)]}'
{
  "log": [
    {
      "commit": "801e6644bf6caad2ad2f09cdd41f2f634ce73570",
      "tree": "ea5b4251763eee40d0349033d0d94d42f30d18fb",
      "parents": [
        "4077ba8a331e8e6979f0a3d4e40a5766f7238630",
        "07e7f33eb5f652d8c6ea87b27c2f406fde97f15c"
      ],
      "author": {
        "name": "Matthew Wilson",
        "email": "mwilson@google.com",
        "time": "Thu Aug 15 12:01:16 2024 -0700"
      },
      "committer": {
        "name": "Matthew Wilson",
        "email": "mwilson@google.com",
        "time": "Thu Aug 15 19:46:01 2024 +0000"
      },
      "message": "Merge remote-tracking branch \u0027spacebeaker/upstream\u0027 into master\n\nChange-Id: Ieb5c6acc9bbafad3813c83058a9e40089fee2e3c\n"
    },
    {
      "commit": "07e7f33eb5f652d8c6ea87b27c2f406fde97f15c",
      "tree": "3b3c83c465a5c37b9991108e831fd2b263bab503",
      "parents": [
        "2b127fd8e4c90228d14561c54812c5369dea2c5d"
      ],
      "author": {
        "name": "Ryan Kuester",
        "email": "kuester@bdti.com",
        "time": "Wed Aug 14 16:58:12 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Aug 14 21:58:12 2024 +0000"
      },
      "message": "feat: add features to tflite::Span (#2664)\n\nTo tflite::Span, add constructors for arrays and std::arrays, equality\nand inequality operators, and tests thereof.\n\nBUG\u003dsee description"
    },
    {
      "commit": "2b127fd8e4c90228d14561c54812c5369dea2c5d",
      "tree": "43d363351ffc6a84ce3ec9dc759cd8092c0c6de4",
      "parents": [
        "95da7a8cd2f5236da0e0d414bd812bc33e4708d5"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Wed Aug 14 10:02:52 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Aug 14 17:02:52 2024 +0000"
      },
      "message": "TFLM compression changes (3rd) (#2658)\n\n@tensorflow/micro\n\nUpdates to support TFLM compression:\n\nMicroContext\nMicroInterpreterContext\nFakeMicroContext\nKernelRunner\n\nbug\u003d#2657"
    },
    {
      "commit": "95da7a8cd2f5236da0e0d414bd812bc33e4708d5",
      "tree": "8cb53bef3350416d8834bea59dc1f5eb045f7b03",
      "parents": [
        "19577c1677d642221895dffc91049b6fea5ed003"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon Aug 12 10:40:54 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Aug 12 17:40:54 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2662)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "19577c1677d642221895dffc91049b6fea5ed003",
      "tree": "f33671a59dd077e1f4c726e87c106f8d584cfb7e",
      "parents": [
        "79ceb4c45c5cd9223dd4bed8701cdf6a0da9cc9f"
      ],
      "author": {
        "name": "Lucas Chollet",
        "email": "lucas.chollet@free.fr",
        "time": "Mon Aug 12 12:36:25 2024 -0400"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Aug 12 16:36:25 2024 +0000"
      },
      "message": "Fix build with recent version of GCC (#2661)\n\nWithout this patch, the definition of the custom deleter would drop the attributes of the `fclose` function.\n\nBUG\u003d#2660"
    },
    {
      "commit": "79ceb4c45c5cd9223dd4bed8701cdf6a0da9cc9f",
      "tree": "3aacf58afba61dc408e54551b1f633516e587b73",
      "parents": [
        "d3475aa77621545b7b0ef585f38fb8634c7ee766"
      ],
      "author": {
        "name": "Lucas Chollet",
        "email": "lucas.chollet@free.fr",
        "time": "Mon Aug 12 11:58:35 2024 -0400"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Aug 12 15:58:35 2024 +0000"
      },
      "message": "build: update python_rules to support Python 3.12 (#2654)\n\nBefore, when trying to use to build the project using Bazel, this would result in multiple occurrence of this error:\n`AttributeError: module \u0027pkgutil\u0027 has no attribute \u0027ImpImporter\u0027. Did you mean: \u0027zipimporter\u0027?`\n\nThis error is due to `rules_python` being too old. While building with Python3.12 still result in an error, the new one clearly indicate is mismatch of the TensorFlow version.\n\nVery similar/related: https://github.com/googleapis/gapic-generator-python/pull/1825\n\nBUG\u003dhttps://github.com/tensorflow/tensorflow/issues/73174\n"
    },
    {
      "commit": "d3475aa77621545b7b0ef585f38fb8634c7ee766",
      "tree": "9e0078638812384408609588bafcbfb30c843bc3",
      "parents": [
        "8f9a923ad306a2d298c5086b57ec9b6caca76dfd"
      ],
      "author": {
        "name": "Ryan Kuester",
        "email": "kuester@bdti.com",
        "time": "Wed Aug 07 11:44:08 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Aug 07 16:44:08 2024 +0000"
      },
      "message": "feat: add fixed-capacity, statically-allocated type tflite::StaticVector (#2642)\n\nAdd a type, tflite::StaticVector, which behaves like std::vector, but\nwhich avoids heap memory allocation.\n\nBUG\u003d#2636"
    },
    {
      "commit": "8f9a923ad306a2d298c5086b57ec9b6caca76dfd",
      "tree": "c521fad9fb7de055e5631a9f33e00a7e56609ec7",
      "parents": [
        "39f9f1912ae646ea91314c305b2d0f3fdfba0109"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Tue Aug 06 10:36:14 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Aug 06 10:36:14 2024 -0700"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2656)\n\n"
    },
    {
      "commit": "39f9f1912ae646ea91314c305b2d0f3fdfba0109",
      "tree": "6e8074a615b32b5b71f907ffe943019c376bcefd",
      "parents": [
        "6c6e9b3ecae2cde5a832ff2363d85ba7267d69de"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Tue Aug 06 10:36:03 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Aug 06 17:36:03 2024 +0000"
      },
      "message": "TFLM compression changes (2nd) (#2653)\n\n@tensorflow/micro\n\nPreparatory changes for upcomming PRs related to TFLM compression changes.\n\nAdd unit testing failure check macro.\n\nbug\u003d#2652"
    },
    {
      "commit": "6c6e9b3ecae2cde5a832ff2363d85ba7267d69de",
      "tree": "8a4d27bed2ffa55f0b714b01513e7ece718347a3",
      "parents": [
        "f309046ae28326b943894a4916f3d7bef65ebe9d"
      ],
      "author": {
        "name": "Ryan Kuester",
        "email": "kuester@bdti.com",
        "time": "Thu Aug 01 16:30:54 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Aug 01 21:30:54 2024 +0000"
      },
      "message": "build: begin using -std\u003dc++17 and c17 in Makefile builds (#2648)\n\nBegin using -std\u003dc++17 and -std\u003dc17 in Makefile builds on\nall platforms. Bazel builds have been using C++17 since 52c9568.\n\nSet `-stdlib\u003dlibc++ on xt-clang` on Xtensa to add C++17 library\nsupport in addition to compiler support. From the xt-clang docs:\n\n    Starting with the RI-2019.1 release, XT-CLANG has included support\n    for C++14 and C++17 language features. The compiler support for\n    C++14 and C++17 is accompanied by the C++ library from the LLVM\n    project. This library can be selected with the -stdlib\u003dlibc++\n    option, and this is strongly recommended when compiling with\n    -std\u003dc++14 or - std\u003dc++17. Starting with the RI-2021.6 release, two\n    additional versions of this C++ library are provided— one that\n    excludes support for exception handling, and one that excludes both\n    exception handling and run-time type identification. These libraries\n    can be selected with -stdlib\u003dlibc ++-e and -stdlib\u003dlibc++-re options\n    respectively.\n\nBased on the `docker run` command in\ntflite-micro/.github/workflows/xtensa_presubmit.yml, our CI is\ncurrently using RI-2020.4.\n\nRefactor the make/ext_libs/xtensa_download.sh script to make it eaiser\nto patch downloads for all Xtensa platforms. The old script made overly\nstrict assumptions about the name of the patch.\n\nPatch the Xtensa vision_p6 platform download xi_tflmlib_vision_p6 for\ncompatibility with the C++ library standard. Use the header \u003cclimits\u003e to\naccess constants such as INT_MAX.\n\nBUG\u003d#2650"
    },
    {
      "commit": "f309046ae28326b943894a4916f3d7bef65ebe9d",
      "tree": "3f157b1ea41e143e67874d2a50310927aef54195",
      "parents": [
        "91fc719fc8bf889a95968347b4ababcf603cf6c5"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Thu Aug 01 11:59:33 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Aug 01 18:59:33 2024 +0000"
      },
      "message": "TFLM compression changes (#2647)\n\n@tensorflow/micro\n\ninitial header file changes/additions.\n\nbug\u003d#2646"
