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

(fix) log cache hits on SpendLogs table

Browse source
This commit is contained in:
ishaan-jaff 2024-01-24 18:51:39 -08:00
parent bf851ef19a
commit 2f3765a03f
2 changed files with 10 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

6
litellm/proxy/proxy_server.py
View file

@ -593,6 +593,12 @@ async def track_cost_callback(
"user_api_key_user_id", None "user_api_key_user_id", None
) )
if kwargs.get("cache_hit", False) == True:
response_cost = 0.0
verbose_proxy_logger.info(
f"Cache Hit: response_cost {response_cost}, for user_id {user_id}"
)
verbose_proxy_logger.info( verbose_proxy_logger.info(
f"response_cost {response_cost}, for user_id {user_id}" f"response_cost {response_cost}, for user_id {user_id}"
) )

4
litellm/proxy/utils.py
View file

@ -1003,6 +1003,10 @@ def get_logging_payload(kwargs, response_obj, start_time, end_time):
cache_key = litellm.cache.get_cache_key(**kwargs) cache_key = litellm.cache.get_cache_key(**kwargs)
else: else:
cache_key = "Cache OFF" cache_key = "Cache OFF"
if cache_hit == True:
import time
id = f"{id}_cache_hit{time.time()}" # SpendLogs does not allow duplicate request_id
payload = { payload = {
"request_id": id, "request_id": id,