Merge branch 'master' into tenglu/gen_onednn_version

.github/workflows/arm-cd.yml

@@ -31,15 +31,10 @@ jobs:
   build:
     if: github.repository == 'tensorflow/tensorflow' # Don't do this in forks
     runs-on: [self-hosted, linux, ARM64]
-    continue-on-error: ${{ matrix.experimental }}
     strategy:
       fail-fast: false
       matrix:
-        pyver: ['3.9', '3.10']
-        experimental: [false]
-        include:
-          - pyver: '3.11'
-            experimental: true
+        pyver: ['3.9', '3.10', '3.11', '3.12']
     steps:
       - name: Stop old running containers (if any)
         shell: bash

.github/workflows/arm-ci-extended.yml

@@ -32,7 +32,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        pyver: ['3.9', '3.10', '3.11']
+        pyver: ['3.9', '3.10', '3.11', '3.12']
     steps:
       - name: Stop old running containers (if any)
         shell: bash

.github/workflows/update-rbe.yml

@@ -117,13 +117,25 @@ jobs:
         map sigbuild-r2.16-clang-python3.10 2.16-python3.10
         map sigbuild-r2.16-clang-python3.11 2.16-python3.11
         map sigbuild-r2.16-clang-python3.12 2.16-python3.12
+        # TF 2.17
+        map sigbuild-r2.17 2.17-python3.11
+        map sigbuild-r2.17-python3.9 2.17-python3.9
+        map sigbuild-r2.17-python3.10 2.17-python3.10
+        map sigbuild-r2.17-python3.11 2.17-python3.11
+        map sigbuild-r2.17-python3.12 2.17-python3.12
+        # TF 2.17 + Clang (containers are the same, but env vars in configs.bzl are different)
+        map sigbuild-r2.17-clang 2.17-python3.11
+        map sigbuild-r2.17-clang-python3.9 2.17-python3.9
+        map sigbuild-r2.17-clang-python3.10 2.17-python3.10
+        map sigbuild-r2.17-clang-python3.11 2.17-python3.11
+        map sigbuild-r2.17-clang-python3.12 2.17-python3.12
     - name: Create Pull Request with changes
       uses: peter-evans/create-pull-request@2b011faafdcbc9ceb11414d64d0573f37c774b04 # v4.2.3
       with:
         title: Update the RBE images to the latest container versions
         committer: TensorFlow Release Automation <[email protected]>
         token: ${{ secrets.JENKINS_TOKEN }}
-        reviewers: angerson,mihaimaruseac,learning-to-play,nitins17
+        reviewers: mihaimaruseac,learning-to-play,nitins17
         body: |
           This PR was created by a GitHub Actions workflow to update all the SIG Build-based RBE containers to the most recent containers. See:
 

CONTRIBUTING.md

@@ -71,7 +71,7 @@ Before sending your pull requests, make sure you do the following:
 
 In a graphical form, the entire lifetime of a PR looks like
 
-![image](https://user-images.githubusercontent.com/323199/229561784-0a2f5509-b731-493f-ad88-bad487688c8d.png)
+![image](https://github.com/tensorflow/tensorflow/assets/52792999/3eea4ca5-daa0-4570-b0b5-2a2b03a724a3)
 
 ### Contributor License Agreements
 

RELEASE.md

@@ -26,6 +26,12 @@
 * <IF A CHANGE CLOSES A GITHUB ISSUE, IT SHOULD BE DOCUMENTED HERE>
 * <NOTES SHOULD BE GROUPED PER AREA>
 
+* GPU
+    * Support for NVIDIA GPUs with compute capability 8.9 (e.g. L4 & L40) has
+      been added to TF binary distributions (Python wheels).
+* Replace `DebuggerOptions` of TensorFlow Quantizer, and migrate to
+  `DebuggerConfig` of StableHLO Quantizer.
+
 ## Keras
 
 <INSERT SMALL BLURB ABOUT RELEASE FOCUS AREA AND POTENTIAL TOOLCHAIN CHANGES>
@@ -34,6 +40,9 @@
 
 * <DOCUMENT BREAKING CHANGES HERE>
 * <THIS SECTION SHOULD CONTAIN API, ABI AND BEHAVIORAL BREAKING CHANGES>
+* GPU
+    * Support for NVIDIA GPUs with compute capability 5.x (Maxwell generation)
+      has been removed from TF binary distributions (Python wheels).
 
