phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-26 11:14:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

refactor(main.py): remove loop run in executor

Browse source 
all litellm routes are async
This commit is contained in:
Krrish Dholakia 2025-04-17 07:00:48 -07:00
parent 8be8022914
commit 7a4d25c45a
1 changed files with 11 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

15
litellm/main.py
View file

@ -24,6 +24,7 @@ from copy import deepcopy
from functools import partial from functools import partial
from typing import ( from typing import (
Any, Any,
Awaitable,
Callable, Callable,
Coroutine, Coroutine,
Dict, Dict,
@ -460,18 +461,22 @@ async def acompletion(
try: try:
# Use a partial function to pass your keyword arguments # Use a partial function to pass your keyword arguments
func = partial(completion, **completion_kwargs, **kwargs) # func = partial(completion, **completion_kwargs, **kwargs)
# Add the context to the function # # Add the context to the function
ctx = contextvars.copy_context() # ctx = contextvars.copy_context()
func_with_context = partial(ctx.run, func) # func_with_context = partial(ctx.run, func)
init_response = await loop.run_in_executor(None, func_with_context) init_response = await cast(
Awaitable[Union[dict, ModelResponse, CustomStreamWrapper]],
completion(**completion_kwargs, **kwargs),
)
if isinstance(init_response, dict) or isinstance( if isinstance(init_response, dict) or isinstance(
init_response, ModelResponse init_response, ModelResponse
): ## CACHING SCENARIO ): ## CACHING SCENARIO
if isinstance(init_response, dict): if isinstance(init_response, dict):
response = ModelResponse(**init_response) response = ModelResponse(**init_response)
else:
response = init_response response = init_response
elif asyncio.iscoroutine(init_response): elif asyncio.iscoroutine(init_response):
response = await init_response response = await init_response