}

TFE_MonitoringBuckets* TFE_MonitoringNewExponentialBuckets(double scale,
                                                           double growth_factor,
                                                           int bucket_count) {
  return new TFE_MonitoringBuckets([scale, growth_factor, bucket_count]() {
    return tensorflow::monitoring::Buckets::Exponential(scale, growth_factor,

  }

  Status Compute(AbstractContext* ctx,
                 absl::Span<AbstractTensorHandle* const> grad_outputs,
                 absl::Span<AbstractTensorHandle*> grad_inputs) override {
    // TODO(vnvo2409): Add shape broadcasting
    /* Given upstream grad U and a Div op: Z = X/Y, the gradients are:
     *
     *    dX = U / Y
     *    dY = -U*X / Y^2 = (X/Y) * -U / Y = -U*Z / Y
     *
     */

    ShapeHandle shape_handle = c.output(i);
    TF_ShapeAndType& shape = output_shapes_result->items[i];
    shape.num_dims = c.Rank(shape_handle);
    if (shape.num_dims == InferenceContext::kUnknownRank) {
      shape.dims = nullptr;
      continue;
    }
    shape.dims = new int64_t[shape.num_dims];
    for (size_t j = 0; j < shape.num_dims; ++j) {
      shape.dims[j] = c.Value(c.Dim(shape_handle, j));

  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());
  const std::vector<std::unique_ptr<ParallelTensor>>& handles = *outputs;
  const std::vector<int64_t>* shape;
  Status s = handles[0]->Shape(&shape);
  ASSERT_TRUE(s.ok());
  EXPECT_EQ(0, shape->size());
}

TEST(PARALLEL_DEVICE_LIB, TestCancelOnError) {
  std::unique_ptr<TF_Status, decltype(&TF_DeleteStatus)> status(
      TF_NewStatus(), TF_DeleteStatus);

TF_CAPI_EXPORT extern void TF_DeleteShapeAndTypeListArray(
    TF_ShapeAndTypeList** shape_list_array, int num_items);

// Infer shapes for the given `op`. The arguments mimic the arguments of the
// `shape_inference::InferenceContext` constructor. Note the following:
//   - The inputs of the `op` are not used for shape inference. So, it is
//     OK to not have the inputs properly set in `op`. See `input_tensors`

cff-version: 1.2.0
message: "If you use TensorFlow in your research, please cite it using these metadata. Software is available from tensorflow.org."
title: TensorFlow, Large-scale machine learning on heterogeneous systems

    python3 -m pip install "$TF_WHEEL"
}

@test "TensorFlow is importable" {
    source /tf/venv/bin/activate
    python3 -c 'import tensorflow as tf; t1=tf.constant([1,2,3,4]); t2=tf.constant([5,6,7,8]); print(tf.add(t1,t2).shape)'
}

# Is this still useful?
@test "TensorFlow has Keras" {
    source /tf/venv/bin/activate
    python3 -c 'import sys; import tensorflow as tf; sys.exit(0 if "keras" in tf.keras.__name__ else 1)'

#include <stack>
#include <unordered_map>
#include <unordered_set>
#include <vector>

#include "tensorflow/core/config/flag_defs.h"
#include "tensorflow/core/config/flags.h"
#include "tensorflow/core/framework/tensor_shape.h"
#include "tensorflow/core/framework/types.h"
#include "tensorflow/core/lib/gtl/array_slice.h"
#include "tensorflow/core/lib/gtl/cleanup.h"
#include "tensorflow/core/lib/gtl/flatmap.h"

build:rbe_linux --linkopt=-lrt
build:rbe_linux --host_linkopt=-lrt
build:rbe_linux --linkopt=-lm
build:rbe_linux --host_linkopt=-lm

build:rbe_linux_cpu --config=rbe_linux
# Linux cpu and cuda builds share the same toolchain now.
build:rbe_linux_cpu --host_crosstool_top="@sigbuild-r2.17-clang_config_cuda//crosstool:toolchain"
build:rbe_linux_cpu --crosstool_top="@sigbuild-r2.17-clang_config_cuda//crosstool:toolchain"

namespace tensorflow {

std::string AbstractTensorHandle::DebugString() const {
  PartialTensorShape shape;
  Status s = Shape(&shape);
  std::string shape_string;
  if (!s.ok()) {
    shape_string = "<error computing shape>";
  } else {
    shape_string = shape.DebugString();
  }
  return absl::StrCat("TensorHandle(shape=", shape_string,
                      ", dtype=", DataType_Name(DataType()),