(TF_ConstOp (getScaledAlpha $features)))))>;



//===----------------------------------------------------------------------===//
// Expm1 op patterns.
//===----------------------------------------------------------------------===//

// Expm1(x) = Exp(x) - 1
def LowerExp1mOp : Pat<
  (TF_Expm1Op $x),
  (TF_SubOp
    (TF_ExpOp $x),
    (TF_ConstOp (GetScalarOfType<1> $x))
  )>;

  %0 = "tf.Exp"(%arg0) : (tensor<2xf32>) -> tensor<2xf32>
  func.return %0 : tensor<2xf32>
}

// -----

// CHECK-LABEL: @expm1
func.func @expm1(%arg0: tensor<2xf32>) -> tensor<2xf32> {
  // CHECK:  mhlo.exponential_minus_one %arg0 : tensor<2xf32>
  %0 = "tf.Expm1"(%arg0) : (tensor<2xf32>) -> tensor<2xf32>
  func.return %0 : tensor<2xf32>
}

// -----

// CHECK-LABEL: func @exp_dynamic

  ];
}

def TF_Expm1Op : TF_Op<"Expm1", [InferTensorType, Pure, TF_SameOperandsAndResultTypeResolveRef]> {
  let summary = "Computes `exp(x) - 1` element-wise.";

  let description = [{
i.e. `exp(x) - 1` or `e^(x) - 1`, where `x` is the input tensor.
  `e` denotes Euler's number and is approximately equal to 2.718281.

  ```python
  x = tf.constant(2.0)
  tf.math.expm1(x) ==> 6.389056

    List<Integer> x = list(1, 2);
    List<String> y = list("3", "4");

    List<Object> exp1 = list((Object) 1, "3");
    List<Object> exp2 = list((Object) 1, "4");
    List<Object> exp3 = list((Object) 2, "3");
    List<Object> exp4 = list((Object) 2, "4");

    assertThat(Lists.<Object>cartesianProduct(x, y))
        .containsExactly(exp1, exp2, exp3, exp4)
        .inOrder();
  }

    List<Integer> x = list(1, 2);
    List<String> y = list("3", "4");

    List<Object> exp1 = list((Object) 1, "3");
    List<Object> exp2 = list((Object) 1, "4");
    List<Object> exp3 = list((Object) 2, "3");
    List<Object> exp4 = list((Object) 2, "4");

    assertThat(Lists.<Object>cartesianProduct(x, y))
        .containsExactly(exp1, exp2, exp3, exp4)
        .inOrder();
  }

in the generated body function. Instead, we allocate fake parameters and
start the body with an assignment. For example:

	for expr1, expr2 = range f {
		...
	}

becomes

	f(func(#p1 T1, #p2 T2) bool {
		expr1, expr2 = #p1, #p2
		...
	})

(All the generated variables have a # at the start to signal that they
are internal variables when looking at the generated code in a

    Set<Integer> x = set(1, 2);
    Set<String> y = set("3", "4");

    List<Object> exp1 = list((Object) 1, "3");
    List<Object> exp2 = list((Object) 1, "4");
    List<Object> exp3 = list((Object) 2, "3");
    List<Object> exp4 = list((Object) 2, "4");

    assertThat(Sets.<Object>cartesianProduct(x, y))
        .containsExactly(exp1, exp2, exp3, exp4)
        .inOrder();
  }

    Set<Integer> x = set(1, 2);
    Set<String> y = set("3", "4");

    List<Object> exp1 = list((Object) 1, "3");
    List<Object> exp2 = list((Object) 1, "4");
    List<Object> exp3 = list((Object) 2, "3");
    List<Object> exp4 = list((Object) 2, "4");

    assertThat(Sets.<Object>cartesianProduct(x, y))
        .containsExactly(exp1, exp2, exp3, exp4)
        .inOrder();
  }

        `tf.math.atan2`, `tf.math.atanh`, `tf.math.betainc`, `tf.math.ceil`,
        `tf.math.cos`, `tf.math.cosh`, `tf.math.digamma`, `tf.math.equal`,
        `tf.math.erfc`, `tf.math.exp`, `tf.math.expm1`, `tf.math.floor`,
        `tf.math.greater`, `tf.math.greater_equal`, `tf.math.igamma`,
        `tf.math.igammac`, `tf.math.invert_permutation`, `tf.math.less`,