Google Git
Sign in
chromium/chromiumos/platform/tflite/refs/heads/firmware-ec-R136-16238.2.B
commit
db487fac17030fe43c100e48bd64178b3a677e6c
[log]
author
Tommy Chiang <ototot@chromium.org>
Thu Mar 20 08:43:03 2025
committer
Chromeos LUCI <chromeos-scoped@luci-project-accounts.iam.gserviceaccount.com>
Wed Mar 26 02:38:41 2025
tree
67b8380b6a0536bba1f0326a9499d15d47e74d07
parent
fcca206d13a964ad5f8b4f84ff383e7a805e8bb1 [diff]
tflite: Add Fp16Eq matcher

Some delegates may internally use fp16 for fp32 arithmetic, making the
default EXPECT_FLOAT_EQ too strict for such cases.
Since we are going to check fp16 results in many places, it would be
nice for us to have a single fp16 matchers similar to EXPECT_FLOAT_EQ.
This CL implements a new Fp16Eq matcher based on the implementation in
tensorflow [1] to provide a more precise and easy comparison.

BUG=b:374245241
TEST=bazelisk run --config='asan' -c dbg --config=host_clang '//common/test_util:fp16_compare_test'

[1] https://github.com/tensorflow/tensorflow/blob/689a4291c92835bc3d7b51adb3540b87432557d0/tensorflow/lite/kernels/test_util.cc#L121-L135

Change-Id: I6660ceab524ef84cae1c6d0e98bced7c2e51b14f
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/platform/tflite/+/6370771
Reviewed-by: Shik Chen <shik@chromium.org>
Tested-by: Tommy Chiang <ototot@google.com>
Commit-Queue: Tommy Chiang <ototot@google.com>
7 files changed
tree: 67b8380b6a0536bba1f0326a9499d15d47e74d07
  1. android/
  2. common/
  3. delegate/
  4. patch/
  5. script/
  6. tool/
  7. udev/
  8. .bazelignore
  9. .bazelrc
  10. .bazelversion
  11. .clang-format
  12. .gitignore
  13. BUILD.bazel
  14. CPPLINT.cfg
  15. DIR_METADATA
  16. OWNERS
  17. PRESUBMIT.cfg
  18. README.md
  19. WORKSPACE.bazel
README.md

ChromeOS TFLite

This repository hosts the core ChromeOS TFLite components, enabling on-device machine learning (ODML) workloads accelerated by NPU.

The corresponding ebuild can be found at: tensorflow-9999.ebuild

TensorFlow Patch Management

Patches are stored in the patch/ directory and explicitly listed in WORKSPACE.bazel. A helper script, ./script/patcher.py, is included to facilitate patch management within a TFLite workspace.

The typical workflow:

  1. Eject (Download) TensorFlow Source Code

    Download the TensorFlow source code into a local git repository with patches applied as individual commits:

    ./script/patcher.py eject

    This creates a new local git repository at tensorflow/.

  2. Modify the TensorFlow Repository

    Make changes to the tensorflow/ repository as needed, following standard git workflows. Optionally, include a PATCH_NAME= tag in commit messages to specify the filename of the corresponding patch.

  3. Seal the Repository

    Regenerate the patch files and update the WORKSPACE.bazel file:

    ./script/patcher.py seal

    This updates the patches in the patch/ directory and reflects the changes in WORKSPACE.bazel.

It's preferred to submit changes to upstream TensorFlow first and cherry-pick them as patches. This helps minimize divergence and makes TensorFlow updates easier.