const std::string new_inputs_attr_str =
        llvm::join(std::move(inputs_attrs), /*Separator=*/",");

    named_attr.setValue(StringAttr::get(ctx, new_inputs_attr_str));
  }

  func_op->setAttr(kTfEntryFunctionAttr,
                   DictionaryAttr::get(ctx, entry_func_attrs));
}
}  // namespace quant

  builder.setInsertionPoint(execute_launch);
  auto copy_with_layout = BuildCopyWithLayout(execute_launch, compile_launch,
                                              get_layout, input, &builder);
  copy_with_layout->setAttr(kDeviceAttr, execute_launch.getDeviceAttr());
  execute.setOperand(execute_arg_index, copy_with_layout);
}

// Performs transformation for replicated inputs. Returns true if this is a

    self._outputs = outputs

  def __call__(self, compose_fn):
    # TODO(fengliuai): more sanity check of the input function and make sure
    # the bounded arguments of the function matches the 'inputs' and 'attrs'.
    setattr(compose_fn, '_tfr_op_name', self._op_name)

                            live_ins);

  func::FuncOp outlined_func =
      BuildFunction(live_ins.getArrayRef(), cluster_op, symbol_table, builder);
  cluster_op->setAttr(
      builder->getStringAttr(kFuncAttr),
      mlir::SymbolRefAttr::get(builder->getContext(), outlined_func.getName()));

  builder->setInsertionPoint(cluster_op);

    SmallVector<Type> result_types(op.getResultTypes().begin(),
                                   op.getResultTypes().end());
    if (!op->getAttr("layout")) {
      auto layout = mlir::GetTPUInfeedLayout(result_types, builder);
      if (failed(layout)) return;

      op->setAttr("layout", layout.value());
    }
  }
};

void InfeedsOpsXlaAdjustLayout::runOnOperation() {
  getOperation().walk(runOnInfeedOp);
}

                            FunctionFilter::FunctionFilterType type,
                            OpBuilder& builder) {
  for (Operation& op : func.front()) {
    if (type == FunctionFilter::SKIP_TARGET_ANNOTATION) {
      op.setAttr(kSkipTargetAnnotation, builder.getUnitAttr());
    } else if (type == FunctionFilter::INCLUDE_TARGET_ANNOTATION) {
      op.removeAttr(kSkipTargetAnnotation);
    }
  }
}

        return WalkResult::advance();
      for (Operation* pred : IslandPredecessors(split)) {
        if (auto colocation_classes =
                pred->getAttrOfType<ArrayAttr>(kClassAttr)) {
          split->setAttr(kClassAttr, colocation_classes);
          return WalkResult::advance();
        }
      }
    }
    return WalkResult::advance();
  });
}

}  // namespace

  return ConstBytesAttr::get(builder->getContext(),
                             StringRef(content.data(), content.size()));
}

}  // namespace

void ConvertNMSPaddedFunc::RewriteFunc() {
  func_->setAttr(kTFImplements,
                 StringAttr::get(func_.getContext(), kTfNMSPadded));
  Value boxes = func_.getArgument(0);
  Value scores = func_.getArgument(1);
  Value max_output_size = func_.getArgument(2);

  import tensorflow_io_gcs_filesystem as _tensorflow_io_gcs_filesystem

# Lazy-load Keras v1.
_tf_uses_legacy_keras = (
    _os.environ.get("TF_USE_LEGACY_KERAS", None) in ("true", "True", "1"))
setattr(_current_module, "keras", _KerasLazyLoader(globals(), mode="v1"))
_module_dir = _module_util.get_parent_dir_for_name("keras._tf_keras.keras")
_current_module.__path__ = [_module_dir] + _current_module.__path__

    int64_t arithmetic_count = arithmetic_count_op.GetArithmeticCount(op);
    auto attr =
        builder.getIntegerAttr(builder.getIntegerType(64), arithmetic_count);
    op->setAttr("_arithmetic_count", attr);
  });
}

}  // namespace

/// Creates an instance of the TensorFlow Lite dialect GetArithmeticCount
/// pass.