    },
    {
      "commit": "91fc719fc8bf889a95968347b4ababcf603cf6c5",
      "tree": "aa9054d43ee27f30fb196393b697fec199d412e7",
      "parents": [
        "6b676567b79053d4d275ae834427b4f1259a710d"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Thu Aug 01 08:47:53 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Aug 01 08:47:53 2024 -0700"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2649)\n\n"
    },
    {
      "commit": "6b676567b79053d4d275ae834427b4f1259a710d",
      "tree": "e66a567f82669a86d8287cfe4460fbb8981decfc",
      "parents": [
        "a57349f0bc3bc3ded105238fe2f14c36888d42b0"
      ],
      "author": {
        "name": "Ibrahim Abdelkader",
        "email": "i.abdalkader@gmail.com",
        "time": "Mon Jul 29 19:53:52 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jul 29 17:53:52 2024 +0000"
      },
      "message": "Fix typo in Ethos-U Makefile. (#2644)\n\nThe core platform sources were overwritten by typo. I tested this patch by triggering the Cortex-M workflow manually, and it completed successfully. See the run here in my fork:\n\nhttps://github.com/iabdalkader/tflite-micro/actions/runs/10113795223\n\nBUG\u003d#2639"
    },
    {
      "commit": "a57349f0bc3bc3ded105238fe2f14c36888d42b0",
      "tree": "1b93440ef2a71053f6e5c836e1bdc4281bb9688f",
      "parents": [
        "fff7b50807c102684f1a97db7f428bd4eb4ed1f2"
      ],
      "author": {
        "name": "Ibrahim Abdelkader",
        "email": "i.abdalkader@gmail.com",
        "time": "Thu Jul 25 21:07:08 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Jul 25 19:07:08 2024 +0000"
      },
      "message": "Fix Ethos-U driver build. (#2643)\n\nThe updated Ethos-U driver v24.05 (introduced in #2626), contains a second set of device functions for the U85. This can cause the linker to link the functions of the U85, instead of the U55-U65, if they\u0027re both included in the driver\u0027s sources files. Therefore a glob pattern can no longer be used to collect the driver\u0027s source files in the Makefile.\n\nBUG\u003d#2639"
    },
    {
      "commit": "fff7b50807c102684f1a97db7f428bd4eb4ed1f2",
      "tree": "b0f793cd6fbe3a69eba82ab98b48f47c64b77e00",
      "parents": [
        "213eb8769b919f722f85877a2affeddd76f45dca"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Thu Jul 25 11:10:06 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Jul 25 18:10:06 2024 +0000"
      },
      "message": "Fix users of NumPy APIs that are removed in NumPy 2.0 (#2640)\n\nBUG\u003dcl/655772545"
    },
    {
      "commit": "213eb8769b919f722f85877a2affeddd76f45dca",
      "tree": "d55275ac7d3c22b16f1806b1f4c0d808e08516e3",
      "parents": [
        "d619ad8ec501a654d5893bd5bdb66f1c07f5d16a"
      ],
      "author": {
        "name": "Ibrahim Abdelkader",
        "email": "i.abdalkader@gmail.com",
        "time": "Wed Jul 24 23:53:31 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Jul 24 21:53:31 2024 +0000"
      },
      "message": "Update Ethos-U driver. (#2626)\n\n* Update Ethos-U driver to the latest (v24.05)\n* Allow passing extra C flags to the driver from the command line.\n* Fixes #2619\n\nNote I tested with `TARGET\u003dcortex_m_corstone_300 test_network_tester_test` and on real hardware as well built with:\n```bash\nmake -j12 -f tensorflow/lite/micro/tools/make/Makefile \\\nTARGET\u003dcortex_m_generic TARGET_ARCH\u003dcortex-m55 \\\nCO_PROCESSOR\u003dethos_u ETHOSU_ARCH\u003du55 OPTIMIZED_KERNEL_DIR\u003dethos_u  \\\nCORE_OPTIMIZATION_LEVEL\u003d-O2 KERNEL_OPTIMIZATION_LEVEL\u003d-O2 \\\nTHIRD_PARTY_KERNEL_OPTIMIZATION_LEVEL\u003d-O2 \\\nTARGET_TOOLCHAIN_ROOT\u003d/opt/arm-none-eabi/bin/ \\ \nTARGET_TOOLCHAIN_PREFIX\u003darm-none-eabi- BUILD_TYPE\u003drelease microlite\n```\n\nBUG\u003d#2619"
    },
    {
      "commit": "d619ad8ec501a654d5893bd5bdb66f1c07f5d16a",
      "tree": "c898690903e7cda35e7696798d2bcb6a81c9082b",
      "parents": [
        "569ed29b9d907b1ec311aa4d413500af23124c3d"
      ],
      "author": {
        "name": "Ryan Kuester",
        "email": "kuester@bdti.com",
        "time": "Wed Jul 24 15:20:44 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Jul 24 20:20:44 2024 +0000"
      },
      "message": "refactor: move tflite::Span to independent header (#2638)\n\nHoist the universally-useful tflite::Span out of\ncodegen/runtime/micro_codegen_context.h and into an independent header\nusable from elsewhere in the project.\n\nBUG\u003d#2636"
    },
    {
      "commit": "569ed29b9d907b1ec311aa4d413500af23124c3d",
      "tree": "d6bde71e96ac8fc5d66bf0b00a529d929ab7859c",
      "parents": [
        "f7d989eb89edc411c3e3a50a1cb8f6c1989c9905"
      ],
      "author": {
        "name": "Ryan Kuester",
        "email": "kuester@bdti.com",
        "time": "Wed Jul 24 13:49:24 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Jul 24 18:49:24 2024 +0000"
      },
      "message": "build(py): add hexdump requirement (#2637)\n\nAdd the Python distribution package `hexdump`, to be used in tests\nand utilities which display raw memory.\n\nBUG\u003d#2636"
    },
    {
      "commit": "f7d989eb89edc411c3e3a50a1cb8f6c1989c9905",
      "tree": "fc39ea40b31aaee48acdf78ce926e85e076cb026",
      "parents": [
        "5c2ac76b80aa231779ba32b14c6194f6948eb56f"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Wed Jul 24 09:44:30 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Jul 24 16:44:30 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2634)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "5c2ac76b80aa231779ba32b14c6194f6948eb56f",
      "tree": "a2c8dd980f0209af3bd41dc255d1108f632818b9",
      "parents": [
        "eec75f069fe576a8633ca6a954fd4f7ad4406343"
      ],
      "author": {
        "name": "Ryan OShea",
        "email": "86965113+ArmRyan@users.noreply.github.com",
        "time": "Mon Jul 22 18:19:40 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jul 22 16:19:40 2024 +0000"
      },
      "message": "Update cmsis prepare functions to exit earlier (#2630)\n\n* Adds additional checks to various cmsis-nn prepare functions\n\nBUG\u003d1121\nAuthored-by: Ryan O\u0027Shea \u003cryan.oshea3@arm.com\u003e\nChange-Id: Ic6481873f064a94a4dd0b4a49790842180d73dd9"
    },
    {
      "commit": "eec75f069fe576a8633ca6a954fd4f7ad4406343",
      "tree": "9d2b89428f8fd8134ea34f29db49472e91ee0ad3",
      "parents": [
        "7a0249686f412551634a5058ddd6d2ec3f224203"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon Jul 22 08:53:21 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jul 22 15:53:21 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2632)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "7a0249686f412551634a5058ddd6d2ec3f224203",
      "tree": "c3bce1d4be62e95d11f70b7916eef88187231dbc",
      "parents": [
        "ff5c090ecac692689506e5598b0f3b8391ef6be0"
      ],
      "author": {
        "name": "Johnney Jung",
        "email": "johnney.jung@samsung.com",
        "time": "Wed Jul 17 02:52:18 2024 +0900"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Jul 16 17:52:18 2024 +0000"
      },
      "message": "Static Code Analysis result Update (#2624)\n\nBUG\u003d\"static code analysis result update\"\n\nBuild test was done with the command below\n  make -f tensorflow/lite/micro/tools/make/Makefile test\n\nThere will be an additional pull request later related to static code analysis "
    },
    {
      "commit": "ff5c090ecac692689506e5598b0f3b8391ef6be0",
      "tree": "061469c9523521fa0b3e605655d49c45fd6f11f4",
      "parents": [
        "0cc2e49fd96eca29761b323a16f41683fe112fc6"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Mon Jul 15 13:04:01 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jul 15 13:04:01 2024 -0700"
      },
      "message": "Fix upstream TF sync (#2625)\n\nThe CI sync from the upstream TF has been failing for the last month due\r\nto the relocation of certain TFLite files. In particular, some schema\r\nfiles were relocated to tensorflow/compiler/mlir/lite subfolder. This PR\r\nmirrors the migration and adds a few redirection headers to ensure\r\nsource compatibility for now.\r\n\r\nThis PR also includes a TF sync as well to get us caught up again. While\r\nthat could have been done separately, this is done together to ensure\r\neverything continues to build with the relocations.\r\n\r\nBUG\u003db/351824449"
