v0.16.0

⚠️ Breaking Changes ⚠️

• Drop support for Python <= 3.9 and TensorFlow <= 2.7 (#807) @Tombana

🎉 Features

• Add Python 3.12 release (#807) @Tombana
• Update Tensorflow to 2.16.1 (#805) @CNugteren

👷‍♂️ Internal Improvements

• Use Hermetic Python in bazel build (#805) @Tombana

⬆️ Dependencies

• Update Android NDK version (#805) @Tombana
• Bump black from 22.3.0 to 24.3.0 (#804) @dependabot
• Bump al-cheb/configure-pagefile-action from 1.3 to 1.4 (#802) @dependabot
• Bump toolmantim/release-drafter from 5.25.0 to 6.0.0 (#801) @dependabot
• Bump actions/cache from 3 to 4 (#799) @dependabot
• Bump DoozyX/clang-format-lint-action from 0.16 to 0.17 (#798) @dependabot
• Bump actions/download-artifact from 3 to 4 (#797) @dependabot
• Bump actions/upload-artifact from 3 to 4 (#796) @dependabot
• Bump actions/setup-python from 4 to 5 (#795) @dependabot
• Bump google-github-actions/auth from 1 to 2 (#794) @dependabot
• Bump toolmantim/release-drafter from 5.24.0 to 5.25.0 (#793) @dependabot
• Bump actions/checkout from 3 to 4 (#792) @dependabot

v0.13.0

⚠️ Breaking Changes ⚠️

• Drop support for Python <=3.8 and TensorFlow <=2.4 (#787) @lgeiger

🎉 Features

• Build wheels for Python 3.11 (#786) @lgeiger
• ⬆️ tensorflow@2.13.0 (#784) @CNugteren

👷‍♂️ Internal Improvements

• Upgrade Android NDK version (#789) @lgeiger
• Use static library in cmake build (#780) @Tombana
• Fix trailing slash in path in CMakeLists.txt (#781) @Tombana
• Use command parameter files when compiling on Windows (#791) @lgeiger

v0.11.1

🎉 Features

• Add support for multiple subgraphs in some of the Python MLIR utils (#776) @CNugteren

👷‍♂️ Internal Improvements

• Remove anonymous namespaces in MLIR transforms (#768) @Tombana

⬆️ Dependencies

• Bump toolmantim/release-drafter from 5.23.0 to 5.24.0 (#774) @dependabot
• Bump DoozyX/clang-format-lint-action from 0.15 to 0.16 (#772) @dependabot
• Bump toolmantim/release-drafter from 5.22.0 to 5.23.0 (#769) @dependabot

v0.11.0

Note that the previous release was 0.8.0, but this one is 0.11.0 instead of 0.9.0 to match the '11' in the TensorFlow version: 2.11. Future releases will also follow this scheme, such that it becomes easy to match with a TensorFlow version that it is based on.

🎉 Features

• ⬆️ tensorflow@2.11.0 (#749, #762) @lgeiger

👷‍♂️ Internal Improvements

• Replace -mmacosx-version-min with MACOSX_DEPLOYMENT_TARGET env var (#767) @lgeiger
• [CI] Use concurrency setting instead of cancel action (#764) @lgeiger
• Fix Windows pip build (#755) @lgeiger
• Increase minimal macOS version requirement (#753) @lgeiger

⬆️ Dependencies

• Bump toolmantim/release-drafter from 5.21.1 to 5.22.0 (#765) @dependabot
• Bump DoozyX/clang-format-lint-action from 0.14 to 0.15 (#763) @dependabot
• Bump google-github-actions/auth from 0 to 1 (#761) @dependabot
• Bump toolmantim/release-drafter from 5.21.0 to 5.21.1 (#759) @dependabot
• Bump al-cheb/configure-pagefile-action from 1.2 to 1.3 (#760) @dependabot
• Bump styfle/cancel-workflow-action from 0.10.0 to 0.10.1 (#757) @dependabot
• Bump toolmantim/release-drafter from 5.20.1 to 5.21.0 (#756) @dependabot

v0.8.0

🎉 Features

• ⬆️ TensorFlow 2.9 (#724) @lgeiger

👷‍♂️ Internal Improvements

• Use Java 8 for building Android benchmark binaries (#752) @lgeiger
• Remove --copt=-O2 from CI (#751) @lgeiger
• Run linting on Python 3.9 (#746) @lgeiger
• Add Python 3.10 to setuptools metadata (#738) @lgeiger
• Remove weak attribute for RegisterSelectedOps (#734) @Tombana
• Add Python 3.10 to release builds (#733) @CNugteren

