bool operator()(TFFakeQuantOp tf_op, AttrType &min_value,
                  AttrType &max_value) const {
    min_value = tf_op.getMinAttr();
    max_value = tf_op.getMaxAttr();
    return true;  // Successfully matched and fetched.
  }
};

template <class TFFakeQuantOp>
struct FetchConstantMinMaxInputs {
  using AttrType = DenseFPElementsAttr;
  bool operator()(TFFakeQuantOp tf_op, AttrType &min_value,

}

def TF_FlushSummaryWriterOp : TF_Op<"FlushSummaryWriter", []> {
  let summary = "Flushes the writer's unwritten events.";

  let description = [{
writer: A handle to the summary writer resource.
  }];

  let arguments = (ins
    Arg<TF_ResourceTensor, "", [TF_SummaryWrite]>:$writer
  );

  let results = (outs);
}

def TF_ImportEventOp : TF_Op<"ImportEvent", []> {

  bool operator()(TFFakeQuantOp tf_op, AttrType &min_value,
                  AttrType &max_value) const {
    min_value = tf_op.getMinAttr();
    max_value = tf_op.getMaxAttr();
    return true;  // Successfully matched and fetched.
  }
};

template <class TFFakeQuantOp>
struct FetchConstantMinMaxInputs {
  using AttrType = DenseFPElementsAttr;
  bool operator()(TFFakeQuantOp tf_op, AttrType &min_value,

  );
}

def TF_AnonymousMemoryCacheOp : TF_Op<"AnonymousMemoryCache", [TF_UniqueResourceAllocation]> {
  let summary = "";

  let arguments = (ins);

  let results = (outs
    Res<TF_ResourceTensor, "", [TF_DatasetMemoryCacheAlloc]>:$handle,
    TF_VariantTensor:$deleter
  );
}

def TF_AnonymousMultiDeviceIteratorOp : TF_Op<"AnonymousMultiDeviceIterator", [TF_UniqueResourceAllocation]> {

    // attribute to create the quantization parameter for the new quantize op.
    rewriter.setInsertionPointAfter(tf_op.getOperation());
    IntegerAttr num_bits = rewriter.getI64IntegerAttr(tf_op.getNumBits());
    BoolAttr narrow_range = rewriter.getBoolAttr(tf_op.getNarrowRange());
    Type res_type = tf_op.getType();
    TypeAttr qtype = quant::GetQuantizedTypeAttr(
        rewriter, res_type, min_value, max_value, quant_dim, num_bits,

          }
          auto attr_value =
              mlir::DenseI32ArrayAttr::get(tf_op->getContext(), values);
          tf_op->setAttr(attr_name, attr_value);
        };
    if (tf_op->hasTrait<mlir::OpTrait::AttrSizedOperandSegments>() ||
        tf_op->hasTrait<mlir::OpTrait::AttrSizedResultSegments>()) {
      // The op has multiple variadic operands or results.

include "mlir/Interfaces/CallInterfaces.td"
include "mlir/Interfaces/InferTypeOpInterface.td"
include "mlir/IR/OpAsmInterface.td"

// TODO(b/230804708): Add hybrid dot general.
def TF_UniformQuantizedDotGeneralOp : TF_Op<"UniformQuantizedDotGeneral", [Pure]> {
  // TODO(b/230804708): Improve the operator description.
  let summary = "Quantized dot general operator.";

  let arguments = (ins
    TensorOf<[TF_Float32, TF_Qint8]>:$lhs,

  "$_builder.getStringAttr("# values #")">;

def LegalizeTensorListReserve : Pat<(TF_TensorListReserveOp:$tf_op $shape, $num_elements),
    (TFL_CustomOp (Size2InputRange $shape, $num_elements),
   (CreateStringAttr<"\"TensorListReserve\"">), (CustomOptions $tf_op))>;

def LegalizeTensorListStack : Pat<(TF_TensorListStackOp $input, $shape, $unused_num_elements),
    (TFL_CustomOp (Size2InputRange $input, $shape),

// implementation available only in TFRT.

#ifndef TFRT_OPS
#define TFRT_OPS

include "tensorflow/compiler/mlir/tensorflow/ir/tf_op_base.td"
include "mlir/IR/OpBase.td"

def TF__TfrtSetResourceOp : TF_Op<"_TfrtSetResource", []> {
  let summary = "Set values as TFRT runtime's static resource.";

  let description = [{
    Setting a tensor value in TFRT runtime's static resource manager, using

#include "tensorflow/compiler/mlir/lite/transforms/generated_legalize_tf.inc"

#define DECL_CONVERT_OP(tf_op)                                               \
  struct ConvertTF##tf_op##Op : public RewritePattern {                      \
    explicit ConvertTF##tf_op##Op(MLIRContext* context)                      \
        : RewritePattern(TF::tf_op##Op::getOperationName(), 1, context) {}   \