fix(rate-limiting): satisfy strict lint gate and regenerate stale schema.d.ts

- Add the missing __init__ return annotation (ANN204) and drop the
  now-unused typing.Any import in favor of a concrete object fallback
  for the optional-otel-Span type alias (TID251), both newly over the
  ruff-strict budget.
- Regenerate ui/litellm-dashboard/src/lib/http/schema.d.ts: the prior
  comment-trimming pass left it out of sync with the trimmed
  TagRateLimitEntry/TagRateLimits docstrings.

litellm/proxy/hooks/tag_rate_limiter.py

@@ -5,7 +5,7 @@ import contextvars
 from collections.abc import Callable
 from dataclasses import dataclass
 from datetime import datetime
-from typing import TYPE_CHECKING, Any, Literal
+from typing import TYPE_CHECKING, Literal
 
 from litellm._logging import verbose_proxy_logger
 from litellm.caching.dual_cache import DualCache
@@ -30,9 +30,9 @@ from litellm.types.utils import StandardLoggingPayload
 if TYPE_CHECKING:
     from opentelemetry.trace import Span as _Span
 
-    Span = _Span | Any
+    Span = _Span
 else:
-    Span = Any
+    Span = object
 
 _LimitUnit = Literal["tokens", "requests", "dollars", "concurrency"]
 _LIMIT_UNITS: tuple[_LimitUnit, ...] = ("tokens", "requests", "dollars", "concurrency")
@@ -414,7 +414,7 @@ class _PROXY_TagRateLimiter(CustomLogger):
         self,
         internal_usage_cache: DualCache,
         time_provider: Callable[[], datetime] | None = None,
-    ):
+    ) -> None:
         self.internal_usage_cache = InternalUsageCache(dual_cache=internal_usage_cache)
         self._v3 = _PROXY_MaxParallelRequestsHandler_v3(self.internal_usage_cache, time_provider=time_provider)
         self._time_provider = time_provider or datetime.now

ui/litellm-dashboard/src/lib/http/schema.d.ts

@@ -35110,25 +35110,7 @@ export interface components {
             /** Tpm Limit */
             tpm_limit?: number | null;
         };
-        /**
-         * TagRateLimitEntry
-         * @description One tag-scoped limit: a caller-supplied tag value (identified by `tag_id`,
-         *     e.g. `end_user_id` in a request tag like `end_user_id:user-123`) is capped
-         *     at `limit` units per rolling `period_seconds`-second window. Bucketing is
-         *     `epoch_second // period_seconds`, so `period_seconds=86400` resets at UTC
-         *     midnight and `period_seconds=60` resets on real clock-minute boundaries.
-         *
-         *     For a `concurrency_limits` entry specifically, `period_seconds` is not a
-         *     window: it is a floor under the safety TTL a reserved in-flight slot
-         *     self-heals after, in case a worker crashes before releasing it (the
-         *     counter, not a window). The effective TTL is at least one hour regardless
-         *     of this value, so a slow but genuinely still-running request never has
-         *     its reservation expire out from under it; set this higher only if an
-         *     even longer self-heal window is wanted. `concurrency_limits` also only
-         *     supports chain-wide entries (declared identically by every deployment
-         *     sharing a `model_name`) -- a divergent per-deployment value is dropped
-         *     with a warning, not silently scoped to a subset of deployments.
-         */
+        /** TagRateLimitEntry */
         TagRateLimitEntry: {
             /** Limit */
             limit: number;
@@ -35152,12 +35134,7 @@ export interface components {
             /** Limits */
             limits?: components["schemas"]["TagRateLimitEntry"][];
         };
-        /**
-         * TagRateLimits
-         * @description Per-chain/model-group tag rate limits, set under a deployment's
-         *     `model_info.tag_rate_limits`. Each entry carries its own `tag_id`, so two
-         *     entries of the same unit on the same chain can key by different tags.
-         */
+        /** TagRateLimits */
         TagRateLimits: {
             concurrency_limits?: components["schemas"]["TagRateLimitGroup"] | null;
             dollar_limits?: components["schemas"]["TagRateLimitGroup"] | null;