while (!tensor_stack.empty()) {
    int64_t tensor_id = tensor_stack.back();
    tensor_stack.pop_back();
    auto op_id_it = tensor_tape.find(tensor_id);
    if (op_id_it == tensor_tape.end()) {
      continue;
    }
    int64_t op_id = op_id_it->second;
    auto op_it = op_tape->find(op_id);
    auto result_op_it = result.op_tape.find(op_id);
    if (op_id == -1 || op_it == op_tape->end() ||

                                                       TF_Status* status);

// On success, enqueues `tensor` into a TF-managed FifoQueue given by
// `tensor_id`, associated with `session`. There must be a graph node named
// "fifo_queue_enqueue_<tensor_id>", to be executed by this API call. It reads
// from a placeholder node "arg_tensor_enqueue_<tensor_id>".
//

  }
  return ret;
}

TF_Tensor* TF_DequeueNamedTensor(TF_Session* session, int tensor_id,
                                 TF_Status* status) {
  assert(session);
  {
    tensorflow::mutex_lock c(session->graph->mu);
    VLOG(1) << "Dequeuing named tensor with id " << tensor_id
            << ", with input graph: "
            << session->graph->graph.ToGraphDefDebug().DebugString();
  }

          # Also get the current date to do cache busting. Assumes one day
          # is an ok range for rebuilds
          echo "DATE=$(date +'%Y-%m-%d')" >> "$GITHUB_OUTPUT"
        id: tf-version
      -
        name: Build and push
        id: docker_build
        uses: docker/build-push-action@c56af957549030174b10d6867f20e78cfd7debc5 # v3.2.0
        with:
          push: true

  tensorflow::DeviceNameUtils::ParseFullName(device_name, &parsed_name);
  std::string device_type = parsed_name.type;
  int device_id = 0;
  if (parsed_name.has_id) {
    device_id = parsed_name.id;
  }

  ctx.device_id = device_id;
  if (device_type == "CPU") {
    ctx.device_type = DLDeviceType::kDLCPU;
  } else if (device_type == "GPU") {
#if TENSORFLOW_USE_ROCM

-   type: input id: Bazel attributes: label: Bazel version description: if
    compiling from source placeholder: validations: required: false

-   type: input id: Compiler attributes: label: GCC/Compiler version
    description: if compiling from source placeholder: validations: required:
    false

-   type: input id: Cuda attributes: label: CUDA/cuDNN version description:
    placeholder: validations: required: false

   pid_t ut_pid;			/* Process ID of login process.  */
-  char ut_line[UT_LINESIZE];	/* Devicename.  */
-  char ut_id[4];		/* Inittab ID.  */
-  char ut_user[UT_NAMESIZE];	/* Username.  */
-  char ut_host[UT_HOSTSIZE];	/* Hostname for remote login.  */
+  char ut_line[UT_LINESIZE]
+    __attribute_nonstring__;	/* Devicename.  */
+  char ut_id[4]
+    __attribute_nonstring__;		/* Inittab ID.  */

  for (int i = 0; i < target_tensor_ids.size(); ++i) {
    int64_t target_id = target_tensor_ids[i];
    if (sources_set.find(target_id) != sources_set.end()) {
      auto tensor = targets[i];
      sources_that_are_targets.insert(
          std::make_pair(target_id, TapeTensor(tensor)));
    }
  }

  TF_RETURN_IF_ERROR(GradientTape::ComputeGradient(

// Wrapper for a tensor output of an operation executing under a tape.
//
// `GetID` returns a unique id for the wrapped tensor which is used to maintain
// a map (`tensorflow::eager::TensorTape`) from the wrapped tensor to the id of
// the op that produced it (or -1 if this tensor was watched using
// `GradientTape::Watch`.) The op_id is simply a unique index assigned to each
// op executed under the tape. A separate map (`tensorflow::eager::OpTape`)

  TF_OperationDescription* id_descr =
      TF_NewOperation(function_graph, "Identity", "id");
  TF_SetAttrType(id_descr, "T", TF_INT32);
  TF_AddInput(id_descr, {arg, 0});
  TF_Operation* id = TF_FinishOperation(id_descr, status);
  ASSERT_TRUE(TF_GetCode(status) == TF_OK) << TF_Message(status);
  TF_Output input{arg, 0};
  TF_Output output{id, 0};
  TF_Function* fn =