mirror of
https://github.com/BerriAI/litellm.git
synced 2025-04-25 18:54:30 +00:00
Merge branch 'main' into litellm_batch_completions
This commit is contained in:
Krish Dholakia
GitHub
commit
e838bd1c79
10 changed files with 1780 additions and 45 deletions
|
|
@ -100,6 +100,13 @@ from litellm.proxy.utils import (
|
|||
encrypt_value,
|
||||
decrypt_value,
|
||||
)
|
||||
from litellm import (
|
||||
CreateBatchRequest,
|
||||
RetrieveBatchRequest,
|
||||
ListBatchRequest,
|
||||
CancelBatchRequest,
|
||||
CreateFileRequest,
|
||||
)
|
||||
from litellm.proxy.secret_managers.google_kms import load_google_kms
|
||||
from litellm.proxy.secret_managers.aws_secret_manager import load_aws_secret_manager
|
||||
import pydantic
|
||||
|
|
@ -142,6 +149,7 @@ from fastapi import (
|
|||
Request,
|
||||
HTTPException,
|
||||
status,
|
||||
Path,
|
||||
Depends,
|
||||
Header,
|
||||
Response,
|
||||
|
|
@ -499,7 +507,7 @@ async def user_api_key_auth(
|
|||
|
||||
if route in LiteLLMRoutes.public_routes.value:
|
||||
# check if public endpoint
|
||||
return UserAPIKeyAuth()
|
||||
return UserAPIKeyAuth(user_role="app_owner")
|
||||
|
||||
if general_settings.get("enable_jwt_auth", False) == True:
|
||||
is_jwt = jwt_handler.is_jwt(token=api_key)
|
||||
|
|
@ -1391,7 +1399,9 @@ async def user_api_key_auth(
|
|||
api_key=api_key, user_role="app_owner", **valid_token_dict
|
||||
)
|
||||
else:
|
||||
return UserAPIKeyAuth(api_key=api_key, **valid_token_dict)
|
||||
return UserAPIKeyAuth(
|
||||
api_key=api_key, user_role="app_owner", **valid_token_dict
|
||||
)
|
||||
else:
|
||||
raise Exception()
|
||||
except Exception as e:
|
||||
|
|
@ -5042,6 +5052,447 @@ async def audio_transcriptions(
|
|||
)
|
||||
|
||||
|
||||
######################################################################
|
||||
|
||||
# /v1/batches Endpoints
|
||||
|
||||
|
||||
######################################################################
|
||||
@router.post(
|
||||
"/v1/batches",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["batch"],
|
||||
)
|
||||
@router.post(
|
||||
"/batches",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["batch"],
|
||||
)
|
||||
async def create_batch(
|
||||
request: Request,
|
||||
fastapi_response: Response,
|
||||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
):
|
||||
"""
|
||||
Create large batches of API requests for asynchronous processing.
|
||||
This is the equivalent of POST https://api.openai.com/v1/batch
|
||||
Supports Identical Params as: https://platform.openai.com/docs/api-reference/batch
|
||||
|
||||
Example Curl
|
||||
```
|
||||
curl http://localhost:4000/v1/batches \
|
||||
-H "Authorization: Bearer sk-1234" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"input_file_id": "file-abc123",
|
||||
"endpoint": "/v1/chat/completions",
|
||||
"completion_window": "24h"
|
||||
}'
|
||||
```
|
||||
"""
|
||||
global proxy_logging_obj
|
||||
data: Dict = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
form_data = await request.form()
|
||||
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
|
||||
"url": str(request.url),
|
||||
"method": request.method,
|
||||
"headers": dict(request.headers),
|
||||
"body": copy.copy(data), # use copy instead of deepcopy
|
||||
}
|
||||
|
||||
if data.get("user", None) is None and user_api_key_dict.user_id is not None:
|
||||
data["user"] = user_api_key_dict.user_id
|
||||
|
||||
if "metadata" not in data:
|
||||
data["metadata"] = {}
|
||||
