func.func @modified(%arg0: tensor<1x224x224x3xf32>) -> tensor<1x401408xf32> attributes {tf.entry_function = {control_outputs = "", inputs = "input", outputs = "output"}} {
  %cst = arith.constant dense<[1, 401408]> : tensor<2xi32>
  %0 = "tfl.quantize"(%arg0) {qtype = tensor<1x224x224x3x!quant.uniform<i8:f32, 7.812500e-03>>} : (tensor<1x224x224x3xf32>) -> tensor<1x224x224x3x!quant.uniform<i8:f32, 7.812500e-03>>

// CHECK-DAG:   [[ITEM_RANK:%.*]] = "tf.Rank"([[ITEM]]) : (tensor<10xf32>) -> tensor<i32>
// CHECK-DAG:   [[ZERO:%cst.*]] = arith.constant dense<0> : tensor<i32>
// CHECK-DAG:   [[VECTOR_RANK:%.*]] = "tf.ExpandDims"([[ITEM_RANK]], [[ZERO]]) : (tensor<i32>, tensor<i32>) -> tensor<1xi32>
// CHECK-DAG:   [[ZERO_1:%cst.*]] = arith.constant dense<0> : tensor<i32>

func.func @main() -> (tensor<1xf32>, tensor<i32>)
    attributes {tf.entry_function = {outputs = "result,pconst"}} {
  %cst = arith.constant dense<1> : tensor<i32> loc("dec")
  %arg0 = arith.constant dense<5> : tensor<i32> loc("N")
  %arg1 = arith.constant dense<3.0> : tensor<1xf32> loc("val")
  %0:3 = "tfl.while"(%arg0, %arg1, %cst) ({
    ^bb0(%arg2: tensor<*xi32>, %arg3: tensor<*xf32>, %arg4: tensor<i32>):

module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 1399 : i32}} {

// CHECK-LABEL: stateless_op
func.func @stateless_op() -> tensor<i32> {
  // CHECK: %cst = "tf.Const"() <{value = dense<1> : tensor<i32>}> : () -> tensor<i32>
  %0 = "tf.StatelessRandomGetAlg"() {device = ""} : () -> tensor<i32>
  return %0 : tensor<i32>
}

}

// DEBUG-LABEL: QuantizeSplit
func.func @QuantizeSplit(%arg: tensor<4x!quant.uniform<u8:f32, 1.0>>, %cst: tensor<i32>) -> (tensor<2x!quant.uniform<u8:f32, 1.0>>,tensor<2x!quant.uniform<u8:f32, 1.0>>) {
  %0 = "tfl.dequantize"(%arg) : (tensor<4x!quant.uniform<u8:f32, 1.0>>) -> tensor<4xf32>
  %1:2 = "tfl.split"(%cst, %0) {num_splits = 2 : i32} : (tensor<i32>, tensor<4xf32>) -> (tensor<2xf32>, tensor<2xf32>)

  // CHECK: return
}

// -----
// This test verifies that the extraction works correctly for evaluation-only models.
module {
  func.func @main() {
    %cst = "tf.Const"() {value = dense<2> : tensor<i32>} : () -> tensor<i32>
    %0 = "tf.While"(%cst) {body = @while_body, cond = @while_cond, is_stateless = false} : (tensor<i32>) -> (tensor<i32>)
    return
  }
  func.func private @while_body(%arg0: tensor<i32>) -> (tensor<i32>) {

  // CHECK: %[[CST:.*]] = "tf.Const"() <{value = dense<0> : tensor<i32>}> : () -> tensor<i32>
  // CHECK: %[[RES0:.*]] = "tf.GatherV2"(%arg0, %arg2, %cst) <{batch_dims = 0 : i64}> {device = "/job:localhost/replica:0/task:0/device:GPU:0"} : (tensor<1600x1x32xf32>, tensor<i32>, tensor<i32>) -> tensor<1x32xf32>

  %cst = arith.constant dense<0> : tensor<i32> loc("Const")
  %0 = "tfl.greater"(%arg0, %cst) : (tensor<*xi32>, tensor<i32>) -> tensor<i1>
  func.return %0 : tensor<i1>
}

func.func @WhileOp_body(%arg0: tensor<*xi32>, %arg1: tensor<*xf32>) -> (tensor<*xi32>, tensor<*xf32>) {
  %cst = arith.constant dense<1> : tensor<i32> loc("Const1")

module attributes {tf.versions = {producer = 888 : i32}} {
func.func @testReshapeShapeInference(%arg0: tensor<3x4xi32>) -> tensor<*xi32> {
  %cst = arith.constant dense<[1, 6, 2]> : tensor<3xi32>
  // CHECK: "tfl.reshape"(%arg0, %cst) : (tensor<3x4xi32>, tensor<3xi32>) -> tensor<1x6x2xi32>
  %0 = "tfl.reshape"(%arg0, %cst) : (tensor<3x4xi32>, tensor<3xi32>) -> tensor<*xi32>
  func.return %0 : tensor<*xi32>
}
}

// -----

  // CHECK-NEXT:  name: "min_runtime_version",
  // CHECK-NEXT:  buffer: 5
  // CHECK-NEXT:  } ]
  // CHECK-NEXT:  signature_defs: [ ]
  // CHECK-NEXT:}

  %cst = "tfl.no_value"() {value = unit} : () -> none
  %0:2 = "tfl.fully_connected"(%arg0, %arg1, %cst) {fused_activation_function = "NONE", keep_num_dims = false, weights_format = "DEFAULT"} : (tensor<40x37xf32>, tensor<40x37xf32>, none) -> (tensor<40x40xf32>, tensor<40x40xf32>)