    },
    {
      "commit": "0cc2e49fd96eca29761b323a16f41683fe112fc6",
      "tree": "0de73f42382f435a8bca973960cee8236078b304",
      "parents": [
        "dc4dcb7433c14897c5668d1b46b4c773b91f4ed1"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Tue Jul 02 10:27:06 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Jul 02 17:27:06 2024 +0000"
      },
      "message": "Disable per-channel quant on requantize test (#2616)\n\nBUG\u003dcleanup"
    },
    {
      "commit": "4077ba8a331e8e6979f0a3d4e40a5766f7238630",
      "tree": "cb6d7ed9aa4502dea4be8c36e6cc6e49d2c15db2",
      "parents": [
        "e1459bc02c2cb9ccfc5a4d80d018054f690bed34"
      ],
      "author": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Tue Jun 25 22:20:25 2024 -0700"
      },
      "committer": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Thu Jun 27 14:31:48 2024 -0700"
      },
      "message": "tflite-micro: add max_pooling int16 kernel for Kelvin\n\nChange-Id: I71cde103ead25bc6ebcf158217008a78a351e666\n"
    },
    {
      "commit": "e1459bc02c2cb9ccfc5a4d80d018054f690bed34",
      "tree": "74d9658f8d1ef77690fe27cfd9027da77c8850c6",
      "parents": [
        "5cb1144314c5794f43514a8bc539eb2c3ba9faa5",
        "c6934d66790124cc36c79146fa92d50219ed85c1"
      ],
      "author": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Thu Jun 27 19:28:27 2024 +0000"
      },
      "committer": {
        "name": "Gerrit Code Review",
        "email": "noreply-gerritcodereview@google.com",
        "time": "Thu Jun 27 19:28:27 2024 +0000"
      },
      "message": "Merge \"sw/tflite-micro: clean up function definitions\""
    },
    {
      "commit": "c6934d66790124cc36c79146fa92d50219ed85c1",
      "tree": "95f44a201b7a4daa4a10e2fe62347f26e1a83db9",
      "parents": [
        "f1bf04912b206a4e4e234bcbf3e71d16fa8dad93"
      ],
      "author": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Wed Jun 26 00:08:11 2024 -0700"
      },
      "committer": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Wed Jun 26 00:12:29 2024 -0700"
      },
      "message": "sw/tflite-micro: clean up function definitions\n\nTry to use a same function definition as in reference\nimplementations. This makes codes cleaner and more expandable.\n\nChange-Id: I8cecbb405eef4ae28d60b94a42abb9c2587d25c4\n"
    },
    {
      "commit": "5cb1144314c5794f43514a8bc539eb2c3ba9faa5",
      "tree": "374a70984928d67b40889ba387d9efc5bd9510b7",
      "parents": [
        "f1bf04912b206a4e4e234bcbf3e71d16fa8dad93"
      ],
      "author": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Tue Jun 25 22:40:17 2024 -0700"
      },
      "committer": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Tue Jun 25 22:43:27 2024 -0700"
      },
      "message": "tflite-micro: add elementwise multiplication kernel for Kelvin\n\nChange-Id: I14946324e5a460f0c345b8c5528f577310ee532f\n"
    },
    {
      "commit": "f1bf04912b206a4e4e234bcbf3e71d16fa8dad93",
      "tree": "992d158ea8982f56d99c958bed48e8622f20f696",
      "parents": [
        "29ee45ced8b79565a9f7ec2cee8e92ebe31f74a8"
      ],
      "author": {
        "name": "Naveen Dodda",
        "email": "ndodda@google.com",
        "time": "Sat Jun 08 16:49:34 2024 +0000"
      },
      "committer": {
        "name": "Naveen Dodda",
        "email": "ndodda@google.com",
        "time": "Mon Jun 24 17:50:17 2024 +0000"
      },
      "message": "Kelvin resize nearest neighbors tfmicro\n\nChange-Id: Ifd38bbb9a5ff66e89d70cc83c69ece71966378ed\n"
    },
    {
      "commit": "dc4dcb7433c14897c5668d1b46b4c773b91f4ed1",
      "tree": "8c6b0606e2f51005e492f09fe0fd904b4f65cb07",
      "parents": [
        "8cd499a14ef801d997156047e73ea4c42748ea2b"
      ],
      "author": {
        "name": "Daniele Parravicini",
        "email": "98016455+DanieleParravicini-Synthara@users.noreply.github.com",
        "time": "Mon Jun 24 15:56:32 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jun 24 13:56:32 2024 +0000"
      },
      "message": "Adds GetCurrentOperatorIndex to micro_interpreter (#2605)\n\nIn this way an AOT can add metadata that could then be accessed with the couple GetCurrentSubgraphIndex, GetCurrentOperatorIndex.\nThese can be accessed from tflite::GetMicroContext(context)-\u003egraph()\n\nRefers to #2593\nBUG\u003d#2593"
    },
    {
      "commit": "8cd499a14ef801d997156047e73ea4c42748ea2b",
      "tree": "72ecb88400acc45a21db0a8790d2fe02a0498122",
      "parents": [
        "8c458fc48ee972ea4fc56ddc08849823b9af7ea8"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Mon Jun 24 09:49:18 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Jun 24 07:49:18 2024 +0000"
      },
      "message": "Fix person detect converter script (#2614)\n\nBUG\u003dhttps://github.com/tensorflow/tflite-micro/issues/2609"
    },
    {
      "commit": "8c458fc48ee972ea4fc56ddc08849823b9af7ea8",
      "tree": "72642ba13db5cc0034630496f45ad80f2dc6a4bc",
      "parents": [
        "f6403469b4ba981833bf09d7ea1d8eb3934f24a6"
      ],
      "author": {
        "name": "asraa",
        "email": "asraa@google.com",
        "time": "Tue Jun 18 15:13:15 2024 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Jun 18 20:13:15 2024 +0000"
      },
      "message": "fix: fix path to train target in hello_world example (#2608)\n\nBUG\u003dcleanup"
    },
    {
      "commit": "f6403469b4ba981833bf09d7ea1d8eb3934f24a6",
      "tree": "dd7c8c7e2f8b48deb7e2fa5e85113e93ca42feb7",
      "parents": [
        "71fba59b5bebfd64b48e167bfcb669fff568ec74"
      ],
      "author": {
        "name": "StrugglingBunny",
        "email": "160554513+StrugglingBunny@users.noreply.github.com",
        "time": "Wed Jun 19 02:45:35 2024 +0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Jun 18 18:45:35 2024 +0000"
      },
      "message": "Update new_platform_support.md (#2603)\n\nChange link of system_setup.h\n\nBUG\u003dcleanup"
    },
    {
      "commit": "29ee45ced8b79565a9f7ec2cee8e92ebe31f74a8",
      "tree": "1932c22830e5e04738cba74380fbc3bdc38f2fc9",
      "parents": [
        "27c820fed31089b3c1857022b6d143575adb782a"
      ],
      "author": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Fri May 31 16:34:02 2024 -0700"
      },
      "committer": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Fri May 31 16:34:02 2024 -0700"
      },
      "message": "Add logistic kernel for Kelvin\n\nChange-Id: I9e36fea5d4b32d6ac03a3d2f78c7362563bd276e\n"
    },
    {
      "commit": "27c820fed31089b3c1857022b6d143575adb782a",
      "tree": "eef76e200da96d5bfd657f2f3c29b44d768d69f7",
      "parents": [
        "aa3fb6a13c9d8bbc8de57420c6c27ccd1f52e9f2"
      ],
      "author": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed May 29 13:28:46 2024 -0700"
      },
      "committer": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed May 29 13:28:46 2024 -0700"
      },
      "message": "Add int16 pack and transpose\n\nChange-Id: I8d63eadd27cb20868045666efb3b0369efc2c1cb\n"
    },
    {
      "commit": "71fba59b5bebfd64b48e167bfcb669fff568ec74",
      "tree": "2924e4b32206617d28dc8dfcc4b8e3007db077f9",
      "parents": [
        "be11bd79e4a8b28c9ee92c6f02ca0e85414fb768"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Tue May 28 21:05:42 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed May 29 04:05:42 2024 +0000"
      },
      "message": "Fail per-channel quantized FullyConnected layers (#2602)\n\nThe TFLiteConverter recently switched over to using per-channel quantization for all Dense/FullyConnected layers. TFLite-Micro does not yet have support for this, and was using incorrect quantization parameters for FullyConnected layers on newly converted models. Unsurprisingly, this leads to invalid output.\n\nWhile we intend to add per-channel quantization support for FullyConnected, this PR adds a runtime check for per-channel quantization until it can be supported by individual kernels. If you encounter this runtime error, you can disable the new Converter behavior by setting:\n\n`TfLiteConverter._experimental_disable_per_channel_quantization_for_dense_layers \u003d True` https://github.com/tensorflow/tensorflow/blob/377f47694fa790e98db6665b9adecde00b5e0d68/tensorflow/lite/python/lite.py#L674\n\nBUG\u003db/324385802"
    },
    {
      "commit": "be11bd79e4a8b28c9ee92c6f02ca0e85414fb768",
      "tree": "38ed57ebeec4da7a9e66e9335ded458091aa6481",
      "parents": [
        "efa694beba9a833a954ee38cf035a990d1c3bea8"
      ],
      "author": {
        "name": "Robert Kamara",
        "email": "73614495+robertkamara@users.noreply.github.com",
        "time": "Fri May 24 10:02:35 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri May 24 17:02:35 2024 +0000"
      },
      "message": "Add support for 16 bit activations in ExpandDims operation. (#2589)\n\nThe change introduces the support for ML models with 16 bit activations and 8 bit weights in the ExpandDims operation.\n\nBUG\u003dfixes #68293"
    },
    {
      "commit": "efa694beba9a833a954ee38cf035a990d1c3bea8",
      "tree": "d4ff667159df39126de72a4e54b0cbdebfbf0d21",
      "parents": [
        "d069ba82992f186d99f71e6224685b6752f16dc3"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Thu May 23 10:14:16 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu May 23 17:14:16 2024 +0000"
      },
      "message": "Import visualize from tensorflow, not tflite-micro (#2595)\n\nWhen running the gen_micro_mutable_op_resolver script as a regular python script without bazel, it cannot find visualize because it is not part of the tflite-micro wheel. When running the script with bazel, bazel can pick up visualize from the path\ntflite_micro/tensorflow/lite/tools as long as tflite-micro wheel is not installed. If it is installed, then bazel will only look at that, which it is not a part of.\n\nThis PR switches the import to just be from tensorflow, which should work regardless of whether it is from the tf package or relative path.\n\nBUG\u003d#2564"
    },
    {
      "commit": "d069ba82992f186d99f71e6224685b6752f16dc3",
      "tree": "fced04005ed7e92cd5281dc8e808539c0dcaffec",
      "parents": [
        "dfdd666ae075e07da288ce8d4a38c60aeecbc07a"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Wed May 22 10:37:31 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed May 22 17:37:31 2024 +0000"
      },
      "message": "Remove unused visibility spec (#2594)\n\nBUG\u003dcl/634722988"
    },
    {
      "commit": "dfdd666ae075e07da288ce8d4a38c60aeecbc07a",
      "tree": "f94463c45bc17f397e5bce027875a6152e981f40",
      "parents": [
        "8e22946b3faa51564df5dd9194f7540b2694892c"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Tue May 21 10:14:57 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue May 21 17:14:57 2024 +0000"
      },
      "message": "more fixes for generic benchmark wrt Corstone-300 (#2592)\n\n@tensorflow/micro\n\nAdd person_detection_vela.tflite to tensorflow/lite/micro/models.\n\nUse the ethos-u co-processor and ethos-u enabled model when running benchmark.\n\nbug\u003dfixes #2591"
    },
    {
      "commit": "8e22946b3faa51564df5dd9194f7540b2694892c",
      "tree": "f780de31e4ab9839536b0c56a889cee0078419ce",
      "parents": [
        "9045e1b012a3b02fc61f770588cc44540aa79766"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon May 20 10:56:45 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon May 20 17:56:45 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2585)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "9045e1b012a3b02fc61f770588cc44540aa79766",
      "tree": "51e14dfc5c162e48221cd84c9e65c01a81708294",
      "parents": [
        "04d071e3cd0f1396cf3afa99d698cc277273ef0d"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Mon May 20 19:21:03 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon May 20 17:21:03 2024 +0000"
      },
      "message": "Patch NDP Hifi3 to allow override of MAX_RFFT_LEN (#2588)\n\nBUG\u003d340206722"
    },
    {
      "commit": "04d071e3cd0f1396cf3afa99d698cc277273ef0d",
      "tree": "67bf6a1c74314f441738494b66aca68fc36dda50",
      "parents": [
        "60284939db58e01f0dc364f49c14cd9c093f5950"
      ],
      "author": {
        "name": "Ryan OShea",
        "email": "86965113+ArmRyan@users.noreply.github.com",
        "time": "Mon May 20 18:36:27 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon May 20 16:36:27 2024 +0000"
      },
      "message": "Fix test_cortex_m_corstone_300 ci job (#2582)\n\n* Update path for new corstone300 downloads\n * Add missing benchmark arena size to generic benchmark test\n \nBUG\u003dCorstone300 Ci Job fails to run\nAuthor: Ryan O\u0027Shea \u003cryan.oshea3@arm.com\u003e"
    },
    {
      "commit": "60284939db58e01f0dc364f49c14cd9c093f5950",
      "tree": "d97cf5519ecdd1f62ee250c9fefc85d0d3a8d8a8",
      "parents": [
        "94e7cbac65ebb1378a95340c7969dfdb54648204"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Fri May 17 17:31:42 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat May 18 00:31:42 2024 +0000"
      },
      "message": "Fixes for generic benchmark wrt corstone-300 (#2581)\n\n@tensorflow/micro\n\nUpdate README to use cortex-m4 (it has instruction cycle timer)\n\nUpdate README sample output to reflect cortex-m4\n\nUpdate test script to not use Ethos-U co-processor (instruction cycle timer non-functional in simulator) when running benchmark.\n\nbug\u003dfixes #2580"
    },
    {
      "commit": "94e7cbac65ebb1378a95340c7969dfdb54648204",
      "tree": "e226a90aeddd130d16b933cbc7761a5bf4925cca",
      "parents": [
        "ae690a1d77edf5c8d3affa14a60e1eba87e0cb13"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Fri May 17 16:11:08 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri May 17 23:11:08 2024 +0000"
      },
      "message": "Fix corstone-300 simulator path (#2579)\n\n@tensorflow/micro\n\nPR#2573 updated the corstone-300 simulator download.  This resulted in the path to the simulator binary changing.  This PR updates the PATH environment.\n\nbug\u003dfixes #2578"
    },
    {
      "commit": "ae690a1d77edf5c8d3affa14a60e1eba87e0cb13",
      "tree": "6abd87efca64e5af5a20b4f5c4a2facd2c1a3e67",
      "parents": [
        "b2cf5b3fe60bbf5e1aa790668fe0b665888f7420"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Fri May 17 19:35:38 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri May 17 17:35:38 2024 +0000"
      },
      "message": "Remove #definitions of MAX_RFFT_PWR, MIN_RFFT_PWR from Xtensa flags (#2577)\n\nFirst, the definition -DMIN_RFFT_PWR\u003dMAX_RFFT_PWR\nis meaningless, because MAX_RFFT_PWR doesn\u0027t evaluate to a number. That\u0027s a bug.\nSecond, both #definitions are only applicable when building Nature DSP for hifi3 and hifimini. Later archs (hifi4, hifi5) stopped using them. That\u0027s confusing to define them for all archs.\nFinally, there are a lot of possible #definitions and there\u0027s no reason to define these particular two in the main Xtensa makefile. A user can add any #define to XTENSA_EXTRA_CFLAGS.\nBUG\u003d340206722"
    },
    {
      "commit": "b2cf5b3fe60bbf5e1aa790668fe0b665888f7420",
      "tree": "0967b4521265b4e47b4fd4deff0a8b6560d1ac93",
      "parents": [
        "1973d40343c44648636146fd1b5a818ff9eacba7"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Fri May 17 08:58:41 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri May 17 15:58:41 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2572)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "1973d40343c44648636146fd1b5a818ff9eacba7",
      "tree": "13f5782d6a5a74326a8d7d55d8ea947a2983da28",
      "parents": [
        "b281a0df5faf8b95aae008e67fefebc2dc5e38b6"
      ],
      "author": {
        "name": "Ryan OShea",
        "email": "86965113+ArmRyan@users.noreply.github.com",
        "time": "Thu May 16 17:40:32 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu May 16 15:40:32 2024 +0000"
      },
      "message": "Add cortex300 download for aarch64 linux host (#2573)\n\n* Updates x86 corstone300 version\n \n BUG\u003dCorstone300 tests were not able to be ran on aarch64 hosts\n Author: Ryan O\u0027Shea \u003cryan.oshea3@arm.com\u003e"
    },
    {
      "commit": "b281a0df5faf8b95aae008e67fefebc2dc5e38b6",
      "tree": "f021b0fb755b77b38568d3421b04f9fa722f1d0f",
      "parents": [