data["metadata"]["user_api_key"] = user_api_key_dict.api_key
|
||||
data["metadata"]["user_api_key_metadata"] = user_api_key_dict.metadata
|
||||
_headers = dict(request.headers)
|
||||
_headers.pop(
|
||||
"authorization", None
|
||||
) # do not store the original `sk-..` api key in the db
|
||||
data["metadata"]["headers"] = _headers
|
||||
data["metadata"]["user_api_key_alias"] = getattr(
|
||||
user_api_key_dict, "key_alias", None
|
||||
)
|
||||
data["metadata"]["user_api_key_user_id"] = user_api_key_dict.user_id
|
||||
data["metadata"]["user_api_key_team_id"] = getattr(
|
||||
user_api_key_dict, "team_id", None
|
||||
)
|
||||
data["metadata"]["global_max_parallel_requests"] = general_settings.get(
|
||||
"global_max_parallel_requests", None
|
||||
)
|
||||
data["metadata"]["user_api_key_team_alias"] = getattr(
|
||||
user_api_key_dict, "team_alias", None
|
||||
)
|
||||
data["metadata"]["endpoint"] = str(request.url)
|
||||
|
||||
### TEAM-SPECIFIC PARAMS ###
|
||||
if user_api_key_dict.team_id is not None:
|
||||
team_config = await proxy_config.load_team_config(
|
||||
team_id=user_api_key_dict.team_id
|
||||
)
|
||||
if len(team_config) == 0:
|
||||
pass
|
||||
else:
|
||||
team_id = team_config.pop("team_id", None)
|
||||
data["metadata"]["team_id"] = team_id
|
||||
data = {
|
||||
**team_config,
|
||||
**data,
|
||||
} # add the team-specific configs to the completion call
|
||||
|
||||
_create_batch_data = CreateBatchRequest(**data)
|
||||
|
||||
# for now use custom_llm_provider=="openai" -> this will change as LiteLLM adds more providers for acreate_batch
|
||||
response = await litellm.acreate_batch(
|
||||
custom_llm_provider="openai", **_create_batch_data
|
||||
)
|
||||
|
||||
### ALERTING ###
|
||||
data["litellm_status"] = "success" # used for alerting
|
||||
|
||||
### RESPONSE HEADERS ###
|
||||
hidden_params = getattr(response, "_hidden_params", {}) or {}
|
||||
model_id = hidden_params.get("model_id", None) or ""
|
||||
cache_key = hidden_params.get("cache_key", None) or ""
|
||||
api_base = hidden_params.get("api_base", None) or ""
|
||||
|
||||
fastapi_response.headers.update(
|
||||
get_custom_headers(
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
model_id=model_id,
|
||||
cache_key=cache_key,
|
||||
api_base=api_base,
|
||||
version=version,
|
||||
model_region=getattr(user_api_key_dict, "allowed_model_region", ""),
|
||||
)
|
||||
)
|
||||
|
||||
return response
|
||||
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, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", str(e.detail)),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", status.HTTP_400_BAD_REQUEST),
|
||||
)
|
||||
else:
|
||||
error_traceback = traceback.format_exc()
|
||||
error_msg = f"{str(e)}"
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", error_msg),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", 500),
|
||||
)
|
||||
|
||||
|
||||
@router.get(
|
||||
"/v1/batches{batch_id}",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["Batch"],
|
||||
)
|
||||
@router.get(
|
||||
"/batches{batch_id}",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["Batch"],
|
||||
)
|
||||
async def retrieve_batch(
|
||||
request: Request,
|
||||
fastapi_response: Response,
|
||||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
batch_id: str = Path(
|
||||
title="Batch ID to retrieve", description="The ID of the batch to retrieve"
|
||||
),
|
||||
):
|
||||
"""
|
||||
Retrieves a batch.
|
||||
This is the equivalent of GET https://api.openai.com/v1/batches/{batch_id}
|
||||
Supports Identical Params as: https://platform.openai.com/docs/api-reference/batch/retrieve
|
||||
|
||||
Example Curl
|
||||
```
|
||||
curl http://localhost:4000/v1/batches/batch_abc123 \
|
||||
-H "Authorization: Bearer sk-1234" \
|
||||
-H "Content-Type: application/json" \
|
||||
|
||||
```
|
||||
"""
|
||||
global proxy_logging_obj
|
||||
data: Dict = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
form_data = await request.form()
|
||||
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
|
||||
"url": str(request.url),
|
||||
"method": request.method,
|
||||
"headers": dict(request.headers),
|
||||
"body": copy.copy(data), # use copy instead of deepcopy
|
||||
}
|
||||
|
||||
if data.get("user", None) is None and user_api_key_dict.user_id is not None:
|
||||
data["user"] = user_api_key_dict.user_id
|
||||
|
||||
if "metadata" not in data:
|
||||
data["metadata"] = {}
|
||||
data["metadata"]["user_api_key"] = user_api_key_dict.api_key
|
||||
data["metadata"]["user_api_key_metadata"] = user_api_key_dict.metadata
|
||||
_headers = dict(request.headers)
|
||||
_headers.pop(
|
||||
"authorization", None
|
||||
) # do not store the original `sk-..` api key in the db
|
||||
data["metadata"]["headers"] = _headers
|
||||
data["metadata"]["user_api_key_alias"] = getattr(
|
||||
user_api_key_dict, "key_alias", None
|
||||
)
|
||||
data["metadata"]["user_api_key_user_id"] = user_api_key_dict.user_id
|
||||
data["metadata"]["user_api_key_team_id"] = getattr(
|
||||
user_api_key_dict, "team_id", None
|
||||
)
|
||||
data["metadata"]["global_max_parallel_requests"] = general_settings.get(
|
||||
"global_max_parallel_requests", None
|
||||
)
|
||||
data["metadata"]["user_api_key_team_alias"] = getattr(
|
||||
user_api_key_dict, "team_alias", None
|
||||
)
|
||||
data["metadata"]["endpoint"] = str(request.url)
|
||||
|
||||
### TEAM-SPECIFIC PARAMS ###
|
||||
if user_api_key_dict.team_id is not None:
|
||||
team_config = await proxy_config.load_team_config(
|
||||
team_id=user_api_key_dict.team_id
|
||||
)
|
||||
if len(team_config) == 0:
|
||||
pass
|
||||
else:
|
||||
team_id = team_config.pop("team_id", None)
|
||||
data["metadata"]["team_id"] = team_id
|
||||
data = {
|
||||
**team_config,
|
||||
**data,
|
||||
} # add the team-specific configs to the completion call
|
||||
|
||||
_retrieve_batch_request = RetrieveBatchRequest(
|
||||
batch_id=batch_id,
|
||||
)
|
||||
|
||||
# for now use custom_llm_provider=="openai" -> this will change as LiteLLM adds more providers for acreate_batch
|
||||
response = await litellm.aretrieve_batch(
|
||||
custom_llm_provider="openai", **_retrieve_batch_request
|
||||
)
|
||||
|
||||
### ALERTING ###
|
||||
data["litellm_status"] = "success" # used for alerting
|
||||
|
||||
### RESPONSE HEADERS ###
|
||||
hidden_params = getattr(response, "_hidden_params", {}) or {}
|
||||
model_id = hidden_params.get("model_id", None) or ""
|
||||
cache_key = hidden_params.get("cache_key", None) or ""
|
||||
api_base = hidden_params.get("api_base", None) or ""
|
||||
|
||||
fastapi_response.headers.update(
|
||||
get_custom_headers(
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
model_id=model_id,
|
||||
cache_key=cache_key,
|
||||
api_base=api_base,
|
||||
version=version,
|
||||
model_region=getattr(user_api_key_dict, "allowed_model_region", ""),
|
||||
)
|
||||
)
|
||||
|
||||
return response
|
||||
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, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", str(e.detail)),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", status.HTTP_400_BAD_REQUEST),
|
||||
)
|
||||
else:
|
||||
error_traceback = traceback.format_exc()
|
||||
error_msg = f"{str(e)}"
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", error_msg),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", 500),
|
||||
)
|
||||
|
||||
|
||||
######################################################################
|
||||
|
||||
# END OF /v1/batches Endpoints Implementation
|
||||
|
||||
######################################################################
|
||||
|
||||
|
||||
######################################################################
