// the quantized inputs, performing the operation in the expressed type, then
// requantizing if a quantized output is required.
//
// The motivation behind these changes is for Dialects that assume only float
// or quantized computation, and do not support a mixture of these types on
// dense operations. Decomposition allows TFLite to be compiled to these
// dialects, such as TOSA.

#include <utility>

	// 		Let F ← log₂(d) and c = 1.
	// 	else
	// 		Let F ← N + L where L is the smallest integer
	// 		such that d ≤ (2^(N+L) mod d) + 2^L.
	// 	end if
	//
	// [1] "Faster Remainder by Direct Computation: Applications to
	// Compilers and Software Libraries" Daniel Lemire, Owen Kaser,
	// Nathan Kurz arXiv:1902.01961
	//
	// To minimize the risk of introducing errors, we implement the

    // as the reduction.
    Value iota = reduce_op.getInputs().back();
    if (!MatchIota(reduce_op.getDimensions(), iota)) return failure();

    // Match the reduction computation.
    const bool is_float = mlir::isa<FloatType>(operand_init.getElementType());
    if (failed(MatchReduceToArgMinMaxType1(reduce_op, is_float, is_argmax)) &&
        failed(MatchReduceToArgMinMaxType2(reduce_op, is_argmax)))

    "means that a shape or dimension argument could not be evaluated at "
    "compile time, usually because the value of the argument depends on a "
    "parameter to the computation, on a variable, or on a stateful operation "
    "such as a random number generator.";

// TODO(b/282188914) remove the operations to skip once tests are fixed.
static const DenseSet<mlir::TypeID>* operations_to_skip =

	}
	return e.result
}

// Get returns the cached result associated with key
// and reports whether there is such a result.
//
// If the result for key is being computed, Get does not wait for the computation to finish.
func (c *Cache[K, V]) Get(key K) (V, bool) {
	entryIface, ok := c.m.Load(key)
	if !ok {
		return *new(V), false
	}
	e := entryIface.(*cacheEntry[V])
	if !e.done.Load() {

      if (!operand_type.hasStaticShape() || !indices_type.hasStaticShape() ||
          !updates_type.hasStaticShape()) {
        return failure();
      }

      // Match the scatter computation against computations supported by TF.
      if (failed(MatchBinaryReduceFunction<BinaryOp>(
              scatter_op.getUpdateComputation()))) {
        return failure();
      }

	})
	WorkloadServices.RegisterBatch(krt.BatchedEventFilter(
		func(a model.ServiceInfo) *workloadapi.Service {
			// Only trigger push if the XDS object changed; the rest is just for computation of others
			return a.Service
		},
		PushXds(a.XDSUpdater, func(i model.ServiceInfo) model.ConfigKey {
			return model.ConfigKey{Kind: kind.Address, Name: i.ResourceName()}
		})), false)

You can see more details in <a href="https://www.starlette.io/background/" class="external-link" target="_blank">Starlette's official docs for Background Tasks</a>.

## Caveat

  let description = [{
    This op is a TensorFlow op that represents "streamed outputs", where
    intermediate results can be returned immediately without waiting for the
    entire signature computation to complete.

    This op takes `args` with their `names` (their cardinality must match) and
    sends the given argument tensors back to the serving controller. This

#define GEN_PASS_DEF_TPUBRIDGEEXECUTORISLANDOUTLININGPASS
#include "tensorflow/compiler/mlir/tensorflow/transforms/tf_passes.h.inc"

// Extract the islands containing a TPU cluster computation into an outlined
// function in a nested module. This will allow to run the usual bridge on this
// nested module which exhibit a more friendly "V2-like" structure.