        "7dbd9d23734c8c36ddf8e15b1d1c8a3346f5264c"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Wed May 15 07:30:20 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed May 15 14:30:20 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2571)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "7dbd9d23734c8c36ddf8e15b1d1c8a3346f5264c",
      "tree": "c78c03ce5b3d25d0c1ef68c0c41bc8ea01802fdf",
      "parents": [
        "c377b745c25a6384bab4f2ba175cdae34b1103fb"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Tue May 14 10:29:11 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue May 14 17:29:11 2024 +0000"
      },
      "message": "Add meta-data to generic benchmark (#2496)\n\n@tensorflow/micro\n\nAdd meta-data strings to the binary.  Meta-data is output each time binary is executed.  See the README for a sample of the output.\n\nIf a Git repo. is not available at ```${TENSORFLOW_ROOT}```, an appropriate message is generated in the meta-data.\n\nChanges to how the CMSISNN and XTENSA NN library download scripts work.\n\nAdds run of generic benchmark with embedded model for default build of x86, Corstone-300, Xtensa to test scripts executed during CI.\n\nbug\u003dfixes #2495"
    },
    {
      "commit": "c377b745c25a6384bab4f2ba175cdae34b1103fb",
      "tree": "662e1ef5eb395579bbf16fb53c99b6221b47d69d",
      "parents": [
        "5e4e51bbe3a04ec1314efa9562e7eacc9c054ef5"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon May 13 20:11:24 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue May 14 03:11:24 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2567)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "5e4e51bbe3a04ec1314efa9562e7eacc9c054ef5",
      "tree": "0dba7fcbe1f3f87af52b1e326ec86d21c4fd650a",
      "parents": [
        "8a1fe96cfd18a45037836891d81b7b3b418dc642"
      ],
      "author": {
        "name": "Dan Suh",
        "email": "kaehops@gmail.com",
        "time": "Thu May 09 13:53:37 2024 +0900"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu May 09 04:53:37 2024 +0000"
      },
      "message": "Check whether the bias tensor is `nullptr` before accessing the type. (#2566)\n\nPrevents program crash due to null pointer dereference.\n\nBUG\u003dnone"
    },
    {
      "commit": "8a1fe96cfd18a45037836891d81b7b3b418dc642",
      "tree": "d1358d686b9711e9f52ca6150c9f6711d542b655",
      "parents": [
        "6b766e35740a9dbe20fc294e842bd29421eb130b"
      ],
      "author": {
        "name": "David Davis",
        "email": "ddavis-2015@users.noreply.github.com",
        "time": "Wed May 08 10:00:33 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed May 08 17:00:33 2024 +0000"
      },
      "message": "Update documentation (#2565)\n\n@tensorflow/micro\n\nUpdate README with additional makefile targets for this example application.\n\nbug\u003dfixes #2561"
    },
    {
      "commit": "aa3fb6a13c9d8bbc8de57420c6c27ccd1f52e9f2",
      "tree": "30a7cfcc93ec087cd28a5fd5d056913f160dc3e3",
      "parents": [
        "ad2d40bcb7c94a9f7d3f710181452bdabc14c954"
      ],
      "author": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Tue May 07 02:12:55 2024 -0700"
      },
      "committer": {
        "name": "Lun Dong",
        "email": "lundong@google.com",
        "time": "Tue May 07 10:09:37 2024 -0700"
      },
      "message": "tensorflow/tflite-micro: refactoring TFLM Kelvin kernel\n\nThis is the first version for refactoring TFLM Kelvin kernel.\n\nChange-Id: Id61d215e4e6df8e3a09f4663cc5a7f23885f071f\n"
    },
    {
      "commit": "ad2d40bcb7c94a9f7d3f710181452bdabc14c954",
      "tree": "bf7b2507123ba544c8917a79d10c84e88d8ad5f6",
      "parents": [
        "b3d19eefba01e091b6fd4829f22c32f3e3bafedc",
        "f50dac75dbf329a70f2d00ea4b6dd7b015dd7ebd"
      ],
      "author": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Tue May 07 16:28:33 2024 +0000"
      },
      "committer": {
        "name": "Gerrit Code Review",
        "email": "noreply-gerritcodereview@google.com",
        "time": "Tue May 07 16:28:33 2024 +0000"
      },
      "message": "Merge \"Delegate kernel selection for int8 conv to our library\""
    },
    {
      "commit": "6b766e35740a9dbe20fc294e842bd29421eb130b",
      "tree": "824da80b19a0f4be12c299bbf0a55dd1f0ea00de",
      "parents": [
        "cf10f942ed0b0556f96e138a7926194bad958284"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Tue May 07 18:15:59 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue May 07 16:15:59 2024 +0000"
      },
      "message": "Update CMSIS-NN download SHA-1 (#2563)\n\nThe latest CMSIS-NN adds MVE support for int4 convolution, depthwise and fully connected.\nBUG\u003djust updating to latest"
    },
    {
      "commit": "b3d19eefba01e091b6fd4829f22c32f3e3bafedc",
      "tree": "b6fb1cb21838b5f4be392c17d6e6f3312f18510c",
      "parents": [
        "e6de9ac3691fba8baa96ac8169eb04675dd74b59"
      ],
      "author": {
        "name": "Naveen Dodda",
        "email": "ndodda@google.com",
        "time": "Wed Apr 17 17:46:51 2024 +0000"
      },
      "committer": {
        "name": "Naveen Dodda",
        "email": "ndodda@google.com",
        "time": "Fri May 03 18:28:10 2024 +0000"
      },
      "message": "depthwise conv for 2d kernels\n\n* draft version of kernel is working\n*Kernel tirggers for pad \u003d0 and int8\n*need to add more kelvin instrcutions\n\nChange-Id: Ifb16c7f7508458cede8bd078c657b0106e348a99\n"
    },
    {
      "commit": "cf10f942ed0b0556f96e138a7926194bad958284",
      "tree": "74e052c1951bd934ecb24fb887345c33abe1ec81",
      "parents": [
        "ed36c46dd5230608f0c59dda789b6c3570991869"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Fri May 03 08:22:39 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri May 03 15:22:39 2024 +0000"
      },
      "message": "Enable optimized dequantize on HIFI3 (#2560)\n\nBUG\u003dnone"
    },
    {
      "commit": "ed36c46dd5230608f0c59dda789b6c3570991869",
      "tree": "ddad242f20946768c31b3b291db4e452673d9508",
      "parents": [
        "a32a33b4944871fb12c8a6cf00867db9b188e8eb"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Thu May 02 17:11:21 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu May 02 15:11:21 2024 +0000"
      },
      "message": "Fix patch file for Nature NDSP library for  hifi3 (#2541)\n\nBUG\u003d327609938"
    },
    {
      "commit": "a32a33b4944871fb12c8a6cf00867db9b188e8eb",
      "tree": "d5d280ca691324b1ff5b1c65ff20299f35d80bd0",
      "parents": [
        "e5e230552525c012c3b6912761c9a01728ad225c"
      ],
      "author": {
        "name": "Jesse Spielman",
        "email": "jesse.spielman@gmail.com",
        "time": "Tue Apr 30 19:35:25 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Apr 30 18:35:25 2024 +0000"
      },
      "message": "Fix misspelling of integer (#2554)\n\nJust noticed this while reading through the code;  obviously not urgent but thought I\u0027d submit a PR just in case :-)\n\nBUG\u003dtypo"
    },
    {
      "commit": "e5e230552525c012c3b6912761c9a01728ad225c",
      "tree": "e48eb9543431daa4d6377071a29fe1a919d8a7f4",
      "parents": [
        "fe4229cb3cb1adec112c0aabb958d904b7d3703d"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Mon Apr 29 19:39:07 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Apr 29 17:39:07 2024 +0000"
      },
      "message": "CMSIS-NN: Add int32 bias support int16xint8 conv (#2557)\n\nBUG\u003dCMSIS-NN glue is updated to support int32bias for int16x8"
    },
    {
      "commit": "fe4229cb3cb1adec112c0aabb958d904b7d3703d",
      "tree": "d96c794f05abe8e27d11deb30e23a435707a5f86",
      "parents": [
        "0fc313bfe7341b499e3d9150a04f4ac6559151b2"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon Apr 29 08:58:57 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Apr 29 15:58:57 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2558)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "0fc313bfe7341b499e3d9150a04f4ac6559151b2",
      "tree": "9c6b75a48ab46c028b94500974cf7768b54b84ca",
      "parents": [
        "a3be93b49bde7856a135497a02643cb382d88a30"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Fri Apr 26 15:05:48 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Apr 26 22:05:48 2024 +0000"
      },
      "message": "Revert \"gen_micro_mutable script: update import path\" (#2556)\n\nReverts tensorflow/tflite-micro#2545\n\nBUG\u003dbroke some internal scripts."
