graph_def->length));
  TF_Buffer graph_def_buffer;
  graph_def_buffer.data = reinterpret_cast<const void*>(&graph_def_proto);
  graph_def_buffer.length = sizeof(tensorflow::GraphDef*);
  TF_ImportGraphDefResults* results =
      TF_GraphImportGraphDefWithResultsNoSerialization(graph, &graph_def_buffer,
                                                       opts, s);

  }
  *result = TF_TensorFromTensor(**status_or, &status->status);
  return status->status.ok();
}

TF_ApiDefMap* TF_NewApiDefMap(TF_Buffer* op_list_buffer, TF_Status* status) {
  tensorflow::OpList op_list;
  if (!op_list.ParseFromArray(op_list_buffer->data, op_list_buffer->length)) {
    status->status = InvalidArgument("Unparseable OpList");
    return nullptr;
  }
  status->status = absl::OkStatus();

_compat.enable_v2_behavior()
_major_api_version = 2


# Load all plugin libraries from site-packages/tensorflow-plugins if we are
# running under pip.
# TODO(gunan): Find a better location for this code snippet.
from tensorflow.python.framework import load_library as _ll
from tensorflow.python.lib.io import file_io as _fi

# Get sitepackages directories for the python installation.

        ],
    }),
)

cc_library(
    name = "tf_buffer_hdrs",
    hdrs = [
        "tf_buffer.h",
    ],
    visibility = ["//visibility:public"],
    deps = [
        ":c_api_macros_hdrs",
    ],
)

cc_library(
    name = "tf_buffer",
    srcs = [
        "tf_buffer.cc",
    ],
    hdrs = [
        "tf_buffer.h",
    ],
    copts = tf_copts(),

// Serialize `attrs` as a tensorflow::NameAttrList protocol buffer (into `buf`),
// containing the op name and a map of its attributes.
TF_CAPI_EXPORT extern void TFE_OpAttrsSerialize(const TFE_OpAttrs* attrs,
                                                TF_Buffer* buf,
                                                TF_Status* status);

// Set an op's attribute from a serialized AttrValue protocol buffer.
//

_current_module.nn.__path__ = [_module_dir] + _current_module.nn.__path__

del importlib

# Load all plugin libraries from site-packages/tensorflow-plugins if we are
# running under pip.
# TODO(gunan): Find a better location for this code snippet.
from tensorflow.python.framework import load_library as _ll
from tensorflow.python.lib.io import file_io as _fi

# Get sitepackages directories for the python installation.

// Obtains key-value from coordination service agent. The returned `TF_Buffer`
// is a newly allocated buffer to hold the string key-value, and caller is
// responsible for managing the lifetime. If error, `status` will be set and a
// nullptr will be returned.
TF_CAPI_EXPORT extern TF_Buffer* TF_CoordinationServiceGetKeyValue(
    const char* key, int64_t key_size, TF_CoordinationServiceAgent* agent,

*   Added a distributed runtime using GRPC
*   Move skflow to `contrib/learn`
*   Better linear optimizer in `contrib/linear_optimizer`
*   Random forest implementation in `contrib/tensor_forest`
*   CTC loss and decoders in `contrib/ctc`
*   Basic support for `half` data type
*   Better support for loading user ops (see examples in `contrib/`)
*   Allow use of (non-blocking) Eigen threadpool with

                                      const char* value) {
  cell->cell.Set({value});
}

const void TFE_MonitoringStringGaugeCellValue(
    TFE_MonitoringStringGaugeCell* cell, TF_Buffer* buf) {
  tensorflow::string value = cell->cell.value();
  void* data = tensorflow::port::Malloc(value.length());
  value.copy(static_cast<char*>(data), value.length(), 0);
  buf->data = data;

    }
    TF_Status c_status;
    std::unique_ptr<TF_Buffer, decltype(&TF_DeleteBuffer)> summary_buffer(
        methods_.summarize(data_, &c_status), TF_DeleteBuffer);
    if (!c_status.status.ok()) {
      return c_status.status;
    }
    summary = std::string(reinterpret_cast<const char*>(summary_buffer->data),
                          summary_buffer->length);
    return absl::OkStatus();
  }

 private: