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

Merge pull request #3655 from BerriAI/litellm_only_show_debug_info_server_side

Browse source 
Fix Proxy Server - only show API base, Model server log exceptions, not on client side
This commit is contained in:
Ishaan Jaff 2024-05-15 14:20:14 -07:00 • committed by GitHub
commit c00d232900
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
4 changed files with 199 additions and 54 deletions
Download patch file Download diff file Expand all files Collapse all files

117
litellm/exceptions.py
View file

@ -15,11 +15,19 @@ from typing import Optional
class AuthenticationError(openai.AuthenticationError): # type: ignore
def __init__(self, message, llm_provider, model, response: httpx.Response):
def __init__(
self,
message,
llm_provider,
model,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 401
self.message = message
self.llm_provider = llm_provider
self.model = model
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
@ -27,11 +35,19 @@ class AuthenticationError(openai.AuthenticationError): # type: ignore
# raise when invalid models passed, example gpt-8
class NotFoundError(openai.NotFoundError): # type: ignore
def __init__(self, message, model, llm_provider, response: httpx.Response):
def __init__(
self,
message,
model,
llm_provider,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 404
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
@ -39,12 +55,18 @@ class NotFoundError(openai.NotFoundError): # type: ignore
class BadRequestError(openai.BadRequestError): # type: ignore
def __init__(
self, message, model, llm_provider, response: Optional[httpx.Response] = None
self,
message,
model,
llm_provider,
response: Optional[httpx.Response] = None,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 400
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
response = response or httpx.Response(
status_code=self.status_code,
request=httpx.Request(
@ -57,18 +79,28 @@ class BadRequestError(openai.BadRequestError): # type: ignore
class UnprocessableEntityError(openai.UnprocessableEntityError): # type: ignore
def __init__(self, message, model, llm_provider, response: httpx.Response):
def __init__(
self,
message,
model,
llm_provider,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 422
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
class Timeout(openai.APITimeoutError): # type: ignore
def __init__(self, message, model, llm_provider):
def __init__(
self, message, model, llm_provider, litellm_debug_info: Optional[str] = None
):
request = httpx.Request(method="POST", url="https://api.openai.com/v1")
super().__init__(
request=request
@ -77,6 +109,7 @@ class Timeout(openai.APITimeoutError): # type: ignore
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
# custom function to convert to str
def __str__(self):
@ -84,22 +117,38 @@ class Timeout(openai.APITimeoutError): # type: ignore
class PermissionDeniedError(openai.PermissionDeniedError): # type:ignore
def __init__(self, message, llm_provider, model, response: httpx.Response):
def __init__(
self,
message,
llm_provider,
model,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 403
self.message = message
self.llm_provider = llm_provider
self.model = model
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
class RateLimitError(openai.RateLimitError): # type: ignore
def __init__(self, message, llm_provider, model, response: httpx.Response):
def __init__(
self,
message,
llm_provider,
model,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 429
self.message = message
self.llm_provider = llm_provider
self.modle = model
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
@ -107,11 +156,19 @@ class RateLimitError(openai.RateLimitError): # type: ignore
# sub class of rate limit error - meant to give more granularity for error handling context window exceeded errors
class ContextWindowExceededError(BadRequestError): # type: ignore
def __init__(self, message, model, llm_provider, response: httpx.Response):
def __init__(
self,
message,
model,
llm_provider,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 400
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
super().__init__(
message=self.message,
model=self.model, # type: ignore
@ -122,11 +179,19 @@ class ContextWindowExceededError(BadRequestError): # type: ignore
class ContentPolicyViolationError(BadRequestError): # type: ignore
# Error code: 400 - {'error': {'code': 'content_policy_violation', 'message': 'Your request was rejected as a result of our safety system. Image descriptions generated from your prompt may contain text that is not allowed by our safety system. If you believe this was done in error, your request may succeed if retried, or by adjusting your prompt.', 'param': None, 'type': 'invalid_request_error'}}
def __init__(self, message, model, llm_provider, response: httpx.Response):
def __init__(
self,
message,
model,
llm_provider,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 400
self.message = message
self.model = model
self.llm_provider = llm_provider
self.litellm_debug_info = litellm_debug_info
super().__init__(
message=self.message,
model=self.model, # type: ignore
@ -136,11 +201,19 @@ class ContentPolicyViolationError(BadRequestError): # type: ignore
class ServiceUnavailableError(openai.APIStatusError): # type: ignore
def __init__(self, message, llm_provider, model, response: httpx.Response):
def __init__(
self,
message,
llm_provider,
model,
response: httpx.Response,
litellm_debug_info: Optional[str] = None,
):
self.status_code = 503
self.message = message
self.llm_provider = llm_provider
self.model = model
self.litellm_debug_info = litellm_debug_info
super().__init__(
self.message, response=response, body=None
) # Call the base class constructor with the parameters it needs
@ -149,33 +222,51 @@ class ServiceUnavailableError(openai.APIStatusError): # type: ignore
# raise this when the API returns an invalid response object - https://github.com/openai/openai-python/blob/1be14ee34a0f8e42d3f9aa5451aa4cb161f1781f/openai/api_requestor.py#L401
class APIError(openai.APIError): # type: ignore
def __init__(
self, status_code, message, llm_provider, model, request: httpx.Request
self,
status_code,
message,
llm_provider,
model,
request: httpx.Request,
litellm_debug_info: Optional[str] = None,
):
self.status_code = status_code
self.message = message
self.llm_provider = llm_provider
self.model = model
self.litellm_debug_info = litellm_debug_info
super().__init__(self.message, request=request, body=None) # type: ignore
# raised if an invalid request (not get, delete, put, post) is made
class APIConnectionError(openai.APIConnectionError): # type: ignore
def __init__(self, message, llm_provider, model, request: httpx.Request):
def __init__(
self,
message,
llm_provider,
model,
request: httpx.Request,
litellm_debug_info: Optional[str] = None,
):
self.message = message
self.llm_provider = llm_provider
self.model = model
self.status_code = 500
self.litellm_debug_info = litellm_debug_info
super().__init__(message=self.message, request=request)
# raised if an invalid request (not get, delete, put, post) is made
class APIResponseValidationError(openai.APIResponseValidationError): # type: ignore
def __init__(self, message, llm_provider, model):
def __init__(
self, message, llm_provider, model, litellm_debug_info: Optional[str] = None
):
self.message = message
self.llm_provider = llm_provider
self.model = model
request = httpx.Request(method="POST", url="https://api.openai.com/v1")
response = httpx.Response(status_code=500, request=request)
self.litellm_debug_info = litellm_debug_info
super().__init__(response=response, body=None, message=message)