 ### Known Caveats
 
@@ -46,6 +55,19 @@
 *   <INSERT MAJOR FEATURE HERE, USING MARKDOWN SYNTAX>
 *   <IF RELEASE CONTAINS MULTIPLE FEATURES FROM SAME AREA, GROUP THEM TOGETHER>
 
+*   Add `is_cpu_target_available`, which indicates whether or not TensorFlow was
+    built with support for a given CPU target. This can be useful for skipping
+    target-specific tests if a target is not supported.
+
+*   `tf.data`
+    * Support `data.experimental.distribued_save`. `distribued_save` uses
+      tf.data service
+      (https://www.tensorflow.org/api_docs/python/tf/data/experimental/service)
+      to write distributed dataset snapshots. The call is non-blocking and
+      returns without waiting for the snapshot to finish. Setting `wait=True` to
+      `tf.data.Dataset.load` allows the snapshots to be read while they are
+      being written.
+
 ### Bug Fixes and Other Changes
 
 * <SIMILAR TO ABOVE SECTION, BUT FOR OTHER IMPORTANT CHANGES / BUG FIXES>
@@ -59,6 +81,26 @@
       schema globally in the converter and inference engine. The new behaviour
       can be disabled via experimental
       flag `converter._experimental_disable_per_channel_quantization_for_dense_layers = True`.
+    * C API:
+        * The experimental `TfLiteRegistrationExternal` type has been renamed as
+          `TfLiteOperator`, and likewise for the corresponding API functions.
+    * The Python TF Lite Interpreter bindings now have an option
+      `experimental_default_delegate_latest_features` to enable all default
+      delegate features.
+
+* `tf.data`
+    * Add `wait` to `tf.data.Dataset.load`. If `True`, for snapshots written
+      with `distributed_save`, it reads the snapshot while it is being written.
+      For snapshots written with regular `save`, it waits for the snapshot until
+      it's finished. The default is `False` for backward compatibility. Users of
+      `distributed_save` are recommended to set it to `True`.
+
+* `tf.tpu.experimental.embedding.TPUEmbeddingV2`
+    * Add `compute_sparse_core_stats` for sparse core users to profile the 
+      data with this API to get the `max_ids` and `max_unique_ids`. These
+      numbers will be needed to configure the sparse core embedding mid level 
+      api.
+    * Remove the `preprocess_features` method since that's no longer needed.
 
 ## Thanks to our Contributors
 
@@ -118,8 +160,9 @@ This release contains contributions from many people at Google, as well as:
     keras
 * **Apple Silicon users:** If you previously installed TensorFlow using
     `pip install tensorflow-macos`, please update your installation method. Use
-    `pip install tensorflow` from now on. Starting with TF 2.17, the
-    `tensorflow-macos` package will no longer receive updates.
+    `pip install tensorflow` from now on.
+* **Mac x86 users:** Mac x86 builds are being deprecated and will no longer be
+  released as a Pip package from TF 2.17 onwards.
 
 ### Known Caveats
 
@@ -154,6 +197,25 @@ This release contains contributions from many people at Google, as well as:
     * Added support for `stablehlo.maximum`.
     * Added support for `stablehlo.minimum`.
     * Added boolean parameter support for `tfl.gather_nd`.
+    * C API:
+        * New API functions:
+            * `tensorflow/lite/c/c_api_experimental.h`:
+                * `TfLiteInterpreterGetVariableTensorCount`
+                * `TfLiteInterpreterGetVariableTensor`
+                * `TfLiteInterpreterGetBufferHandle`
+                * `TfLiteInterpreterSetBufferHandle`
+            * `tensorflow/lite/c/c_api_opaque.h`:
+                * `TfLiteOpaqueTensorSetAllocationTypeToDynamic`
+        * API functions promoted from experimental to stable:
+            * `tensorflow/lite/c/c_api.h`:
+                * `TfLiteInterpreterOptionsEnableCancellation`
+                * `TfLiteInterpreterCancel`
+    * C++ API:
+        * New virtual methods in the `tflite::SimpleDelegateInterface` class in `tensorflow/lite/delegates/utils/simple_delegate.h`,
+          and likewise in the `tflite::SimpleOpaqueDelegateInterface` class in `tensorflow/lite/delegates/utils/simple_opaque_delegate.h`:
+            * `CopyFromBufferHandle`
+            * `CopyToBufferHandle`
+            * `FreeBufferHandle`
 
