TensorFlow via the command `import tensorflow as tf`.

The primary function of this module is to import all of the public TensorFlow
interfaces into a single place. The interfaces themselves are located in
sub-modules, as described below.

Note that the file `__init__.py` in the TensorFlow source code tree is actually
only a placeholder to enable test cases to run. The TensorFlow build replaces

init files creation process and is subsequently stored in the wheel file after
the build. It also contains a few paths that cannot be directly imported. These
paths point to attributes or sub-modules within a module's namespace, but they
don't correspond to an actual file or directory on the filesystem. The list of
such paths is stored in the packages_for_skip variable and will be skipped
during the test.

  ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
}

TEST_F(CApiGradientsTest, GradientsPrefix_2ndGradientsAsSubScopeOf1st) {
  BuildGraphAndAddGradientsWithPrefixes("gradients", "gradients/sub");
  ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
}

TEST_F(CApiGradientsTest, GradientsPrefix_PrefixMatchesExistingNodeName) {
  BuildGraphAndAddGradientsWithPrefixes("Const_0");

}

Status SubModel(AbstractContext* ctx,
                absl::Span<AbstractTensorHandle* const> inputs,
                absl::Span<AbstractTensorHandle*> outputs) {
  return ops::Sub(ctx, inputs[0], inputs[1], &outputs[0], "Sub");
}

Status MulModel(AbstractContext* ctx,
                absl::Span<AbstractTensorHandle* const> inputs,
                absl::Span<AbstractTensorHandle*> outputs) {

    super().setUp()
    self.api_packages_v2 = _get_api_packages_v2()
    # Some paths in the api_packages_path file cannot be directly imported.
    # These paths may point to attributes or sub-modules within a module's
    # namespace, but they don't correspond to an actual file
    # or directory on the filesystem.
    self.packages_for_skip = [
        "tensorflow.distribute.cluster_resolver",

It takes in a file with a set of dependencies, and produces a more detailed
requirements file for each version, with hashes specified for each
dependency required, as well as their sub-dependencies.

### How to update/add requirements

By default, the name of the base requirements file is `requirements.in`, but it
can be set using the `REQUIREMENTS_FILE_NAME` variable. \
For example:

```

    /* Given upstream grad U and a Sub op A-B, the gradients are:
     *
     *    dA =  U
     *    dB = -U
     *
     */

    // Grad for A
    DCHECK(grad_outputs[0]);
    grad_inputs[0] = grad_outputs[0];
    grad_inputs[0]->Ref();

    // Grad for B
    // negate the upstream grad
    std::string name = "Neg_Sub_Grad_B";
    TF_RETURN_IF_ERROR(

*   `tf.data`:

    *   Moved option `warm_start` from `tf.data.experimental.OptimizationOptions` to `tf.data.Options`.

*   `tf.lite`:

    *   `sub_op` and `mul_op` support broadcasting up to 6 dimensions.

      # debugging, but breaks this "number of failures" counter because it's
      # different for repetitions of the same test. We use re.sub(r"0x\w+")
      # to remove it.
      key = re.sub(r"0x\w+", "", p.getparent().get("name", "")) + p.text
      if key in seen:
        testsuite._elem.remove(p.getparent())
      seen[key] += 1

    AbstractTensorHandlePtr fMinus(f_outputs[0]);

    // Take Difference of both estimates: (f(theta + eps) - f(theta - eps)).
    TF_RETURN_IF_ERROR(
        ops::Sub(ctx, fPlus.get(), fMinus.get(), f_outputs, "sub_top"));
    AbstractTensorHandlePtr fDiff(f_outputs[0]);

    // Calculate using the difference quotient definition:
    // (f(theta + eps) - f(theta - eps)) / (2 * eps).
    TF_RETURN_IF_ERROR(