forked from phoenix/litellm-mirror
Merge pull request #3506 from BerriAI/litellm_reintegrate_langfuse_url_slack_alert
feat(slack_alerting.py): reintegrate langfuse trace url for slack alerts
This commit is contained in:
Krish Dholakia
GitHub
commit
2aaaa5e1b4
6 changed files with 164 additions and 122 deletions
|
|
@ -3164,7 +3164,9 @@ def data_generator(response):
|
|||
yield f"data: {json.dumps(chunk)}\n\n"
|
||||
|
||||
|
||||
async def async_data_generator(response, user_api_key_dict):
|
||||
async def async_data_generator(
|
||||
response, user_api_key_dict: UserAPIKeyAuth, request_data: dict
|
||||
):
|
||||
verbose_proxy_logger.debug("inside generator")
|
||||
try:
|
||||
start_time = time.time()
|
||||
|
|
@ -3181,7 +3183,9 @@ async def async_data_generator(response, user_api_key_dict):
|
|||
except Exception as e:
|
||||
traceback.print_exc()
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
original_exception=e,
|
||||
request_data=request_data,
|
||||
)
|
||||
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`"
|
||||
|
|
@ -3206,8 +3210,14 @@ async def async_data_generator(response, user_api_key_dict):
|
|||
yield f"data: {error_returned}\n\n"
|
||||
|
||||
|
||||
def select_data_generator(response, user_api_key_dict):
|
||||
return async_data_generator(response=response, user_api_key_dict=user_api_key_dict)
|
||||
def select_data_generator(
|
||||
response, user_api_key_dict: UserAPIKeyAuth, request_data: dict
|
||||
):
|
||||
return async_data_generator(
|
||||
response=response,
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
request_data=request_data,
|
||||
)
|
||||
|
||||
|
||||
def get_litellm_model_info(model: dict = {}):
|
||||
|
|
@ -3502,9 +3512,8 @@ async def chat_completion(
|
|||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
):
|
||||
global general_settings, user_debug, proxy_logging_obj, llm_model_list
|
||||
data = {}
|
||||
try:
|
||||
# async with llm_router.sem
|
||||
data = {}
|
||||
body = await request.body()
|
||||
body_str = body.decode()
|
||||
try:
|
||||
|
|
@ -3695,7 +3704,9 @@ async def chat_completion(
|
|||
"x-litellm-model-api-base": api_base,
|
||||
}
|
||||
selected_data_generator = select_data_generator(
|
||||
response=response, user_api_key_dict=user_api_key_dict
|
||||
response=response,
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
request_data=data,
|
||||
)
|
||||
return StreamingResponse(
|
||||
selected_data_generator,
|
||||
|
|
@ -3717,7 +3728,7 @@ async def chat_completion(
|
|||
data["litellm_status"] = "fail" # used for alerting
|
||||
traceback.print_exc()
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
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`"
|
||||
|
|
@ -3879,7 +3890,9 @@ async def completion(
|
|||
"x-litellm-model-id": model_id,
|
||||
}
|
||||
selected_data_generator = select_data_generator(
|
||||
response=response, user_api_key_dict=user_api_key_dict
|
||||
response=response,
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
request_data=data,
|
||||
)
|
||||
|
||||
return StreamingResponse(
|
||||
|
|
@ -3932,6 +3945,7 @@ async def embeddings(
|
|||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
):
|
||||
global proxy_logging_obj
|
||||
data: Any = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
body = await request.body()
|
||||
|
|
@ -4077,7 +4091,7 @@ async def embeddings(
|
|||
except Exception as e:
|
||||
data["litellm_status"] = "fail" # used for alerting
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
|
|
@ -4114,6 +4128,7 @@ async def image_generation(
|
|||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
):
|
||||
global proxy_logging_obj
|
||||
data = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
body = await request.body()
|
||||
|
|
@ -4233,7 +4248,7 @@ async def image_generation(
|
|||
except Exception as e:
|
||||
data["litellm_status"] = "fail" # used for alerting
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
|
|
@ -4274,10 +4289,11 @@ async def audio_transcriptions(
|
|||
https://platform.openai.com/docs/api-reference/audio/createTranscription?lang=curl
|
||||
"""
|
||||
global proxy_logging_obj
|
||||
data: Dict = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
form_data = await request.form()
|
||||
data: Dict = {key: value for key, value in form_data.items() if key != "file"}
|
||||
data = {key: value for key, value in form_data.items() if key != "file"}
|
||||
|
||||
# Include original request and headers in the data
|
||||
data["proxy_server_request"] = { # type: ignore
|
||||
|
|
@ -4412,7 +4428,7 @@ async def audio_transcriptions(
|
|||
except Exception as e:
|
||||
data["litellm_status"] = "fail" # used for alerting
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
|
|
@ -4461,6 +4477,7 @@ async def moderations(
|
|||
```
|
||||
"""
|
||||
global proxy_logging_obj
|
||||
data: Dict = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
body = await request.body()
|
||||
|
|
@ -4574,7 +4591,7 @@ async def moderations(
|
|||
except Exception as e:
|
||||
data["litellm_status"] = "fail" # used for alerting
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
|
|
@ -8006,8 +8023,8 @@ async def async_queue_request(
|
|||
|
||||
Now using a FastAPI background task + /chat/completions compatible endpoint
|
||||
"""
|
||||
data = {}
|
||||
try:
|
||||
data = {}
|
||||
data = await request.json() # type: ignore
|
||||
|
||||
# Include original request and headers in the data
|
||||
|
|
@ -8072,7 +8089,9 @@ async def async_queue_request(
|
|||
): # use generate_responses to stream responses
|
||||
return StreamingResponse(
|
||||
async_data_generator(
|
||||
user_api_key_dict=user_api_key_dict, response=response
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
response=response,
|
||||
request_data=data,
|
||||
),
|
||||
media_type="text/event-stream",
|
||||
)
|
||||
|
|
@ -8080,7 +8099,7 @@ async def async_queue_request(
|
|||
return response
|
||||
except Exception as e:
|
||||
await proxy_logging_obj.post_call_failure_hook(
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e
|
||||
user_api_key_dict=user_api_key_dict, original_exception=e, request_data=data
|
||||
)
|
||||
if isinstance(e, HTTPException):
|
||||
raise ProxyException(
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue