mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-06 08:16:43 +00:00
chore(techdebt): clear fresh debt from the 2026-08-31 window
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
This commit is contained in:
Devin AI
parent
ec3f8183c3
commit
d4fc54a11d
11 changed files with 35 additions and 76 deletions
|
|
@ -57,7 +57,7 @@
|
|||
"limit": 5607
|
||||
},
|
||||
"reportMissingTypeArgument": {
|
||||
"limit": 15310
|
||||
"limit": 15308
|
||||
},
|
||||
"reportMissingTypeStubs": {
|
||||
"limit": 40
|
||||
|
|
@ -105,13 +105,13 @@
|
|||
"limit": 109
|
||||
},
|
||||
"reportUnknownMemberType": {
|
||||
"limit": 38368
|
||||
"limit": 38367
|
||||
},
|
||||
"reportUnknownParameterType": {
|
||||
"limit": 19633
|
||||
},
|
||||
"reportUnknownVariableType": {
|
||||
"limit": 29908
|
||||
"limit": 29906
|
||||
},
|
||||
"reportUnnecessaryCast": {
|
||||
"limit": 111
|
||||
|
|
|
|||
|
|
@ -8,6 +8,7 @@ Based on official GigaChat SDK authentication flow.
|
|||
import time
|
||||
import uuid
|
||||
from collections.abc import Mapping
|
||||
from types import MappingProxyType
|
||||
from typing import Final
|
||||
|
||||
import httpx
|
||||
|
|
@ -32,8 +33,8 @@ GIGACHAT_SCOPE: Final = "GIGACHAT_API_PERS"
|
|||
# Token expiry buffer in milliseconds (refresh token 60s before expiry)
|
||||
TOKEN_EXPIRY_BUFFER_MS: Final = 60000
|
||||
|
||||
# Cache for access tokens
|
||||
_token_cache: Final = InMemoryCache()
|
||||
_NO_LITELLM_PARAMS: Final[Mapping[str, object]] = MappingProxyType({})
|
||||
|
||||
|
||||
class GigaChatAuthError(BaseLLMException):
|
||||
|
|
@ -80,10 +81,9 @@ def get_access_token(
|
|||
Raises:
|
||||
GigaChatAuthError: If authentication fails
|
||||
"""
|
||||
if not litellm_params:
|
||||
litellm_params = {} # mutable-ok: empty dict default; rebind-ok: provide default
|
||||
params: Final = litellm_params or _NO_LITELLM_PARAMS
|
||||
|
||||
access_token: Final = litellm_params.get("gigachat_access_token") or get_secret_str("GIGACHAT_ACCESS_TOKEN")
|
||||
access_token: Final = params.get("gigachat_access_token") or get_secret_str("GIGACHAT_ACCESS_TOKEN")
|
||||
if access_token:
|
||||
return access_token
|
||||
|
||||
|
|
@ -94,24 +94,20 @@ def get_access_token(
|
|||
message="GigaChat credentials not provided. Set GIGACHAT_CREDENTIALS or GIGACHAT_API_KEY environment variable.",
|
||||
)
|
||||
|
||||
effective_scope: Final = scope or litellm_params.get("gigachat_scope") or _get_scope()
|
||||
effective_auth_url: Final = auth_url or litellm_params.get("gigachat_auth_url") or _get_auth_url()
|
||||
effective_scope: Final = scope or params.get("gigachat_scope") or _get_scope()
|
||||
effective_auth_url: Final = auth_url or params.get("gigachat_auth_url") or _get_auth_url()
|
||||
|
||||
# Check cache
|
||||
cache_key: Final = f"gigachat_token:{effective_credentials[:16]}"
|
||||
cached: Final = _token_cache.get_cache(cache_key)
|
||||
if cached:
|
||||
_token, _expires_at = cached
|
||||
# Check if token is still valid (with buffer)
|
||||
if time.time() * 1000 < _expires_at - TOKEN_EXPIRY_BUFFER_MS:
|
||||
verbose_logger.debug("Using cached GigaChat access token")
|
||||
return _token
|
||||
|
||||
# Request new token
|
||||
new_token, new_expires_at = _request_token_sync(effective_credentials, effective_scope, effective_auth_url) # pyright: ignore[reportArgumentType] # credential keys may be broader than str
|
||||
|
||||
if new_expires_at:
|
||||
# Cache token
|
||||
ttl_seconds: Final = max(0, (new_expires_at - TOKEN_EXPIRY_BUFFER_MS - time.time() * 1000) / 1000)
|
||||
if ttl_seconds > 0:
|
||||
_token_cache.set_cache(cache_key, (new_token, new_expires_at), ttl=ttl_seconds)
|
||||
|
|
@ -126,10 +122,9 @@ async def get_access_token_async(
|
|||
litellm_params: Mapping[str, object] | None = None,
|
||||
) -> str:
|
||||
"""Async version of get_access_token."""
|
||||
if not litellm_params:
|
||||
litellm_params = {} # mutable-ok: empty dict default; rebind-ok: provide default
|
||||
params: Final = litellm_params or _NO_LITELLM_PARAMS
|
||||
|
||||
access_token: Final = litellm_params.get("gigachat_access_token") or get_secret_str("GIGACHAT_ACCESS_TOKEN")
|
||||
access_token: Final = params.get("gigachat_access_token") or get_secret_str("GIGACHAT_ACCESS_TOKEN")
|
||||
if access_token:
|
||||
return access_token
|
||||
|
||||
|
|
@ -140,10 +135,9 @@ async def get_access_token_async(
|
|||
message="GigaChat credentials not provided. Set GIGACHAT_CREDENTIALS or GIGACHAT_API_KEY environment variable.",
|
||||
)
|
||||
|
||||
effective_scope: Final = scope or litellm_params.get("gigachat_scope") or _get_scope()
|
||||
effective_auth_url: Final = auth_url or litellm_params.get("gigachat_auth_url") or _get_auth_url()
|
||||
effective_scope: Final = scope or params.get("gigachat_scope") or _get_scope()
|
||||
effective_auth_url: Final = auth_url or params.get("gigachat_auth_url") or _get_auth_url()
|
||||
|
||||
# Check cache
|
||||
cache_key: Final = f"gigachat_token:{effective_credentials[:16]}"
|
||||
cached: Final = _token_cache.get_cache(cache_key)
|
||||
if cached:
|
||||
|
|
@ -152,11 +146,9 @@ async def get_access_token_async(
|
|||
verbose_logger.debug("Using cached GigaChat access token")
|
||||
return _token
|
||||
|
||||
# Request new token
|
||||
new_token, new_expires_at = await _request_token_async(effective_credentials, effective_scope, effective_auth_url) # pyright: ignore[reportArgumentType] # credential keys may be broader than str
|
||||
|
||||
if new_expires_at:
|
||||
# Cache token
|
||||
ttl_seconds: Final = max(0, (new_expires_at - TOKEN_EXPIRY_BUFFER_MS - time.time() * 1000) / 1000)
|
||||
if ttl_seconds > 0:
|
||||
_token_cache.set_cache(cache_key, (new_token, new_expires_at), ttl=ttl_seconds)
|
||||
|
|
|
|||
|
|
@ -52,7 +52,6 @@ class GigaChatModelResponseIterator:
|
|||
tool_use: ChatCompletionToolCallChunk | None = None # rebind-ok: conditionally assigned on function_call
|
||||
finish_reason: str | None = chunk_finish_reason
|
||||
|
||||
# Handle function_call in stream
|
||||
raw_function_call: Final = delta.get("function_call")
|
||||
if chunk_finish_reason == "function_call" and isinstance(raw_function_call, Mapping) and raw_function_call:
|
||||
func_call: Final[Mapping[str, object]] = raw_function_call
|
||||
|
|
|
|||
|
|
@ -111,11 +111,9 @@ class GigaChatConfig(BaseConfig):
|
|||
"""
|
||||
Set up headers with OAuth token.
|
||||
"""
|
||||
# Get access token
|
||||
credentials: Final = api_key or get_secret_str("GIGACHAT_CREDENTIALS") or get_secret_str("GIGACHAT_API_KEY")
|
||||
access_token: Final = get_access_token(credentials=credentials, litellm_params=litellm_params)
|
||||
|
||||
# Store credentials for image uploads
|
||||
self._current_credentials = credentials
|
||||
self._current_api_base = api_base
|
||||
|
||||
|
|
@ -208,18 +206,16 @@ class GigaChatConfig(BaseConfig):
|
|||
|
||||
def _convert_tools_to_functions(self, tools: Sequence) -> Sequence[dict]:
|
||||
"""Convert OpenAI tools format to GigaChat functions format."""
|
||||
functions: Final[list[dict]] = [] # mutable-ok: accumulator for building functions list
|
||||
for tool in tools:
|
||||
if isinstance(tool, dict) and tool.get("type") == "function":
|
||||
func = tool.get("function", {})
|
||||
functions.append(
|
||||
{
|
||||
"name": func.get("name", ""),
|
||||
"description": func.get("description", ""),
|
||||
"parameters": func.get("parameters", {}),
|
||||
}
|
||||
)
|
||||
return functions
|
||||
return [
|
||||
{
|
||||
"name": function.get("name", ""),
|
||||
"description": function.get("description", ""),
|
||||
"parameters": function.get("parameters", {}),
|
||||
}
|
||||
for function in (
|
||||
tool.get("function", {}) for tool in tools if isinstance(tool, dict) and tool.get("type") == "function"
|
||||
)
|
||||
]
|
||||
|
||||
def _map_tool_choice(self, tool_choice: str | Mapping[str, object]) -> str | Mapping[str, object] | None:
|
||||
"""
|
||||
|
|
@ -299,7 +295,6 @@ class GigaChatConfig(BaseConfig):
|
|||
if part.get("type") == "text":
|
||||
texts.append(part.get("text", ""))
|
||||
elif part.get("type") == "image_url":
|
||||
# Extract image URL and upload to GigaChat
|
||||
image_url: object = part.get("image_url", {})
|
||||
upload_url: str
|
||||
if isinstance(image_url, str):
|
||||
|
|
@ -322,16 +317,13 @@ class GigaChatConfig(BaseConfig):
|
|||
headers: Mapping[str, object],
|
||||
) -> dict: # mutable-ok: request payload sent to httpx
|
||||
"""Transform OpenAI request to GigaChat format."""
|
||||
# Transform messages
|
||||
giga_messages: Final = self._transform_messages(messages)
|
||||
|
||||
# Build request
|
||||
request_data: Final[dict[str, object]] = {
|
||||
"model": model.replace("gigachat/", ""),
|
||||
"messages": giga_messages,
|
||||
}
|
||||
|
||||
# Add optional params
|
||||
for key in [
|
||||
"temperature",
|
||||
"top_p",
|
||||
|
|
@ -343,7 +335,6 @@ class GigaChatConfig(BaseConfig):
|
|||
if key in optional_params:
|
||||
request_data[key] = optional_params[key]
|
||||
|
||||
# Add functions if present
|
||||
if "functions" in optional_params:
|
||||
request_data["functions"] = optional_params["functions"]
|
||||
if "function_call" in optional_params:
|
||||
|
|
@ -358,10 +349,8 @@ class GigaChatConfig(BaseConfig):
|
|||
for i, msg in enumerate(messages):
|
||||
message = dict(msg)
|
||||
|
||||
# Remove unsupported fields
|
||||
message.pop("name", None)
|
||||
|
||||
# Transform roles
|
||||
role = message.get("role", "user")
|
||||
if role == "developer":
|
||||
message["role"] = "system"
|
||||
|
|
@ -374,18 +363,15 @@ class GigaChatConfig(BaseConfig):
|
|||
if not isinstance(content, str) or not is_valid_json(content):
|
||||
message["content"] = json.dumps(content, ensure_ascii=False)
|
||||
|
||||
# Handle None content
|
||||
if message.get("content") is None:
|
||||
message["content"] = ""
|
||||
|
||||
# Handle list content (multimodal) - extract text and images
|
||||
content = message.get("content")
|
||||
if isinstance(content, list):
|
||||
message["content"], attachments = self._transform_list_content(content)
|
||||
if attachments:
|
||||
message["attachments"] = attachments
|
||||
|
||||
# Transform tool_calls to function_call
|
||||
tool_calls = message.get("tool_calls")
|
||||
if tool_calls and isinstance(tool_calls, list) and len(tool_calls) > 0:
|
||||
tool_call = tool_calls[0]
|
||||
|
|
@ -436,13 +422,11 @@ class GigaChatConfig(BaseConfig):
|
|||
message_data = choice.get("message", {})
|
||||
finish_reason = choice.get("finish_reason", "stop")
|
||||
|
||||
# Transform function_call to tool_calls or content
|
||||
if finish_reason == "function_call" and message_data.get("function_call"):
|
||||
func_call = message_data["function_call"]
|
||||
args = func_call.get("arguments", {})
|
||||
|
||||
if is_structured_output:
|
||||
# Convert to content for structured output
|
||||
if isinstance(args, dict):
|
||||
content = json.dumps(args, ensure_ascii=False)
|
||||
else:
|
||||
|
|
@ -452,7 +436,6 @@ class GigaChatConfig(BaseConfig):
|
|||
message_data.pop("functions_state_id", None)
|
||||
finish_reason = "stop"
|
||||
else:
|
||||
# Convert to tool_calls format
|
||||
if isinstance(args, dict):
|
||||
args = json.dumps(args, ensure_ascii=False)
|
||||
message_data["tool_calls"] = [
|
||||
|
|
@ -468,7 +451,6 @@ class GigaChatConfig(BaseConfig):
|
|||
message_data.pop("function_call", None)
|
||||
finish_reason = "tool_calls"
|
||||
|
||||
# Clean up GigaChat-specific fields
|
||||
message_data.pop("functions_state_id", None)
|
||||
|
||||
choices.append(
|
||||
|
|
|
|||
|
|
@ -112,18 +112,10 @@ class GigaChatEmbeddingConfig(BaseEmbeddingConfig):
|
|||
"input": ["text1", "text2", ...]
|
||||
}
|
||||
"""
|
||||
# Normalize input to list
|
||||
if isinstance(input, str):
|
||||
input_list: list = [input] # rebind-ok: locally scoped conversion
|
||||
else:
|
||||
input_list = input
|
||||
|
||||
# Remove gigachat/ prefix from model if present
|
||||
model = model.removeprefix("gigachat/") # rebind-ok: parameter reassignment for normalization
|
||||
|
||||
normalized_input: Final = [input] if isinstance(input, str) else input # mutable-ok: preserve list API
|
||||
return {
|
||||
"model": model,
|
||||
"input": input_list,
|
||||
"model": model.removeprefix("gigachat/"),
|
||||
"input": normalized_input,
|
||||
}
|
||||
|
||||
def transform_embedding_response(
|
||||
|
|
|
|||
|
|
@ -60,7 +60,6 @@ class GigaChatPassthroughConfig(BasePassthroughConfig):
|
|||
"""
|
||||
Set up headers with OAuth token.
|
||||
"""
|
||||
# Get access token
|
||||
access_token: Final = get_access_token(credentials=api_key, litellm_params=litellm_params)
|
||||
|
||||
headers["Authorization"] = f"Bearer {access_token}" # rebind-ok: mutating for OAuth setup
|
||||
|
|
@ -82,7 +81,6 @@ class GigaChatPassthroughConfig(BasePassthroughConfig):
|
|||
from litellm.types.utils import LlmProviders, ModelResponse
|
||||
from litellm.utils import ProviderConfigManager
|
||||
|
||||
# cost tracking only for completions and embeddings
|
||||
if "completions" in endpoint:
|
||||
provider_chat_config: Final = ProviderConfigManager.get_provider_chat_config(
|
||||
provider=LlmProviders(custom_llm_provider),
|
||||
|
|
|
|||
|
|
@ -4,7 +4,6 @@ from typing import Final
|
|||
from litellm.secret_managers.main import get_secret_str
|
||||
from litellm.types.utils import PromptTokensDetailsWrapper, Usage
|
||||
|
||||
# GigaChat API endpoint
|
||||
GIGACHAT_BASE_URL: Final = "https://gigachat.devices.sberbank.ru/api/v1"
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -113,10 +113,8 @@ class AsyncPassthroughStreamingResponse(AsyncGenerator[Any, Any]):
|
|||
)
|
||||
)
|
||||
|
||||
# Compliant: Save a strong reference to prevent GC
|
||||
self._background_tasks.add(task)
|
||||
|
||||
# Remove the task from the set when it finishes to avoid memory leaks
|
||||
task.add_done_callback(self._background_tasks.discard)
|
||||
except Exception as e: # noqa: BLE001 # Safe catch-all for verbose logging
|
||||
verbose_logger.exception(
|
||||
|
|
@ -578,7 +576,6 @@ def llm_passthrough_route(
|
|||
else:
|
||||
return response
|
||||
except Exception as e:
|
||||
# provider_config is guaranteed non-None here due to the earlier guard
|
||||
assert provider_config is not None
|
||||
raise base_llm_http_handler._handle_error(
|
||||
e=e,
|
||||
|
|
|
|||
|
|
@ -1731,7 +1731,7 @@ def get_vertex_ai_allowed_incoming_headers(request: Request) -> dict:
|
|||
|
||||
|
||||
def get_vertex_pass_through_handler(
|
||||
call_type: Literal["discovery", "aiplatform"], # noqa: UP037
|
||||
call_type: Literal["discovery", "aiplatform"], # noqa: UP037 # ruff reports quoted Literal values here
|
||||
) -> BaseVertexAIPassThroughHandler:
|
||||
if call_type == "discovery":
|
||||
return VertexAIDiscoveryPassThroughHandler()
|
||||
|
|
@ -2961,7 +2961,6 @@ async def handle_gigachat_passthrough_router_model(
|
|||
"""
|
||||
from litellm.proxy.common_request_processing import ProxyBaseLLMRequestProcessing
|
||||
|
||||
# Detect streaming based on request body
|
||||
is_streaming: Final = request_body.get("stream", False) # pyright: ignore[reportUnknownVariableType] # request_body is dict[Unknown, Unknown]
|
||||
|
||||
data: dict[str, Any] = await _read_request_body(
|
||||
|
|
@ -2997,7 +2996,6 @@ async def handle_gigachat_passthrough_router_model(
|
|||
data["json"] = request_body
|
||||
data["custom_llm_provider"] = "gigachat"
|
||||
|
||||
# Remove sensitive keys from data
|
||||
keys: Final = [ # mutable-ok: list of keys to remove from data
|
||||
"gigachat_auth_url",
|
||||
"gigachat_access_token",
|
||||
|
|
|
|||
|
|
@ -10268,7 +10268,8 @@ class TestContextWindowEscalation:
|
|||
litellm_router_instance=_windowed_router(_SMALL, _BIG),
|
||||
complexity_router_config=_tier_config(session_affinity=True),
|
||||
)
|
||||
session_kwargs = lambda: {"metadata": {"session_id": "s-1", "user_api_key_hash": "k-1"}} # noqa: E731
|
||||
def session_kwargs() -> dict[str, object]:
|
||||
return {"metadata": {"session_id": "s-1", "user_api_key_hash": "k-1"}}
|
||||
|
||||
first = await router.async_pre_routing_hook(
|
||||
model="test-router", request_kwargs=session_kwargs(), messages=_OVERSIZED_TURNS
|
||||
|
|
@ -10291,7 +10292,8 @@ class TestContextWindowEscalation:
|
|||
litellm_router_instance=_windowed_router(_SMALL, _BIG),
|
||||
complexity_router_config=_tier_config(session_affinity=True),
|
||||
)
|
||||
session_kwargs = lambda: {"metadata": {"session_id": "s-2", "user_api_key_hash": "k-2"}} # noqa: E731
|
||||
def session_kwargs() -> dict[str, object]:
|
||||
return {"metadata": {"session_id": "s-2", "user_api_key_hash": "k-2"}}
|
||||
|
||||
pinned = await router.async_pre_routing_hook(
|
||||
model="test-router", request_kwargs=session_kwargs(), messages=[{"role": "user", "content": "ok continue"}]
|
||||
|
|
|
|||
|
|
@ -1,12 +1,12 @@
|
|||
{
|
||||
"LIT001": {
|
||||
"limit": 22403
|
||||
"limit": 22402
|
||||
},
|
||||
"LIT002": {
|
||||
"limit": 26780
|
||||
},
|
||||
"LIT003": {
|
||||
"limit": 269
|
||||
"limit": 268
|
||||
},
|
||||
"LIT004": {
|
||||
"limit": 40
|
||||
|
|
@ -27,10 +27,10 @@
|
|||
"limit": 0
|
||||
},
|
||||
"LIT010": {
|
||||
"limit": 16512
|
||||
"limit": 16511
|
||||
},
|
||||
"LIT011": {
|
||||
"limit": 5537
|
||||
"limit": 5535
|
||||
},
|
||||
"LIT012": {
|
||||
"limit": 4495
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue