-
        name: Checkout
        uses: actions/checkout@755da8c3cf115ac066823e79a1e1788f8940201b # v3.2.0
      -
        name: Set up Docker Buildx
        uses: docker/setup-buildx-action@8c0edbc76e98fa90f69d9a2c020dcb50019dc325 # v2.2.1
      -
        name: Login to GCR
        if: contains(github.event.pull_request.labels.*.name, 'build and push to gcr.io for staging')

    if: |
      github.event.pull_request.draft == false &&
      github.event.sender.type == 'User'
    steps:
      - name: Checkout repo
        uses: actions/checkout@755da8c3cf115ac066823e79a1e1788f8940201b # v3.2.0
      - name: Trusted-Partners-PR
        uses: actions/github-script@d556feaca394842dc55e4734bf3bb9f685482fa0 # v6.3.3
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          script: |

      - name: Checkout repository for nightly (skipped for releases)
        if: ${{ github.event_name == 'schedule' }}
        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1
        with:
          ref: 'nightly'
      - name: Checkout repository
        if: ${{ github.event_name == 'push' }}
        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1

are working with. When loading untrusted checkpoints, the values of the traced
variables from your model are also going to be untrusted. That means that if
your code interacts with the filesystem, network, etc. and uses checkpointed
variables as part of those interactions (ex: using a string variable to build a
filesystem path), a maliciously created checkpoint might be able to change the
targets of those operations, which could result in arbitrary

            `nodes_and_registration()` method, both of which were previously
            documented as temporary and to be removed.
            *   Uses of `tensors()` can be replaced by calling the existing
                methods `tensors_size()` and `tensor(int)`.
            *   Uses of the non-const overload of `nodes_and_registration` can
                be replaced by calling the existing methods `nodes_size()` and

  }

 private:
  TFE_Context* context_;
  TFE_CustomDevice device_;
  void* info_;
  string name_;
};

// An adapter which wraps the shape/data produced by C custom devices and uses
// it to implement custom device methods.
class CAPICustomDeviceTensorHandle
    : public tensorflow::CustomDeviceTensorHandle {
 public:
  CAPICustomDeviceTensorHandle(tensorflow::ImmediateExecutionContext* context,

            "c_api_unified_experimental_internal.h",
            "graph_function.h",
            "graph_function.cc",
            "gradient_checker.cc",
            "gradient_checker.h",
            "gradients.cc",  # Uses RTTI.
            "tracing_utils.h",
            "tracing_utils.cc",
            "*test*",
            "*dlpack*",
        ],
    ),
    visibility = ["//visibility:public"],

    }),
)

# -------------------------------------------
# New rules should be added above this target.
# -------------------------------------------

# TensorFlow uses several libraries that may also be used by applications
# linking against the C and C++ APIs (such as libjpeg).  When we create
# the shared library, only export the core TF API functions to avoid

  ASSERT_TRUE(worker_server2->Start().ok());
  const char dev2_name[] = "/job:localhost/replica:0/task:2/device:CPU:0";

  if (inject_error) {
    // Inject a function optimization pass failure when it sees the
    // 'read0_maybe_with_graph_error' op having a requested device `dev2_name`.
    // During execution:
    //   * task:0 processes main function `VariableAddFunctionWithGraphError`