    },
    {
      "commit": "a3be93b49bde7856a135497a02643cb382d88a30",
      "tree": "6b15cc6350a64c6c1b93def4f8d281876de65a49",
      "parents": [
        "fc498a80493b26226fc8a48b53dff669104bbd74"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Fri Apr 26 12:29:45 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Apr 26 19:29:45 2024 +0000"
      },
      "message": "Enable hifi5 optimized kernels (#2555)\n\nSeveral of the xtensa optimized kernels were guarded with #ifdefs for HIFI3 and HIFI4, but were missing HIFI5. Those kernels have also been enabled via the hifi5 nnlib, so we can add the defines.\n\nBUG\u003dnone"
    },
    {
      "commit": "f50dac75dbf329a70f2d00ea4b6dd7b015dd7ebd",
      "tree": "799e34240365dbd89377d2e030924b420b16ec48",
      "parents": [
        "e6de9ac3691fba8baa96ac8169eb04675dd74b59"
      ],
      "author": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed Apr 24 15:06:54 2024 -0700"
      },
      "committer": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed Apr 24 15:06:54 2024 -0700"
      },
      "message": "Delegate kernel selection for int8 conv to our library\n\n- kelvin::opt::conv_per_channel_b8 now selects from several optimized\n  implementations, or falls back to the generic scalar implementation.\n\nChange-Id: I26ea73368ea43728575ff1557330326399ebacf7\n"
    },
    {
      "commit": "fc498a80493b26226fc8a48b53dff669104bbd74",
      "tree": "24a1b9055667522de08c1f258561a6318af26a5d",
      "parents": [
        "9741578dfae97f3504b5fcccf938796a4750abae"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Wed Apr 24 07:45:16 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Apr 24 14:45:16 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2553)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "9741578dfae97f3504b5fcccf938796a4750abae",
      "tree": "f3209014cb29e8e846308b56f35cad526ded4ca1",
      "parents": [
        "a17682de2038418028d2d1f6674c6d1235f902d3"
      ],
      "author": {
        "name": "Adrian Lundell",
        "email": "36153706+AdrianLundell@users.noreply.github.com",
        "time": "Tue Apr 23 17:01:44 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Apr 23 15:01:44 2024 +0000"
      },
      "message": "Enable int16_t support for CMSIS-NN LSTM kernel (#2551)\n\nBUG\u003dNew feature in CMSIS-NN"
    },
    {
      "commit": "a17682de2038418028d2d1f6674c6d1235f902d3",
      "tree": "bd3fbddd19a6cfd467de8bfce01ae2a146515500",
      "parents": [
        "26fbc4b6c5cc1c4ea3f78292f310c500c3cf3366"
      ],
      "author": {
        "name": "cad-audio",
        "email": "86048415+cad-audio@users.noreply.github.com",
        "time": "Thu Apr 18 08:04:35 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Apr 18 15:04:35 2024 +0000"
      },
      "message": "Quantize, dequantize optimizations for HiFi targets (#2544)\n\nBUG\u003dnone"
    },
    {
      "commit": "26fbc4b6c5cc1c4ea3f78292f310c500c3cf3366",
      "tree": "5df401d2434061f5b8bca2ad863219bfb503929c",
      "parents": [
        "deec0d81ed94ff9489b42a159faa2425638ea5d5"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Wed Apr 17 16:49:56 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Apr 17 23:49:56 2024 +0000"
      },
      "message": "Replace test with invalid-enum-load (#2546)\n\nThe TestTypeSizeOf was casting an invalid value for TfLiteType into a variable of that type. This is undefined behavior and triggered ubsan. This PR replaces this test case with using kTfLiteNoType, which will still trigger the default failure case for TfLiteTypeSizeOf.\n\nBUG\u003db/333708102"
    },
    {
      "commit": "deec0d81ed94ff9489b42a159faa2425638ea5d5",
      "tree": "1af0589bae4aa74fa8ef2a2f26f7ccaa599067f3",
      "parents": [
        "f99fcf84ed0555341f7247fc782a6a9b994569c3"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Thu Apr 18 00:17:34 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Apr 17 22:17:34 2024 +0000"
      },
      "message": "gen_micro_mutable script: update import path (#2545)\n\nBUG\u003dhttps://github.com/tensorflow/tflite-micro/issues/1973"
    },
    {
      "commit": "f99fcf84ed0555341f7247fc782a6a9b994569c3",
      "tree": "cf2e0d8bb88dbb5a8bd04b7183a5b9394ca27cd2",
      "parents": [
        "7aa0d93dcef1bba137c742836aed444cf7c853ef"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Fri Apr 12 17:46:43 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Apr 12 15:46:43 2024 +0000"
      },
      "message": "Use scratch memory instead of perm buffer for the  RFFT kernel\u0027s work area. (#2542)\n\nBUG\u003d333876877"
    },
    {
      "commit": "7aa0d93dcef1bba137c742836aed444cf7c853ef",
      "tree": "ed822bc8e4db66f2bcf7db8a8857e39bc2063ed7",
      "parents": [
        "bb4fc83d1083e75c2cae0dbd59332a29d7a8afe7"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Wed Apr 03 19:02:10 2024 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Apr 03 17:02:10 2024 +0000"
      },
      "message": "Basic makefile for MIPS architecture (#2536)\n\nBUG\u003d331000775"
    },
    {
      "commit": "e6de9ac3691fba8baa96ac8169eb04675dd74b59",
      "tree": "a2a869c07c2b23ce04344961b6b7b07c7f60758b",
      "parents": [
        "07c6b8a8097f22b39651052c66809434c0e647b6"
      ],
      "author": {
        "name": "Michael Brooks",
        "email": "michaelbrooks@google.com",
        "time": "Fri Mar 29 17:40:29 2024 +0000"
      },
      "committer": {
        "name": "Michael Brooks",
        "email": "michaelbrooks@google.com",
        "time": "Fri Mar 29 17:40:29 2024 +0000"
      },
      "message": "Use Kelvin Cycle Count for micro_time\n\nUses the Kelvin cycle count to indicate ticks. The ticks per second\nis ommited (meaning profile.Log() won\u0027t work as expected) because\nthe values in device.h aren\u0027t set for the simulator.\n\nTested:\nsoundstream benchmark with profile\u003dTrue/False\n\nChange-Id: I65a29c06fb4dbc74393e892f5c192a325c87ab89\n"
    },
    {
      "commit": "bb4fc83d1083e75c2cae0dbd59332a29d7a8afe7",
      "tree": "ada08137d8db1c9ea8afd4999941c6fb008aa517",
      "parents": [
        "1b2245f0387c0f299db1daaef2f71a824e57cbf8"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Thu Mar 28 11:56:17 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Mar 28 18:56:17 2024 +0000"
      },
      "message": "Pass cc_op_defs as deps, not srcs, to underlying cc_library rule (#2522)\n\nThis is because the targets are cc_libraries, not C++ source files.\n\nBUG\u003dcl/619288666"
    },
    {
      "commit": "1b2245f0387c0f299db1daaef2f71a824e57cbf8",
      "tree": "65aaf0f8779c9921f347c7b1c016cf11d37e0152",
      "parents": [
        "7005d60ede074762f01c1d0fd24ec58240af89b5"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Thu Mar 28 02:57:40 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Mar 28 01:57:40 2024 +0000"
      },
      "message": "Download CMSIS for ref kernels as well for cortex_m_generic (#2525)\n\nBUG\u003dhttps://github.com/tensorflow/tflite-micro/issues/2523"
    },
    {
      "commit": "7005d60ede074762f01c1d0fd24ec58240af89b5",
      "tree": "175ed070ab341602dc47102a3bfa058cace0f53b",
      "parents": [
        "0ef3452312fec819fa9591844ced836901cd88ff"
      ],
      "author": {
        "name": "Adrian Lundell",
        "email": "36153706+AdrianLundell@users.noreply.github.com",
        "time": "Tue Mar 26 17:55:17 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Mar 26 16:55:17 2024 +0000"
      },
      "message": "Compile with -ffp-mode\u003dfull for armclang (#2517)\n\nThis is needed due to multiple uses of NaN and Infinity which is not supported with the default -ffp-mode\u003dstd.\n\nBUG\u003d#2518"
    },
    {
      "commit": "0ef3452312fec819fa9591844ced836901cd88ff",
      "tree": "a9de8c746e37494e1f194c9e18b7c7b6552b6df5",
      "parents": [