|
||||
|
||||
# /v1/files Endpoints
|
||||
|
||||
|
||||
######################################################################
|
||||
@router.post(
|
||||
"/v1/files",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["files"],
|
||||
)
|
||||
@router.post(
|
||||
"/files",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
tags=["files"],
|
||||
)
|
||||
async def create_file(
|
||||
request: Request,
|
||||
fastapi_response: Response,
|
||||
user_api_key_dict: UserAPIKeyAuth = Depends(user_api_key_auth),
|
||||
):
|
||||
"""
|
||||
Upload a file that can be used across - Assistants API, Batch API
|
||||
This is the equivalent of POST https://api.openai.com/v1/files
|
||||
|
||||
Supports Identical Params as: https://platform.openai.com/docs/api-reference/files/create
|
||||
|
||||
Example Curl
|
||||
```
|
||||
curl https://api.openai.com/v1/files \
|
||||
-H "Authorization: Bearer sk-1234" \
|
||||
-F purpose="batch" \
|
||||
-F file="@mydata.jsonl"
|
||||
|
||||
```
|
||||
"""
|
||||
global proxy_logging_obj
|
||||
data: Dict = {}
|
||||
try:
|
||||
# Use orjson to parse JSON data, orjson speeds up requests significantly
|
||||
form_data = await request.form()
|
||||
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
|
||||
"url": str(request.url),
|
||||
"method": request.method,
|
||||
"headers": dict(request.headers),
|
||||
"body": copy.copy(data), # use copy instead of deepcopy
|
||||
}
|
||||
|
||||
if data.get("user", None) is None and user_api_key_dict.user_id is not None:
|
||||
data["user"] = user_api_key_dict.user_id
|
||||
|
||||
if "metadata" not in data:
|
||||
data["metadata"] = {}
|
||||
data["metadata"]["user_api_key"] = user_api_key_dict.api_key
|
||||
data["metadata"]["user_api_key_metadata"] = user_api_key_dict.metadata
|
||||
_headers = dict(request.headers)
|
||||
_headers.pop(
|
||||
"authorization", None
|
||||
) # do not store the original `sk-..` api key in the db
|
||||
data["metadata"]["headers"] = _headers
|
||||
data["metadata"]["user_api_key_alias"] = getattr(
|
||||
user_api_key_dict, "key_alias", None
|
||||
)
|
||||
data["metadata"]["user_api_key_user_id"] = user_api_key_dict.user_id
|
||||
data["metadata"]["user_api_key_team_id"] = getattr(
|
||||
user_api_key_dict, "team_id", None
|
||||
)
|
||||
data["metadata"]["global_max_parallel_requests"] = general_settings.get(
|
||||
"global_max_parallel_requests", None
|
||||
)
|
||||
data["metadata"]["user_api_key_team_alias"] = getattr(
|
||||
user_api_key_dict, "team_alias", None
|
||||
)
|
||||
data["metadata"]["endpoint"] = str(request.url)
|
||||
|
||||
### TEAM-SPECIFIC PARAMS ###
|
||||
if user_api_key_dict.team_id is not None:
|
||||
team_config = await proxy_config.load_team_config(
|
||||
team_id=user_api_key_dict.team_id
|
||||
)
|
||||
if len(team_config) == 0:
|
||||
pass
|
||||
else:
|
||||
team_id = team_config.pop("team_id", None)
|
||||
data["metadata"]["team_id"] = team_id
|
||||
data = {
|
||||
**team_config,
|
||||
**data,
|
||||
} # add the team-specific configs to the completion call
|
||||
|
||||
_create_file_request = CreateFileRequest()
|
||||
|
||||
# for now use custom_llm_provider=="openai" -> this will change as LiteLLM adds more providers for acreate_batch
|
||||
response = await litellm.acreate_file(
|
||||
custom_llm_provider="openai", **_create_file_request
|
||||
)
|
||||
|
||||
### ALERTING ###
|
||||
data["litellm_status"] = "success" # used for alerting
|
||||
|
||||
### RESPONSE HEADERS ###
|
||||
hidden_params = getattr(response, "_hidden_params", {}) or {}
|
||||
model_id = hidden_params.get("model_id", None) or ""
|
||||
cache_key = hidden_params.get("cache_key", None) or ""
|
||||
api_base = hidden_params.get("api_base", None) or ""
|
||||
|
||||
fastapi_response.headers.update(
|
||||
get_custom_headers(
|
||||
user_api_key_dict=user_api_key_dict,
|
||||
model_id=model_id,
|
||||
cache_key=cache_key,
|
||||
api_base=api_base,
|
||||
version=version,
|
||||
model_region=getattr(user_api_key_dict, "allowed_model_region", ""),
|
||||
)
|
||||
)
|
||||
|
||||
return response
|
||||
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, request_data=data
|
||||
)
|
||||
traceback.print_exc()
|
||||
if isinstance(e, HTTPException):
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", str(e.detail)),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", status.HTTP_400_BAD_REQUEST),
|
||||
)
|
||||
else:
|
||||
error_traceback = traceback.format_exc()
|
||||
error_msg = f"{str(e)}"
|
||||
raise ProxyException(
|
||||
message=getattr(e, "message", error_msg),
|
||||
type=getattr(e, "type", "None"),
|
||||
param=getattr(e, "param", "None"),
|
||||
code=getattr(e, "status_code", 500),
|
||||
)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/v1/moderations",
|
||||
dependencies=[Depends(user_api_key_auth)],
|
||||
|
|
@ -9596,28 +10047,54 @@ async def model_streaming_metrics(
|
|||
startTime = startTime or datetime.now() - timedelta(days=7) # show over past week
|
||||
endTime = endTime or datetime.now()
|
||||
|
||||
sql_query = """
|
||||
SELECT
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
DATE_TRUNC('day', "startTime")::DATE AS day,
|
||||
AVG(EXTRACT(epoch FROM ("completionStartTime" - "startTime"))) AS time_to_first_token
|
||||
FROM
|
||||
"LiteLLM_SpendLogs"
|
||||
WHERE
|
||||
"startTime" BETWEEN $2::timestamp AND $3::timestamp
|
||||
AND "model_group" = $1 AND "cache_hit" != 'True'
|
||||
AND "completionStartTime" IS NOT NULL
|
||||
AND "completionStartTime" != "endTime"
|
||||
GROUP BY
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
day
|
||||
ORDER BY
|
||||
time_to_first_token DESC;
|
||||
"""
|
||||
is_same_day = startTime.date() == endTime.date()
|
||||
if is_same_day:
|
||||
sql_query = """
|
||||
SELECT
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
"startTime",
|
||||
request_id,
|
||||
EXTRACT(epoch FROM ("completionStartTime" - "startTime")) AS time_to_first_token
|
||||
FROM
|
||||
"LiteLLM_SpendLogs"
|
||||
WHERE
|
||||
"model_group" = $1 AND "cache_hit" != 'True'
|
||||
AND "completionStartTime" IS NOT NULL
|
||||
AND "completionStartTime" != "endTime"
|
||||
AND DATE("startTime") = DATE($2::timestamp)
|
||||
GROUP BY
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
request_id
|
||||
ORDER BY
|
||||
time_to_first_token DESC;
|
||||
"""
|
||||
else:
|
||||
sql_query = """
|
||||
SELECT
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
DATE_TRUNC('day', "startTime")::DATE AS day,
|
||||
AVG(EXTRACT(epoch FROM ("completionStartTime" - "startTime"))) AS time_to_first_token
|
||||
FROM
|
||||
"LiteLLM_SpendLogs"
|
||||
WHERE
|
||||
"startTime" BETWEEN $2::timestamp AND $3::timestamp
|
||||
AND "model_group" = $1 AND "cache_hit" != 'True'
|
||||
AND "completionStartTime" IS NOT NULL
|
||||
AND "completionStartTime" != "endTime"
|
||||
GROUP BY
|
||||
api_base,
|
||||
model_group,
|
||||
model,
|
||||
day
|
||||
ORDER BY
|
||||
time_to_first_token DESC;
|
||||
"""
|
||||
|
||||
_all_api_bases = set()
|
||||
db_response = await prisma_client.db.query_raw(
|
||||
|
|
@ -9628,10 +10105,19 @@ async def model_streaming_metrics(
|
|||
for model_data in db_response:
|
||||
_api_base = model_data["api_base"]
|
||||
_model = model_data["model"]
|
||||
_day = model_data["day"]
|
||||
time_to_first_token = model_data["time_to_first_token"]
|
||||
if _day not in _daily_entries:
|
||||
_daily_entries[_day] = {}
|
||||
unique_key = ""
|
||||
if is_same_day:
|
||||
_request_id = model_data["request_id"]
|
||||
unique_key = _request_id
|
||||
if _request_id not in _daily_entries:
|
||||
_daily_entries[_request_id] = {}
|
||||
else:
|
||||
_day = model_data["day"]
|
||||
unique_key = _day
|
||||
time_to_first_token = model_data["time_to_first_token"]
|
||||
if _day not in _daily_entries:
|
||||
_daily_entries[_day] = {}
|
||||
_combined_model_name = str(_model)
|
||||
if "https://" in _api_base:
|
||||
_combined_model_name = str(_api_base)
|
||||
|
|
@ -9639,7 +10125,8 @@ async def model_streaming_metrics(
|
|||
_combined_model_name = _combined_model_name.split("/openai/")[0]
|
||||
|
||||
_all_api_bases.add(_combined_model_name)
|
||||
_daily_entries[_day][_combined_model_name] = time_to_first_token
|
||||
|
||||
_daily_entries[unique_key][_combined_model_name] = time_to_first_token
|
||||
|
||||
"""
|
||||
each entry needs to be like this:
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue