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

fix(utils.py): fix azure completion cost calculation

Browse source
This commit is contained in:
Krrish Dholakia 2023-11-30 09:19:35 -08:00
parent 7ee089b5ca
commit b4b7acdb72
2 changed files with 2 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

2
litellm/proxy/proxy_server.py
View file

@ -863,7 +863,7 @@ async def model_info(request: Request):
model_name = m["model_name"]
model_params = {}
for k,v in m["litellm_params"].items():
if k == "api_key": # don't send the api key
if k == "api_key" or k == "api_base": # don't send the api key or api base
continue
if k == "model":

4
litellm/utils.py
View file

@ -1660,8 +1660,6 @@ def cost_per_token(model="", prompt_tokens=0, completion_tokens=0):
"gpt-35-turbo-16k": "azure/gpt-3.5-turbo-16k",
"gpt-35-turbo-instruct": "azure/gpt-3.5-turbo-instruct"
}
if "azure/" in model:
model = model.replace("azure/", "")
if model in model_cost_ref:
prompt_tokens_cost_usd_dollar = (
model_cost_ref[model]["input_cost_per_token"] * prompt_tokens
@ -1741,7 +1739,7 @@ def completion_cost(
# Handle Inputs to completion_cost
prompt_tokens = 0
completion_tokens = 0
if completion_response != None:
if completion_response is not None:
# get input/output tokens from completion_response
prompt_tokens = completion_response['usage']['prompt_tokens']
completion_tokens = completion_response['usage']['completion_tokens']