}
    inputs_.push_back(t->output_);
    *output = tensorflow::down_cast<TracingTensorHandle*>(outputs[0]);
    return absl::OkStatus();
  }

  absl::Status Finalize(OutputList* outputs, AbstractFunction** f) override {
    std::vector<TF_Output> graph_outputs;
    graph_outputs.reserve(outputs->outputs.size());
    for (auto* abstract_output : outputs->outputs) {
      GraphTensor* output = dyn_cast<GraphTensor>(abstract_output);

   *
   *   A =[[0, 1],[1, 0]]
   *   B =[[1, 0],[0, 1]]
   *
   *   output1, output2, output3 = two_adds_and_mm(A, B)
   *
   * We expect outputs:
   *
   *   output1 = [[1, 1],[1, 1]]
   *   output2 = [[2, 0],[0, 2]]
   *   output3 = [[2, 2],[2, 2]]
   *
   */

  // Build eager context.
  TFE_ContextOptions* opts = TFE_NewContextOptions();

Jacek Ewertowski <******@****.***> 1730145173 +0100

Jacek Ewertowski <******@****.***> 1730145173 +0100

  output_tensors->reserve(noutputs);
  for (int i = 0; i < noutputs; ++i) {
    Node* node = outputs[i].oper ? &outputs[i].oper->node : nullptr;
    int idx = outputs[i].index;
    TF_RETURN_WITH_CONTEXT_IF_ERROR(
        fn_body->graph.IsValidOutputTensor(node, idx),
        "Encountered while processing output ", i, " from function '", fn_name,
        "'");
    TF_RETURN_WITH_CONTEXT_IF_ERROR(ValidateNonRefOutput(node, idx),

  }
  return outputs;
}

void TFOutputsFromOutputs(const std::vector<tensorflow::Output>& outputs,
                          TF_Output* tf_outputs) {
  for (int i = 0; i < outputs.size(); i++) {
    tf_outputs[i].oper = ToOperation(outputs[i].node());
    tf_outputs[i].index = outputs[i].index();
  }
}
#endif  // !defined(IS_MOBILE_PLATFORM) && !defined(IS_SLIM_BUILD)

  VLOG(1) << "Running the dequeue op";
  TF_Output output{dequeue_op, 0};
  TF_Tensor* ret;
  TF_SessionRun(session, /*run_options*/ nullptr,
                // input related parameters
                /*inputs*/ nullptr, /*input_values*/ nullptr, /*ninputs*/ 0,
                // output related parameters
                /*outputs*/ &output, /*output_values*/ &ret,
                /*noutputs*/ 1,

    // Convert back to a C++ Tensor and ensure we get expected output.
    Tensor output;
    ASSERT_EQ(absl::OkStatus(), TF_TensorToTensor(dst, &output)) << line;
    ASSERT_EQ(src.NumElements(), output.NumElements()) << line;
    for (int64_t i = 0; i < src.NumElements(); ++i) {
      ASSERT_EQ(data[i], output.flat<tstring>()(i)) << line;
    }

    TF_DeleteTensor(dst);
  }
}

  int num_dims_out = TF_NumDims(model_out_tensor);
  TF_DeleteTensor(model_out_tensor);

  // If the output is a scalar, then return the scalar output
  if (num_dims_out == 0) {
    outputs[0] = model_out.release();
    return absl::OkStatus();
  }

  // Else, reduce sum the output to get a scalar

  // Will sum all dimensions, so get a Tensor containing [0,...,num_dims_out-1].
  AbstractTensorHandlePtr sum_dims;

  //
  // This is a no-op if the tensor is already being watched either from an
  // earlier call to `GradientTape::Watch` or being an output of an op with
  // watched inputs.
  void Watch(const AbstractTensorHandle*);
  // Records an operation with given inputs and outputs
  // on the tape and marks all its outputs as watched if at
  // least one input of the op is watched and has a trainable dtype.