// Name
      "AssertAndCollective",
      // Args
      {"x: float", "condition: bool"},
      // Return values
      {"y: float"},
      // Attr def
      {},
      // Nodes
      {
          {{"assert"},
           "Assert",
           {"condition", "x"},
           {{"T", std::vector<DataType>{DT_FLOAT}}}},
          {{"y"},
           "CollectiveReduce",
           {"x"},

6.5.0

abstract: TensorFlow is a machine learning system that operates at large scale and in heterogeneous environments. TensorFlow uses dataflow graphs to represent computation, shared state, and the operations that mutate that state. It maps the nodes of a dataflow graph across many machines in a cluster, and within a machine across multiple computational devices, including multicore CPUs, general purpose GPUs, and custom-designed ASICs known as Tensor Processing Units (TPUs). This architecture...

    echo "# Size of $TF_WHEEL is $WHEEL_MEGABYTES / $LARGEST_OK_SIZE megabytes." >&3
    test "$WHEEL_MEGABYTES" -le "$LARGEST_OK_SIZE"
}

# Note: this runs before the tests further down the file, so TF is installed in
# the venv and the venv is active when those tests run. The venv gets cleaned
# up in teardown_file() above.
@test "Wheel is installable" {

          # returned by the API, but a sha256sum of the entire chunk of image
          # metadata. gcr.io helpfully includes it in the header of the response
          # as docker-content-digest: sha256:[digest]. Note we use egrep to
          # match exactly sha256:<hash> because curl may include a ^M symbol at
          # the end of the line.

-#  endif
-# endif
-
-__STRING_INLINE size_t __strspn_c1 (const char *__s, int __accept);
-__STRING_INLINE size_t
-__strspn_c1 (const char *__s, int __accept)
-{
-  size_t __result = 0;
-  /* Please note that __accept never can be '\0'.  */
-  while (__s[__result] == __accept)
-    ++__result;
-  return __result;
-}
-
-__STRING_INLINE size_t __strspn_c2 (const char *__s, int __accept1,
-				    int __accept2);

          'errors.\n' %
          (android_ndk_home_path, ndk_version, _SUPPORTED_ANDROID_NDK_VERSIONS))
  write_action_env_to_bazelrc('ANDROID_NDK_VERSION', ndk_version)

  # Now grab the NDK API level to use. Note that this is different from the
  # SDK API level, as the NDK API level is effectively the *min* target SDK
  # version.
  meta = open(os.path.join(android_ndk_home_path, 'meta/platforms.json'))
  platforms = json.load(meta)

build:monolithic --define framework_shared_object=false
build:monolithic --define tsl_protobuf_header_only=false
build:monolithic --experimental_link_static_libraries_once=false  # b/229868128

# Please note that MKL on MacOS is still not supported.
# If you would like to use a local MKL instead of downloading, please set the
# environment variable "TF_MKL_ROOT" every time before build.

    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`

    [Environment]::SetEnvironmentVariable(\"PATH\", $old_path + \";C:\VS\VC\Tools\MSVC\14.33.31629\bin\Hostx64\x64;C:\VS\Common7\Tools;C:\VS\MSBuild\Current\Bin\", \"Machine\");

# Add signtool.exe to the PATH. Note this path may need to be edited if updates
# are made to the Windows 10 SDK.
RUN $old_path = [Environment]::GetEnvironmentVariable(\"PATH\", \"Machine\"); \