Datasets:

---

Isuru0x01

/

tensorflow-issues

like 0

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 1097545817, "node_id": "MDU6TGFiZWwxMDk3NTQ1ODE3", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:apis", "name": "comp:apis", "color": "0052cc", "default": false, "description": "Highlevel API related issues" }, { "id": 4257391797, "node_id": "LA_kwDOArmXAs79wqS1", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/wsl2", "name": "wsl2", "color": "A352E6", "default": false, "description": "Windows Subsystem for Linux" }, { "id": 4511033337, "node_id": "LA_kwDOArmXAs8AAAABDODn-Q", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.10", "name": "TF 2.10", "color": "C15088", "default": false, "description": "" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "@abhinavchawla7,\r\nTensorFlow 2.10 was the last TensorFlow release that supported GPUs on native-Windows. Could you please provide any specific reason you are using WSL2 for 2.10, you can directly install the v2.12 on native-Windows which was supported. \r\ndoes the tf.test.is_gpu_available return False? If that's the case, then it definitely means the GPU is not configured correctly.\r\n\r\nAlso, the error message somehow indicating that the GPU device is not there:\r\n\r\nRegistered devices: [CPU,XLA_CPU,XLA_GPU]\r\n\r\nXLA_GPU is not same as GPU.\r\n\r\nMay I know have you tried to execute the same code with the CPU only if yes is it working fine with the CPU ? If not please try it to run on either with CPU only or on Google Colab and check whether it is working as expected or not ? Thank you!", "I am on windows 11 and i have installed CUDA on native windows. \r\n\r\nAs per the link: https://www.tensorflow.org/install/pip#windows-native, it is said that TensorFlow 2.10 was the last TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow 2.11, you will need to install [TensorFlow in WSL2](https://tensorflow.org/install/pip#windows-wsl2), or install tensorflow or tensorflow-cpu and, optionally, try the [TensorFlow-DirectML-Plugin](https://github.com/microsoft/tensorflow-directml-plugin#tensorflow-directml-plugin-).\r\n\r\nso when i install tensorflow >2.10.* i am unable to see the gpu on wsl2.....and on installing directml-plugin i am unable to run RNN model only CNN model works.....I have NVIDIA 16gb GPU (8gb dedicated) and 32GB CPU.....RNN model works fiine with CPU but kernel crashes on GPU....while CNN works easily on both ", "Awaiting for reply", "Still no reply", "Sorry, I don't think I can help much here. Assigning this one back to you @sachinprasadhs. " ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473184161, "node_id": "MDU6TGFiZWw0NzMxODQxNjE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:support", "name": "type:support", "color": "159b2e", "default": false, "description": "Support issues" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 2671339633, "node_id": "MDU6TGFiZWwyNjcxMzM5NjMz", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteGpuDelegate", "name": "TFLiteGpuDelegate", "color": "F71F04", "default": false, "description": "TFLite Gpu delegate issue" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "sirakiin", "id": 4479310, "node_id": "MDQ6VXNlcjQ0NzkzMTA=", "avatar_url": "https://avatars.githubusercontent.com/u/4479310?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sirakiin", "html_url": "https://github.com/sirakiin", "followers_url": "https://api.github.com/users/sirakiin/followers", "following_url": "https://api.github.com/users/sirakiin/following{/other_user}", "gists_url": "https://api.github.com/users/sirakiin/gists{/gist_id}", "starred_url": "https://api.github.com/users/sirakiin/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sirakiin/subscriptions", "organizations_url": "https://api.github.com/users/sirakiin/orgs", "repos_url": "https://api.github.com/users/sirakiin/repos", "events_url": "https://api.github.com/users/sirakiin/events{/privacy}", "received_events_url": "https://api.github.com/users/sirakiin/received_events", "type": "User", "site_admin": false }, { "login": "grantjensen", "id": 58962643, "node_id": "MDQ6VXNlcjU4OTYyNjQz", "avatar_url": "https://avatars.githubusercontent.com/u/58962643?v=4", "gravatar_id": "", "url": "https://api.github.com/users/grantjensen", "html_url": "https://github.com/grantjensen", "followers_url": "https://api.github.com/users/grantjensen/followers", "following_url": "https://api.github.com/users/grantjensen/following{/other_user}", "gists_url": "https://api.github.com/users/grantjensen/gists{/gist_id}", "starred_url": "https://api.github.com/users/grantjensen/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/grantjensen/subscriptions", "organizations_url": "https://api.github.com/users/grantjensen/orgs", "repos_url": "https://api.github.com/users/grantjensen/repos", "events_url": "https://api.github.com/users/grantjensen/events{/privacy}", "received_events_url": "https://api.github.com/users/grantjensen/received_events", "type": "User", "site_admin": false }, { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @ldfandian \r\n\r\nCan you please add the following to the Android manifest in order to detect GPU delegate.\r\n\r\n```\r\n<uses-library android:name=\"libOpenCL.so\"\r\n android:required=\"false\"/>\r\n\r\n<uses-library android:name=\"libOpenCL-pixel.so\"\r\n android:required=\"false\"/>\r\n```\r\n\r\nThanks.\r\n", "> Hi @ldfandian\r\n> \r\n> Can you please add the following to the Android manifest in order to detect GPU delegate.\r\n> \r\n> ```\r\n> <uses-library android:name=\"libOpenCL.so\"\r\n> android:required=\"false\"/>\r\n> \r\n> <uses-library android:name=\"libOpenCL-pixel.so\"\r\n> android:required=\"false\"/>\r\n> ```\r\n> \r\n> Thanks.\r\n\r\nThanks for the quick response.\r\n\r\nI tried to add the following code in AndroidManifest.xml, but it does not work.\r\n\r\n```\r\n <application\r\n android:allowBackup=\"true\"\r\n android:icon=\"@mipmap/ic_launcher\"\r\n android:label=\"@string/app_name\"\r\n android:roundIcon=\"@mipmap/ic_launcher_round\"\r\n android:supportsRtl=\"true\"\r\n android:taskAffinity=\"\"\r\n tools:ignore=\"AllowBackup\">\r\n\r\n <uses-native-library\r\n android:name=\"libOpenCL.so\"\r\n android:required=\"false\" />\r\n <uses-native-library\r\n android:name=\"libOpenCL-pixel.so\"\r\n android:required=\"false\" />\r\n <uses-native-library\r\n android:name=\"libGLESv2.so\"\r\n android:required=\"false\" />\r\n <uses-native-library\r\n android:name=\"libGLESv3.so\"\r\n android:required=\"false\" />\r\n <uses-native-library\r\n android:name=\"libvulkan.so\"\r\n android:required=\"false\" />\r\n <uses-native-library\r\n android:name=\"libneuralnetworks.so\"\r\n android:required=\"false\" />\r\n\r\n <uses-library\r\n android:name=\"libOpenCL.so\"\r\n android:required=\"false\" />\r\n <uses-library\r\n android:name=\"libOpenCL-pixel.so\"\r\n android:required=\"false\" />\r\n <uses-library\r\n android:name=\"libGLESv2.so\"\r\n android:required=\"false\" />\r\n <uses-library\r\n android:name=\"libGLESv3.so\"\r\n android:required=\"false\" />\r\n <uses-library\r\n android:name=\"libvulkan.so\"\r\n android:required=\"false\" />\r\n <uses-library\r\n android:name=\"libneuralnetworks.so\"\r\n android:required=\"false\" />\r\n ...\r\n```\r\n\r\nBtw, on the Xiaomi 12s Ultra, I checked that the file \"/system/vendor/lib64/libOpenCL.so\" does exist.\r\nHowever, it looks like it does not show in the \"/system/etc/public.libraries.txt\".\r\nMaybe it is the cause? Meaning that the mobile vendor blocks the GPU delegate?\r\n\r\nThe content of \"/system/etc/public.libraries.txt\" and \"/etc/public.libraries.txt\" is like:\r\n```\r\n# See https://android.googlesource.com/platform/ndk/+/master/docs/PlatformApis.md\r\nlibandroid.so\r\nlibaaudio.so\r\nlibamidi.so\r\nlibbinder_ndk.so\r\nlibc.so\r\nlibcamera2ndk.so\r\nlibdl.so\r\nlibEGL.so\r\nlibGLESv1_CM.so\r\nlibGLESv2.so\r\nlibGLESv3.so\r\nlibicu.so\r\nlibicui18n.so\r\nlibicuuc.so\r\nlibjnigraphics.so\r\nliblog.so\r\nlibmediandk.so\r\nlibm.so\r\nlibnativehelper.so\r\nlibnativewindow.so\r\nliblistenjni.qti.so\r\nliblistensoundmodel2.qti.so\r\nlibneuralnetworks.so nopreload\r\nlibOpenMAXAL.so\r\nlibOpenSLES.so\r\nlibRS.so\r\nlibstdc++.so\r\nlibsync.so\r\nlibvulkan.so\r\nlibwebviewchromium_plat_support.so\r\nlibz.so\r\n```\r\n\r\nThe content of \"/vendor/etc/public.libraries.txt\" and \"/system/vendor/etc/public.libraries.txt\" is like:\r\n```\r\nlibqti-perfd-client.so\r\nlibadsprpc.so\r\nlibcdsprpc.so\r\nlibsdsprpc.so\r\nlibfastcvopt.so\r\nlibOpenCL.so\r\nlibSNPE.so\r\nlibmialgo_ai_vision.so\r\nlibmialgo_utils.so\r\nlibxmi_slow_motion_mein.so\r\n```\r\n\r\n\r\nHere is the logcat info:\r\n```\r\n2023-06-10 01:19:29.175 30415-22047 AdrenoGLES-0 org....examples.imageclassification I QUALCOMM build : 5e81ec0141, Icc5bd9b9d5\r\n Build Date : 11/07/22\r\n OpenGL ES Shader Compiler Version: EV031.36.08.11\r\n Local Branch : \r\n Remote Branch : \r\n Remote Branch : \r\n Reconstruct Branch : \r\n2023-06-10 01:19:29.175 30415-22047 AdrenoGLES-0 org....examples.imageclassification I Build Config : S P 12.1.1 AArch64\r\n2023-06-10 01:19:29.175 30415-22047 AdrenoGLES-0 org....examples.imageclassification I Driver Path : /vendor/lib64/egl/libGLESv2_adreno.so\r\n2023-06-10 01:19:29.175 30415-22047 AdrenoGLES-0 org....examples.imageclassification I Driver Version : 0615.50\r\n2023-06-10 01:19:29.181 30415-22047 AdrenoGLES-0 org....examples.imageclassification I PFP: 0x01730155, ME: 0x00000000\r\n\r\n...\r\n\r\n2023-06-10 01:30:26.249 6749-25290 libEGL org....examples.imageclassification E pre_cache appList: com.sina.weibo,com.ss.android.article.news,com.taobao.taobao,com.smile.gifmaker,com.ss.android.ugc.aweme,com.tencent.mm,tv.danmaku.bili,\r\n...\r\n2023-06-10 01:26:11.511 6746-24015 libEGL org....examples.imageclassification E call to OpenGL ES API with no current context (logged once per thread)\r\n...\r\n\r\n\r\n```\r\n", "An update, after adding those uses-library/uses-native-library settings in AndroidManifest.xml, OnePlus 8T works but the other three devices still don't work...\r\n\r\nI compared the their content of /system/etc/public.libraries.txt, but find nothing different among these devices.\r\n\r\nWhat's the magic here?", "Hi @ldfandian,\r\n\r\nHave you tried using our acceleration service to see which configuration might be optimal for each of your phones? https://www.tensorflow.org/lite/android/acceleration_service\r\n\r\nHave you checked the documentation here: https://www.tensorflow.org/lite/android/delegates/gpu_native ?\r\n\r\nAlso for more information, are you using an emulator or the actual phones?", "> Hi @ldfandian,\r\n> \r\n> Have you tried using our acceleration service to see which configuration might be optimal for each of your phones? https://www.tensorflow.org/lite/android/acceleration_service\r\n> \r\n> Have you checked the documentation here: https://www.tensorflow.org/lite/android/delegates/gpu_native ?\r\n> \r\n> Also for more information, are you using an emulator or the actual phones?\r\n\r\nThanks for the advice.\r\n\r\nGood to know it, but my phone does not have a Google play service installed~", "No worries, @ldfandian did the documentation help at all? Additionally please let us know if you are using emulators or actual phones or some combination of them. Thanks.", "> No worries, @ldfandian did the documentation help at all? Additionally please let us know if you are using emulators or actual phones or some combination of them. Thanks.\r\n\r\nAll of these (Xiaomi 12s Ultra, OnePlus 8, Honor Nova 10, Oppo Reno 9) are real phones sold in China.", "Hi @sirakiin, can you please take a look?", "Triage to @grantjensen . Could you help on this? Thanks!", "Hi @ldfandian, looking at the GPU's used in these devices (Adreno 730, Adreno 642L) we have full support. I was able to confirm this via the following commands:\r\n\r\n`bazel build -c opt --config=android_arm64 --copt=-DCL_DELEGATE_NO_GL third_party/tensorflow/lite/tools/benchmark:benchmark_model `\r\n\r\n`adb push blaze-bin/third_party/tensorflow/lite/tools/benchmark/benchmark_model /data/local/tmp`\r\n\r\n`adb push test.tflite /data/local/tmp`\r\n\r\n`adb shell /data/local/tmp/benchmark_model --use_gpu=true --graph=/data/local/tmp/test.tflite`\r\n\r\nThis way we can confirm whether your issue is confined to TFL code or something broader. Please run the above code & let me know how it goes, thanks.\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60825\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60825\">No</a>\n", "Actually, I have similar issue too. I cannot do GPU prediction on Phone. \r\n\r\nTensorflow : 2.9\r\nAndroind : 13\r\nPhone - Samsung A73 5G\r\n\r\nI am getting this flag as False\r\ncompatList.isDelegateSupportedOnThisDevice: false" ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" }, { "id": 1178505529, "node_id": "MDU6TGFiZWwxMTc4NTA1NTI5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/prtype:bugfix", "name": "prtype:bugfix", "color": "159b2e", "default": false, "description": "PR to fix a bug" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Hi @kenfranko Can you please review this PR ? Thank you!", "Hi @SuryanarayanaY Can you please check @kenfranko's [comments](https://github.com/tensorflow/tensorflow/pull/60824#discussion_r1239106026) and keep us posted? Thank you!", "Done the suggested changes." ]

[ { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" }, { "id": 1478826728, "node_id": "MDU6TGFiZWwxNDc4ODI2NzI4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:core", "name": "comp:core", "color": "024391", "default": false, "description": "issues related to core part of tensorflow" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "@cantonios, Thanks for the review. I was not able to find a testcase in the framework, (that I could test and enable) for bfloat16. Do you want me to add a python testcase or C++ testcase for this?. If you can give me a reference testcase it will be great. I tried updating and running this testcase tensorflow/python/ops/image_ops_test.py (only forImageProjectiveTransformV2) . It fails for different versions of protobuf (latest : 4.23, and based on comments in failure: 3.19, 3.20).\r\n", "> @cantonios, Thanks for the review. I was not able to find a testcase in the framework, (that I could test and enable) for bfloat16. Do you want me to add a python testcase or C++ testcase for this?. If you can give me a reference testcase it will be great. I tried updating and running this testcase tensorflow/python/ops/image_ops_test.py (only forImageProjectiveTransformV2) . It fails for different versions of protobuf (latest : 4.23, and based on comments in failure: 3.19, 3.20).\r\n\r\nYes, modifying the existing `ImagProjectiveTransformV2` python test is fine. I don't understand your comment about protobuf issues... the test is currently enabled, so should be passing. It sounds like a configuration error on your end.", "Hi @jojivk73 Any update on this PR? Please. Thank you!", "Hi @gbaned , Changes under internal review. Will upstream soon. Thanks", "This PR is stale because it has been open for 14 days with no activity. It will be closed if no further activity occurs. Thank you.", "This PR was closed because it has been inactive for 14 days since being marked as stale. Please reopen if you'd like to work on this further." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1114343535, "node_id": "MDU6TGFiZWwxMTE0MzQzNTM1", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:data", "name": "comp:data", "color": "0052cc", "default": false, "description": "tf.data related issues" }, { "id": 1463677878, "node_id": "MDU6TGFiZWwxNDYzNjc3ODc4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:performance", "name": "type:performance", "color": "159b2e", "default": false, "description": "Performance Issue" }, { "id": 3531398540, "node_id": "LA_kwDOArmXAs7SfN2M", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.7", "name": "TF 2.7", "color": "77237D", "default": false, "description": "Issues related to TF 2.7.0" } ]