 * `tf.train.CheckpointOptions` and `tf.saved_model.SaveOptions`
     * These now take in a new argument called `experimental_sharding_callback`.
@@ -229,7 +291,7 @@ This release contains contributions from many people at Google, as well as:
 
 This release contains contributions from many people at Google, as well as:
 
-<INSERT>, <NAME>, <HERE>, <USING>, <GITHUB>, <HANDLE>
+RoboTux, <INSERT>, <NAME>, <HERE>, <USING>, <GITHUB>, <HANDLE>
 
 # Release 2.15.0.post1
 

WORKSPACE

@@ -1,5 +1,9 @@
+# buildifier: disable=load-on-top
+
 workspace(name = "org_tensorflow")
 
+# buildifier: disable=load-on-top
+
 # We must initialize hermetic python first.
 load("@bazel_tools//tools/build_defs/repo:http.bzl", "http_archive")
 
@@ -12,18 +16,25 @@ http_archive(
     ],
 )
 
+http_archive(
+    name = "rules_java",
+    sha256 = "c73336802d0b4882e40770666ad055212df4ea62cfa6edf9cb0f9d29828a0934",
+    url = "https://github.com/bazelbuild/rules_java/releases/download/5.3.5/rules_java-5.3.5.tar.gz",
+)
+
 http_archive(
     name = "rules_python",
     sha256 = "9d04041ac92a0985e344235f5d946f71ac543f1b1565f2cdbc9a2aaee8adf55b",
     strip_prefix = "rules_python-0.26.0",
     url = "https://github.com/bazelbuild/rules_python/releases/download/0.26.0/rules_python-0.26.0.tar.gz",
 )
 
+# buildifier: disable=same-origin-load
 load("@rules_python//python:repositories.bzl", "py_repositories")
 
 py_repositories()
 
-load("@rules_python//python:repositories.bzl", "python_register_toolchains")
+load("@rules_python//python:repositories.bzl", "python_register_toolchains")  # buildifier: disable=same-origin-load
 load(
     "//tensorflow/tools/toolchains/python:python_repo.bzl",
     "python_repository",

ci/official/envs/ci_default

@@ -64,5 +64,6 @@ TFCI_PYTHON_VERSION=
 TFCI_WHL_AUDIT_ENABLE=
 TFCI_WHL_AUDIT_PLAT=
 TFCI_WHL_BAZEL_TEST_ENABLE=
+TFCI_WHL_IMPORT_TEST_ENABLE=1
 TFCI_WHL_SIZE_LIMIT=
 TFCI_WHL_SIZE_LIMIT_ENABLE=

ci/official/envs/linux_x86_tpu

@@ -18,5 +18,6 @@ TFCI_BAZEL_TARGET_SELECTING_CONFIG_PREFIX=linux_tpu
 TFCI_BUILD_PIP_PACKAGE_ARGS="--repo_env=WHEEL_NAME=tensorflow_tpu"
 TFCI_LIB_SUFFIX="-tpu-linux-x86_64"
 TFCI_WHL_BAZEL_TEST_ENABLE=0
+TFCI_WHL_IMPORT_TEST_ENABLE=0
 TFCI_WHL_SIZE_LIMIT=580M
 TFCI_PYTHON_VERIFY_PIP_INSTALL_ARGS="-f https://storage.googleapis.com/libtpu-tf-releases/index.html"

ci/official/requirements_updater/BUILD.bazel

@@ -13,10 +13,10 @@
 # limitations under the License.
 # ==============================================================================
 