⬆️ Dependencies

• Bump styfle/cancel-workflow-action from 0.9.1 to 0.10.0 (#743) @dependabot
• Bump actions/setup-python from 3 to 4 (#741) @dependabot
• Bump DoozyX/clang-format-lint-action from 0.13 to 0.14 (#740) @dependabot
• Bump toolmantim/release-drafter from 5.19.0 to 5.20.1 (#737, #747) @dependabot

v0.7.0

⚠️ Breaking Changes ⚠️

• Drop Python 3.6 pip metadata (#701) @lgeiger
• Enable bitpacked activation optimisations by default. (#597) @AdamHillier

🎉 Features

• ⬆️ tensorflow@2.8.0 (#687, #722) @CNugteren, @lgeiger
• LCE benchmark and interpreter flags (#717) @simonmaurer
• Add a pass to translate from TFL CustomOp to the relevant Larq ops (#698) @panickal-xmos

🐛 Bug Fixes

• Fix segfaults in benchmark binaries when setting num_threads > 1 (#727) @CNugteren

📖 Documentation

• Update QuickNet documentation. (#663) @AdamHillier

👷‍♂️ Internal Improvements

• Fix build cache for windows release build (#731) @Tombana
• Deal with non-deterministic input order for interpreter test (#729) @CNugteren
• Add bazel cache for release jobs (#728) @Tombana
• Make assembly compatible with clang (#720) @Tombana
• Fuse padding layers into (Depthwise) Convolutions (#712) @Tombana
• Remove unused environment variable in wheel build (#704) @lgeiger
• Simplify Python Linux wheel release scripts (#700) @CNugteren
• Converter internals unification (#699) @CNugteren
• Add td_library target to be able to include lce_ops.td (#694) @panickal-xmos
• Switch to google-github-actions/auth (#695) @lgeiger
• Add SetBatchSize pass to MLIR converter (#690) @Tombana

⬆️ Dependencies

• Bump actions/download-artifact from 2 to 3 (#726) @dependabot
• Bump actions/upload-artifact from 2 to 3 (#725) @dependabot
• Update lint package versions (#721) @Tombana
• Bump actions/cache from 2.1.6 to 3 (#689, #719) @dependabot
• Bump actions/checkout from 2 to 3 (#716) @dependabot
• Bump actions/setup-python from 2 to 3 (#715) @dependabot
• Bump toolmantim/release-drafter from 5.15.0 to 5.19.0 (#702, #703, #708, #709, #718) @dependabot
• Bump google-github-actions/setup-gcloud from 0.2.1 to 0.3 (#691) @dependabot
• Bump DoozyX/clang-format-lint-action from 0.12 to 0.13 (#684) @dependabot

v0.6.2

🎉 Features

• QuantConv2D binarized activations with tf.int32 bitpacked output (#611) @simonmaurer
• Automatically build macOS ARM wheels on CI (#681) @lgeiger

📖 Documentation

• Fix docstring of Python interpreter (#677) @Tombana

👷‍♂️ Internal Improvements

• Split Python interpreter in two classes (#675) @Tombana

⬆️ Dependencies

• Bump styfle/cancel-workflow-action from 0.9.0 to 0.9.1 (#678) @dependabot

v0.6.1

🎉 Features

• Extend quantiser support so as to accelerate more binary models. (#668) @AdamHillier
• Support for macos_arm64 platform (#664) @simonmaurer