        "d20aa314eaa188e9ac3a468066869090c3119145"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Tue Mar 26 16:11:11 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Mar 26 15:11:11 2024 +0000"
      },
      "message": "Add support for architecture armv8.2-a in Cortex-A target (#2519)\n\nmakefile\n\nBUG\u003d331000775"
    },
    {
      "commit": "d20aa314eaa188e9ac3a468066869090c3119145",
      "tree": "653833df3a88c6dd2eb34ea2e367b91bab677136",
      "parents": [
        "cf0a5609a834895695250518a2a0ac8525e96ccf"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon Mar 25 08:41:27 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Mar 25 15:41:27 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2516)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "cf0a5609a834895695250518a2a0ac8525e96ccf",
      "tree": "8b49162ba647142b71f2d35a0820003b729ec93f",
      "parents": [
        "15092dfdba4007dc7475cdf56f0dfa72ac280355"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Fri Mar 22 17:32:56 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Mar 22 16:32:56 2024 +0000"
      },
      "message": "Add GNU Make rule to print GENDIR (#2514)\n\nUseful when the TFLM makefile is called recursively and build artifacts used.\n\nBUG\u003db/327609938"
    },
    {
      "commit": "15092dfdba4007dc7475cdf56f0dfa72ac280355",
      "tree": "a94dfa0db77cea9ad8eb5c94f6f610d69c78e43f",
      "parents": [
        "79e3f35f3ce03c4f24fc9dfef40fcff5827b0112"
      ],
      "author": {
        "name": "suleshahid",
        "email": "110432064+suleshahid@users.noreply.github.com",
        "time": "Thu Mar 21 21:54:02 2024 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Mar 21 21:54:02 2024 +0000"
      },
      "message": "IF kernel nulltpr fix (#2515)\n\nIn some cases if you IF op has 0 outputs, it can reference a nullptr in trying to access node-\u003eoutputs-\u003esize(). This is in the case if your model was flatbuffer_aligned which most likely optimizes away the empty array to nullptr.\n\nBUG\u003d[328527454](https://buganizer.corp.google.com/issues/328527454)"
    },
    {
      "commit": "79e3f35f3ce03c4f24fc9dfef40fcff5827b0112",
      "tree": "bd9d82d7d75dd8efd653c480a3fe238a1169cf75",
      "parents": [
        "5e17f076bc3780e6e3df614279f7bb898a779a08"
      ],
      "author": {
        "name": "Steven Toribio",
        "email": "34755817+turbotoribio@users.noreply.github.com",
        "time": "Tue Mar 19 14:56:48 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Mar 19 21:56:48 2024 +0000"
      },
      "message": "Remove the `output_to_genfiles \u003d True` parameter from Starlark rules (#2512)\n\ncorresponding [g3 cl](https://critique.corp.google.com/cl/616227359)\n\nBUG\u003dN/A"
    },
    {
      "commit": "5e17f076bc3780e6e3df614279f7bb898a779a08",
      "tree": "aa8b3be2bba15e98a74db5f296cc864d91d61e40",
      "parents": [
        "178cfebaaedcc2321cc896ab7d6d14045c9d826a"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Tue Mar 19 16:23:31 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Mar 19 15:23:31 2024 +0000"
      },
      "message": "Add option to rename symbols in Nature DSP library (#2509)\n\nBUG\u003db/327609938"
    },
    {
      "commit": "178cfebaaedcc2321cc896ab7d6d14045c9d826a",
      "tree": "a22dcb7da787c7795cbfe3f27eb902856fce9f51",
      "parents": [
        "98de88594ab5b59182021225b8a38c12f76b48ce"
      ],
      "author": {
        "name": "Adrian Lundell",
        "email": "36153706+AdrianLundell@users.noreply.github.com",
        "time": "Mon Mar 18 23:56:24 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Mar 18 22:56:24 2024 +0000"
      },
      "message": "CMSIS-NN LSTM issue fixes (#2504)\n\n- Remove read of non-initialized buffer\n- Adds back integer unidirectional_sequence_lstm_test for CMSIS-NN\n- Adds back dtln example using CMSIS-NN\n- Reverts \"Add intermediate tensors to LSTM unit test to enable CMSIS-NN (#1996)\"\n\nBUG\u003d#2449"
    },
    {
      "commit": "98de88594ab5b59182021225b8a38c12f76b48ce",
      "tree": "6920f708c6aa98c2b73cc79264bfe454f479d1b8",
      "parents": [
        "d3088479753fb0963dd72d75bddbc5ff5bf64aec"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Mon Mar 18 15:09:28 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Mar 18 22:09:28 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2510)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "d3088479753fb0963dd72d75bddbc5ff5bf64aec",
      "tree": "545f810c93ad316b3e6ea29381b0e4483faf5087",
      "parents": [
        "f9044b03343d282e9f7dc8db489e6f02dc36278c"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Fri Mar 15 23:29:09 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Mar 15 22:29:09 2024 +0000"
      },
      "message": "Update examples and documentation for Arm(R) Corstone(TM)-300 FVP (#2503)\n\n* Updates benchmark documentation.\n* Updates benchmarks and network tester example.\n* Select only needed ops for memory measurements.\n* Only pip install if needed.\n* Also gen folder output will be created differently depending on also toolchain and type of kernels\n\nBUG\u003ddocumentation for the benchmark application is not correct"
    },
    {
      "commit": "f9044b03343d282e9f7dc8db489e6f02dc36278c",
      "tree": "dcaf8b9c583c613d4c36383718e812618fe0ab87",
      "parents": [
        "32876bce6933d91d6f7e07131d31bb22ad9dbb0c"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Fri Mar 15 14:04:11 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Mar 15 14:04:11 2024 -0700"
      },
      "message": "Remove Wave as dependency (#2508)\n\nThe Wave pypi package was recently updated, and the new version caused\r\nfailures in our CI for pip installing it. From what I can tell, this\r\npackage is no longer being used, so this PR simply removes it as a build\r\ndependency.\r\n\r\nBUG\u003db/329887147"
    },
    {
      "commit": "32876bce6933d91d6f7e07131d31bb22ad9dbb0c",
      "tree": "b468174540aaa9aae0b17cfa8f24d6378cb94a3c",
      "parents": [
        "8f143b5fe098e7b2ba965e76da205039a45fe8f8"
      ],
      "author": {
        "name": "TFLM-bot",
        "email": "tflm-github-bot@google.com",
        "time": "Fri Mar 15 10:11:08 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Mar 15 17:11:08 2024 +0000"
      },
      "message": "Automated sync from github.com/tensorflow/tensorflow (#2502)\n\nBUG\u003dautomated sync from upstream\nNO_CHECK_TFLITE_FILES\u003dautomated sync from upstream"
    },
    {
      "commit": "8f143b5fe098e7b2ba965e76da205039a45fe8f8",
      "tree": "9be755c2c5759f545e7138fa332c41beff57b056",
      "parents": [
        "74c4632d4d29ca7c8242729e6f3aac97c488558d"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Fri Mar 15 09:27:30 2024 -0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Mar 15 09:27:30 2024 -0700"
      },
      "message": "Add workaround for newer GitHub runners (#2505)\n\nAsan in llvm 14 provided in ubuntu 22.04 is incompatible with\r\nhigh-entropy ASLR in much newer kernels that GitHub runners are using\r\nleading to random crashes: https://reviews.llvm.org/D148280\r\n\r\nThis workaround comes from the discussion in:\r\nhttps://github.com/actions/runner-images/issues/9491\r\n\r\nBUG\u003d#2501"
    },
    {
      "commit": "74c4632d4d29ca7c8242729e6f3aac97c488558d",
      "tree": "52a6d7c800168bcb9229c5aecda61b63449d9e8a",
      "parents": [
        "ec086d23f7ff0c8790c690cb1dbe85e09a0fea7c"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Mon Mar 11 22:19:08 2024 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Mar 11 22:19:08 2024 +0000"
      },
      "message": "Unique Init, Prepare, Eval functions of signal kernels. (#2497)\n\nRefactor init, prepare , and eval functions to be unique names for signal/ kernels\n\nBUG\u003db/327655885"
    },
    {
      "commit": "ec086d23f7ff0c8790c690cb1dbe85e09a0fea7c",
      "tree": "7df4aeb2a828cf36295914ebae8308760533de43",
      "parents": [
        "1f7f54b95ef4fc215855b850358e8770aed89287"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Wed Mar 06 16:15:11 2024 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Mar 06 16:15:11 2024 -0800"
      },
      "message": "Fix tests_entry \u0026 tests_post workflows (#2493)\n\nThe tests_entry workflow was invalid due to parsing error when a `\u003d\u003d`\r\nwas expected instead of `\u003d`. The tests_post needed to remove the\r\nci:run_full label.\r\n\r\nBUG\u003db/327686165"
    },
    {
      "commit": "1f7f54b95ef4fc215855b850358e8770aed89287",
      "tree": "f912b0441c51e4d4c2aacf29cefec78e72104c8e",
      "parents": [
        "9adee084ecf1244a55bca2b45d5a5b6e36b7ec6e"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Mon Mar 04 15:56:23 2024 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Mar 04 23:56:23 2024 +0000"
      },
      "message": "Fix GitHub workflow conditionals (#2489)\n\nBUG\u003db/327686165"
    },
    {
      "commit": "07c6b8a8097f22b39651052c66809434c0e647b6",
      "tree": "a38d72b9e60fbc39219c1eacade95b05ab9595af",
      "parents": [
        "6e40c788f32279917e360d133150280b24e9a020"
      ],
      "author": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed Feb 28 11:06:15 2024 -0800"
      },
      "committer": {
        "name": "Alex Van Damme",
        "email": "atv@google.com",
        "time": "Wed Feb 28 11:06:15 2024 -0800"
      },
      "message": "Add a second option for kelvin\u0027s platform select\n\nChange-Id: I8a8b41b0108bad30248d4af4c17c209500f7fc81\n"
    },
    {
      "commit": "9adee084ecf1244a55bca2b45d5a5b6e36b7ec6e",
      "tree": "9866621a732288db3d3f2b6109d25aa07ff6c89e",
      "parents": [
        "20fd5b23ac305918cf391158e931573c3507b101"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Wed Feb 28 10:45:56 2024 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Feb 28 18:45:56 2024 +0000"
      },
      "message": "Remove experimental flag for per-channel FC quantization (#2482)\n\nPer-channel quantization in fully connected layers are still not supported by TFLM, but the converter now has proper support so we can remove the flag.\n\nBUG\u003dcl/610755484"
    },
    {
      "commit": "20fd5b23ac305918cf391158e931573c3507b101",
      "tree": "fdb08bcfd4ac2e06ed5041896f88ede74a5c566d",
      "parents": [
        "b991e726b01393580fafc192d6e18ef7447180fb"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Wed Feb 28 10:20:50 2024 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Feb 28 10:20:50 2024 -0800"
      },
      "message": "Prevent runtime_shape.cc removal during sync (#2484)\n\nDespite existing in TFLite, the runtime_shape.h has long differed\r\nbetween TFLite and TFLM. The file is not copied during the sync and the\r\nsync script does a `git checkout` on the file to ensure that the\r\nexisting version in the TFLM tree remains.\r\n\r\nIn PR #2476, we needed to add a runtime_shape.cc file. This PR ensures\r\nthat the runtime_shape.cc file will not be removed during the sync by\r\nperforming a `git checkout` on the existing file.\r\n\r\nBUG\u003d323856831"
    },
    {
      "commit": "b991e726b01393580fafc192d6e18ef7447180fb",
      "tree": "7c0313711a74b1a2c7d67c428939aec584f6a897",
      "parents": [
        "0445accf1104cd4583f107a17072b1ca92d853fb"
      ],
      "author": {
        "name": "Shlomi Regev",
        "email": "shlmregev@users.noreply.github.com",
        "time": "Tue Feb 27 14:35:14 2024 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Feb 27 14:35:14 2024 +0000"
      },
      "message": "Fix a linker error: undefined kMaxSmallSize (#2476)\n\nWhen building with a recent version of xt-clang and -std\u003dc++11, the linker errs on missing definition of a static constexpr class member. In C++11, static class members still had to be defined in a C++11 file and TFLM code is expected to be compatible with C++11.\n\nBUG\u003d323856831"
    },
    {
      "commit": "0445accf1104cd4583f107a17072b1ca92d853fb",
      "tree": "8833b9bf1a095bba4278e172650259868efb9f18",
      "parents": [
        "cfa4c91d1b36c37c7c104b9c664615e59f1abfe3"
      ],
      "author": {
        "name": "RJ Ascani",
        "email": "rjascani@google.com",
        "time": "Mon Feb 26 19:19:11 2024 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Feb 27 03:19:11 2024 +0000"
      },
      "message": "Use absl::Status instead of tsl::Status (#2477)\n\ntsl::Status has been deprecated and is just an alias for absl::Status. This PR updates the usage to just use absl::Status directly.\n\nBUG\u003dcl/609910156"
    },
    {
      "commit": "6e40c788f32279917e360d133150280b24e9a020",
      "tree": "3c7c8eda88e0dd4487e91bfff9fe28f815b3eccb",
      "parents": [
        "4b56c865eb1ccb40bc48319e56837584508893fe"
      ],
      "author": {
        "name": "Derek Chow",
        "email": "derekjchow@google.com",
        "time": "Mon Feb 26 15:55:30 2024 -0800"
      },
      "committer": {
        "name": "Derek Chow",
        "email": "derekjchow@google.com",
        "time": "Mon Feb 26 15:55:30 2024 -0800"
      },
      "message": "Prefix platform check with kelvin_sw.\n\nChange-Id: I27aafd64c229e219ab2c9953f5ac783cb49cae45\n"
    },
    {
      "commit": "cfa4c91d1b36c37c7c104b9c664615e59f1abfe3",
      "tree": "1806762d442c096a30fa7b8f5749190f29f86c37",
      "parents": [
        "8085cecefcc2daefbc4c4480c644b59052841b95"
      ],
      "author": {
        "name": "Alpha BAO",
        "email": "zhuijun.bao@outlook.com",
        "time": "Sat Feb 24 13:40:47 2024 +0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Feb 24 05:40:47 2024 +0000"
      },
      "message": "Fix array out-of-bounds access in WideDynamicFuncLut (#2468)\n\n### Problem description:\n\nIn the original code, pointer arithmetic of gain_lut and the assignment of gain_lut[4 * interval + 3] could potentially lead to out-of-bounds array access.\nOn certain architectures (e.g., macOS ARM64), this out-of-bounds access causes the program to crash.\nBUG\u003dNone, reported issue#2464\n\n### Solution:\n\nIncrease the size of the gain_lut_storage array by 1 to provide an extra buffer and prevent overflow during the calculation within the loop.\n\n### Risks and considerations:\n\nIncreasing the array size will slightly increase memory usage.\nIn extremely resource-constrained systems, alternative algorithm implementations may need to be considered."
    },
    {
      "commit": "8085cecefcc2daefbc4c4480c644b59052841b95",
      "tree": "400876d3bad9ca820fdcde47ec8639e1a277da71",
      "parents": [
        "f5e498b68bddce1ed18c96f8f4b5e866fdd0dcac"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Fri Feb 23 19:03:57 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Feb 23 18:03:57 2024 +0000"
      },
      "message": "Add kernel build flag for prioritizing speed or size (#2408)\n\nAdds a build flag that can be used by any kernel to provide a different implementation depending on use case.\nAdds a first use case for cmsis-nn transpose conv.\n\nThe background for this PR is in https://github.com/tensorflow/tflite-micro/pull/2345\n\nBUG\u003dnone "
    },
    {
      "commit": "f5e498b68bddce1ed18c96f8f4b5e866fdd0dcac",
      "tree": "76e38e5116cc2e5565b6e8d1a5313bb3f532f718",
      "parents": [
        "29526e413bdd7d27594e6fe261de368fb5689731"
      ],
      "author": {
        "name": "Måns Nilsson",
        "email": "mans.nilsson@arm.com",
        "time": "Fri Feb 23 18:41:53 2024 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Feb 23 17:41:53 2024 +0000"
      },
      "message": "cmsis-nn: update fully connected int8 (#2469)\n\n- Adds non zero filter offset support.\n- Adds support for batch matmul like behavior where weights are like input, i.e. not initialized before eval.\n\nBUG\u003dnon zero filter offset not supported for CMSIS-NN"
    }
  ],
  "next": "29526e413bdd7d27594e6fe261de368fb5689731"
}