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

(fix) langfuse logging + openai streaming when chunk = [DONE}

Browse source
This commit is contained in:
ishaan-jaff 2023-11-16 10:44:09 -08:00
parent bb51216846
commit c84fe1af5e
1 changed files with 14 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

18
litellm/utils.py
View file

@ -694,7 +694,7 @@ class Logging:
## BUILD COMPLETE STREAMED RESPONSE ## BUILD COMPLETE STREAMED RESPONSE
if self.stream: if self.stream:
if result.choices[0].finish_reason: # if it's the last chunk if result.choices[0].finish_reason is not None: # if it's the last chunk
self.streaming_chunks.append(result) self.streaming_chunks.append(result)
complete_streaming_response = litellm.stream_chunk_builder(self.streaming_chunks) complete_streaming_response = litellm.stream_chunk_builder(self.streaming_chunks)
else: else:
@ -832,12 +832,20 @@ class Logging:
) )
if callback == "langfuse": if callback == "langfuse":
print_verbose("reaches langfuse for logging!") print_verbose("reaches langfuse for logging!")
deep_copy = {} kwargs = {}
for k, v in self.model_call_details.items(): for k, v in self.model_call_details.items():
if k != "original_response": # copy.deepcopy raises errors as this could be a coroutine if k != "original_response": # copy.deepcopy raises errors as this could be a coroutine
deep_copy[k] = v kwargs[k] = v
# this only logs streaming once, complete_streaming_response exists i.e when stream ends
if self.stream:
if "complete_streaming_response" not in kwargs:
return
else:
print_verbose("reaches langfuse for streaming logging!")
result = kwargs["complete_streaming_response"]
langFuseLogger.log_event( langFuseLogger.log_event(
kwargs=deep_copy, kwargs=kwargs,
response_obj=result, response_obj=result,
start_time=start_time, start_time=start_time,
end_time=end_time, end_time=end_time,
@ -4738,6 +4746,8 @@ class CustomStreamWrapper:
model_response.choices[0].finish_reason = response_obj["finish_reason"] model_response.choices[0].finish_reason = response_obj["finish_reason"]
else: # openai chat model else: # openai chat model
response_obj = self.handle_openai_chat_completion_chunk(chunk) response_obj = self.handle_openai_chat_completion_chunk(chunk)
if response_obj == None:
return
completion_obj["content"] = response_obj["text"] completion_obj["content"] = response_obj["text"]
print_verbose(f"completion obj content: {completion_obj['content']}") print_verbose(f"completion obj content: {completion_obj['content']}")
print_verbose(f"len(completion_obj['content']: {len(completion_obj['content'])}") print_verbose(f"len(completion_obj['content']: {len(completion_obj['content'])}")