# and don't print duplicates
  awk '/Streaming build results to/ {print $NF}' "$TFCI_OUTPUT_DIR/script.log" | uniq
}

# Print out any ResultStore URLs for Bazel invocations' results.
# Each failed target there will have its own representation, making failures
# easier to find and read.
function resultstore_extract {
  local \
    XML_PATH="$TFCI_OUTPUT_DIR/Bazel_Test_and_Build_Results/sponge_log.xml"

  python3 \

  /// `max_age` is also 0, the cache will not be populated).
  ExpiringLRUCache(uint64_t max_age, size_t max_entries,
                   std::function<uint64_t()> timer_seconds = TF_NowSeconds)
      : max_age_(max_age),
        max_entries_(max_entries),
        timer_seconds_(timer_seconds) {}

  /// Insert `value` with key `key`. This will replace any previous entry with
  /// the same key.

# ==============================================================================

# script to run pip-compile for keras, tensorboard, estimator deps.
# if there is a change in requirements.in then all lock files will be updated
# accordingly.

# -e: abort script if one command fails
# -u: error if undefined variable used
# -o pipefail: entire command fails if pipe fails. watch out for yes | ...
# -o history: record shell history

# and click "Run Workflow." Leave "Use Workflow From" set to "master", then
# input the branch name and paste the cherry-pick commit and click Run. A PR
# will be created.

name: Release Branch Cherrypick
on:
  workflow_dispatch:
    inputs:
      # We use this instead of the "run on branch" argument because GitHub looks

  // TODO(srbs): It seems like this is used only for performance optimization
  // and not for correctness. The only downside of keeping this 1 seems to be
  // that the gradient accumulation is unbounded and we will never
  // aggressively aggregate accumulated gradients to recover memory.
  // Revisit and fix.
  return 1;
}

// Consumes references to the tensors in the gradient_tensors list and returns

lit ~= 17.0.2
opt_einsum == 3.3.0
astunparse == 1.6.3
dill == 0.3.7
astor == 0.7.1
typing_extensions == 4.8.0
gast == 0.4.0
termcolor == 2.3.0
wrapt == 1.16.0
tblib == 2.0.0

# Install tensorboard, and keras
# Note that here we want the latest version that matches TF major.minor version
# Note that we must use nightly here as these are used in nightly jobs
# For release jobs, we will pin these on the release branch

enum ContextDevicePlacementPolicy {
  // Running operations with input tensors on the wrong device will fail.
  DEVICE_PLACEMENT_EXPLICIT = 0,
  // Copy the tensor to the right device but log a warning.
  DEVICE_PLACEMENT_WARN = 1,
  // Silently copy the tensor, which has a performance cost since the operation
  // will be blocked till the copy completes. This is the default policy.
  DEVICE_PLACEMENT_SILENT = 2,

      gcs_file->buffer.resize(gcs_file->buffer_size);
      auto read_fill_buffer = gcs_file->read_fn(
          gcs_file->path, gcs_file->buffer_start, gcs_file->buffer_size,
          &(gcs_file->buffer[0]), status);
      gcs_file->buffer_end_is_past_eof =
          (TF_GetCode(status) == TF_OUT_OF_RANGE);
      if (read_fill_buffer >= 0) gcs_file->buffer.resize(read_fill_buffer);
      if (TF_GetCode(status) != TF_OK &&

/// pointed to by the `void*` members is always owned by the plugin. The plugin
/// will provide functions to call to allocate and deallocate this data (see
/// next sections) and core TensorFlow ensures to call these at the proper time.
///
/// Plugins will never receive a `TF_*` pointer that is `nullptr`. Core
/// TensorFlow will never touch the `void*` wrapped by these structures, except
/// to initialize it as `nullptr`.

    update your imports accordingly, the old files will be removed in Release
    2.11.
*   `tf.keras.optimizers.experimental.Optimizer` will graduate in Release 2.11,
    which means `tf.keras.optimizers.Optimizer` will be an alias of
    `tf.keras.optimizers.experimental.Optimizer`. The current
    `tf.keras.optimizers.Optimizer` will continue to be supported as
    `tf.keras.optimizers.legacy.Optimizer`,