lru_list_.push_front(key);
  lra_list_.push_front(key);
  new_entry->lru_iterator = lru_list_.begin();
  new_entry->lra_iterator = lra_list_.begin();
  new_entry->timestamp = timer_seconds_();
  block_map_.emplace(std::make_pair(key, new_entry));
  return new_entry;
}

// Remove blocks from the cache until we do not exceed our maximum size.
void RamFileBlockCache::Trim() {
  while (!lru_list_.empty() && cache_size_ > max_bytes_) {

      if (!status.ok()) out_tensor->reset();
    }
  }
  if (!status.ok()) {
    tsl::Set_TF_Status_from_Status(out_status, status);
  }
}

std::pair<std::unique_ptr<TensorSliceReader::VarToShapeMap>,
          std::unique_ptr<TensorSliceReader::VarToDataTypeMap>>
CheckpointReader::BuildV2VarMaps() {
  CHECK(v2_reader_ != nullptr);
  CHECK(v2_reader_->status().ok());

 private:
  // Uses "v2_reader_" to build "var name -> shape" and "var name -> data type"
  // maps; both owned by caller.
  // REQUIRES: "v2_reader_ != nullptr && v2_reader_.status().ok()".
  std::pair<std::unique_ptr<TensorSliceReader::VarToShapeMap>,
            std::unique_ptr<TensorSliceReader::VarToDataTypeMap> >
  BuildV2VarMaps();

  // Invariant: exactly one of "reader_" and "v2_reader_" is non-null.

## Our Responsibilities

Project maintainers are responsible for clarifying the standards of acceptable behavior and are expected to take appropriate and fair corrective action in response to any instances of unacceptable behavior.

  tensorflow::ImportGraphDefResults results;
  TF_RETURN_IF_ERROR(
      ImportGraphDef(opts, gdef, dst_graph, dst_refiner, &results));

  for (const auto& pair : results.return_tensors) {
    return_nodes->emplace_back(pair.first, pair.second);
  }
  return absl::OkStatus();
}

bool ValidateConstWhileParams(const TF_WhileParams& params, TF_Status* s) {

// Sets `output` based on `proto`, which should be a serialized
// CppShapeInferenceResult::HandleData proto. `output` should be a resource
// or variant tensor.
// NOTE(skyewm): `proto` is passed a void*/size_t pair instead of a std::string
// because I couldn't get SWIG to work otherwise.
TF_CAPI_EXPORT extern void TF_SetHandleShapeAndType(TF_Graph* graph,
                                                    TF_Output output,

  // users want to compute multiple gradients over the same tape.
  explicit GradientTape(bool persistent) : persistent_(persistent) {}
  ~GradientTape() {
    for (const auto& pair : op_tape_) {
      pair.second.backward_function_deleter(pair.second.backward_function);
    }
  }

  // Returns whether any tensor in a list of tensors is being watched and has
  // a trainable dtype.

  if (ret) ret = attr_value->ParseFromArray(buffer->data, buffer->length);
  TF_DeleteBuffer(buffer);
  return ret;
}

std::vector<std::pair<string, string>> GetGradDefs(
    const tensorflow::GraphDef& graph_def) {
  std::vector<std::pair<string, string>> grads;
  for (const tensorflow::GradientDef& grad : graph_def.library().gradient()) {
    grads.emplace_back(grad.function_name(), grad.gradient_func());
  }

      ABSL_EXCLUSIVE_LOCKS_REQUIRED(mu_) {
    lru_list_.push_front(key);
    Entry entry{timer_seconds_(), value, lru_list_.begin()};
    auto insert = cache_.insert(std::make_pair(key, entry));
    if (!insert.second) {
      lru_list_.erase(insert.first->second.lru_iterator);
      insert.first->second = entry;
    } else if (max_entries_ > 0 && cache_.size() > max_entries_) {

    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(
      vspace, target_tensor_ids, source_tensor_ids, sources_that_are_targets,