}

  void TestGradientsSuccess(bool grad_inputs_provided) {
    TF_Output inputs[2];
    TF_Output outputs[1];
    TF_Output grad_outputs[2];
    TF_Output expected_grad_outputs[2];

    BuildSuccessGraph(inputs, outputs);
    BuildExpectedGraph(grad_inputs_provided, expected_grad_outputs);

    AddGradients(grad_inputs_provided, nullptr, inputs, 2, outputs, 1,
                 grad_outputs);

  // Set up distributed execution environment on local and remote tasks.
  // When `reset_context` is true, initialize new cluster context state based
  // on cluster configurations provided in `server_def`; otherwise, update
  // existing context state with the provided `server_def`. Contexts created
  // on remote tasks will be considered stale and garbage collected after
  // `keep_alive_secs` of inactivity.

  AbstractTensorHandle* AggregateGradients(
      gtl::ArraySlice<AbstractTensorHandle*> gradient_tensors) const override;

  // Calls the passed-in backward function.
  // op_type is the op's name provided in RecordOperation.
  Status CallBackwardFunction(
      const string& op_type, GradientFunction* gradient_function,
      const std::vector<int64_t>& unneeded_gradients,

TF_CAPI_EXPORT extern TF_Device* TF_GetDevice(TF_OpKernelContext* ctx);

// --------------------------  Resource  ---------------------------------------
// Create a `tensorflow::PluginResource` to the ResourceMgr provided by the
// `ctx`. The `tensorflow::PluginResource` wraps a resource by plugin (as a
// opaque pointer, since TensorFlow cannot parse it). `delete_func` is needed
// for ResourceMgr to clean up the resource. `status` will be set.

      if (n == nullptr) continue;

      // Adding the gradients to the graph can alter the prefix to prevent
      // name collisions only if this prefix has not been provided explicitly
      // by the user. If it was provided, assert that it remained intact.
      if (prefix != nullptr && !absl::StartsWith(n->name(), prefix_cmp)) {
        status->status = tensorflow::errors::Internal(

                                    var_default):
  """Get var_name either from env, or user or default.

  If var_name has been set as environment variable, use the preset value, else
  ask for user input. If no input is provided, the default is used.

  Args:
    environ_cp: copy of the os.environ.
    var_name: string for name of environment variable, e.g. "TF_NEED_CUDA".
    ask_for_var: string for how to ask for user input.

setup_file() {
    bazel version  # Start the bazel server
}

# Do a bazel query specifically for the licenses checker. It searches for
# targets matching the provided query, which start with // or @ but not
# //tensorflow (so it looks for //third_party, //external, etc.), and then
# gathers the list of all packages (i.e. directories) which contain those
# targets.
license_query() {

        for more details.
    *   Added `dataset_id` to `tf.data.experimental.service.register_dataset`.
        If provided, `tf.data` service will use the provided ID for the dataset.
        If the dataset ID already exists, no new dataset will be registered.
        This is useful if multiple training jobs need to use the same dataset

                      {x.get()}, absl::MakeSpan(outputs),
                      /*use_function=*/!std::get<2>(GetParam()));
  ASSERT_EQ(error::INVALID_ARGUMENT, s.code());
  ASSERT_EQ(
      "Provided null gradient_function for 'Neg'.\nIf the intent is to treat "
      "this op as non-differentiable consider using RegisterNotDifferentiable "
      "or NotDifferentiableGradientFunction.",
      s.message());

  // A vector to keep track of newly created `tf::Tensor` objects.
  std::vector<Tensor> all_input_tensors;
  // Update the vector with information from `input_tensors` if provided.
  if (input_tensors != nullptr) {
    // Note that we take the address of the elements in `all_input_tensors`
    // below. Allocate enough space so that no reallocation happens, which will
    // make the pointers invalid.