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): azure streaming initial format

Browse source
This commit is contained in:
Krrish Dholakia 2023-11-15 13:29:58 -08:00
parent ae7731b4dc
commit 4348fd6435
1 changed files with 23 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

37
litellm/utils.py
View file

@ -18,7 +18,7 @@ import tiktoken
import uuid import uuid
import aiohttp import aiohttp
import logging import logging
import asyncio import asyncio, httpx
import copy import copy
from tokenizers import Tokenizer from tokenizers import Tokenizer
from dataclasses import ( from dataclasses import (
@ -4089,14 +4089,22 @@ def exception_type(
llm_provider=custom_llm_provider, llm_provider=custom_llm_provider,
response=original_exception.response response=original_exception.response
) )
else: else: # ensure generic errors always return APIConnectionError
exception_mapping_worked = True exception_mapping_worked = True
raise APIConnectionError( if hasattr(original_exception, "request"):
message=f"{str(original_exception)}", raise APIConnectionError(
llm_provider=custom_llm_provider, message=f"{str(original_exception)}",
model=model, llm_provider=custom_llm_provider,
request=original_exception.request model=model,
) request=original_exception.request
)
else:
raise APIConnectionError(
message=f"{str(original_exception)}",
llm_provider=custom_llm_provider,
model=model,
request= httpx.Request(method="POST", url="https://api.openai.com/v1/") # stub the request
)
except Exception as e: except Exception as e:
# LOGGING # LOGGING
exception_logging( exception_logging(
@ -4400,10 +4408,11 @@ class CustomStreamWrapper:
elif chunk.startswith("data:"): elif chunk.startswith("data:"):
data_json = json.loads(chunk[5:]) # chunk.startswith("data:"): data_json = json.loads(chunk[5:]) # chunk.startswith("data:"):
try: try:
text = data_json["choices"][0]["delta"].get("content", "") if len(data_json["choices"]) > 0:
if data_json["choices"][0].get("finish_reason", None): text = data_json["choices"][0]["delta"].get("content", "")
is_finished = True if data_json["choices"][0].get("finish_reason", None):
finish_reason = data_json["choices"][0]["finish_reason"] is_finished = True
finish_reason = data_json["choices"][0]["finish_reason"]
print_verbose(f"text: {text}; is_finished: {is_finished}; finish_reason: {finish_reason}") print_verbose(f"text: {text}; is_finished: {is_finished}; finish_reason: {finish_reason}")
return {"text": text, "is_finished": is_finished, "finish_reason": finish_reason} return {"text": text, "is_finished": is_finished, "finish_reason": finish_reason}
except: except:
@ -4725,7 +4734,7 @@ class CustomStreamWrapper:
e.message = str(e) e.message = str(e)
# LOG FAILURE - handle streaming failure logging in the _next_ object, remove `handle_failure` once it's deprecated # LOG FAILURE - handle streaming failure logging in the _next_ object, remove `handle_failure` once it's deprecated
threading.Thread(target=self.logging_obj.failure_handler, args=(e, traceback_exception)).start() threading.Thread(target=self.logging_obj.failure_handler, args=(e, traceback_exception)).start()
return exception_type(model=self.model, custom_llm_provider=self.custom_llm_provider, original_exception=e) raise exception_type(model=self.model, custom_llm_provider=self.custom_llm_provider, original_exception=e)
## needs to handle the empty string case (even starting chunk can be an empty string) ## needs to handle the empty string case (even starting chunk can be an empty string)
def __next__(self): def __next__(self):
@ -4746,7 +4755,7 @@ class CustomStreamWrapper:
raise # Re-raise StopIteration raise # Re-raise StopIteration
except Exception as e: except Exception as e:
# Handle other exceptions if needed # Handle other exceptions if needed
pass raise e