status_ =
          BuildImmediateExecutionContext(std::get<1>(GetParam()), &ctx_raw);
      ASSERT_EQ(errors::OK, status_.code()) << status_.message();
      immediate_execution_ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically
    // unstable. Some forward pass tests also fail with TensorFloat-32 due to
    // low tolerances

    ASSERT_EQ(errors::OK, s.code()) << s.message();
    ctx.reset(ctx_raw);
  }

  AbstractTensorHandlePtr x;
  {
    AbstractTensorHandle* x_raw = nullptr;
    Status s = TestScalarTensorHandle<float, TF_FLOAT>(ctx.get(), 2.0f, &x_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();
    x.reset(x_raw);
  }

  Status s = RunModel(TestScalarShape, ctx.get(),

    ASSERT_EQ(errors::OK, status_.code()) << status_.message();
    x1.reset(x1_raw);
  }

  AbstractTensorHandlePtr x2;
  {
    AbstractTensorHandle* x2_raw = nullptr;
    status_ = TestScalarTensorHandle<float, TF_FLOAT>(
        immediate_execution_ctx_.get(), 1.0f, &x2_raw);
    ASSERT_EQ(errors::OK, status_.code()) << status_.message();
    x2.reset(x2_raw);
  }

 public:
  explicit GraphOperation(TF_Graph* g) : TracingOperation(kGraph), g_(g) {}
  void Release() override { delete this; }
  Status Reset(const char* op, const char* raw_device_name) override {
    if (op_) {
      return errors::FailedPrecondition("Reset called on already built op.");
    }
    if (raw_device_name) {
      device_name_ = raw_device_name;
    }
    op_type_ = op;

                             ->Finalize(&output_list, &func));
      scoped_func.reset(func);
      for (auto output : output_list.outputs) {
        output->Unref();
      }
      TF_RETURN_IF_ERROR(ctx->RegisterFunction(func));
    }

    AbstractOperationPtr fn_op(ctx->CreateOperation());
    TF_RETURN_IF_ERROR(fn_op->Reset(fn_name, /*raw_device_name=*/nullptr));
    for (auto input : inputs) {

    ASSERT_EQ(errors::OK, s.code()) << s.message();
    ctx.reset(ctx_raw);
  }

  AbstractTensorHandlePtr x;
  {
    AbstractTensorHandle* x_raw = nullptr;
    Status s = TestScalarTensorHandle<float, TF_FLOAT>(ctx.get(), 1.0f, &x_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();
    x.reset(x_raw);
  }

  // Pseudo-code:
  //
  // tape.watch(x)
  // y = exp(x)

using tensorflow::string;

void TFE_OpReset(TFE_Op* op_to_reset, const char* op_or_function_name,
                 const char* raw_device_name, TF_Status* status) {
  if (op_to_reset) {
    tensorflow::ImmediateExecutionOperation* op =
        tensorflow::unwrap(op_to_reset);
    op->Clear();
    status->status = op->Reset(op_or_function_name, raw_device_name);
  } else {
    TF_SetStatus(status, TF_INVALID_ARGUMENT,

}

void StringVectorToArrays(const std::vector<string>& v,
                          std::unique_ptr<const void*[]>* ptrs,
                          std::unique_ptr<size_t[]>* lens) {
  ptrs->reset(new const void*[v.size()]);
  lens->reset(new size_t[v.size()]);
  for (size_t i = 0; i < v.size(); ++i) {
    (*ptrs)[i] = v[i].data();
    (*lens)[i] = v[i].size();
  }
}

                            /*input_index=*/i, use_function,
                            &numerical_grad_raw);
      ASSERT_EQ(errors::OK, s.code()) << s.message();
      numerical_grad.reset(numerical_grad_raw);
    }

    TF_Tensor* numerical_tensor;
    s = GetValue(numerical_grad.get(), &numerical_tensor);
    ASSERT_EQ(errors::OK, s.code()) << s.message();

  for (int i = 0; i < ncontainers; ++i) {
    container_names[i] = containers[i];
  }

  status->status = Reset(opt->options, container_names);
}

extern "C" {

void TF_Reset(const TF_SessionOptions* opt, const char** containers,
              int ncontainers, TF_Status* status) {
  TF_Reset_Helper(opt, containers, ncontainers, status);
}

}  // end extern "C"

namespace tensorflow {