🐛 Bug Fixes

• Fix default ranges in saved model converter (#671) @Tombana

👷‍♂️ Internal Improvements

• Remove manylinux container from CI (#670) @Tombana
• Add quantization properties to python interpreter (#667) @Tombana
• Disable mkl for windows builds (#666) @AdamHillier

v0.6.0

🎉 Features

• Add support for saved model conversion (#655) @lgeiger
• Add grouped binary convolution support (3/3): indirect BGEMM kernel. (#551) @AdamHillier
• Fuse tfl.Quantize(lq.Dequantize(x)) -> lq.Dequantize(x) (#644) @lgeiger

🐛 Bug Fixes

• Fix windows build of grouped convolutions. (#656) @Tombana
• Prevent spurious int8 Quantize/Dequantize ops (#646) @lgeiger
• Statically link GLIBC in benchmark binaries (#610) @lgeiger
• Update QuantOp check (#608) @lgeiger

📖 Documentation

• Remove spectrum badge since we moved to GitHub discussions (#616) @lgeiger

👷‍♂️ Internal Improvements

• Upload release wheels even if some jobs fail. (#660) @AdamHillier
• Add constant folding for lq.quantize and lq.dequantize (#654) @lgeiger
• Add test-requirements.txt file (#652) @lgeiger
• Add pytype linting to CI (#651) @lgeiger
• Fork ConvertTFExecutorToTFLOrFlatbuffer to re-enable argname truncation (#649) @lgeiger
• Add check for float32 in IsBinaryTensor (#642) @Tombana
• Configure the Android SDK/NDK interactively in configure.py. (#627) @AdamHillier
• Simplify bazel build flags on CI (#626) @lgeiger
• Fix lce_benchmark_model linking on Aarch32 (#623) @Tombana
• Add -lm link option to lce_benchmark_model (#622) @Tombana
• ⬆️ CI dependencies (#618) @lgeiger
• Run AAR build on Ubuntu 18.04 (#617) @lgeiger
• Raise an error if mixed-precision models are being converted (#607) @CNugteren
• Automatically upload AAR package to release (#603) @lgeiger
• Update release.yml to actually push Windows wheels. (#596) @AdamHillier

⬆️ Dependencies

• Bump actions/cache from 2.1.5 to 2.1.6 (#650) @dependabot
• Pin flatbuffers to <2.0 (#648) @Tombana
• Upgrade TF dependency to v2.5. (#628) @AdamHillier
• Bump DoozyX/clang-format-lint-action from v0.11 to v0.12 (#634) @dependabot
• Bump actions/cache from v2.1.4 to v2.1.5 (#630) @dependabot
• Bump styfle/cancel-workflow-action from 0.8.0 to 0.9.0 (#629) @dependabot
• Bump toolmantim/release-drafter from v5.14.0 to v5.15.0 (#621) @dependabot
• Bump google-github-actions/setup-gcloud from v0.2.0 to v0.2.1 (#612) @dependabot
• Bump styfle/cancel-workflow-action from 0.7.0 to 0.8.0 (#605) @dependabot
• Bump toolmantim/release-drafter from v5.13.0 to v5.14.0 (#606) @dependabot
• Bump actions/cache from v2 to v2.1.4 (#600) @dependabot

v0.5.0

🎉 Features

• Add preliminary support for the xcore target. (#585) @AdamHillier
• Add zero-padding support to the reference kernel (#571) @Tombana
• Add grouped binary convolution support (2/3): reference kernel. (#550) @AdamHillier
• Add grouped binary convolution support (1/3): the converter. (#549) @AdamHillier

🚀 Performance

• Slightly optimise the loads in the 8x4x2 indirect BGEMM kernel (#521) @AdamHillier
• Add optimised indirect bgemm kernels for Aarch64. (#518) @AdamHillier

🐛 Bug Fixes

• Fix Windows release builds. (#590) @AdamHillier
• Run LCE optimize pass before and after TFLite optimize pass (#525) @lgeiger

📖 Documentation

• Update LCE python interpreter docstring with efficiency warning (#582) @jneeven

