_get_local_provider_call should return an AsyncOpenAI client when appropriate

[yml]

Description

_get_local_provider_call should return a client appropriate to the context of execution sync or async like it is the case for openAI for example.

This is probably not the best solution but illustrate the discussion.

 diff --git a/mirascope/llm/_call.py b/mirascope/llm/_call.py
index 48dc2098..ebec6992 100644
--- a/mirascope/llm/_call.py
+++ b/mirascope/llm/_call.py
@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+import asyncio
 from collections.abc import AsyncIterable, Awaitable, Callable, Iterable
 from enum import Enum
 from functools import wraps
@@ -56,9 +57,13 @@ def _get_local_provider_call(
 
         if client:
             return openai_call, client
-        from openai import OpenAI
+        from openai import AsyncOpenAI, OpenAI
 
-        client = OpenAI(api_key="ollama", base_url="http://localhost:11434/v1")
+        try:
+            asyncio.get_running_loop()
+            client = AsyncOpenAI(api_key="ollama", base_url="http://localhost:11434/v1")
+        except RuntimeError:
+            client = OpenAI(api_key="ollama", base_url="http://localhost:11434/v1")
         return openai_call, client
     else:  # provider == "vllm"
         from ..core.openai import openai_call
(END)

Python, Mirascope & OS Versions, related packages (not required)

The issue occurs to me using Ollama but exactly the same issue probably exists with Vllm

[willbakst]

Fix released in v1.21.1! Thanks for pointing this out