15
litellm/proxy/proxy_server.py
View file

@ -3857,8 +3857,11 @@ async def chat_completion(
await proxy_logging_obj.post_call_failure_hook(
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
)
litellm_debug_info = getattr(e, "litellm_debug_info", "")
verbose_proxy_logger.debug(
f"\033[1;31mAn error occurred: {e}\n\n Debug this by setting `--debug`, e.g. `litellm --model gpt-3.5-turbo --debug`"
"\033[1;31mAn error occurred: %s %s\n\n Debug this by setting `--debug`, e.g. `litellm --model gpt-3.5-turbo --debug`",
e,
litellm_debug_info,
)
router_model_names = llm_router.model_names if llm_router is not None else []
if user_debug:
@ -4049,9 +4052,11 @@ async def completion(
except Exception as e:
data["litellm_status"] = "fail" # used for alerting
verbose_proxy_logger.debug("EXCEPTION RAISED IN PROXY MAIN.PY")
litellm_debug_info = getattr(e, "litellm_debug_info", "")
verbose_proxy_logger.debug(
"\033[1;31mAn error occurred: %s\n\n Debug this by setting `--debug`, e.g. `litellm --model gpt-3.5-turbo --debug`",
"\033[1;31mAn error occurred: %s %s\n\n Debug this by setting `--debug`, e.g. `litellm --model gpt-3.5-turbo --debug`",
e,
litellm_debug_info,
)
traceback.print_exc()
error_traceback = traceback.format_exc()
@ -4254,6 +4259,12 @@ async def embeddings(
await proxy_logging_obj.post_call_failure_hook(
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
)
litellm_debug_info = getattr(e, "litellm_debug_info", "")
verbose_proxy_logger.debug(
"\033[1;31mAn error occurred: %s %s\n\n Debug this by setting `--debug`, e.g. `litellm --model gpt-3.5-turbo --debug`",
e,
litellm_debug_info,
)
traceback.print_exc()
if isinstance(e, HTTPException):
raise ProxyException(

