phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-26 03:04:13 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix(openai.py): switch back to using requests instead of httpx

Browse source
This commit is contained in:
Krrish Dholakia 2023-11-15 18:25:14 -08:00
parent 5990d046c4
commit fc8b54eb73
4 changed files with 3 additions and 37 deletions
Download patch file Download diff file Expand all files Collapse all files

2
litellm/__init__.py
View file

@ -19,7 +19,7 @@ telemetry = True
max_tokens = 256 # OpenAI Defaults max_tokens = 256 # OpenAI Defaults
drop_params = False drop_params = False
retry = True retry = True
request_timeout: Optional[float] = 600 request_timeout: Optional[float] = 6000
api_key: Optional[str] = None api_key: Optional[str] = None
openai_key: Optional[str] = None openai_key: Optional[str] = None
azure_key: Optional[str] = None azure_key: Optional[str] = None

5
litellm/llms/openai.py
View file

@ -4,7 +4,7 @@ import httpx
from .base import BaseLLM from .base import BaseLLM
from litellm.utils import ModelResponse, Choices, Message, CustomStreamWrapper, convert_to_model_response_object, Usage from litellm.utils import ModelResponse, Choices, Message, CustomStreamWrapper, convert_to_model_response_object, Usage
from typing import Callable, Optional from typing import Callable, Optional
import aiohttp import aiohttp, requests
import litellm import litellm
class OpenAIError(Exception): class OpenAIError(Exception):
@ -356,7 +356,7 @@ class OpenAIChatCompletion(BaseLLM):
additional_args={"complete_input_dict": data}, additional_args={"complete_input_dict": data},
) )
## COMPLETION CALL ## COMPLETION CALL
response = httpx.post( response = requests.post(
api_base, headers=headers, json=data, timeout=litellm.request_timeout api_base, headers=headers, json=data, timeout=litellm.request_timeout
) )
## LOGGING ## LOGGING
@ -488,7 +488,6 @@ class OpenAITextCompletion(BaseLLM):
url=f"{api_base}", url=f"{api_base}",
json=data, json=data,
headers=headers, headers=headers,
timeout=litellm.request_timeout
) )
if response.status_code != 200: if response.status_code != 200:
raise OpenAIError(status_code=response.status_code, message=response.text) raise OpenAIError(status_code=response.status_code, message=response.text)

5
litellm/tests/test_custom_logger.py
View file

@ -38,11 +38,6 @@ def test_chat_openai():
for chunk in response: for chunk in response:
# print(chunk) # print(chunk)
continue continue
response = completion(model="gpt-3.5-turbo",
messages=[{
"role": "user",
"content": "Hi 👋 - i'm openai"
}])
# print(response) # print(response)
except Exception as e: except Exception as e:

28
litellm/tests/test_model_fallback.py
View file

@ -1,28 +0,0 @@
#### What this tests ####
# This tests error handling + logging (esp. for sentry breadcrumbs)
import sys, os
import traceback
sys.path.insert(
0, os.path.abspath("../..")
) # Adds the parent directory to the system path
import litellm
from litellm import embedding, completion
litellm.success_callback = ["posthog"]
litellm.failure_callback = ["sentry", "posthog"]
litellm.set_verbose = True
model_fallback_list = ["claude-instant-1", "gpt-3.5-turbo", "chatgpt-test"]
user_message = "Hello, how are you?"
messages = [{"content": user_message, "role": "user"}]
for model in model_fallback_list:
try:
response = embedding(model="text-embedding-ada-002", input=[user_message])
response = completion(model=model, messages=messages)
except Exception as e:
print(f"error occurred: {traceback.format_exc()}")