENTRIES)
self.config = config
self._client = None
self._clients: dict[asyncio.AbstractEventLoop, AsyncClient] = {}
self._openai_client = None
@property
def client(self) -> AsyncClient:
if self._client is None:
self._client = AsyncClient(host=self.config.url)
return self._client
# ollama client attaches itself to the current event loop (sadly?)
loop = asyncio.get_running_loop()
if loop not in self._clients:
self._clients[loop] = AsyncClient(host=self.config.url)
return self._clients[loop]
@property
def openai_client(self) -> AsyncOpenAI:
@ -191,8 +194,7 @@ class OllamaInferenceAdapter(
return HealthResponse(status=HealthStatus.ERROR, message=f"Health check failed: {str(e)}")
async def shutdown(self) -> None:
self._client = None
self._openai_client = None
self._clients.clear()
async def unregister_model(self, model_id: str) -> None:
pass