// CHECK-DAG: %[[RES0:.*]] = "tfl.pseudo_const"() <{value = dense<{{.*}}> : tensor<1x4xf32>}> {tfl.is_variable} : () -> tensor<1x4xf32>
// CHECK-DAG: %[[RES1:.*]] = "tfl.pseudo_const"() <{value = dense<{{.*}}> : tensor<1x4xf32>}> {tfl.is_variable} : () -> tensor<1x4xf32>

// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {
// CHECK-NEXT:       shape: [ 1, 4 ],
// CHECK-NEXT:       name: "Const1",
// CHECK-NEXT:       quantization: {
// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {

// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {
// CHECK-NEXT:       shape: [ 1, 4 ],
// CHECK-NEXT:       name: "Const1",
// CHECK-NEXT:       quantization: {
// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {

// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {
// CHECK-NEXT:       shape: [ 4, 4 ],
// CHECK-NEXT:       name: "Const1",
// CHECK-NEXT:       quantization: {
// CHECK-EMPTY:
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {

            return LogicalResult::failure();
          }

          if (is_variable) {
            input.emitOpError() << "unexpected multiple TPUReplicatedInputOp "
                                << "for single argument";
            return LogicalResult::failure();
          }
          is_variable = true;
        } else {
          is_non_variable = true;
        }
      }

// CHECK-NEXT:       name: "arg19",
// CHECK-NEXT:       quantization: {
// CHECK-NEXT:         scale: [ 0.096711 ],
// CHECK-NEXT:         zero_point: [ 10 ]
// CHECK-NEXT:       },
// CHECK-NEXT:       is_variable: true,
// CHECK-NEXT:       has_rank: true
// CHECK-NEXT:     }, {
// CHECK-NEXT:       shape: [ 1, 2048 ],
// CHECK-NEXT:       type: INT16,
// CHECK-NEXT:       name: "arg20",

// as tensor names may have changed via MLIR constant folding process.
bool ExpectEqualTensor(const Tensor* tensor, const Tensor* expected_tensor) {
  return (tensor->is_variable() == expected_tensor->is_variable()) &&
         (GetAsVector(tensor->shape()) ==
          GetAsVector(expected_tensor->shape()));
}

// Finds the match of the quantized tensor from the possible tensors. Each

    )
    return model

  def _create_gather_model(self, input_type, use_variable) -> module.Module:
    class GatherModel(module.Module):
      """A simple model with a single gather."""

      def __init__(self, use_variable):
        """Initializes a GatherModel.

        Args:
          use_variable: If True, creates a variable for weight.
        """
        super().__init__()

                }
            }
        }

        if (descriptor is IrBasedVariableDescriptor && owner is IrVariable) {
            val name = owner.name
            val isMutated = false // TODO capture the usage somehow

            if (owner.origin == IrDeclarationOrigin.PROPERTY_DELEGATE) {