7
litellm/proxy/utils.py
View file

@ -418,9 +418,14 @@ class ProxyLogging:
Related issue - https://github.com/BerriAI/litellm/issues/3395
"""
litellm_debug_info = getattr(original_exception, "litellm_debug_info", None)
exception_str = str(original_exception)
if litellm_debug_info is not None:
exception_str += litellm_debug_info
asyncio.create_task(
self.alerting_handler(
message=f"LLM API call failed: {str(original_exception)}",
message=f"LLM API call failed: {exception_str}",
level="High",
alert_type="llm_exceptions",
request_data=request_data,

114
litellm/utils.py
View file

@ -8194,9 +8194,10 @@ def exception_type(
if "Request Timeout Error" in error_str or "Request timed out" in error_str:
exception_mapping_worked = True
raise Timeout(
message=f"APITimeoutError - Request timed out. {extra_information} \n error_str: {error_str}",
message=f"APITimeoutError - Request timed out. \nerror_str: {error_str}",
model=model,
llm_provider=custom_llm_provider,
litellm_debug_info=extra_information,
)
if (
@ -8226,10 +8227,11 @@ def exception_type(
if "This model's maximum context length is" in error_str:
exception_mapping_worked = True
raise ContextWindowExceededError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif (
"invalid_request_error" in error_str
@ -8237,10 +8239,11 @@ def exception_type(
):
exception_mapping_worked = True
raise NotFoundError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif (
"invalid_request_error" in error_str
@ -8248,10 +8251,11 @@ def exception_type(
):
exception_mapping_worked = True
raise ContentPolicyViolationError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif (
"invalid_request_error" in error_str
@ -8259,17 +8263,19 @@ def exception_type(
):
exception_mapping_worked = True
raise BadRequestError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif "Request too large" in error_str:
raise RateLimitError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif (
"The api_key client option must be set either by passing api_key to the client or by setting the OPENAI_API_KEY environment variable"
@ -8277,10 +8283,11 @@ def exception_type(
):
exception_mapping_worked = True
raise AuthenticationError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif "Mistral API raised a streaming error" in error_str:
exception_mapping_worked = True
@ -8289,82 +8296,92 @@ def exception_type(
)
raise APIError(
status_code=500,
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
request=_request,
litellm_debug_info=extra_information,
)
elif hasattr(original_exception, "status_code"):
exception_mapping_worked = True
if original_exception.status_code == 401:
exception_mapping_worked = True
raise AuthenticationError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 404:
exception_mapping_worked = True
raise NotFoundError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 408:
exception_mapping_worked = True
raise Timeout(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 422:
exception_mapping_worked = True
raise BadRequestError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 429:
exception_mapping_worked = True
raise RateLimitError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 503:
exception_mapping_worked = True
raise ServiceUnavailableError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif original_exception.status_code == 504: # gateway timeout error
exception_mapping_worked = True
raise Timeout(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
model=model,
llm_provider=custom_llm_provider,
litellm_debug_info=extra_information,
)
else:
exception_mapping_worked = True
raise APIError(
status_code=original_exception.status_code,
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
request=original_exception.request,
litellm_debug_info=extra_information,
)
else:
# if no status code then it is an APIConnectionError: https://github.com/openai/openai-python#handling-errors
raise APIConnectionError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider=custom_llm_provider,
model=model,
litellm_debug_info=extra_information,
request=httpx.Request(
method="POST", url="https://api.openai.com/v1/"
),
@ -8666,10 +8683,11 @@ def exception_type(
):
exception_mapping_worked = True
raise BadRequestError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
model=model,
llm_provider="vertex_ai",
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif (
"None Unknown Error." in error_str
@ -8677,26 +8695,29 @@ def exception_type(
):
exception_mapping_worked = True
raise APIError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
status_code=500,
model=model,
llm_provider="vertex_ai",
request=original_exception.request,
litellm_debug_info=extra_information,
)
elif "403" in error_str:
exception_mapping_worked = True
raise BadRequestError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
model=model,
llm_provider="vertex_ai",
response=original_exception.response,
litellm_debug_info=extra_information,
)
elif "The response was blocked." in error_str:
exception_mapping_worked = True
raise UnprocessableEntityError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
model=model,
llm_provider="vertex_ai",
litellm_debug_info=extra_information,
response=httpx.Response(
status_code=429,
request=httpx.Request(
@ -8713,9 +8734,10 @@ def exception_type(
):
exception_mapping_worked = True
raise RateLimitError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
model=model,
llm_provider="vertex_ai",
litellm_debug_info=extra_information,
response=httpx.Response(
status_code=429,
request=httpx.Request(
@ -8728,18 +8750,20 @@ def exception_type(
if original_exception.status_code == 400:
exception_mapping_worked = True
raise BadRequestError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
model=model,
llm_provider="vertex_ai",
litellm_debug_info=extra_information,
response=original_exception.response,
)
if original_exception.status_code == 500:
exception_mapping_worked = True
raise APIError(
message=f"VertexAIException - {error_str} {extra_information}",
message=f"VertexAIException - {error_str}",
status_code=500,
model=model,
llm_provider="vertex_ai",
litellm_debug_info=extra_information,
request=original_exception.request,
)
elif custom_llm_provider == "palm" or custom_llm_provider == "gemini":
@ -9340,25 +9364,28 @@ def exception_type(
exception_mapping_worked = True
raise APIError(
status_code=500,
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
request=httpx.Request(method="POST", url="https://openai.com/"),
)
elif "This model's maximum context length is" in error_str:
exception_mapping_worked = True
raise ContextWindowExceededError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif "DeploymentNotFound" in error_str:
exception_mapping_worked = True
raise NotFoundError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif (
@ -9370,17 +9397,19 @@ def exception_type(
):
exception_mapping_worked = True
raise ContentPolicyViolationError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif "invalid_request_error" in error_str:
exception_mapping_worked = True
raise BadRequestError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif (
@ -9389,9 +9418,10 @@ def exception_type(
):
exception_mapping_worked = True
raise AuthenticationError(
message=f"{exception_provider} - {original_exception.message} {extra_information}",
message=f"{exception_provider} - {original_exception.message}",
llm_provider=custom_llm_provider,
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif hasattr(original_exception, "status_code"):
@ -9399,55 +9429,62 @@ def exception_type(
if original_exception.status_code == 401:
exception_mapping_worked = True
raise AuthenticationError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif original_exception.status_code == 408:
exception_mapping_worked = True
raise Timeout(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
model=model,
litellm_debug_info=extra_information,
llm_provider="azure",
)
if original_exception.status_code == 422:
exception_mapping_worked = True
raise BadRequestError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
model=model,
llm_provider="azure",
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif original_exception.status_code == 429:
exception_mapping_worked = True
raise RateLimitError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
model=model,
llm_provider="azure",
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif original_exception.status_code == 503:
exception_mapping_worked = True
raise ServiceUnavailableError(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
model=model,
llm_provider="azure",
litellm_debug_info=extra_information,
response=original_exception.response,
)
elif original_exception.status_code == 504: # gateway timeout error
exception_mapping_worked = True
raise Timeout(
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
model=model,
litellm_debug_info=extra_information,
llm_provider="azure",
)
else:
exception_mapping_worked = True
raise APIError(
status_code=original_exception.status_code,
message=f"AzureException - {original_exception.message} {extra_information}",
message=f"AzureException - {original_exception.message}",
llm_provider="azure",
litellm_debug_info=extra_information,
model=model,
request=httpx.Request(
method="POST", url="https://openai.com/"
@ -9456,9 +9493,10 @@ def exception_type(
else:
# if no status code then it is an APIConnectionError: https://github.com/openai/openai-python#handling-errors
raise APIConnectionError(
message=f"{exception_provider} - {message} {extra_information}",
message=f"{exception_provider} - {message}",
llm_provider="azure",
model=model,
litellm_debug_info=extra_information,
request=httpx.Request(method="POST", url="https://openai.com/"),
)
if (