*
   * <p><b>Performance note:</b> the instance returned is a singleton.
   */
  @SuppressWarnings("unchecked")
  // unsafe, comparator() returns a comparator on the specified type
  // TODO(kevinb): evaluate whether or not of().comparator() should return null
  public static <K, V> ImmutableSortedMap<K, V> of() {
    return (ImmutableSortedMap<K, V>) NATURAL_EMPTY_MAP;
  }

// `device` is copied and has no lifetime requirements.
TF_CAPI_EXPORT extern void TF_ImportGraphDefOptionsSetDefaultDevice(
    TF_ImportGraphDefOptions* opts, const char* device);

// Set whether to uniquify imported operation names. If true, imported operation
// names will be modified if their name already exists in the graph. If false,
// conflicting names will be treated as an error. Note that this option has no

## Changelog since v1.30.0

## Urgent Upgrade Notes 

### (No, really, you MUST read this before you upgrade)

- Added support to the scheduler to start using QueueingHint registered for Pod/Updated event to determine whether unschedulable Pods update make them schedulable, when the feature gate `SchedulerQueueingHints` is enabled.

  }

  /**
   * We had a bug where we'd perform infinite retries of route that fail with connection shutdown
   * errors. The problem was that the logic that decided whether to reuse a route didn't track
   * certain HTTP/2 errors. https://github.com/square/okhttp/issues/5547
   */
  @ParameterizedTest
  @ArgumentsSource(ProtocolParamProvider::class)

   * into cache. This can be useful in testing, or to disable caching temporarily.
   *
   * <p>Note that weight is only used to determine whether the cache is over capacity; it has no
   * effect on selecting which entry should be evicted next.
   *
   * <p>This feature cannot be used in conjunction with {@link #maximumSize}.
   *

* We change the default attach_detach_controller sync period to 1 minute to reduce the query frequency through cloud provider to check whether volumes are attached or not.  ([#41363](https://github.com/kubernetes/kubernetes/pull/41363), [@jingxu97](https://github.com/jingxu97))

## Changelog since v1.5.2

### Other notable changes

* We change the default attach_detach_controller sync period to 1 minute to reduce the query frequency through cloud provider to check whether volumes are attached or not.  ([#41363](https://github.com/kubernetes/kubernetes/pull/41363), [@jingxu97](https://github.com/jingxu97))

### Major Features and Improvements

*   Add `is_cpu_target_available`, which indicates whether or not TensorFlow was built with support for a given CPU target. This can be useful for skipping target-specific tests if a target is not supported.

*   `tf.data`

		// Log any error in input session policy config.
		iamLogIf(GlobalContext, err)
		return
	}

	// SPECIAL CASE: For service accounts, any valid JSON is allowed as a
	// policy, regardless of whether the number of statements is 0, this
	// includes `null`, `{}` and `{"Statement": null}`. In fact, MinIO Console
	// sends `null` when no policy is set and the intended behavior is that the

- In a scheduler with `Permit` plugins, when a Pod is rejected during `WaitOnPermit`, the scheduler records the plugin.
  The scheduler will use the record to honor cluster events and queueing `hints registered` for the plugin, to inform whether to retry the pod. ([#119785](https://github.com/kubernetes/kubernetes/pull/119785), [@sanposhiho](https://github.com/sanposhiho))