phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-25 18:54:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix(utils.py): fix cost calculation to handle tool input

Browse source
This commit is contained in:
Krrish Dholakia 2023-12-08 09:53:46 -08:00
parent 9494c2cd9e
commit 1b35736797
3 changed files with 77 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

7
litellm/proxy/proxy_server.py
View file

@ -366,6 +366,7 @@ async def track_cost_callback(
global prisma_client
try:
# check if it has collected an entire stream response
print(f"kwargs stream: {kwargs.get('stream', None)} + complete streaming response: {kwargs.get('complete_streaming_response', None)}")
if "complete_streaming_response" in kwargs:
# for tracking streaming cost we pass the "messages" and the output_text to litellm.completion_cost
completion_response=kwargs["complete_streaming_response"]
@ -377,16 +378,16 @@ async def track_cost_callback(
completion=output_text
)
print("streaming response_cost", response_cost)
# for non streaming responses
elif kwargs["stream"] is False: # regular response
elif kwargs["stream"] == False: # for non streaming responses
input_text = kwargs.get("messages", "")
print(f"type of input_text: {type(input_text)}")
if isinstance(input_text, list):
response_cost = litellm.completion_cost(completion_response=completion_response, messages=input_text)
elif isinstance(input_text, str):
response_cost = litellm.completion_cost(completion_response=completion_response, prompt=input_text)
print(f"received completion response: {completion_response}")
print("regular response_cost", response_cost)
print(f"regular response_cost: {response_cost}")
user_api_key = kwargs["litellm_params"]["metadata"].get("user_api_key", None)
print(f"user_api_key - {user_api_key}; prisma_client - {prisma_client}")
if user_api_key and prisma_client: