PartialTensorShape shape;
  TF_RETURN_IF_ERROR(inputs[0]->Shape(&shape));
  if (shape.dims() != 0) {
    return errors::InvalidArgument(
        "Tensor expected to have scalar shape found rank: ", shape.dims());
  }
  return absl::OkStatus();
}

TEST_P(UnifiedAPI, TestTensorShapeScalar) {
  if (UseFunction() && UseMlir()) {
    // TODO(b/173074167): Remove this.

COPY cuda.packages.txt /cuda.packages.txt
RUN /setup.sources.sh && /setup.packages.sh /devel.packages.txt

# Install various tools.
# - bats: bash unit testing framework
#         NOTE: v1.6.0 seems to have a bug that made "git" in setup_file break
# - bazelisk: always use the correct bazel version
# - buildifier: clean bazel build deps
# - buildozer: clean bazel build deps

}  // end namespace tensorflow

// =============================================================================
// Public C API entry points
//
// These are only the generic entry points for the C API. This file does not
// have any visibility into the graph/eager implementation and is only providing
// C bindings to the abstract classes defined in the
// c_api_unified_experimental_internal.h header.
//

build --linkopt="-Wl,--undefined-version"

# Prevent double-compilation of some TF code, ref. b/183279666 (internal)
# > TF's gen_api_init_files has a genrule to run the core TensorFlow code
# > on the host machine. If we don't have --distinct_host_configuration=false,
# > the core TensorFlow code will be built once for the host and once for the
# > target platform.

// Causes the calling thread to block till all ops dispatched in this executor
// have been executed. Note that "execution" here refers to kernel execution /
// scheduling of copies, etc. Similar to sync execution, it doesn't guarantee
// that lower level device queues (like GPU streams) have been flushed.
//
// This call may not block for execution of ops enqueued concurrently with this
// call.

fi
# VERY basic check to ensure the [and-cuda] package variant is installable.
# Checks TFCI_BAZEL_COMMON_ARGS for "gpu" or "cuda", implying that the test is
# relevant. All of the GPU test machines have CUDA installed via other means,
# so I am not sure how to verify that the dependencies themselves are valid for
# the moment.
if [[ "$TFCI_BAZEL_COMMON_ARGS" =~ gpu|cuda ]]; then

  return true;
}

// Copies `src_graph` into `dst_graph`. Any node in `src_graph` with input
// `src_inputs[i]` will have that input replaced with `dst_inputs[i]`.  `prefix`
// will be prepended to copied node names. `control_deps` are nodes in
// `dst_graph` that the copied `src_graph` nodes will have control dependencies
// on. `return_nodes` are nodes in `src_graph`, and the new corresponding nodes

    // track of indices in the model's outputs are nullptr in this set.
    // The FunctionDef only outputs the non-null tensors. We later pad the
    // function op outputs to have nullptrs at the `null_indices`.
    absl::flat_hash_set<int> null_indices;
    {
      AbstractContextPtr func_ctx(BuildFunction(fn_name));
      std::vector<AbstractTensorHandle*> func_inputs;

        `allow_build_at_runtime`.
    *   Added a new parameter called `save_gpu_specific_engines` to the
        `.save()` function inside `TrtGraphConverterV2`. When `False`, the
        `.save()` function won't save any TRT engines that have been built. When
        `True` (default), the original behavior is preserved.

// have all gradients available. This set usually is a subset of the set of
// targets (not all since targets which have outputs in the tape will not have
// gradients available initially).
//
// Then we repeatedly pop an entry from the stack, run its backprop, and update
// the gradients of its inputs. Once we have computed all gradients for a single