👷‍♂️ Internal Improvements

• Fix release version number for non-manual releases. (#595) @AdamHillier
• Use a Python script for Bazel configuration, like TensorFlow does. (#589) @AdamHillier
• Add TF 2.4 to CI matrix (#579) @lgeiger
• Fix linebreaks in release script (#577) @lgeiger
• Use GitHub actions workflow dispatch to trigger nightly releases (#576) @lgeiger
• Fix windows build (#575) @lgeiger
• Only upload cache on master branch (#574) @lgeiger
• Do not pin official GitHub actions (#572) @lgeiger
• Do not pin official GitHub actions (#570) @lgeiger
• Set --linkopt=-dead_strip when building mac wheels (#566) @lgeiger
• Remove unused dependencies (#565) @lgeiger
• Fix filter tensor shapes in bconv2d 'death' tests. (#561) @AdamHillier
• Create a lightweight BConv2DParams struct in /core. (#560) @AdamHillier
• Fix manylinux release build (#558) @lgeiger
• Cancel workflows of outdated commits (#556) @lgeiger
• Speedup end2end tests (#552) @lgeiger
• ⬆️ larq_zoo@v2 (#547) @lgeiger
• Auto-generate the TFLite Python schema. (#546) @AdamHillier
• Add version property. (#543) @AdamHillier
• Use test sharding to parallelise the tests. (#544) @AdamHillier
• Update Flatbuffer schema for TF 2.4 (#536) @lgeiger
• Move padding conversion function to kernels/utils.h (#534) @lgeiger
• Add macros to help GCC with branch prediction (#530) @lgeiger
• Prebuild the benchmark binary with -O3 on Android (#532) @lgeiger
• Change qemu test scripts by using a new bazel rule (#533) @Tombana
• Upgrade TensorFlow to latest version (#529) @lgeiger
• Inline kernel utils functions that convert padding and activation enums (#528) @lgeiger
• Initialise batch norm stats for e2e models (#524) @AdamHillier
• Add support for batched inputs in the indirect BGEMM kernels (#519) @AdamHillier
• Add optimised 'Indirect BGEMM' binary convolution kernels. (#516) @AdamHillier
• Simplify the bconv2d padding functor. (#515) @AdamHillier
• Reorganise /core directory. (#514) @AdamHillier

⬆️ Dependencies

• ⬆️ TensorFlow v2.4.1. (#587) @AdamHillier
• Bump package versions on CI. (#586) @AdamHillier
• Bump styfle/cancel-workflow-action from 0.6.0 to 0.7.0 (#584) @dependabot
• Bump toolmantim/release-drafter from v5.12.1 to v5.13.0 (#580) @dependabot
• ⬆️ tensorflow@2.4 (#578) @lgeiger
• Bump GoogleCloudPlatform/github-actions from 0.1.3 to v0.2.0 (#569) @dependabot
• Bump actions/upload-artifact from v2.2.0 to v2.2.1 (#568) @dependabot
• :arrow-up: TensorFlow submodule (#564) @lgeiger
• ⬆️ tensorflow@2.4.0rc1 (#563) @lgeiger
• Update bazelisk version on CI and in containers. (#559) @AdamHillier
• ⬆️ tensorflow@v2.4.0-rc0 (#557) @lgeiger
• ⬆️ lint dependencies (#555) @lgeiger
• ⬆️ tensorflow@r2.4 (#545) @lgeiger
• Bump toolmantim/release-drafter from v5.12.0 to v5.12.1 (#539) @dependabot
• Bump toolmantim/release-drafter from v5.11.0 to v5.12.0 (#538) @dependabot
• Bump DoozyX/clang-format-lint-action from v0.10 to v0.11 (#537) @dependabot
• Bump actions/setup-python from v2.1.3 to v2.1.4 (#531) @dependabot
• Bump actions/upload-artifact from v2.1.4 to v2.2.0 (#522) @dependabot
• Bump actions/setup-python from v2.1.2 to v2.1.3 (#520) @dependabot
• ⬆️ tensorflow@2.3.1 (#517) @lgeiger