-load("@python//3.9:defs.bzl", compile_pip_requirements_3_9 = "compile_pip_requirements")
 load("@python//3.10:defs.bzl", compile_pip_requirements_3_10 = "compile_pip_requirements")
 load("@python//3.11:defs.bzl", compile_pip_requirements_3_11 = "compile_pip_requirements")
 load("@python//3.12:defs.bzl", compile_pip_requirements_3_12 = "compile_pip_requirements")
+load("@python//3.9:defs.bzl", compile_pip_requirements_3_9 = "compile_pip_requirements")
 load("@updater_config_repository//:updater_config_repository.bzl", "REQUIREMENTS_FILE_NAME")
 
 compile_pip_requirements_3_9(

ci/official/requirements_updater/WORKSPACE

@@ -1,5 +1,9 @@
+# buildifier: disable=load-on-top
+
 workspace(name = "requirements_updater")
 
+# buildifier: disable=load-on-top
+
 load("@bazel_tools//tools/build_defs/repo:http.bzl", "http_archive")
 
 http_archive(
@@ -18,11 +22,12 @@ http_archive(
     url = "https://github.com/bazelbuild/rules_python/releases/download/0.26.0/rules_python-0.26.0.tar.gz",
 )
 
+# buildifier: disable=same-origin-load
 load("@rules_python//python:repositories.bzl", "py_repositories")
 
 py_repositories()
 
-load("@rules_python//python:repositories.bzl", "python_register_multi_toolchains")
+load("@rules_python//python:repositories.bzl", "python_register_multi_toolchains")  # buildifier: disable=same-origin-load
 load("@rules_python//python/pip_install:repositories.bzl", "pip_install_dependencies")
 
 default_python_version = "3.10"

ci/official/utilities/code_check_full.bats

@@ -101,7 +101,7 @@ EOF
 
 @test "Pip package generated license includes all dependencies' licenses" {
   do_external_licenses_check \
-    "//tensorflow/tools/pip_package:build_pip_package" \
+    "//tensorflow/tools/pip_package:wheel" \
     "//tensorflow/tools/pip_package:licenses"
 }
 
@@ -172,7 +172,7 @@ EOF
 
   # Get the full list of files and targets which get included into the pip
   # package
-  bazel cquery --keep_going 'deps(//tensorflow/tools/pip_package:build_pip_package)' | sort -u > $BATS_TEST_TMPDIR/pip_deps
+  bazel cquery --keep_going 'deps(//tensorflow/tools/pip_package:wheel)' | sort -u > $BATS_TEST_TMPDIR/pip_deps
   # Find all Python py_test targets not tagged "no_pip" or "manual", excluding
   # any targets in ignored packages. Combine this list of targets into a bazel
   # query list (e.g. the list becomes "target+target2+target3")
@@ -215,7 +215,7 @@ EOF
   bazel cquery \
     --experimental_cc_shared_library \
     --@local_config_cuda//:enable_cuda \
-    "somepath(//tensorflow/tools/pip_package:build_pip_package, " \
+    "somepath(//tensorflow/tools/pip_package:wheel, " \
     "@local_config_cuda//cuda:cudart + "\
     "@local_config_cuda//cuda:cudart + "\
     "@local_config_cuda//cuda:cuda_driver + "\
@@ -225,7 +225,7 @@ EOF
     "@local_config_tensorrt//:tensorrt)" --keep_going > $BATS_TEST_TMPDIR/out
 
   cat <<EOF
-There was a path found connecting //tensorflow/tools/pip_package:build_pip_package
+There was a path found connecting //tensorflow/tools/pip_package:wheel
 to a banned CUDA dependency. Here's the output from bazel query:
 EOF
   cat $BATS_TEST_TMPDIR/out
@@ -237,7 +237,7 @@ EOF
     --experimental_cc_shared_library \
     --@local_config_cuda//:enable_cuda \
     --define framework_shared_object=false \
-    "somepath(//tensorflow/tools/pip_package:build_pip_package, " \
+    "somepath(//tensorflow/tools/pip_package:wheel, " \
     "@local_config_cuda//cuda:cudart + "\
     "@local_config_cuda//cuda:cudart + "\
     "@local_config_cuda//cuda:cuda_driver + "\
@@ -247,7 +247,7 @@ EOF
     "@local_config_tensorrt//:tensorrt)" --keep_going > $BATS_TEST_TMPDIR/out
 
   cat <<EOF
-There was a path found connecting //tensorflow/tools/pip_package:build_pip_package
+There was a path found connecting //tensorflow/tools/pip_package:wheel
 to a banned CUDA dependency when '--define framework_shared_object=false' is set.
 This means that a CUDA target was probably included via an is_static condition,
 used when targeting platforms like Windows where we build statically instead

ci/official/utilities/rename_and_verify_wheels.sh

@@ -58,8 +58,10 @@ venv=$(mktemp -d)
 "python${TFCI_PYTHON_VERSION}" -m venv "$venv"
 python="$venv/bin/python3"
 "$python" -m pip install *.whl $TFCI_PYTHON_VERIFY_PIP_INSTALL_ARGS
-"$python" -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)'
-"$python" -c 'import sys; import tensorflow as tf; sys.exit(0 if "keras" in tf.keras.__name__ else 1)'
+if [[ "$TFCI_WHL_IMPORT_TEST_ENABLE" == "1" ]]; then
+  "$python" -c 'import tensorflow as tf; t1=tf.constant([1,2,3,4]); t2=tf.constant([5,6,7,8]); print(tf.add(t1,t2).shape)'
+  "$python" -c 'import sys; import tensorflow as tf; sys.exit(0 if "keras" in tf.keras.__name__ else 1)'
+fi
 # VERY basic check to ensure the [and-cuda] package variant is installable.
 # Checks TFCI_BAZEL_COMMON_ARGS for "gpu" or "cuda", implying that the test is
 # relevant. All of the GPU test machines have CUDA installed via other means,

ci/official/wheel.sh

@@ -27,7 +27,7 @@ if [[ "$TFCI_NIGHTLY_UPDATE_VERSION_ENABLE" == 1 ]]; then
   export TFCI_BUILD_PIP_PACKAGE_ARGS="$(echo $TFCI_BUILD_PIP_PACKAGE_ARGS | sed 's/tensorflow/tf_nightly/')"
 fi
 
-tfrun bazel build $TFCI_BAZEL_COMMON_ARGS //tensorflow/tools/pip_package/v2:wheel $TFCI_BUILD_PIP_PACKAGE_ARGS
+tfrun bazel build $TFCI_BAZEL_COMMON_ARGS //tensorflow/tools/pip_package:wheel $TFCI_BUILD_PIP_PACKAGE_ARGS
 tfrun find ./bazel-bin/tensorflow/tools/pip_package -iname "*.whl" -exec cp {} $TFCI_OUTPUT_DIR \;
 tfrun ./ci/official/utilities/rename_and_verify_wheels.sh
 

ci/official/wheel_test/WORKSPACE

@@ -1,5 +1,9 @@
+# buildifier: disable=load-on-top
+
 workspace(name = "wheel_test")
 
+# buildifier: disable=load-on-top
+
 load("@bazel_tools//tools/build_defs/repo:http.bzl", "http_archive")
 
 http_archive(
@@ -18,6 +22,7 @@ http_archive(
     url = "https://github.com/bazelbuild/rules_python/releases/download/0.26.0/rules_python-0.26.0.tar.gz",
 )
 
+# buildifier: disable=same-origin-load
 load("@rules_python//python:repositories.bzl", "py_repositories")
 
 py_repositories()
@@ -38,7 +43,7 @@ python_repository(name = "python_version_repo")
 load("@python_version_repo//:py_version.bzl", "TF_PYTHON_VERSION")
 
 # Register multi toolchains
-load("@rules_python//python:repositories.bzl", "python_register_toolchains")
+load("@rules_python//python:repositories.bzl", "python_register_toolchains")  # buildifier: disable=same-origin-load
 
 python_register_toolchains(
     name = "python",