phoenix/litellm
1
0
Fork 0
forked from phoenix/litellm-mirror
Code Pull requests Activity

Merge pull request #2942 from BerriAI/litellm_fix_router_loading

Browse source 
Router Async Improvements
This commit is contained in:
Krish Dholakia 2024-04-10 20:16:53 -07:00 • committed by GitHub
commit 83e7ed94ce
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
10 changed files with 746 additions and 60 deletions
Download patch file Download diff file Expand all files Collapse all files

14
litellm/utils.py
View file

@ -1990,9 +1990,6 @@ class Logging:
else:
litellm.cache.add_cache(result, **kwargs)
if isinstance(callback, CustomLogger): # custom logger class
print_verbose(
f"Running Async success callback: {callback}; self.stream: {self.stream}; async_complete_streaming_response: {self.model_call_details.get('async_complete_streaming_response', None)} result={result}"
)
if self.stream == True:
if (
"async_complete_streaming_response"
@ -2376,7 +2373,6 @@ def client(original_function):
if litellm.use_client or (
"use_client" in kwargs and kwargs["use_client"] == True
):
print_verbose(f"litedebugger initialized")
if "lite_debugger" not in litellm.input_callback:
litellm.input_callback.append("lite_debugger")
if "lite_debugger" not in litellm.success_callback:
@ -5912,6 +5908,16 @@ def get_api_key(llm_provider: str, dynamic_api_key: Optional[str]):
return api_key
def get_utc_datetime():
import datetime as dt
from datetime import datetime
if hasattr(dt, "UTC"):
return datetime.now(dt.UTC) # type: ignore
else:
return datetime.utcnow() # type: ignore
def get_max_tokens(model: str):
"""
Get the maximum number of output tokens allowed for a given model.