phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-24 18:24:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

(feat) router track total, success, failed calls per model

Browse source
This commit is contained in:
ishaan-jaff 2023-11-28 12:56:48 -08:00
parent 20380804b3
commit 2a69cab550
1 changed files with 9 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

9
litellm/router.py
View file

@ -15,6 +15,7 @@ from litellm.caching import RedisCache, InMemoryCache, DualCache
import logging, asyncio
import inspect, concurrent
from openai import AsyncOpenAI
from collections import defaultdict
class Router:
"""
@ -88,6 +89,10 @@ class Router:
self.fallbacks = fallbacks or litellm.fallbacks
self.context_window_fallbacks = context_window_fallbacks or litellm.context_window_fallbacks
self.model_exception_map: dict = {} # dict to store model: list exceptions. self.exceptions = {"gpt-3.5": ["API KEY Error", "Rate Limit Error", "good morning error"]}
self.total_calls: defaultdict = defaultdict(int) # dict to store total calls made to each model
self.fail_calls: defaultdict = defaultdict(int) # dict to store fail_calls made to each model
self.success_calls: defaultdict = defaultdict(int) # dict to store success_calls made to each model
# make Router.chat.completions.create compatible for openai.chat.completions.create
self.chat = litellm.Chat(params=default_litellm_params)
@ -226,9 +231,13 @@ class Router:
data["model"] = original_model_string[:index_of_model_id]
else:
data["model"] = original_model_string
self.total_calls[original_model_string] +=1
response = await litellm.acompletion(**{**data, "messages": messages, "caching": self.cache_responses, **kwargs})
self.success_calls[original_model_string] +=1
return response
except Exception as e:
if original_model_string is not None:
self.fail_calls[original_model_string] +=1
raise e
def text_completion(self,