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 linting

Browse source
This commit is contained in:
ishaan-jaff 2023-09-11 18:11:06 -07:00
parent be5a92c40a
commit d70cff8713
1 changed files with 3 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

8
litellm/utils.py
View file

@ -2445,7 +2445,7 @@ def shorten_message_to_fit_limit(
def safe_messages(
messages,
model = None,
system_message = None,
system_message = None, # str of user system message
trim_ratio: float = 0.75,
return_response_tokens: bool = False,
max_tokens = None
@ -2487,13 +2487,11 @@ def safe_messages(
#### Trimming messages if current_tokens > max_tokens
print_verbose(f"Need to trim input messages: {messages}, current_tokens{current_tokens}, max_tokens: {max_tokens}")
if system_message:
system_message_event, max_tokens = process_system_message(messages=messages, max_tokens=max_tokens, model=model)
system_message_event, max_tokens = process_system_message(system_message=system_message, max_tokens=max_tokens, model=model)
messages = messages + [system_message_event]
final_messages = process_messages(messages=messages, max_tokens=max_tokens, model=model)
if system_message:
final_messages = [system_message_event] + final_messages
if return_response_tokens: # if user wants token count with new trimmed messages
response_tokens = max_tokens - get_token_count(final_messages, model)
return final_messages, response_tokens