phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-27 03:34:10 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

(fix) remove print statements from completion fallbacks, make them print verbose

Browse source
This commit is contained in:
ishaan-jaff 2023-10-27 17:54:48 -07:00
parent effd08f597
commit a88ed12a9a
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
litellm/utils.py
View file

@ -4060,7 +4060,6 @@ def completion_with_split_tests(models={}, messages=[], use_client=False, overri
return litellm.completion(model=selected_llm, messages=messages, use_client=use_client, **kwargs)
def completion_with_fallbacks(**kwargs):
print(f"kwargs inside completion_with_fallbacks: {kwargs}")
nested_kwargs = kwargs.pop("kwargs", {})
response = None
rate_limited_models = set()
@ -4071,6 +4070,7 @@ def completion_with_fallbacks(**kwargs):
if "fallbacks" in nested_kwargs:
del nested_kwargs["fallbacks"] # remove fallbacks so it's not recursive
# max time to process a request with fallbacks: default 45s
while response == None and time.time() - start_time < 45:
for model in fallbacks:
# loop thru all models
@ -4097,10 +4097,10 @@ def completion_with_fallbacks(**kwargs):
if kwargs.get("model"):
del kwargs["model"]
print(f"trying to make completion call with model: {model}")
print_verbose(f"trying to make completion call with model: {model}")
kwargs = {**kwargs, **nested_kwargs} # combine the openai + litellm params at the same level
response = litellm.completion(**kwargs, model=model)
print(f"response: {response}")
print_verbose(f"response: {response}")
if response != None:
return response