[ { "login": "sushreebarsa", "id": 84765720, "node_id": "MDQ6VXNlcjg0NzY1NzIw", "avatar_url": "https://avatars.githubusercontent.com/u/84765720?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sushreebarsa", "html_url": "https://github.com/sushreebarsa", "followers_url": "https://api.github.com/users/sushreebarsa/followers", "following_url": "https://api.github.com/users/sushreebarsa/following{/other_user}", "gists_url": "https://api.github.com/users/sushreebarsa/gists{/gist_id}", "starred_url": "https://api.github.com/users/sushreebarsa/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sushreebarsa/subscriptions", "organizations_url": "https://api.github.com/users/sushreebarsa/orgs", "repos_url": "https://api.github.com/users/sushreebarsa/repos", "events_url": "https://api.github.com/users/sushreebarsa/events{/privacy}", "received_events_url": "https://api.github.com/users/sushreebarsa/received_events", "type": "User", "site_admin": false } ]

[ "Hi, @chnzhangrui! I tried this provided code in TF v[2.8](https://colab.research.google.com/gist/sushreebarsa/12b8d058b114457a93847bdccc7ea87c/60822-2-12.ipynb#scrollTo=tuajrZc2nTS1), TF v[2.12](https://colab.research.google.com/gist/sushreebarsa/c1ad2acaa810562a063e26bce51a2a49/60822-2-12.ipynb) and tf-[nightly](https://colab.research.google.com/gist/sushreebarsa/10f739e75d7dc06531aca5155e40fcef/untitled793.ipynb#scrollTo=wTUkrK8AnQ4D) using colab. The issue is not replicating as described here. Could you please have a look at these gists and confirm the same. Thank you!", "Hi @sushreebarsa, thanks for the quick response. I checked your gists and I can also confirm your (fast) result. However, it is not clear to me if it was running on GPU or CPU for you. The problem persist only if I run on GPU and I don't know if I can connect to a GPU for test on colab. Let's focus on v2.8.", "Hi @chnzhangrui I have tested the issue on colab using GPU runtime with TF v2.8 as well which is also not replicating the issue reported here. Please find the gist [here](https://colab.research.google.com/gist/sushreebarsa/661982f1752ab6fe6f1b4c78bd35fabf/60822-2-12.ipynb) for reference. Thank you!", "Hi @sushreebarsa, thanks for the confirmation. I am trying to confirm this with my system again and it currently encountered an issue. Will update once I have some idea. I am a little concerned that the fast result you see is really from GPU or just CPU, because this code runs very fast on CPU. I may need to come up with a code that doesn't work on CPU.", "@chnzhangrui Thank you for your response. Please let us know if there is any new update?\r\nThank you!", "Sorry for the silence. I am still waiting for response from our IT support.", "@chnzhangrui Is there any update on this issue?\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60822\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60822\">No</a>\n" ]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" }, { "id": 1478826728, "node_id": "MDU6TGFiZWwxNDc4ODI2NzI4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:core", "name": "comp:core", "color": "024391", "default": false, "description": "issues related to core part of tensorflow" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "/cc @penpornk ", "Do you want to keep the old setting as well, and maybe branch depending on zendnn version? Or will everyone be using ZenDNNv4.0?", "@cantonios We do not want to keep the old setting. Everyone will be using ZenDNNv4.0." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 2671339633, "node_id": "MDU6TGFiZWwyNjcxMzM5NjMz", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteGpuDelegate", "name": "TFLiteGpuDelegate", "color": "F71F04", "default": false, "description": "TFLite Gpu delegate issue" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "Hi @ldfandian \r\n\r\nCan you please try in TF Nightly and let us know if you are still facing the issue?\r\n\r\n`implementation 'org.tensorflow:tensorflow-lite-*:0.0.0-nightly-SNAPSHOT'`\r\n\r\nThanks.", "> Hi @ldfandian\r\n> \r\n> Can you please try in TF Nightly and let us know if you are still facing the issue?\r\n> \r\n> `implementation 'org.tensorflow:tensorflow-lite-*:0.0.0-nightly-SNAPSHOT'`\r\n> \r\n> Thanks.\r\n\r\nOK. I will try it.\r\nAfter some tuning on the ops, now STRIDED_SLICE op is not complained any more.\r\n\r\nBut I get a new error message like this. What does it mean?\r\n\r\n```\r\n01:58:27.544 I isKeyboardTypeChanged: false\r\n01:58:27.544 I isKeyboardTypeChanged: false\r\n01:58:27.666 E call to OpenGL ES API with no current context (logged once per thread)\r\n--------- beginning of crash\r\n01:58:27.685 I Created 0 GPU delegate kernels.\r\n01:58:27.690 E FATAL EXCEPTION: Thread-23\r\n Process: com.xsmart.recall.android, PID: 22340\r\n java.lang.IllegalArgumentException: Internal error: Failed to apply delegate: Unrecognized Read selector\r\n Falling back to OpenGL\r\n TfLiteGpuDelegate Init: Batch size mismatch, expected 1 but got 49\r\n TfLiteGpuDelegate Prepare: delegate is not initialized\r\n Node number 409 (TfLiteGpuDelegateV2) failed to prepare.\r\n Restored original execution plan after delegate application failure.\r\n \tat org.tensorflow.lite.NativeInterpreterWrapper.createInterpreter(Native Method)\r\n \tat org.tensorflow.lite.NativeInterpreterWrapper.init(NativeInterpreterWrapper.java:110)\r\n \tat org.tensorflow.lite.NativeInterpreterWrapper.<init>(NativeInterpreterWrapper.java:73)\r\n \tat org.tensorflow.lite.NativeInterpreterWrapperExperimental.<init>(NativeInterpreterWrapperExperimental.java:36)\r\n \tat org.tensorflow.lite.Interpreter.<init>(Interpreter.java:214)\r\n\r\n```\r\n\r\nthe related node is like:\r\n<img width=\"300\" alt=\"image\" src=\"https://github.com/tensorflow/tensorflow/assets/5018331/e5a71a84-c510-4808-a333-0db918039442\">\r\n", "Hi @ldfandian \r\n\r\nAs suggested [here](https://github.com/tensorflow/tensorflow/issues/56545#issuecomment-1190924808), the whole path with SHAPE ->STRIDED_SLICE -> PACK is a static thing you can replace with a static tensor to support GPU delegate.\r\n\r\nThanks.", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60817\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60817\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 5508003926, "node_id": "LA_kwDOArmXAs8AAAABSE14Vg", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.13", "name": "TF 2.13", "color": "B13ACB", "default": false, "description": "For issues related to Tensorflow 2.13" } ]

[ { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @ramonhollands,\r\n\r\nThe release will be notified here: https://cocoapods.org/pods/TensorFlowLiteSwift. Unfortunately, we don't have a publicly available planned release date currently. Please close if this answers your question.", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1463677878, "node_id": "MDU6TGFiZWwxNDYzNjc3ODc4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:performance", "name": "type:performance", "color": "159b2e", "default": false, "description": "Performance Issue" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @akrapukhin, go ahead and make the PR stating that it fixes this issue. We will review internally and if we need a more general solution it will likely be part of the review (or of course if you figure it out feel free to update the PR). We appreciate your contribution!", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "> \r\n\r\nOk, will make a PR then, thanks", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60815\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60815\">No</a>\n" ]

[ { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 2671339633, "node_id": "MDU6TGFiZWwyNjcxMzM5NjMz", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteGpuDelegate", "name": "TFLiteGpuDelegate", "color": "F71F04", "default": false, "description": "TFLite Gpu delegate issue" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @OleksandrGrument, in order to keep TFLite light, we have to compromise on some design choices such as dynamic input shapes. Currently we do not plan to implement this feature in the near future. Apologies we can't help you here soon." ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 1097546578, "node_id": "MDU6TGFiZWwxMDk3NTQ2NTc4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:keras", "name": "comp:keras", "color": "0052cc", "default": false, "description": "Keras related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Hi @Yuanjimengmengda ,\r\n\r\nAs per documentation of [model.predict](https://www.tensorflow.org/api_docs/python/tf/keras/Model#predict) the argument 'x' can be:\r\n\r\n`A generator or [keras.utils.Sequence](https://www.tensorflow.org/api_docs/python/tf/keras/utils/Sequence) returning (inputs, targets) or (inputs, targets, sample_weights)`\r\n\r\nThe generator should confirms the above behaviour. Can you please cross check whether the generator you are testing confirms the same.\r\n\r\nAlso unfortunately I am unable to find reproducible code snippet. Please submit the same for looking into the issue.\r\n\r\nThanks!", "> Hi @Yuanjimengmengda ,\r\n> \r\n> As per documentation of [model.predict](https://www.tensorflow.org/api_docs/python/tf/keras/Model#predict) the argument 'x' can be:\r\n> \r\n> `A generator or [keras.utils.Sequence](https://www.tensorflow.org/api_docs/python/tf/keras/utils/Sequence) returning (inputs, targets) or (inputs, targets, sample_weights)`\r\n> \r\n> The generator should confirms the above behaviour. Can you please cross check whether the generator you are testing confirms the same.\r\n> \r\n> Also unfortunately I am unable to find reproducible code snippet. Please submit the same for looking into the issue.\r\n> \r\n> Thanks!\r\n\r\nHi @SuryanarayanaY,\r\nThx for your reply. I updated the test codes with a sample triggers the error.\r\n\r\nI can confirm the generator is legitmate as I have been using it for years on tf==2.4.0. Moreover, on tf==2.12.0, the generator works well in `model.fit` if a `y` is specified(I'm adding this to the sample codes so you could make sure the generator is legitmate).\r\nI also tried the same code on version tf==2.11.1, no exceptions occurred.\r\nTherefore, I think the problem lies in the changes between 2.11.1 and 2.12.0.\r\n\r\nThanks.", "Hi @Yuanjimengmengda ,\r\n\r\nIt seems with the updated code I can able to replicate the error. The code fails with TF2.12v and success with TF2.11 version.\r\nAttached [gist](https://colab.research.google.com/gist/SuryanarayanaY/24bbd76ff2cad6125edd946f686b23d2/60813-tf2-12v-and-2-11v.ipynb) for reference.\r\n\r\nWe need to dig more further for commenting on this issue. Thanks!\r\n", "The same problem exists in [tf-nightly ](https://colab.research.google.com/gist/SuryanarayanaY/1acf0dfe05e502f86aa8d0c8485c2346/60813-nightly-2-14.ipynb#scrollTo=G6P4bilN4OK2) also." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1097547147, "node_id": "MDU6TGFiZWwxMDk3NTQ3MTQ3", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:ops", "name": "comp:ops", "color": "0052cc", "default": false, "description": "OPs related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "tilakrayal", "id": 81610181, "node_id": "MDQ6VXNlcjgxNjEwMTgx", "avatar_url": "https://avatars.githubusercontent.com/u/81610181?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tilakrayal", "html_url": "https://github.com/tilakrayal", "followers_url": "https://api.github.com/users/tilakrayal/followers", "following_url": "https://api.github.com/users/tilakrayal/following{/other_user}", "gists_url": "https://api.github.com/users/tilakrayal/gists{/gist_id}", "starred_url": "https://api.github.com/users/tilakrayal/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tilakrayal/subscriptions", "organizations_url": "https://api.github.com/users/tilakrayal/orgs", "repos_url": "https://api.github.com/users/tilakrayal/repos", "events_url": "https://api.github.com/users/tilakrayal/events{/privacy}", "received_events_url": "https://api.github.com/users/tilakrayal/received_events", "type": "User", "site_admin": false } ]

[ "@citizenkeynes,\r\nBasically, RaggedTensorVariant objects should never be copied to GPU, because we can't do anything useful with them there. But Placer isn't currently smart enough to figure that out (it just sees a Variant tensor, and doesn't know what kind of value it contains). \r\nhttps://github.com/keras-team/tf-keras/issues/128\r\nhttps://github.com/keras-team/tf-keras/issues/638\r\nThank you!", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60812\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60812\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 1097547538, "node_id": "MDU6TGFiZWwxMDk3NTQ3NTM4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:gpu", "name": "comp:gpu", "color": "0052cc", "default": false, "description": "GPU related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "reedwm", "id": 6510203, "node_id": "MDQ6VXNlcjY1MTAyMDM=", "avatar_url": "https://avatars.githubusercontent.com/u/6510203?v=4", "gravatar_id": "", "url": "https://api.github.com/users/reedwm", "html_url": "https://github.com/reedwm", "followers_url": "https://api.github.com/users/reedwm/followers", "following_url": "https://api.github.com/users/reedwm/following{/other_user}", "gists_url": "https://api.github.com/users/reedwm/gists{/gist_id}", "starred_url": "https://api.github.com/users/reedwm/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/reedwm/subscriptions", "organizations_url": "https://api.github.com/users/reedwm/orgs", "repos_url": "https://api.github.com/users/reedwm/repos", "events_url": "https://api.github.com/users/reedwm/events{/privacy}", "received_events_url": "https://api.github.com/users/reedwm/received_events", "type": "User", "site_admin": false }, { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "@michaeldiener,\r\nThe **memory_limit** in **tf.config.LogicalDeviceConfiguration** API mainly funtions for Maximum memory (in MB) to allocate on the virtual device. Currently it is only supported for GPUs. \r\nCould you please confirm whether it is having the problem whenever you are running inside a docker container and also provide if there is any error log? Thank you!", "@tilakrayal \r\nThanks for responding!\r\nYes, this is for a GPU and Tensorflow is running inside a docker container. \r\n\r\nThis is how I call docker:\r\n`docker run --gpus all -it --rm -v ~/tmp2:/exttmp -v ~/jupyter:/extjupyter --shm-size=1g --ulimit memlock=-1 -p 8889:8889 nvcr.io/nvidia/tensorflow:23.05-tf2-py3`\r\n\r\nThis is the output from setting the memory_limit. No log after setting it:\r\n```\r\nPython 3.10.6 (main, Mar 10 2023, 10:55:28) [GCC 11.3.0] on linux\r\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\r\n>>> import tensorflow as tf\r\n2023-06-12 07:33:41.210045: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.\r\n2023-06-12 07:33:41.232097: I tensorflow/core/platform/cpu_feature_guard.cc:183] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\r\nTo enable the following instructions: SSE3 SSE4.1 SSE4.2 AVX, in other operations, rebuild TensorFlow with the appropriate compiler flags.\r\n>>> gpus = tf.config.experimental.list_physical_devices('GPU')\r\n2023-06-12 07:34:12.541732: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:1013] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\r\n2023-06-12 07:34:12.545833: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:1013] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\r\n2023-06-12 07:34:12.545905: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:1013] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\r\n>>> tf.config.set_logical_device_configuration(\r\n... gpus[0],\r\n... [tf.config.LogicalDeviceConfiguration(memory_limit=4096)]\r\n... )\r\n>>> \r\n```", "@michaeldiener,\r\nCurrently we have --per_process_gpu_memory_fraction to limit the memory usage of model server and there is no such method available to limit GPU usage on the model level.\r\n\r\n\r\nExample code to run model server image with memory limit enabled:\r\n\r\n`docker run --runtime=nvidia -p 8501:8501 \\ --mount type=bind,\\ source=/tmp/tfserving/serving/tensorflow_serving/servables/tensorflow/testdata/saved_model_half_plus_two_gpu,\\ target=/models/half_plus_two \\ -e MODEL_NAME=half_plus_two -t tensorflow/serving:latest-gpu --per_process_gpu_memory_fraction=0.5\r\n`\r\n\r\nhttps://github.com/keras-team/tf-keras/issues/125\r\nThank you!\r\n", "Thanks, didn't know about the model server. This might come handy eventually, but I need to be able to limit GPU memory while training in Jupyter inside docker. There is no way to do that? Again, setting set_logical_device_configuration does not have any effect.", "I cannot reproduce on a Titan V. I started the docker container with `docker run --privileged --gpus all -it --rm --shm-size=1g --ulimit memlock=-1 -p 8889:8889 nvcr.io/nvidia/tensorflow:23.05-tf2-py3`.\r\n\r\nThen I ran the following program in the container to set the memory limit, initialize the GPU and print the memory usage with `nvidia-smi`\r\n\r\n```python\r\nimport tensorflow as tf\r\nimport subprocess\r\n\r\ngpus = tf.config.experimental.list_physical_devices('GPU')\r\ntf.config.set_logical_device_configuration(\r\n gpus[0],\r\n [tf.config.LogicalDeviceConfiguration(memory_limit=4096)]\r\n)\r\n\r\ntf.constant(0.) # Run GPU initialization\r\nsubprocess.run(['nvidia-smi'])\r\n```\r\n\r\nnvidia-smi showed only a bit more than 4096 MiB of memory being used.\r\n\r\n@michaeldiener are there any extra steps needed to reproduce the issue for you? I doubt the fact I have a different GPU is causing `set_logical_device_configuration` to work correctly only for me.", "@reedwm Strangely enough, this works now! \r\nNot sure if there was an update in the meantime for any of the involved components, but either way, problem solved. Thanks for looking into this, I appreciate it.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60811\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60811\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 1097547538, "node_id": "MDU6TGFiZWwxMDk3NTQ3NTM4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:gpu", "name": "comp:gpu", "color": "0052cc", "default": false, "description": "GPU related issues" }, { "id": 4829271983, "node_id": "LA_kwDOArmXAs8AAAABH9jXrw", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.11", "name": "TF 2.11", "color": "46B4D7", "default": false, "description": "Issues related to TF 2.11" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Hi @mmseerrttt ,\r\n\r\nIt seems you are using TF1.x related code which is not supported now. I have gone through the [horovod](https://horovod.readthedocs.io/en/stable/tensorflow.html) documentation and found below notes.\r\n\r\n<img width=\"821\" alt=\"Screenshot 2023-06-08 at 9 50 23 AM\" src=\"https://github.com/tensorflow/tensorflow/assets/116063290/e0dc1706-ee31-4638-85f6-faa55ff57373\">\r\n\r\nAlso [Estimator](https://www.tensorflow.org/guide/estimator) API is deprecated now.Please refer below note for same.\r\n\r\nWarning: Estimators are not recommended for new code. Estimators run [v1.Session](https://www.tensorflow.org/api_docs/python/tf/compat/v1/Session)-style code which is more difficult to write correctly, and can behave unexpectedly, especially when combined with TF 2 code. Estimators do fall under our [compatibility guarantees](https://tensorflow.org/guide/versions), but will receive no fixes other than security vulnerabilities. See the [migration guide](https://tensorflow.org/guide/migrate) for details.\r\n\r\n\r\n", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60810\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60810\">No</a>\n" ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" } ]

[ { "login": "haozha111", "id": 6316921, "node_id": "MDQ6VXNlcjYzMTY5MjE=", "avatar_url": "https://avatars.githubusercontent.com/u/6316921?v=4", "gravatar_id": "", "url": "https://api.github.com/users/haozha111", "html_url": "https://github.com/haozha111", "followers_url": "https://api.github.com/users/haozha111/followers", "following_url": "https://api.github.com/users/haozha111/following{/other_user}", "gists_url": "https://api.github.com/users/haozha111/gists{/gist_id}", "starred_url": "https://api.github.com/users/haozha111/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/haozha111/subscriptions", "organizations_url": "https://api.github.com/users/haozha111/orgs", "repos_url": "https://api.github.com/users/haozha111/repos", "events_url": "https://api.github.com/users/haozha111/events{/privacy}", "received_events_url": "https://api.github.com/users/haozha111/received_events", "type": "User", "site_admin": false }, { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @Jerry-Ge, \r\n\r\nI tried both the commands but I'm getting different errors... did you upload the right files?\r\n\r\n```\r\n../tensorflow/bazel-bin/tensorflow/compiler/mlir/tf-mlir-translate --graphdef-to-mlir --tf-enable-shape-inference-on-import --tf-output-arrays=result erf_model.pb -o test_tf.preopt.mlir --tf-input-arrays=placeholder_0 --tf-input-shapes=1\r\n2023-06-08 23:20:41.144947: E tensorflow/compiler/mlir/tensorflow/utils/import_utils.cc:48] Error parsing Protobuf\r\n2023-06-08 23:20:41.145026: E tensorflow/compiler/mlir/tensorflow/translate/tf_mlir_translate.cc:129] Graph import failed: INVALID_ARGUMENT: Could not parse input proto\r\n```\r\n\r\n```\r\n../tensorflow/bazel-bin/tensorflow/compiler/mlir/lite/flatbuffer_translate --tflite-flatbuffer-to-mlir erf_model.tflite --output-arrays=PartitionedCall:0 -o test_tflite.preopt.mlir\r\nERROR: The model is not a valid Flatbuffer buffer\r\nerf_model.tflite:0:0: error: couldn't parse flatbuffer\r\n```\r\n\r\nalternatively, did I translate your commands incorrectly to my environment?", "thanks @pkgoogle for helping on this. I cloned the model repo again and still couldn't get those Protobuff/Flatbuffer errors. Also double-checked your command and that looks fine. \r\n\r\n### A few options: \r\n- Maybe something wrong when you're downloading the model files which is causing some corruptions\r\n- It's a very dummy model and it you can easily generate those by the following definition\r\n```\r\n @tf.function(input_signature=[tf.TensorSpec(shape=[1, ], dtype=tf.float32)])\r\n def erf(self, x):\r\n return tf.math.erf(x)\r\n```\r\n\r\n### Some more progress/issues for TF\r\nIn the tf model generated, there seems missing a result there and I could get the following errors: \r\n```\r\nINVALID_ARGUMENT: Output result was not found in graph\r\n```\r\nBy running this command: \r\n```\r\ntensorflow/bazel-bin/tensorflow/compiler/mlir/tf-mlir-translate --graphdef-to-mlir --tf-enable-shape-inference-on-import --tf-output-arrays result erf_model.pb -o test_tf.preopt.mlir --tf-input-arrays placeholder_0 --tf-input-shapes=1\r\n```\r\nIf we remove `--tf-output-arrays result` flag, this gonna work and I can see the tf.Erf in the generated ir\r\n\r\n### For TFL\r\nthe same error as described on the top. ", "@Jerry-Ge, What branch are you working out of? Master? Nightly? A release branch?", "> @Jerry-Ge, What branch are you working out of? Master? Nightly? A release branch?\r\n\r\nit's an internal mirror branch from master. The lash commit is: 45f08a0fcc90145b9c2b7057310762d6b0ebae85", "So I went to the exact same commit and recompiled those binaries\r\n\r\nI was able to get the tflite command to work:\r\n```\r\n../tensorflow/bazel-bin/tensorflow/compiler/mlir/lite/flatbuffer_translate --tflite-flatbuffer-to-mlir erf.tflite --output-arrays=PartitionedCall:0 -o test_tflite.preopt.mlir\r\n```\r\ntest_tflite.preopt.mlir:\r\n```\r\nmodule attributes {tf_saved_model.semantics, tfl.description = \"MLIR Converted.\", tfl.schema_version = 3 : i32} {\r\n func.func @main(%arg0: tensor<1xf32> {tf_saved_model.index_path = [\"x\"]}) -> (tensor<1xf32> {tf_saved_model.index_path = [\"output_0\"]}) attributes {tf.entry_function = {inputs = \"serving_defa\r\nult_x:0\", outputs = \"PartitionedCall:0\"}, tf_saved_model.exported_names = [\"serving_default\"]} {\r\n %0 = \"tfl.custom\"(%arg0) {custom_code = \"FlexErf\", custom_option = #tfl<const_bytes : \"0x03457266001212034572661A002A070A0154120230013200000219151414042801\">} : (tensor<1xf32>) -> tensor<1x\r\nf32>\r\n return %0 : tensor<1xf32>\r\n }\r\n}\r\n```\r\n\r\nI'm still getting the previous error with the \"regular\" TF version:\r\n```\r\n../tensorflow/bazel-bin/tensorflow/compiler/mlir/tf-mlir-translate --graphdef-to-mlir --tf-enable-shape-inference-on-import --tf-output-arrays=result saved_model.pb -o test_tf.preopt.mlir --tf-input-arrays=placeholder_0 --tf-input-shapes=1\r\n2023-06-09 21:00:13.685096: E tensorflow/compiler/mlir/tensorflow/utils/import_utils.cc:48] Error parsing Protobuf\r\n2023-06-09 21:00:13.685157: E tensorflow/compiler/mlir/tensorflow/translate/tf_mlir_translate.cc:129] Graph import failed: INVALID_ARGUMENT: Could not parse input proto\r\n```\r\n\r\nMy exact code to produce the .pb and .tflite files:\r\ncreate_model.py:\r\n```\r\nimport tensorflow as tf\r\n\r\n\r\nclass Erfer(tf.Module):\r\n @tf.function(input_signature=[tf.TensorSpec(shape=[1, ], dtype=tf.float32)])\r\n def erf(self, x):\r\n return tf.math.erf(x)\r\n\r\n\r\nmodel = Erfer()\r\ntf.saved_model.save(model, 'saved_model')\r\n```\r\nand\r\nconvert_model.py:\r\n```\r\nimport tensorflow as tf\r\n\r\n# Path to the saved model directory\r\nsaved_model_dir = 'saved_model'\r\n\r\n# Convert the saved model to TFLite format\r\nconverter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)\r\nconverter.target_spec.supported_ops = [\r\n tf.lite.OpsSet.TFLITE_BUILTINS, # enable TensorFlow Lite ops.\r\n tf.lite.OpsSet.SELECT_TF_OPS # enable TensorFlow ops.\r\n]\r\ntflite_model = converter.convert()\r\n\r\n# Save the TFLite model to a file\r\ntflite_model_file = 'erf.tflite'\r\nwith open(tflite_model_file, 'wb') as f:\r\n f.write(tflite_model)\r\n```\r\n\r\nSo I believe the flatbuffer to MLIR is expected given that this is a FlexOp i.e. it is not natively built for TFLite (It essentially just makes it work with the original TF implementation)\r\n\r\nIs this ticket a feature request to make ERF a \"BUILTIN\" TFLite Op or to just make ERF \"work\"? i.e. do you run into any errors when attempting to use the .tflite model? Let me you know if you need any clarity on what I'm asking.", "# TF \r\nIt's quite weird for TF. my code looks very similar except I'm using something like this\r\n```\r\nmodel = Model()\r\nconcrete_function = model.erf.get_concrete_function()\r\ntf.io.write_graph(concrete_function.graph, \".\", \"model_file_name\", True)\r\n```\r\n# TFL \r\nFor TFLite, I'm requesting to make ERF a \"BUILTIN\" Op if I understood that correctly. i.e., I want to see the tfl.erf in the mlir file which currently it's a customOp\r\n", "Hi @haozha111, would you know if/when ERF will make it into TFLite as a BUILTIN op? Thanks!", "Hi all, bring this up again and see if there're any updates. ", "Hi all, bring this up again and see if there're any updates.\r\n\r\n", "Hi all, bring this up again and see if there're any updates.\r\n\r\n", "@pkgoogle seems no responses for a long time. ", "Fixed it. " ]

[ { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" }, { "id": 1178505529, "node_id": "MDU6TGFiZWwxMTc4NTA1NTI5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/prtype:bugfix", "name": "prtype:bugfix", "color": "159b2e", "default": false, "description": "PR to fix a bug" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Hi @jpienaar / @rdzhabarov Can you please review this PR ? Thank you!", "Same caveat as the other one, @changm may be able to help coordinate ", "I think this may be a weird case where it's ok to update the generated file as I can't find how else it's done. Still looping in @rohan100jain for API changes.", "Actually I think you'd have to update [math_ops.cc](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/ops/math_ops.cc#L525) and run the instructions at the top of tf_generated.td to do the right thing.", "Hi @sachinprasadhs Can you please take a look on the above comment from @changm. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "Hi @sachinprasadhs Any update on this PR? Please. Thank you!", "This PR is stale because it has been open for 14 days with no activity. It will be closed if no further activity occurs. Thank you.", "This PR was closed because it has been inactive for 14 days since being marked as stale. Please reopen if you'd like to work on this further." ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1133285679, "node_id": "MDU6TGFiZWwxMTMzMjg1Njc5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:xla", "name": "comp:xla", "color": "0052cc", "default": false, "description": "XLA" }, { "id": 1173072136, "node_id": "MDU6TGFiZWwxMTczMDcyMTM2", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XL", "name": "size:XL", "color": "adafea", "default": false, "description": "CL Change Size:Extra Large" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "CC @reedwm, @nluehr.", "Can you resolve conflicts?" ]

[]

[]

[]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 2691123225, "node_id": "MDU6TGFiZWwyNjkxMTIzMjI1", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:tf.function", "name": "comp:tf.function", "color": "0052cc", "default": false, "description": "tf.function related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "@sachinprasadhs,\r\nI was able to reproduce the issue on tensorflow v2.12 and tf-nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/f4758eb0ec3f8d66c124cb67f0988566/untitled1193.ipynb).", "I found something in the docs that explained this:\r\n\r\n\r\n\r\n\r\n\r\nhttps://www.tensorflow.org/guide/function\r\n\r\nhttps://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/autograph/g3doc/reference/control_flow.md#effects-of-the-tracing-process\r\n\r\nIn a @tf.function, the code executes each branch of the conditional block whether the condition is true or not, which gives you this error.", "@0Waffle I appreciate the comment, but I am not sure your connclusion is accurate. As I initially posted, the issue does not happen when the shape of the input spec is left completely undefined (with `@tf.function(input_signature=[tf.TensorSpec(None, tf.int32)])`), which suggests the operation is not really being executed. I think (although I'm not completely sure) that the error may come from some shape-checking logic that is executed once the inputs are known (but which does not kick in with fully undefined shapes for some reason).\r\n\r\nBesides that, it would be inefficient to execute both branches of the conditional in all cases. It would also be rather limiting, it was easy to work around this case, but it may not be so in other cases. It was also a pain to find the cause, as the problematic operation was much deeper in the call stack in my case.\r\n\r\nAs an example, consider this other function which works as expected in all cases, even when the inactive conditional branch cannot be executed.\r\n\r\n```python\r\nimport tensorflow as tf\r\n\r\[email protected](input_signature=[tf.TensorSpec([None], tf.int32), tf.TensorSpec([None], tf.int32)])\r\ndef f(x, y):\r\n return tf.cond(tf.size(y) > 0, lambda: x[y[0]], lambda: x[0])\r\n\r\n# Works\r\ntf.print(f(tf.constant([1, 2, 3]), tf.constant([1])))\r\n# 2\r\n\r\n# Works too, even though true branch of conditional cannot be executed with these inputs\r\ntf.print(f(tf.constant([1, 2, 3]), tf.constant([], tf.int32)))\r\n# 1\r\n```", "I'm sorry I was inaccurate. From debugging, in case None it executes lambda: x at the pywrap_tfe line, while in case [None] it executes lambda: tf.fill(...) at the pywrap_tfe line and I don't know where in the call stack it decides which one to execute since it should depend on tf.size(x)==1. It doesn't seem to be an autograph problem because it gives the error with autograph=False in the @tf.function parameters.\r\n\r\nIt gives the error with case None and jit_compile=True, so the behavior might have something to do with the way it compiles functions non-XLA." ]

[ { "id": 1169365494, "node_id": "MDU6TGFiZWwxMTY5MzY1NDk0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:M", "name": "size:M", "color": "adafea", "default": false, "description": "CL Change Size: Medium" }, { "id": 1593512946, "node_id": "MDU6TGFiZWwxNTkzNTEyOTQ2", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/invalid", "name": "invalid", "color": "db6f57", "default": true, "description": "Hacktoberfest spam PR" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Errors on the ARM CI seem to be:\r\n```\r\nERROR: /workspace/bazel_pip/tensorflow/dtensor/python/tests/BUILD:22:15: in py_library rule //bazel_pip/tensorflow/dtensor/python/tests:test_util: target '//tensorflow/dtensor/python:api' is not visible from target '//bazel_pip/tensorflow/dtensor/python/tests:test_util'. Check the visibility declaration of the former target if you think the dependency is legitimate\r\n(...)\r\nERROR: command succeeded, but not all targets were analyzed\r\n```\r\n\r\nAll tests built passed though:\r\n`Executed 1359 out of 1359 tests: 1359 tests pass.`", "Ping?" ]

[ { "id": 1169365494, "node_id": "MDU6TGFiZWwxMTY5MzY1NDk0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:M", "name": "size:M", "color": "adafea", "default": false, "description": "CL Change Size: Medium" }, { "id": 1593512946, "node_id": "MDU6TGFiZWwxNTkzNTEyOTQ2", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/invalid", "name": "invalid", "color": "db6f57", "default": true, "description": "Hacktoberfest spam PR" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[]

[ { "id": 1169365494, "node_id": "MDU6TGFiZWwxMTY5MzY1NDk0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:M", "name": "size:M", "color": "adafea", "default": false, "description": "CL Change Size: Medium" }, { "id": 1593512946, "node_id": "MDU6TGFiZWwxNTkzNTEyOTQ2", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/invalid", "name": "invalid", "color": "db6f57", "default": true, "description": "Hacktoberfest spam PR" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).\n\nView this [failed invocation](https://github.com/tensorflow/tensorflow/pull/60801/checks?check_run_id=14072016623) of the CLA check for more information.\n\nFor the most up to date status, view the checks section at the bottom of the pull request.", "Hi, thanks for taking the time to make a PR, but we can't start adding translations of files. Reviews & Maintenance would be impossible. " ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1205765054, "node_id": "MDU6TGFiZWwxMjA1NzY1MDU0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/subtype:macOS", "name": "subtype:macOS", "color": "b619ea", "default": false, "description": "macOS Build/Installation issues" }, { "id": 4032183365, "node_id": "LA_kwDOArmXAs7wVjxF", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.9", "name": "TF 2.9", "color": "1CF842", "default": false, "description": "Issues found in the TF 2.9 release (or RCs)" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Try to update TensorFlow version with \"pip install --upgrade tensorflow\"", "I did, and the problem still appears with tensorflow `2.12`.", "Hi @hovavalon ,\r\n\r\nIt seems the issue is Mac specific and Mac M1 related issues has to be addressed in the Apple metal plugin forum.\r\n\r\nBut I observed earlier there is some issues related to the optimizer. Can you please replace the optimizer with legacy optimizer as below and test it with TF2.9, TF2.11, TF2.12 and let us know the outcome.\r\n\r\n`opt = tf.keras.optimizers.legacy.SGD()`", "Hi @SuryanarayanaY, I am already using the legacy optimizer, at some point I got an error recommending that I do so.", "@hovavalon ,\r\n\r\nI tried to replicate the issue on Mac with tf-nightly(2.14.0-dev20230515) with `epochs=100` with the same test code you provided and its running fine.Please refer to logs below. \r\n\r\n```\r\n(tf-metal) suryanarayanay-macbookpro:~ suryanarayanay$ python Downloads/60800.py\r\n2.14.0-dev20230515\r\nMetal device set to: Apple M1 Pro\r\n\r\nsystemMemory: 16.00 GB\r\nmaxCacheSize: 5.33 GB\r\n\r\nEpoch 1/100\r\n10/10 [==============================] - 1s 3ms/step - loss: 7.0479\r\nEpoch 2/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 4.2836\r\nEpoch 3/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 2.6105\r\nEpoch 4/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 1.5956\r\nEpoch 5/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.9820\r\nEpoch 6/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.6105\r\nEpoch 7/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.3856\r\nEpoch 8/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.2490\r\nEpoch 9/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.1656\r\nEpoch 10/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.1148\r\nEpoch 11/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0843\r\nEpoch 12/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0653\r\nEpoch 13/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0537\r\nEpoch 14/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0463\r\nEpoch 15/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0416\r\nEpoch 16/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0385\r\nEpoch 17/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0364\r\nEpoch 18/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0348\r\nEpoch 19/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0337\r\nEpoch 20/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0328\r\nEpoch 21/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0320\r\nEpoch 22/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0313\r\nEpoch 23/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0306\r\nEpoch 24/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0300\r\nEpoch 25/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0295\r\nEpoch 26/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0290\r\nEpoch 27/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0285\r\nEpoch 28/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0280\r\nEpoch 29/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0275\r\nEpoch 30/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0271\r\nEpoch 31/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0266\r\nEpoch 32/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0261\r\nEpoch 33/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0257\r\nEpoch 34/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0253\r\nEpoch 35/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0249\r\nEpoch 36/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0245\r\nEpoch 37/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0241\r\nEpoch 38/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0237\r\nEpoch 39/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0234\r\nEpoch 40/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0230\r\nEpoch 41/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0226\r\nEpoch 42/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0223\r\nEpoch 43/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0220\r\nEpoch 44/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0216\r\nEpoch 45/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0213\r\nEpoch 46/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0211\r\nEpoch 47/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0208\r\nEpoch 48/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0205\r\nEpoch 49/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0202\r\nEpoch 50/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0199\r\nEpoch 51/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0197\r\nEpoch 52/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0194\r\nEpoch 53/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0191\r\nEpoch 54/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0189\r\nEpoch 55/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0187\r\nEpoch 56/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0184\r\nEpoch 57/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0182\r\nEpoch 58/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0180\r\nEpoch 59/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0178\r\nEpoch 60/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0176\r\nEpoch 61/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0173\r\nEpoch 62/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0172\r\nEpoch 63/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0170\r\nEpoch 64/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0168\r\nEpoch 65/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0166\r\nEpoch 66/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0164\r\nEpoch 67/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0163\r\nEpoch 68/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0161\r\nEpoch 69/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0159\r\nEpoch 70/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0158\r\nEpoch 71/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0156\r\nEpoch 72/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0155\r\nEpoch 73/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0153\r\nEpoch 74/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0152\r\nEpoch 75/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0150\r\nEpoch 76/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0149\r\nEpoch 77/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0148\r\nEpoch 78/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0147\r\nEpoch 79/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0145\r\nEpoch 80/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0144\r\nEpoch 81/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0143\r\nEpoch 82/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0142\r\nEpoch 83/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0140\r\nEpoch 84/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0139\r\nEpoch 85/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0138\r\nEpoch 86/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0137\r\nEpoch 87/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0136\r\nEpoch 88/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0136\r\nEpoch 89/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0134\r\nEpoch 90/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0134\r\nEpoch 91/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0132\r\nEpoch 92/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0132\r\nEpoch 93/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0131\r\nEpoch 94/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0130\r\nEpoch 95/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0129\r\nEpoch 96/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0128\r\nEpoch 97/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0128\r\nEpoch 98/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0127\r\nEpoch 99/100\r\n10/10 [==============================] - 0s 2ms/step - loss: 0.0126\r\nEpoch 100/100\r\n10/10 [==============================] - 0s 3ms/step - loss: 0.0125\r\n(tf-metal) suryanarayanay-macbookpro:~ suryanarayanay$ \r\n```\r\nCan you try with latest nightly version and please confirm the behaviour.\r\n\r\nAlso please confirm you have followed the Apple [metal](https://developer.apple.com/metal/tensorflow-plugin/) plugin instructions correctly.Please confirm you have installed tensorflow-metal and try retesting the code in a Fresh environment.\r\n\r\nThanks!\r\n\r\n", "I have followed the instructions given in the link when installing the application.\r\nEven on my computer 100 short epochs are not necessarily enough to replicate (even though at some times it is enough). ", "Hi @hovavalon ,\r\n\r\nThanks for confirmation. I tried with 1000 epohs and executed 3 times successfully without the reported behaviour.Please refer to attached logs below.\r\n\r\n[#60800_logs.txt](https://github.com/tensorflow/tensorflow/files/11742629/60800_logs.txt)\r\n\r\nCould you try with fresh environment and with tf-nightly version.\r\n\r\nThanks!\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60800\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60800\">No</a>\n" ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1097543484, "node_id": "MDU6TGFiZWwxMDk3NTQzNDg0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:runtime", "name": "comp:runtime", "color": "0052cc", "default": false, "description": "c++ runtime, performance issues (cpu)" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" }, { "id": 2671351731, "node_id": "MDU6TGFiZWwyNjcxMzUxNzMx", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ModelOptimizationToolkit", "name": "ModelOptimizationToolkit", "color": "BFD629", "default": false, "description": "TF Model Optimization Toolkit" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "arfaian", "id": 1106365, "node_id": "MDQ6VXNlcjExMDYzNjU=", "avatar_url": "https://avatars.githubusercontent.com/u/1106365?v=4", "gravatar_id": "", "url": "https://api.github.com/users/arfaian", "html_url": "https://github.com/arfaian", "followers_url": "https://api.github.com/users/arfaian/followers", "following_url": "https://api.github.com/users/arfaian/following{/other_user}", "gists_url": "https://api.github.com/users/arfaian/gists{/gist_id}", "starred_url": "https://api.github.com/users/arfaian/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/arfaian/subscriptions", "organizations_url": "https://api.github.com/users/arfaian/orgs", "repos_url": "https://api.github.com/users/arfaian/repos", "events_url": "https://api.github.com/users/arfaian/events{/privacy}", "received_events_url": "https://api.github.com/users/arfaian/received_events", "type": "User", "site_admin": false }, { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @danielr55 \r\n\r\nI was able to reproduce this issue. Please find the gist [here](https://colab.research.google.com/gist/pjpratik/56093b4a375c85812ab245637bbb5d74/60799.ipynb).\r\n\r\nSeems like int16x8 support is not available for `tf.minimum` built-in op which might be causing the issue.\r\n\r\nYou can try `tf.lite.OpsSet.TFLITE_BUILTINS_INT8` if it works for your case.\r\n\r\nThanks.", "Thanks for your response, @pjpratik.\r\n\r\nIn line with your comment, I've reworked the `CircularBufferLayer` to avoid using the `tf.minimum` operation.\r\n\r\nHowever, when I attempted to convert the `StreamingModel`, I encountered this error:\r\n\r\n`RuntimeError: Max and min for dynamic tensors should be recorded during calibration: Failed for tensor streaming_model_3/circular_buffer_layer_3/add\r\nEmpty min/max for tensor streaming_model_3/circular_buffer_layer_3/add`\r\n\r\nIt seems strange to me that even the `tf.add` built-in operation is not supported by the tflite converter.\r\n\r\nCould it be possible that I'm incorrectly implementing the `CircularBufferLayer`? I'd appreciate your insight.\r\n\r\nThanks!\r\n\r\n-------------------------------------\r\nHere's the revised implementation for the `CircularBufferLayer`:\r\n\r\n```python\r\nclass CircularBufferLayer(tf.keras.layers.Layer):\r\n def __init__(self, num_features, buffer_size, stride, **kwargs):\r\n super().__init__(**kwargs)\r\n self.num_features = num_features\r\n self.buffer_size = buffer_size\r\n self.stride = stride\r\n self.buffer = self.add_weight(name='buffer', shape=(1, buffer_size, self.num_features),\r\n initializer='zeros', trainable=False, dtype=tf.float32)\r\n self.call_count = self.add_weight(name='call_count', shape=(), initializer='zeros',\r\n dtype=tf.int32, trainable=False)\r\n # total count, this count will never reset\r\n self.total_call_count = self.add_weight(name='total_call_count', shape=(), initializer='zeros',\r\n dtype=tf.int32, trainable=False)\r\n\r\n def call(self, inputs, **kwargs):\r\n # inputs should be reshaped to (1, 1, num_features) to match the buffer shape\r\n inputs = tf.reshape(inputs, [1, 1, self.num_features])\r\n\r\n # Update the buffer with the new data\r\n self.buffer.assign(tf.concat([self.buffer[:, 1:], inputs], axis=1))\r\n\r\n # Update the call count \r\n self.call_count.assign(\r\n tf.cond(\r\n tf.greater_equal(self.call_count + 1, self.stride),\r\n true_fn=lambda: self.stride,\r\n false_fn=lambda: self.call_count + 1\r\n )\r\n )\r\n\r\n # Update the total call count\r\n self.total_call_count.assign(\r\n tf.cond(\r\n tf.greater_equal(self.total_call_count + 1, self.buffer_size),\r\n true_fn=lambda: self.buffer_size,\r\n false_fn=lambda: self.total_call_count + 1\r\n )\r\n )\r\n\r\n # If-else condition\r\n self.call_count.assign(\r\n tf.cond(tf.logical_and(tf.equal(self.call_count, self.stride), tf.greater_equal(self.total_call_count, self.buffer_size)),\r\n true_fn=lambda: 0,\r\n false_fn=lambda: self.call_count,\r\n )\r\n )\r\n\r\n # Create a boolean flag indicating if self.call_count is 0\r\n # and the total number of calls to this layer is at least self.buffer_size\r\n flag = tf.equal(self.call_count, 0)\r\n\r\n # Return the buffer data and the flag\r\n return [self.buffer, flag]\r\n\r\n def reset(self):\r\n self.buffer.assign(tf.zeros_like(self.buffer))\r\n self.call_count.assign(tf.zeros_like(self.call_count))\r\n self.total_call_count.assign(tf.zeros_like(self.total_call_count))\r\n\r\n def get_config(self):\r\n config = {\r\n 'buffer': self.buffer,\r\n 'total_call_count': self.total_call_count,\r\n 'call_count': self.call_count\r\n }\r\n base_config = super(CircularBufferLayer, self).get_config()\r\n return dict(list(base_config.items()) + list(config.items()))\r\n```", "Hi @danielr55, currently this looks like a bug with the new experimental feature. My current guess is that it is somehow not calibrating on the representative data set correctly for this type of conversion.\r\n\r\n@arfaian, can you please take a look?\r\n\r\nI am able to replicate on tf-nightly w/ the following script (model save path modified):\r\nAlso as a [gist](https://colab.sandbox.google.com/gist/pkgoogle/d68636dfe54697b9eb1866c48707245d/60799.ipynb)\r\n```python\r\nimport tensorflow as tf\r\nimport os\r\nimport numpy as np\r\nimport pandas as pd\r\nimport time\r\n\r\n\r\n# Set up the environment to use CPU\r\ntf.config.set_visible_devices([], 'GPU')\r\n\r\n\r\n#######################\r\n# CircularBuffer\r\n#######################\r\nclass CircularBufferLayer(tf.keras.layers.Layer):\r\n def __init__(self, num_features, buffer_size, stride, **kwargs):\r\n super().__init__(**kwargs)\r\n self.num_features = num_features\r\n self.buffer_size = buffer_size\r\n self.stride = stride\r\n self.buffer = self.add_weight(name='buffer', shape=(1, buffer_size, self.num_features),\r\n initializer='zeros', trainable=False, dtype=tf.float32)\r\n self.call_count = self.add_weight(name='call_count', shape=(), initializer='zeros',\r\n dtype=tf.int32, trainable=False)\r\n # total count, this count will never reset\r\n self.total_call_count = self.add_weight(name='total_call_count', shape=(), initializer='zeros',\r\n dtype=tf.int32, trainable=False)\r\n\r\n def call(self, inputs, **kwargs):\r\n # inputs should be reshaped to (1, 1, num_features) to match the buffer shape\r\n inputs = tf.reshape(inputs, [1, 1, self.num_features])\r\n\r\n # Update the buffer with the new data\r\n self.buffer.assign(tf.concat([self.buffer[:, 1:], inputs], axis=1))\r\n\r\n # Update the call count \r\n self.call_count.assign(\r\n tf.cond(\r\n tf.greater_equal(self.call_count + 1, self.stride),\r\n true_fn=lambda: self.stride,\r\n false_fn=lambda: self.call_count + 1\r\n )\r\n )\r\n\r\n # Update the total call count\r\n self.total_call_count.assign(\r\n tf.cond(\r\n tf.greater_equal(self.total_call_count + 1, self.buffer_size),\r\n true_fn=lambda: self.buffer_size,\r\n false_fn=lambda: self.total_call_count + 1\r\n )\r\n )\r\n\r\n # If-else condition\r\n self.call_count.assign(\r\n tf.cond(tf.logical_and(tf.equal(self.call_count, self.stride), tf.greater_equal(self.total_call_count, self.buffer_size)),\r\n true_fn=lambda: 0,\r\n false_fn=lambda: self.call_count,\r\n )\r\n )\r\n\r\n # Create a boolean flag indicating if self.call_count is 0\r\n # and the total number of calls to this layer is at least self.buffer_size\r\n flag = tf.equal(self.call_count, 0)\r\n\r\n # Return the buffer data and the flag\r\n return [self.buffer, flag]\r\n\r\n def reset(self):\r\n self.buffer.assign(tf.zeros_like(self.buffer))\r\n self.call_count.assign(tf.zeros_like(self.call_count))\r\n self.total_call_count.assign(tf.zeros_like(self.total_call_count))\r\n\r\n def get_config(self):\r\n config = {\r\n 'buffer': self.buffer,\r\n 'total_call_count': self.total_call_count,\r\n 'call_count': self.call_count\r\n }\r\n base_config = super(CircularBufferLayer, self).get_config()\r\n return dict(list(base_config.items()) + list(config.items()))\r\n\r\n\r\n#######################\r\n# Streaming Model\r\n#######################\r\nclass StreamingModel(tf.keras.Model):\r\n def __init__(self, input_channel, output_channel, kernel_size, stride, **kwargs):\r\n super().__init__(**kwargs)\r\n\r\n self.input_channel = input_channel\r\n self.output_channel = output_channel\r\n self.kernel_size = kernel_size\r\n self.stride = stride\r\n\r\n # set acoustic model buffer\r\n self.buffer = CircularBufferLayer(\r\n num_features=input_channel,\r\n buffer_size=kernel_size,\r\n stride=stride\r\n )\r\n\r\n # set filters\r\n self.conv1d = tf.keras.layers.Conv1D(\r\n output_channel,\r\n kernel_size=kernel_size,\r\n strides=1,\r\n use_bias=False,\r\n padding='valid',\r\n data_format='channels_last'\r\n )\r\n\r\n def call(self, inputs, **kwargs):\r\n # buffer:\r\n [x, flag] = self.buffer(inputs) # output shape = [1, kernel_size, input_channel]\r\n\r\n x = tf.cond(flag, # output shape = [1, 1, output_channel]\r\n true_fn=lambda: self.conv1d(x),\r\n false_fn=lambda: tf.zeros([1, 1, self.output_channel])\r\n )\r\n\r\n x = tf.reshape(x, [1, self.output_channel]) # output shape = [1, output_channel]\r\n\r\n return [x, flag]\r\n\r\n def reset(self):\r\n self.buffer.reset()\r\n\r\nif __name__ == '__main__':\r\n model = StreamingModel(\r\n input_channel=32,\r\n output_channel=64,\r\n kernel_size=5,\r\n stride=2\r\n )\r\n\r\n # create some dummy data with the correct shapes\r\n seq_len = 50\r\n input_data = tf.random.normal([1, seq_len, model.input_channel])\r\n\r\n # call the model on the dummy data - crucial to build all sub-graphs\r\n for t in range(seq_len):\r\n output = model(input_data[:, t])\r\n\r\n # reset the buffer after executing\r\n model.reset()\r\n\r\n # print the model's summary weights\r\n model.summary()\r\n\r\n #####################################\r\n # TFLite Conversion\r\n #####################################\r\n tflite_path = 'here'\r\n saved_model_dir = os.path.expanduser(tflite_path)\r\n tf.saved_model.save(obj=model, export_dir=saved_model_dir)\r\n converter = tf.lite.TFLiteConverter.from_saved_model(saved_model_dir)\r\n converter.experimental_enable_resource_variables = True\r\n\r\n # Quantize the model to 16x8\r\n def representative_data_gen():\r\n input_channel = 32\r\n seq_len = 50\r\n for t in range(seq_len):\r\n x = tf.random.normal([1, input_channel])\r\n yield [x]\r\n\r\n converter.inference_input_type = tf.float32\r\n converter.inference_output_type = tf.float32\r\n converter.optimizations = [tf.lite.Optimize.DEFAULT]\r\n converter.target_spec.supported_ops = [tf.lite.OpsSet.EXPERIMENTAL_TFLITE_BUILTINS_ACTIVATIONS_INT16_WEIGHTS_INT8]\r\n converter.representative_dataset = representative_data_gen\r\n converter.allow_custom_ops = True\r\n tflite_model = converter.convert()\r\n\r\n # Save the TFLite model.\r\n with tf.io.gfile.GFile(tflite_path + '.tflite', 'wb') as f:\r\n f.write(tflite_model)\r\n```" ]

[ { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 1463677878, "node_id": "MDU6TGFiZWwxNDYzNjc3ODc4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:performance", "name": "type:performance", "color": "159b2e", "default": false, "description": "Performance Issue" }, { "id": 4032183365, "node_id": "LA_kwDOArmXAs7wVjxF", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.9", "name": "TF 2.9", "color": "1CF842", "default": false, "description": "Issues found in the TF 2.9 release (or RCs)" } ]

[ { "login": "tilakrayal", "id": 81610181, "node_id": "MDQ6VXNlcjgxNjEwMTgx", "avatar_url": "https://avatars.githubusercontent.com/u/81610181?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tilakrayal", "html_url": "https://github.com/tilakrayal", "followers_url": "https://api.github.com/users/tilakrayal/followers", "following_url": "https://api.github.com/users/tilakrayal/following{/other_user}", "gists_url": "https://api.github.com/users/tilakrayal/gists{/gist_id}", "starred_url": "https://api.github.com/users/tilakrayal/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tilakrayal/subscriptions", "organizations_url": "https://api.github.com/users/tilakrayal/orgs", "repos_url": "https://api.github.com/users/tilakrayal/repos", "events_url": "https://api.github.com/users/tilakrayal/events{/privacy}", "received_events_url": "https://api.github.com/users/tilakrayal/received_events", "type": "User", "site_admin": false } ]

[ "can you show the Error?\r\n", "Hi @fizmatbfu, please follow the steps mentioned in https://www.tensorflow.org/install/source_windows.\r\nI ran the build on my end and didn't get the above error.", "@fizmatbfu,\r\nCould you please follow the steps mentioned on the official website of [Tensorflow](https://www.tensorflow.org/install/source_windows) to build Tensorflow from source. Tensorflow suggests install all the pre-requisite packages to avoid errors. Thank you!", "Ok, I fixed the errors as follows: instead of using the --copt flag, I added the following lines to the .bazelrc file:\r\n```\r\nbuild: windows --cxxopt=/MT\r\nbuild: windows --host_copt=/MT\r\n```\r\nI also replaced MD with MT in all .bzl and .bzl.tpl files. But the /MD flag still appears in the generated .params files. How can I build a library with the /MT flag?\r\n\r\n(Build is successful, but DLL still requires vcruntime libraries).", "I rechecked. The /MT flag I added appears in the .params files after the /MD flag. The compiler issues a message that the /MT flag overrides /MD (Command line warning D9025 : overriding '/MD' with '/MT'). But still the DLL requires the vcruntime library to run.", "In order for the DLL to not require the vcruntime library to run, you also need to remove the `/DEFAULTLIB:msvcrt.lib` line from the tensorflow.dll-2.params file before linking.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60798\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60798\">No</a>\n" ]

[ { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Here are the internal errors, @radekzc can you please verify ? Thank you!\r\n\r\nld: error: undefined reference due to --no-allow-shlib-undefined: util_graphics_giza::BlendRgbaToRgbaIcon::DrawScanlines(int, int, int, int, int, util_graphics_giza::GizaOperator, Callback5<int, int, int, unsigned int, util_graphics_giza::GizaOperator>*) const\r\n>>> referenced by blaze-out/k8-opt/bin/_solib_k8/libutil_Sgraphics_Sgiza_Sinternal_Slibgiza.ifso\r\n\r\nld: error: undefined reference due to --no-allow-shlib-undefined: util_graphics_giza::BlendRgbaToRgbaIcon::DrawIntoIndexcolorBuffer(int, int, int, int, int, int, unsigned char const*, unsigned char*) const\r\n>>> referenced by blaze-out/k8-opt/bin/_solib_k8/libutil_Sgraphics_Sgiza_Sinternal_Slibgiza.ifso\r\n\r\nld: error: undefined reference due to --no-allow-shlib-undefined: util_graphics_giza::BlendRgbaToRgbaIcon::DrawIntoRgbaBuffer(int, int, int, int, int, int, util_graphics_giza::GizaOperator, unsigned char*) const\r\n>>> referenced by blaze-out/k8-opt/bin/_solib_k8/libutil_Sgraphics_Sgiza_Sinternal_Slibgiza.ifso\r\nclang: error: linker command failed with exit code 1 (use -v to see invocation)", "Seems not relevant to the commit.", "Hi @radekzc,\r\n\r\nCould you provide the steps to reproduce the issue (breakages under Windows build)?", "Will try to prepare minimal example, but it may take time. \r\n\r\nIf you are unsure what this PR fixes, its like if you wrote code like\r\n\r\n```\r\nint * a = new int;\r\n...\r\nfree(a); // free instead of delete\r\n```\r\n\r\nwhich would \"work\" if the compiler used\r\n\r\n```\r\nvoid free(void *ptr)\r\n{\r\n delete(ptr);\r\n}\r\n```\r\n\r\nwhich is often the case, but there is zero guarantee it works always, because the behavior is undefined.\r\n\r\n\r\n\r\n\r\n", "Hi @radekzc Any update on this PR? Please. Thank you!", "Hi @radekzc,\r\n\r\nAccording to Google's policy, we are deprecating the usage of absl::any_cast and absl::variant in our codebase. To avoid the scenario you mentioned, Is it ok to fix all uses of absl::variant/any_cast to std::variant/any_cast?", "Yes, perfectly ok.", "Hi @radekzc Any update on the above [comments](https://github.com/tensorflow/tensorflow/pull/60797#issuecomment-1690990772)? Please. Thank you!", "Hi @radekzc Any update on the above https://github.com/tensorflow/tensorflow/pull/60797#issuecomment-1709240974? Please. Thank you!", "Hi @radekzc Any update on the above https://github.com/tensorflow/tensorflow/pull/60797#issuecomment-1709240974? Please. Thank you!", "Hi @radekzc Any update on the above https://github.com/tensorflow/tensorflow/pull/60797#issuecomment-1709240974? Please. Thank you!", "Hi @radekzc Any update on the above https://github.com/tensorflow/tensorflow/pull/60797#issuecomment-1709240974? Please. Thank you!", "Hello\r\nPlease also delete \r\n#include <variant> \r\non top of header\r\n\"tensorflow/lite/delegates/gpu/common/operations.h\"\r\n\r\nIt's broking builds with old compilers (require c++17). \r\n", "> Hello Please also delete #include on top of header \"tensorflow/lite/delegates/gpu/common/operations.h\"\r\n> \r\n> It's broking builds with old compilers (require c++17).\r\n\r\nBased on previous comments the agreed upon way is to completely remove absl and switch to std instead, which is the direct opposite of this commit. I can look into this when our next backend upgrade task (switching to newer TF version in our product) begins, which is due in April. If the absl/std hell still exist, I will refactor it to std and sent a new pull request." ]

[ { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).\n\nView this [failed invocation](https://github.com/tensorflow/tensorflow/pull/60796/checks?check_run_id=14066704200) of the CLA check for more information.\n\nFor the most up to date status, view the checks section at the bottom of the pull request." ]

[ { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 2012480497, "node_id": "MDU6TGFiZWwyMDEyNDgwNDk3", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:docs-feature", "name": "type:docs-feature", "color": "159b2e", "default": false, "description": "Doc issues for new feature, or clarifications about functionality" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Hi @hirwa-nshuti ,\r\n\r\nThanks for your observation and time for bringing this to our attention. Yeah the Numpy version that the Tensorflow API is referring is older and the link can be updated to latest/stable version.\r\n\r\nWe will discuss internally and let you know because this is external link which may change with time. But for time being the provided link in TF documentation has note along another link pointed to latest stable release of numpy.", "I noticed that all numpy functions have same issue. But every function has note and link redirecting to latest release.\r\n", "@hirwa-nshuti ,\r\n\r\nAn internal PR has been done for the requested changes and will be updated soon. Thanks!" ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1173072136, "node_id": "MDU6TGFiZWwxMTczMDcyMTM2", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XL", "name": "size:XL", "color": "adafea", "default": false, "description": "CL Change Size:Extra Large" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).\n\nView this [failed invocation](https://github.com/tensorflow/tensorflow/pull/60794/checks?check_run_id=14063992701) of the CLA check for more information.\n\nFor the most up to date status, view the checks section at the bottom of the pull request.", "Hi @mgokulkrish Can you please sign CLA. Thank you!", "> Hi @mgokulkrish Can you please sign CLA. Thank you!\r\n\r\nHi @gbaned, I have already signed CLA in this PR, should i do it again ? #60749 ", "Hi @yijie-yang, could you please review this PR. Doing open-source contribution for the first time, would love to learn from the mistakes and improve.", "> > Hi @mgokulkrish Can you please sign CLA. Thank you!\r\n> \r\n> Hi @gbaned, I have already signed CLA in this PR, should i do it again ? #60749\r\n\r\nHi @mgokulkrish Sorry for the delay, It shows cla check failed. Can you please make sure to use same GitHub username and email-id associated with it. Thank you!", "Hi @gbaned, I think the cla now is approved.", "> Hi @gbaned, I think the cla now is approved.\r\n\r\nHi @mgokulkrish Yes, it is good now. Thank you!", "Come on, somebody needs check the this merging problem. \r\nBtw @mgokulkrish did u have interest to make a repo to show how to use GPU to acclerate sd model on Android? sounds very interest for me", "> Come on, somebody needs check the this merging problem. Btw @mgokulkrish did u have interest to make a repo to show how to use GPU to acclerate sd model on Android? sounds very interest for me\r\n\r\nHi @fuerpy, sorry I do not have open-source setup for it. The main issue for the main model executing via gpu delegate is 5 dimensional tensors due to group normalization, current opensource tflite gpu delegate does not support it. If you are able to combine the whole block on group norm into a single block through MLIR changes in TFLiteConvertor, you can use the kernel in this commit to execute it through custom layer features.", "@mgokulkrish hello，If I run tflite with the project https://github.com/mgokulkrish/tensorflow\r\n\r\ndoes it mean that this model can also use the GPU on Android\r\n\r\nhttps://huggingface.co/keras-sd/diffusion-model-tflite/tree/main", "> Hi @mgokulkrish, the recommended way of adding custom op (only supported in C++) is documented in https://www.tensorflow.org/lite/guide/ops_custom.\r\n> \r\n> This PR makes a lot of changes to the TFLite kernel, which we cannot approve. Please follow the above guide and if there's any issue with it, please create a GitHub issue to follow up.\r\n> \r\n> Thanks.\r\n\r\nThanks for the review @JunyoungLim , based on the guide of custom op you shared, I think CPU side changes for the kernel should be fine. \r\n\r\nMaybe you could point out which part of GPU changes are wrong or where it is too big for kernel changes, I could learn from it.\r\nSummary of my changes:-\r\n1. added a transformation - adding connections for two sub kernels to calculate mean and variance sum.\r\n2. Added three kernels (tasks) for group norm, group mean sum and group var sum.\r\n3. Added parser and selectors for it (as a custom layer, similar to MaxUnpooing2D which was done similarly before).\r\n\r\nI have not modified any existing kernels in my changes. I have not included any changes with respect to MLIR, tflite convertor or tf.experimental_implements ( https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/models/convert/operation_fusion.md ), that I used to merge group normalization to single block. The change just has additional kernels.\r\n\r\nThanks again for reviewing the change, I would like to learn and make contribution in GPU delegate. So, a deep review will be a good learning and review process for me.", "Hi @sirakiin Can you please review this PR ? Thank you!", "Hi @sirakiin Can you please review this PR ? Thank you!", "> Hi @sirakiin Can you please review this PR ? Thank you!\r\n\r\n@gbaned Hello, do you have the release aar of this pr? I tried to custom build the release, but it failed. I am trying to develop the Android version of stable diffusion. This is the Android version of stable diffusion I made using onnxruntime https://github.com/ZTMIDGO/Android-Stable-diffusion-ONNX . I hope to develop the Android version of TensorFlow.\r\n\r\n", "Hi @sirakiin Can you please assist on above comments from @ZTMIDGO. Thank you!", "Hi @JunyoungLim Can you please assist on above comments from @ZTMIDGO. Thank you!", "Hi @JunyoungLim Any update on this PR? Please. Thank you!", "Hi @JunyoungLim Any update on this PR? Please. Thank you!", "Hi @JunyoungLim Any update on this PR? Please. Thank you!", "Hi @JunyoungLim Any update on this PR? Please. Thank you!", "Hi @sirakiin Any update on this PR? Please. Thank you!", "Hi @sirakiin Any update on this PR? Please. Thank you!", "Hi @sirakiin Any update on this PR? Please. Thank you!" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473184161, "node_id": "MDU6TGFiZWw0NzMxODQxNjE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:support", "name": "type:support", "color": "159b2e", "default": false, "description": "Support issues" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "@Ramees025,\r\nCould you please provide complete code and more information about the issue which was mentioned. It helps to analyse the issue in an effective way. Thank you!", "\r\n\r\nAttached is the single layer model causing the issue when running in GPU.", "Model run fine in CPU. But while running with GPU, it showing the error \"RESHAPE: OP is supported, but tensor type/shape isn't compatible\" and fallback to CPU.\r\nBasically the reshape will convert 5D tensor to 4D tensor.", "@Ramees025,\r\nUnfortunately with the above details provided, we were not able to provide the correct information. Could you please provide the complete code or the colab gist which helps us to debug the issue in an effective way. Thank you!", "TfLiteIntArray* GetOpsToReplace(\r\n TfLiteContext* context, bool allow_quant_ops, int max_delegated_partitions,\r\n const absl::flat_hash_set<TfLiteBuiltinOperator>* excluded_ops) {\r\n delegates::IsNodeSupportedFn node_supported_fn =\r\n [=](TfLiteContext* context, TfLiteNode* node,\r\n TfLiteRegistration* registration,\r\n std::string* unsupported_details) -> bool {\r\n const auto status =\r\n IsSupported(context, node, registration, allow_quant_ops, excluded_ops);\r\n if (!status.ok()) {\r\n if (unsupported_details) {\r\n *unsupported_details = std::string(status.message());\r\n }\r\n return false;\r\n }\r\n std::vector<TfLiteType> allowed_in_types = {kTfLiteFloat32, kTfLiteFloat16};\r\n std::vector<TfLiteType> allowed_out_types = {kTfLiteFloat32,\r\n kTfLiteFloat16};\r\n if (!IsAllAllowedTensors(context, node->inputs, allowed_in_types) ||\r\n !IsAllAllowedTensors(context, node->outputs, allowed_out_types)) {\r\n if (unsupported_details) {\r\n *unsupported_details =\r\n **\"OP is supported, but tensor type/shape isn't compatible.\";**\r\n }\r\n return false;\r\n }\r\n return true;\r\n };\r\n}\r\n\r\n\r\ncode:\r\n\r\n\\tensorflow\\lite\\delegates\\gpu\\common\\model_builder.cc\r\n\r\n", "Hi @Ramees025 \r\n\r\nThe Reshape op might not be supporting the 5D tensors on GPU delegate which is why the error is thrown. \r\n\r\nThanks.", "Hi, is there any plan/roadmap to support 5D RESHAPE ops in tflite gpu delegate?", "Hi @Ramees025 \r\n\r\nThe the support for 5D might be on roadmap of Reshape.\r\n\r\n You can also use [Model Analyzer](https://www.tensorflow.org/lite/guide/model_analyzer) to check the [GPU delegate](https://www.tensorflow.org/lite/performance/gpu) compatibility of the given model by providing `gpu_compatibility=True` option.\r\n\r\nThanks.\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60793\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60793\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473184161, "node_id": "MDU6TGFiZWw0NzMxODQxNjE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:support", "name": "type:support", "color": "159b2e", "default": false, "description": "Support issues" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "Hi:\r\nWhat I would like to add is how to significantly reduce tflite shared libraries when using cmake", "Hi @panhu \r\n\r\nTensorFlow Lite enables you to reduce model binary sizes by using selective builds. Selective builds skip unused operations in your model set and produce a compact library with just the runtime and the op kernels required for the model to run on your mobile device.\r\n\r\nPlease refer to this [document](https://www.tensorflow.org/lite/guide/reduce_binary_size) on different ways of reducing TFLite binary size.\r\n\r\nThanks.", "Hi@pjpratik\r\n\r\nThank you for your reply. What I would like to know is how to reduce tflite time when using cmake. The document you provided mentioned using bazel.\r\n\r\nThanks.", "Hi @panhu \r\n\r\nThe current instructions are provided using `bazel` which is recommended. You can try to build [C API](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/c/README.md#using-the-c-api) using bazel as well and try reducing the binary size using above instructions.\r\n\r\nAlso, you can build custom c library using bazel by changing the build file.\r\n\r\nPlease check this thread [#48068](https://github.com/tensorflow/tensorflow/issues/48068) for reference.\r\n\r\nThanks.\r\n\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60792\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60792\">No</a>\n" ]

[]

[]

[ "Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).\n\nView this [failed invocation](https://github.com/tensorflow/tensorflow/pull/60791/checks?check_run_id=14058598571) of the CLA check for more information.\n\nFor the most up to date status, view the checks section at the bottom of the pull request." ]

[ { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 1097547538, "node_id": "MDU6TGFiZWwxMDk3NTQ3NTM4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:gpu", "name": "comp:gpu", "color": "0052cc", "default": false, "description": "GPU related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "gowthamkpr", "id": 47574994, "node_id": "MDQ6VXNlcjQ3NTc0OTk0", "avatar_url": "https://avatars.githubusercontent.com/u/47574994?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gowthamkpr", "html_url": "https://github.com/gowthamkpr", "followers_url": "https://api.github.com/users/gowthamkpr/followers", "following_url": "https://api.github.com/users/gowthamkpr/following{/other_user}", "gists_url": "https://api.github.com/users/gowthamkpr/gists{/gist_id}", "starred_url": "https://api.github.com/users/gowthamkpr/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gowthamkpr/subscriptions", "organizations_url": "https://api.github.com/users/gowthamkpr/orgs", "repos_url": "https://api.github.com/users/gowthamkpr/repos", "events_url": "https://api.github.com/users/gowthamkpr/events{/privacy}", "received_events_url": "https://api.github.com/users/gowthamkpr/received_events", "type": "User", "site_admin": false } ]

[ "ther could be few reasons but the main is to Ensure that GPU drivers are installed und make sur you have the appropriate GPU drivers on your system.", "Hi @gg4u ,\r\n\r\nCould you please confirm the OS you are using. Could you please confirm whether GPU drivers and nvidia toolkit installed and path variables set properly ? Refer to the documentation source [here](https://www.tensorflow.org/install/pip#step-by-step_instructions) and confirm this was followed.\r\n\r\nPlease share the output of `nvidia-smi` command and `python3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"` from Jupyter. \r\n\r\nThe error seems to be related to CUDA path setting. Can you try fresh installation of conda environment on Jupyter notebook and let us know the outcome.", "OS => I don't know, I am using a university cluster, it must be linux but i don't know which.\r\n\r\nNote that I was eventually able to run on jupyter as well, but only if :\r\nI reset :\r\n```\r\nCUDNN_PATH=$(dirname $(python -c \"import nvidia.cudnn;print(nvidia.cudnn.__file__)\"))\r\nexport LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib\r\n```\r\n\r\nand\r\n```\r\n# Install NVCC\r\nconda install -c nvidia cuda-nvcc=11.3.58\r\n# Configure the XLA cuda directory\r\nmkdir -p $CONDA_PREFIX/etc/conda/activate.d\r\nprintf 'export XLA_FLAGS=--xla_gpu_cuda_data_dir=$CONDA_PREFIX/lib/\\n' >> $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh\r\nsource $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh\r\n# Copy libdevice file to the required path\r\nmkdir -p $CONDA_PREFIX/lib/nvvm/libdevice\r\ncp $CONDA_PREFIX/lib/libdevice.10.bc $CONDA_PREFIX/lib/nvvm/libdevice/\r\n```\r\n\r\nand when I launch jupyter lab, it must be launched from the same conda environment I re-set up those script.\r\nKernel selection won't work anyway (in other words, I must launch jupyter lab from the same conda environment I intend to use the kernel, otherwise environment variables makes confusion and on jupyter Cuda won't be found).\r\n\r\n\r\nnvidia-smi:\r\n\r\n```\r\n+-----------------------------------------------------------------------------+\r\n| NVIDIA-SMI 525.89.02 Driver Version: 525.89.02 CUDA Version: 12.0 |\r\n|-------------------------------+----------------------+----------------------+\r\n| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\r\n| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\r\n| | | MIG M. |\r\n|===============================+======================+======================|\r\n| 0 NVIDIA GeForce ... On | 00000000:03:00.0 Off | N/A |\r\n| 28% 30C P8 15W / 250W | 10451MiB / 11264MiB | 0% Default |\r\n| | | N/A |\r\n+-------------------------------+----------------------+----------------------+\r\n| 1 NVIDIA GeForce ... On | 00000000:21:00.0 Off | N/A |\r\n| 28% 27C P8 16W / 250W | 9961MiB / 11264MiB | 0% Default |\r\n| | | N/A |\r\n+-------------------------------+----------------------+----------------------+\r\n| 2 NVIDIA GeForce ... On | 00000000:41:00.0 Off | N/A |\r\n| 28% 28C P8 4W / 250W | 9961MiB / 11264MiB | 0% Default |\r\n| | | N/A |\r\n+-------------------------------+----------------------+----------------------+\r\n| 3 NVIDIA GeForce ... On | 00000000:61:00.0 Off | N/A |\r\n| 28% 28C P8 20W / 250W | 9961MiB / 11264MiB | 0% Default |\r\n| | | N/A |\r\n+-------------------------------+----------------------+----------------------+\r\n \r\n+-----------------------------------------------------------------------------+\r\n| Processes: |\r\n| GPU GI CI PID Type Process name GPU Memory |\r\n| ID ID Usage |\r\n|=============================================================================|\r\n| 0 N/A N/A 3115300 C ...3/envs/conda01/bin/python 10448MiB |\r\n| 1 N/A N/A 3115300 C ...3/envs/conda01/bin/python 9958MiB |\r\n| 2 N/A N/A 3115300 C ...3/envs/conda01/bin/python 9958MiB |\r\n| 3 N/A N/A 3115300 C ...3/envs/conda01/bin/python 9958MiB |\r\n+-----------------------------------------------------------------------------+\r\n```\r\n\r\nFrom jupyter:\r\n```\r\n\r\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\r\n\r\n2023-06-07 15:11:35.922385: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\r\nTo enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\r\n2023-06-07 15:11:36.765039: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\r\n[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU'), PhysicalDevice(name='/physical_device:GPU:1', device_type='GPU'), PhysicalDevice(name='/physical_device:GPU:2', device_type='GPU'), PhysicalDevice(name='/physical_device:GPU:3', device_type='GPU')]\r\n\r\n\r\n\r\n```\r\n\r\n> Can you try fresh installation of conda environment on Jupyter notebook and let us know the outcome.\r\n\r\nI did it all yesterday, I spent a whole day, and did so many times.\r\n\r\nI won't touch again the installation.\r\n\r\nBut to reproduce, I did the following:\r\n\r\n- create a conda env (as per instruction) and CUDA drivers\r\n- installed ikernel (to select the kernel in jupyter)\r\n- on terminal, TF read the GPU => OK!\r\n- on jupyter lab, I could select a different kernel, BUT inspecting on `!conda -info` and `!which python` it was keep referring to the base from where jupyter was launched, and that I pervously set the environment variables automatically (see '# script' below from https://www.tensorflow.org/install/pip). \r\n\r\n\r\nIt was messy to debug, because there were conflicts on `libdevice` in jupter, but not in terminal; probably because the environment variable instructing on the cuda libraries could not be found. Maybe some issues with `ikernel` - I don't have clue only guessing.\r\n\r\nI eventually got it to work without automating the variables :\r\n\r\n```\r\n# script\r\n\r\nmkdir -p $CONDA_PREFIX/etc/conda/activate.d\r\necho 'CUDNN_PATH=$(dirname $(python -c \"import nvidia.cudnn;print(nvidia.cudnn.__file__)\"))' >> $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh\r\necho 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib' >> $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh\r\n\r\n```\r\nbut setting them manually, and then only launch the jupter lab from the conda environemnt corresponding to the kernel I will work with.\r\n\r\n\r\n", "The attached commands from Documentation should fix system paths to be automatically configured when you activate the conda environment.\r\n\r\nI hope once you have created the conda environment and followed all the steps mentioned in documentation including the commands for Automatic enabling of system paths every time conda environment activated should fix the issue. This should be one time exercise only and then it should work every time when activated the respective conda environment.\r\n\r\n", "Unfortunately not, I see a mismatch between the conda or python (`which python`) listed in the jupyter notebook, and the selected kernel.\r\n\r\nAwkwardly, I was able to get TF see the GPUs on Jupyter, only if I created a virtual environment in python (`python -m venv ./env`), and then tried to install cuda drivers from conda , but all other libraries via pip.\r\n\r\nHowever, there are always challenges, because, for example, some libraries, like ones using `cudf`, will fail (I ve just discovered that with aweful regret). \r\n\r\nIT would be highly beneficial, in my opinion, if you could add a guide to install jupyter lab with multiple kernels, so to minismize error for users approaching TF on rermote cloud for first time. Assume people using remote servers. At my university, the help desk even asked me if it was necessary to use jupyter or conda virtual environment (which I actually would be .. knida necessary) to minimise error, and just keep one base - the default. But I am afraid that if I messy up something, I m kinda done.\r\n", "What challenges are you facing with ```cuDF```? For resolving dependency issues with TF installed in a conda environment you can try creating a new environment with an earlier version of Python (```conda create -n EnvNameHere python=3.8```) or using tools like Libmamba (see https://www.anaconda.com/blog/a-faster-conda-for-a-growing-community/). Also, try running ```lsb_release -a``` or ```uname -a``` from root outside of any virtual environment. Knowing more about your OS is crucial to determining whether this is a CUDA path issue.", "@gg4u ,\r\n\r\nCould you please provide the OS details also.Thanks!", "Hi @ggfu,\r\n\r\nYou might need to use `conda install nb_conda_kernels` and may be also `conda install ipykernel` additionally for Jupyter.\r\nCould you please follow the instructions mentioned here at [SO1](https://stackoverflow.com/questions/76101948/tensorflow-gpu-recognized-in-the-terminal-but-not-in-the-jupyter-notebook) and [SO2](https://stackoverflow.com/questions/57332001/import-tensorflow-working-in-terminal-but-not-in-jupyter-notebook) and let us know it works.\r\n\r\n", "Hi @gg4u this is an issue with using Jupyter Notebook itself. You can take a look at this [issue](https://stackoverflow.com/questions/76101948/tensorflow-gpu-recognized-in-the-terminal-but-not-in-the-jupyter-notebook) on why jupyter doesn't call GPU and follow the work around. As for the documentation, tensorflow doesn't officially support Jupyter notebooks and I don't think this will be documented anywhere officially. Thanks!!", "Hi, thanks for the support. \r\n\r\nI can't summarize exactly what causes are responsible of the issues, I think the problem is conflicting environment variables.\r\n\r\nI had this constrains:\r\n- using a university cluster, the university discouraging in using conda env, just installed on the dedicated env\r\n- tensorflow encorages in using conda\r\n- external libraries complained if on conda (installs with pip conflicted).\r\n\r\nI temporarilly have resolved in creating a pip envionrment, and launching jupyter lab from there.\r\nJupyter sees the GPUs, while if it was launched from conda, then TF was not seeing the GPU.\r\nPytorch saw the GPUs on both.\r\n\r\nI followed :\r\nhttps://stackoverflow.com/questions/76101948/tensorflow-gpu-recognized-in-the-terminal-but-not-in-the-jupyter-notebook\r\n\r\nbut for some reasons, `which python` on jupyter opened in a chosen enviroment, was anyway pointing to the environment from where jupyter lab server was launched.", "Hi @gg4u, Unfortunately there is not much that can be done about this from our end. Please follow the process as mentioned above and this issue should be resolved." ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" }, { "id": 2671351731, "node_id": "MDU6TGFiZWwyNjcxMzUxNzMx", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ModelOptimizationToolkit", "name": "ModelOptimizationToolkit", "color": "BFD629", "default": false, "description": "TF Model Optimization Toolkit" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "miaout17", "id": 22063, "node_id": "MDQ6VXNlcjIyMDYz", "avatar_url": "https://avatars.githubusercontent.com/u/22063?v=4", "gravatar_id": "", "url": "https://api.github.com/users/miaout17", "html_url": "https://github.com/miaout17", "followers_url": "https://api.github.com/users/miaout17/followers", "following_url": "https://api.github.com/users/miaout17/following{/other_user}", "gists_url": "https://api.github.com/users/miaout17/gists{/gist_id}", "starred_url": "https://api.github.com/users/miaout17/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/miaout17/subscriptions", "organizations_url": "https://api.github.com/users/miaout17/orgs", "repos_url": "https://api.github.com/users/miaout17/repos", "events_url": "https://api.github.com/users/miaout17/events{/privacy}", "received_events_url": "https://api.github.com/users/miaout17/received_events", "type": "User", "site_admin": false }, { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @NickLucche \r\n\r\nThanks for reporting this issue.\r\n\r\n@pkgoogle I was able to reproduce this issue. Please find the gist [here](https://colab.research.google.com/gist/pjpratik/0a3f11e436e1e7d73bccf3fc8c57afe9/60789.ipynb).\r\n\r\nCould you please look into this if this is intended behaviour?\r\n\r\nThanks.", "Hi @NickLucche, thanks for reporting the issue. The documentation https://www.tensorflow.org/lite/performance/post_training_quantization#integer_only suggests it should throw an error than a forced conversion. We will work on this as a bug.", "thank you!", "Hi @miaout17, can you please take a look? Thanks." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1097541661, "node_id": "MDU6TGFiZWwxMDk3NTQxNjYx", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:tpus", "name": "comp:tpus", "color": "0052cc", "default": false, "description": "tpu, tpuestimator" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "Hi @MegaCreater \r\n\r\nBased on the error message you provided earlier, it seems that the issue is not related to the specific dataset being used (MNIST vs Fashion MNIST), but rather with how the data is being accessed from the TPU. The error message “File system scheme ‘[local]’ not implemented” suggests that the code is trying to access a local file system, but this is not supported by the TPU.\r\n\r\nTPUs are designed to work with cloud storage services such as Google Cloud Storage (GCS), and cannot directly access data stored on the local file system of the machine running the code. \"If you are using TPU Nodes, you need to store all data files read by the TensorFlow Dataset in Google Cloud Storage (GCS) buckets. If you are using TPU VMs, you can store data wherever you like. For more information on TPU Nodes and TPU VMs, refer to the TPU System Architecture documentation.\"\r\n\r\nYou can create a TPU VM, but it is not for free. You can use it by following this tutorial: “https://colab.research.google.com/github/tensorflow/tpu/blob/master/models/experimental/mnist_jupyter/Cloud-TPU-Demo.ipynb”. If you try changing the runtime to GPU, it should work correctly.\r\nOne other approach you can try is to access data stored in a Google Cloud Storage bucket using Colab. Here is a link :\"https://colab.research.google.com/github/tensorflow/tpu/blob/master/tools/colab/keras_mnist_tpu.ipynb\"", "@MegaCreater,\r\nIt looks like model_dir is set to a local directory in /tmp/, can you double check to make sure that it is a path to a GCS bucket?\r\nIt could be possible that it's creating a directory within data_dir which is a GCS path. Both data_dir and model_dir should be set to GCS buckets.\r\n \r\nYou would need to have your data stored inside a GCS bucket and the Cloud TPUs should have access to that. Also can you simply look up the IP address of your VM (something like **!curl ifconfig.me** in your Colab cell) and then look up which geolocation that IP belongs to.", "@medmabcf it's dataset related only .... \r\n```python3\r\n# this works fine (NO ERROR)\r\n(ds_train, ds_test), ds_info = tfds.load('mnist', split=['train', 'test'], shuffle_files=True, as_supervised=True, with_info=True, try_gcs = True)\r\n```\r\n```python3 \r\n# This create error ! (ERROR)\r\n(ds_train, ds_test), ds_info = tfds.load('fashion_mnist', split=['train', 'test'], shuffle_files=True, as_supervised=True, with_info=True,try_gcs = True)\r\n```\r\n\"https://colab.research.google.com/github/tensorflow/tpu/blob/master/tools/colab/keras_mnist_tpu.ipynb\" this example is not useful because my code works fine on 'minst' dataset. ", "@tilakrayal are you talking about something like this ? \r\n\r\n```python3\r\nimport tensorflow as tf\r\nimport tensorflow_datasets as tfds\r\n# Set TPU address (if using Google Cloud TPU)\r\ntpu_address = 'grpc://<tpu-ip-address>'\r\ntpu_resolver = tf.distribute.cluster_resolver.TPUClusterResolver(tpu=tpu_address)\r\ntf.config.experimental_connect_to_cluster(tpu_resolver)\r\ntf.tpu.experimental.initialize_tpu_system(tpu_resolver)\r\ndata_dir = 'gs://your-bucket/data' # GCS bucket path for data\r\nmodel_dir = 'gs://your-bucket/model' # GCS bucket path for model\r\nbuilder = tfds.builder('your_dataset_name', data_dir=data_dir)\r\nbuilder.download_and_prepare()\r\n\r\n# Define the training and validation datasets\r\ntrain_dataset = builder.as_dataset(split='train', shuffle_files=True)\r\nval_dataset = builder.as_dataset(split='validation', shuffle_files=True)\r\n\r\n# Enable batching and shuffling\r\ntrain_dataset = train_dataset.shuffle(1024).batch(batch_size)\r\nval_dataset = val_dataset.batch(batch_size)\r\nstrategy = tf.distribute.experimental.TPUStrategy(tpu_resolver)\r\nwith strategy.scope():\r\n # Define and compile your model\r\n model = ...\r\n model.compile(...)\r\ncallbacks = [\r\n tf.keras.callbacks.ModelCheckpoint(\r\n filepath=model_dir, # Save the model to the GCS bucket\r\n save_best_only=True,\r\n save_weights_only=True\r\n ),\r\n # Other callbacks...\r\n]\r\n\r\n# Train the model\r\nmodel.fit(train_dataset, validation_data=val_dataset, callbacks=callbacks, ...)\r\n```\r\nHow can I access \"gs://your-bucket/data\" ? I mean how can I get my bucket address ? (what will be the \"gs://your-bucket/\" ?)", "@MegaCreater ,\r\nYou can access a bucket in the Cloud Dataprep UI by manually entering the Google Cloud Storage path. \r\nhttps://cloud.google.com/storage/docs/access-public-data\r\nhttps://cloud.google.com/storage/docs/cloud-console\r\n\r\nThank you!", "@tilakrayal its not working. I don't have cloud storage. And cloud bucket. I don't have any paid or pro access to google cloud. I am not using any cloud VM. I am using google colab TPU. So, how can I do all things in colab TPU environment? ", "Hi @MegaCreater \r\n\r\nThe very first answer mentions that the error is due to TPUs being able to pull data only from GCS. \r\n\r\nIndeed `mnist` dataset exists in `gs://tfds-data/datasets/mnist/` and hence everything works, but `fashion_mnist` is unfortunately not available through public TFDS GCS. You should create your own GCS bucket and store `fashion_mnist` there and then pull it using `tfds.load(data_dir=<your GCS bucket>)`. ", "@fineguy I am using **colab (Colab's TPU)**. Can I use Google drive as cloud space ? ", "@MegaCreater , Yes you can mount your Google Drive and import the data.\r\n\r\n```\r\nfrom google.colab import drive\r\ndrive.mount('/content/drive')\r\n```\r\n\r\n```\r\nfrom tensorflow import keras\r\nimport os\r\n\r\n# Path to the MNIST dataset folder in your Google Drive\r\nmnist_path = '/content/drive/MyDrive/MNIST/'\r\n```", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60788\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60788\">No</a>\n" ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 4511033337, "node_id": "LA_kwDOArmXAs8AAAABDODn-Q", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.10", "name": "TF 2.10", "color": "C15088", "default": false, "description": "" } ]

[ { "login": "alankelly", "id": 5112267, "node_id": "MDQ6VXNlcjUxMTIyNjc=", "avatar_url": "https://avatars.githubusercontent.com/u/5112267?v=4", "gravatar_id": "", "url": "https://api.github.com/users/alankelly", "html_url": "https://github.com/alankelly", "followers_url": "https://api.github.com/users/alankelly/followers", "following_url": "https://api.github.com/users/alankelly/following{/other_user}", "gists_url": "https://api.github.com/users/alankelly/gists{/gist_id}", "starred_url": "https://api.github.com/users/alankelly/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/alankelly/subscriptions", "organizations_url": "https://api.github.com/users/alankelly/orgs", "repos_url": "https://api.github.com/users/alankelly/repos", "events_url": "https://api.github.com/users/alankelly/events{/privacy}", "received_events_url": "https://api.github.com/users/alankelly/received_events", "type": "User", "site_admin": false }, { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @misterBart \r\n\r\nThe TFLite includes the TFLite logger options which can be used to set the logging severity levels.\r\n\r\nhttps://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/logger.h\r\n\r\n\r\nhttps://github.com/tensorflow/tensorflow/blob/2f1a31a27ea0a21e421e77a957cfd573973f2a65/tensorflow/lite/logger.h#L31-L32\r\n\r\nDoes this help in your use case?\r\n\r\nthanks.", "No, this does not help at all.\r\nI posted about a memory bug in TfLite.\r\nThe TfLite logger options pertain to output logging, which is completely unrelated to memory bugs. Changing the logger option will not make a memory bug disappear.", "Hi @alankelly, can you take a look? Thanks.", "More than a month passed. Any news?\r\nMeanwhile I tested the example program in my opening post with TfLite **2.13**, and it gives the same Valgrind messages. So it seems that also in TfLite 2.13, the function `xnn_deinitialize` is never called.", "I can't re-produce this from the latest commit in tensorflow.\r\n\r\n`bazel build -c dbg leak`\r\n\r\n```\r\n==2088258== Memcheck, a memory error detector\r\n==2088258== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.\r\n==2088258== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info\r\n==2088258== Command: ../../bazel-bin/tensorflow/lite/leak\r\n==2088258== \r\nINFO: Initialized TensorFlow Lite runtime.\r\nINFO: Applying 1 TensorFlow Lite delegate(s) lazily.\r\nINFO: Created TensorFlow Lite XNNPACK delegate for CPU.\r\nVERBOSE: Replacing 88 out of 119 node(s) with delegate (TfLiteXNNPackDelegate) node, yielding 42 partitions for the whole graph.\r\nINFO: Successfully applied the default TensorFlow Lite delegate indexed at 0.\r\n *NOTE*: because a delegate has been applied, the precision of computations should be unchanged, but the exact output tensor values may have changed. If such output values are checked in your code, like in your tests etc., please consider increasing error tolerance for the check.\r\n==2088258== \r\n==2088258== HEAP SUMMARY:\r\n==2088258== in use at exit: 2,224 bytes in 10 blocks\r\n==2088258== total heap usage: 4,853 allocs, 4,843 frees, 5,656,799 bytes allocated\r\n==2088258== \r\n==2088258== LEAK SUMMARY:\r\n==2088258== definitely lost: 0 bytes in 0 blocks\r\n==2088258== indirectly lost: 0 bytes in 0 blocks\r\n==2088258== possibly lost: 0 bytes in 0 blocks\r\n==2088258== still reachable: 2,224 bytes in 10 blocks\r\n==2088258== suppressed: 0 bytes in 0 blocks\r\n==2088258== Rerun with --leak-check=full to see details of leaked memory\r\n==2088258== \r\n==2088258== For lists of detected and suppressed errors, rerun with: -s\r\n==2088258== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)\r\n\r\n```\r\n\r\nIt is not an issue with `xnn_deinitialize` - this function simply returns xnn_status_success: https://github.com/google/XNNPACK/blob/645035286fe31d47eeb07900450f4f6540b75c2c/src/init.c#L83\r\n\r\nPerhaps the issue only appears with your model. Have you tried with multiple models?", "Hi @misterBart, let us know if you have tried multiple models to help us look into the problem further.", "@alankelly \r\nBut you are reproducing my issue, your log also mentions\r\n`still reachable: 2,224 bytes in 10 blocks`\r\n\r\nFrom https://github.com/google/XNNPACK/blob/master/include/xnnpack.h\r\n\"To avoid memory and resource leaks, users must call xnn_deinitialize once for each successful xnn_initialize call.\"\r\nIf `xnn_deinitialize` does nothing, then the above statement is probably outdated.\r\n```\r\n$ git clone https://github.com/tensorflow/tensorflow tensorflow_src\r\n$ grep -r 'xnn_deinitialize' tensorflow_src/\r\ntensorflow_src/tensorflow/lite/delegates/xnnpack/xnnpack_delegate.cc: xnn_deinitialize();\r\n$\r\n```\r\nThe function is never called.\r\n\r\nI also tested my issue with the master branch and the .tflite model from https://www.tensorflow.org/lite/examples/segmentation/overview Then `valgrind --leak-check=full ./app` still displays `still reachable: 704 bytes in 11 blocks`\r\n\r\nBut if I disable the XNNPACK delegate, by replacing\r\n`const tflite::ops::builtin::BuiltinOpResolver op_resolver`\r\nwith \r\n`const tflite::ops::builtin::BuiltinOpResolverWithoutDefaultDelegates op_resolver;`\r\nRecompile the program, and then run `valgrind --leak-check=full ./app`, then I obtain:\r\n```\r\n$ valgrind --leak-check=full ./app\r\n==924== Memcheck, a memory error detector\r\n==924== Copyright (C) 2002-2022, and GNU GPL'd, by Julian Seward et al.\r\n==924== Using Valgrind-3.19.0 and LibVEX; rerun with -h for copyright info\r\n==924== Command: ./app\r\n==924==\r\n==924== error calling PR_SET_PTRACER, vgdb might block\r\n==924==\r\n==924== HEAP SUMMARY:\r\n==924== in use at exit: 0 bytes in 0 blocks\r\n==924== total heap usage: 1,875 allocs, 1,875 frees, 6,714,008 bytes allocated\r\n==924==\r\n==924== All heap blocks were freed -- no leaks are possible\r\n==924==\r\n==924== For lists of detected and suppressed errors, rerun with: -s\r\n==924== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)\r\n```\r\nThen it is correct: `All heap blocks were freed -- no leaks are possible`. All `still reachable` messages have disappeared.\r\n\r\nConcludingly, TfLite + XNNPACK delegate allocates memory at the start but does not deallocate this memory.\r\n\r\nMinimal-working example:\r\n```\r\n#include <tensorflow/lite/model.h>\r\n#include <tensorflow/lite/kernels/register.h>\r\n#include <tensorflow/lite/interpreter.h>\r\n\r\nusing namespace tflite;\r\nusing namespace std;\r\n\r\nint main() {\r\n\tunique_ptr<tflite::FlatBufferModel> model = tflite::FlatBufferModel::BuildFromFile(\"lite-model_deeplabv3_1_metadata_2.tflite\"); //This line generates one \"still reachable\" message\r\n\t//const tflite::ops::builtin::BuiltinOpResolver op_resolver; //Use XNNPACK delegate\r\n\tconst tflite::ops::builtin::BuiltinOpResolverWithoutDefaultDelegates op_resolver;\r\n\ttflite::InterpreterBuilder interpreter_builder(*model, op_resolver);\r\n\tinterpreter_builder.SetNumThreads(1);\r\n\r\n\tunique_ptr<tflite::Interpreter> interpreter;\r\n\tinterpreter_builder(&interpreter);\r\n\tinterpreter->AllocateTensors(); //This line generates many \"still reachable\" messages if XNNPACK is used\r\n\r\n\tdelete model->error_reporter(); //This will get rid off the first \"still reachable\" message\r\n\treturn EXIT_SUCCESS;\r\n}\r\n```\r\n", "@alankelly Have you looked at the matter again after my above reply?", "Several weeks have passed. It would be nice if we could resolve this issue.", "Yes you're right I did reproduce the issue. Sorry for the late reply, I was on holidays.\r\n\r\nToday xnn_deinitialize does nothing but someday it may be required to free memory so documenting that it should be called may prevent problems for users of XNNPACK down the road.\r\n\r\nThis stack overflow answer gives a good explanation of different types of memory leak, especially what still reachable means: https://stackoverflow.com/a/3857638\r\n\r\nLet's look at the still reachable blocks in your report:\r\n\r\ntflite::DefaultErrorReporter: This is a static object which is allocated once per instance. The memory is cleaned up when the program is terminated. This looks like it is working as intended.\r\nhttps://github.com/tensorflow/tensorflow/blob/48043f5993743a9bba3a7b4adc9155cb66bbe6b2/tensorflow/lite/stderr_reporter.cc#L30\r\n\r\nThe others come from cpuinfo: https://github.com/pytorch/cpuinfo/blob/959002f82d7962a473d8bf301845f2af720e0aa4/src/x86/linux/init.c#L181\r\n\r\nLooking at the code we see that these functions are only ever called once and the allocated memory lives for the program lifetime: https://github.com/pytorch/cpuinfo/blob/959002f82d7962a473d8bf301845f2af720e0aa4/src/init.c#L24\r\n\r\nThis again is working as intended. Valgrind correctly identifies that the memory is not freed but still accessible. \r\n\r\n\r\n\r\n", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60787\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60787\">No</a>\n", "Ok, we all do need a holiday :) Thanks for the code review. Still (perhaps from purist perspective), many developers try to keep a clean Valgrind report, which precludes them from checking that these \"still reachable bytes\" come from TfLite/cpuinfo and are not something they introduced themselves.\r\n\r\nI checked TfLite for `cpuinfo_deinitialize()`:\r\n```\r\npersonau@desktop1:~/Temp/tensorflow_src/tensorflow/lite$ grep -r 'cpuinfo_deinitialize' .\r\n./kernels/fully_connected.cc: // paths we can take. Note the we do not call `cpuinfo_deinitialize` in\r\n./kernels/cpu_backend_context.cc: cpuinfo_deinitialize();\r\n```\r\nIn ./kernels/fully_connected.cc I read:\r\n```\r\n// We ensure that cpuinfo is initialized to correctly detect the optimized\r\n// paths we can take. Note the we do not call `cpuinfo_deinitialize` in\r\n// `Free`: that operation is currently a no-op AND we want to avoid\r\n// deinitializing cpuinfo for other parts of the program that could need it\r\n// after we free the op if it ever does perform something.\r\n```\r\nThe rationale is not entirely clear to me, but it at least explains you intentionally do not deallocate cpuinfo memory.\r\nIn ./kernels/cpu_backend_context.cc I read;\r\n```\r\nCpuBackendContext::CpuInfo::~CpuInfo() {\r\n if (init_status_ == InitStatus::kInitialized) {\r\n cpuinfo_deinitialize();\r\n }\r\n}\r\n```\r\nSo it seems possible to call `cpuinfo_deinitialize()`.\r\nMy final question, before letting this issue rest, is: How can I call this `CpuBackendContext::CpuInfo::~CpuInfo()` in my minimal TfLite program?\r\n" ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Hi @gbaned can you help to check the PR process? It's approved but the states is still \"Assigned Reviewer\".", "Hi @Zantares Sorry for the delay, can you please take a look on below internal errors? Thank you!\r\n\r\nTraceback (most recent call last):\r\n File \"/tensorflow/lite/python/analyzer_test.py\", line 224, in testTxtWithEinsum\r\n self.assertIn('Op#0 RESHAPE(T#1, T#4[512, 512]) -> [T#5]', txt)\r\nAssertionError: 'Op#0 RESHAPE(T#1, T#4[512, 512]) -> [T#5]' not found in \"=== TFLite ModelAnalyzer ===\\n\\nYour TFLite model has '1' subgraph(s). In the subgraph description below,\\nT# represents the Tensor numbers. For example, in Subgraph#0, the RESHAPE op takes\\ntensor #1 and tensor #3 as input and produces tensor #4 as output.\\n\\nSubgraph#0 main(T#0, T#1) -> [T#6]\\n Op#0 RESHAPE(T#1, T#3[512, 512]) -> [T#4]\\n Op#1 BATCH_MATMUL(T#0, T#4) -> [T#5]\\n Op#2 RESHAPE(T#5, T#2[1, 100, 8, 64]) -> [T#6]\\n\\nTensors of Subgraph#0\\n T#0(lhs) shape:[1, 100, 512], type:FLOAT32\\n T#1(rhs) shape:[512, 8, 64], type:FLOAT32\\n T#2(einsum/Einsum) shape:[4], type:INT32 RO 16 bytes, buffer: 3, data:[1, 100, 8, 64]\\n T#3(einsum/Einsum1) shape:[2], type:INT32 RO 8 bytes, buffer: 4, data:[512, 512]\\n T#4(einsum/Einsum2) shape:[512, 512], type:FLOAT32\\n T#5(einsum/Einsum3) shape:[1, 100, 512], type:FLOAT32\\n T#6(Identity) shape:[1, 100, 8, 64], type:FLOAT32\\n\\n---------------------------------------------------------------\\n Model size: 1204 bytes\\n Non-data buffer size: 1080 bytes (89.70 %)\\n Total data buffer size: 124 bytes (10.30 %)\\n (Zero value buffers): 0 bytes (00.00 %)\\n\\n* Buffers of TFLite model are mostly used for constant tensors.\\n And zero value buffers are buffers filled with zeros.\\n Non-data buffers area are used to store operators, subgraphs and etc.\\n You can find more details from https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/schema/schema.fbs\\n\\n\"", "> Hi @Zantares Sorry for the delay, can you please take a look on below internal errors? Thank you!\r\n> \r\n> Traceback (most recent call last): File \"/tensorflow/lite/python/analyzer_test.py\", line 224, in testTxtWithEinsum self.assertIn('Op#0 RESHAPE(T#1, T#4[512, 512]) -> [T#5]', txt) AssertionError: 'Op#0 RESHAPE(T#1, T#4[512, 512]) -> [T#5]' not found in \"=== TFLite ModelAnalyzer ===\\n\\nYour TFLite model has '1' subgraph(s). In the subgraph description below,\\nT# represents the Tensor numbers. For example, in Subgraph#0, the RESHAPE op takes\\ntensor #1 and tensor #3 as input and produces tensor #4 as output.\\n\\nSubgraph#0 main(T#0, T#1) -> [T#6]\\n Op#0 RESHAPE(T#1, T#3[512, 512]) -> [T#4]\\n Op#1 BATCH_MATMUL(T#0, T#4) -> [T#5]\\n Op#2 RESHAPE(T#5, T#2[1, 100, 8, 64]) -> [T#6]\\n\\nTensors of Subgraph#0\\n T#0(lhs) shape:[1, 100, 512], type:FLOAT32\\n T#1(rhs) shape:[512, 8, 64], type:FLOAT32\\n T#2(einsum/Einsum) shape:[4], type:INT32 RO 16 bytes, buffer: 3, data:[1, 100, 8, 64]\\n T#3(einsum/Einsum1) shape:[2], type:INT32 RO 8 bytes, buffer: 4, data:[512, 512]\\n T#4(einsum/Einsum2) shape:[512, 512], type:FLOAT32\\n T#5(einsum/Einsum3) shape:[1, 100, 512], type:FLOAT32\\n T#6(Identity) shape:[1, 100, 8, 64], type:FLOAT32\\n\\n---------------------------------------------------------------\\n Model size: 1204 bytes\\n Non-data buffer size: 1080 bytes (89.70 %)\\n Total data buffer size: 124 bytes (10.30 %)\\n (Zero value buffers): 0 bytes (00.00 %)\\n\\n* Buffers of TFLite model are mostly used for constant tensors.\\n And zero value buffers are buffers filled with zeros.\\n Non-data buffers area are used to store operators, subgraphs and etc.\\n You can find more details from [https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/schema/schema.fbs\\n\\n](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/schema/schema.fbs%5Cn%5Cn)\"\r\n\r\nSeems this internal error is not related to this PR, I have merged the newest master to see if it can be fixed, thanks!", "Hi @penpornk may you reapprove this PR? ", "Hi @penpornk Can you please review this PR ? Thank you!" ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 1097543484, "node_id": "MDU6TGFiZWwxMDk3NTQzNDg0", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:runtime", "name": "comp:runtime", "color": "0052cc", "default": false, "description": "c++ runtime, performance issues (cpu)" }, { "id": 1133285679, "node_id": "MDU6TGFiZWwxMTMzMjg1Njc5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:xla", "name": "comp:xla", "color": "0052cc", "default": false, "description": "XLA" }, { "id": 4032183365, "node_id": "LA_kwDOArmXAs7wVjxF", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.9", "name": "TF 2.9", "color": "1CF842", "default": false, "description": "Issues found in the TF 2.9 release (or RCs)" } ]

[ { "login": "kenfranko", "id": 56562020, "node_id": "MDQ6VXNlcjU2NTYyMDIw", "avatar_url": "https://avatars.githubusercontent.com/u/56562020?v=4", "gravatar_id": "", "url": "https://api.github.com/users/kenfranko", "html_url": "https://github.com/kenfranko", "followers_url": "https://api.github.com/users/kenfranko/followers", "following_url": "https://api.github.com/users/kenfranko/following{/other_user}", "gists_url": "https://api.github.com/users/kenfranko/gists{/gist_id}", "starred_url": "https://api.github.com/users/kenfranko/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/kenfranko/subscriptions", "organizations_url": "https://api.github.com/users/kenfranko/orgs", "repos_url": "https://api.github.com/users/kenfranko/repos", "events_url": "https://api.github.com/users/kenfranko/events{/privacy}", "received_events_url": "https://api.github.com/users/kenfranko/received_events", "type": "User", "site_admin": false }, { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "@Weili17,\r\n\r\nCould you please try to follow the below instructions and let us know if it helps in this case.\r\n\r\n1. Create a `nvvm/libdevice` folder in the Conda environment lib folder.\r\n\r\n2. Copy the `libdevice.10.bc` file to the directory `nvvm/libdevice`\r\n This file may be found in your system like the below path:\r\n\r\n```\r\nminiconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/recipe/NVIDIA_EULA:libdevice.10.bc\r\nminiconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/files:lib/libdevice.10.bc\r\nminiconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/licenses/NVIDIA_EULA:libdevice.10.bc\r\n```\r\ntry the command: **export XLA_FLAGS=–xla_gpu_cuda_data_dir=/home/miniconda3/envs/lib**\r\n\r\nThe path `/home/miniconda3/envs/lib` may be different in your case, it should trace for absolute path of lib folder available in miniconda3 like **miniconda3/envs/lib** or **miniconda/lib**\r\n\r\nThank you!", "> @Weili17,\r\n> \r\n> Could you please try to follow the below instructions and let us know if it helps in this case.\r\n> \r\n> 1. Create a `nvvm/libdevice` folder in the Conda environment lib folder.\r\n> 2. Copy the `libdevice.10.bc` file to the directory `nvvm/libdevice`\r\n> This file may be found in your system like the below path:\r\n> \r\n> ```\r\n> miniconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/recipe/NVIDIA_EULA:libdevice.10.bc\r\n> miniconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/files:lib/libdevice.10.bc\r\n> miniconda3/pkgs/cudatoolkit-11.2.2-hbe64b41_10/info/licenses/NVIDIA_EULA:libdevice.10.bc\r\n> ```\r\n> \r\n> try the command: **export XLA_FLAGS=–xla_gpu_cuda_data_dir=/home/miniconda3/envs/lib**\r\n> \r\n> The path `/home/miniconda3/envs/lib` may be different in your case, it should trace for absolute path of lib folder available in miniconda3 like **miniconda3/envs/lib** or **miniconda/lib**\r\n> \r\n> Thank you!\r\n\r\nIn my case, the code is running with C++ in a docker environment while cuda is installed in /usr/local/cuda. I have tested put the libdevice.10.bc to /usr/local/lib, it does not work. so what should I do? ", "Any update on this issue? I'm running into the same error while doing something very similar." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "Hi @dsbyprateekg ,\r\n\r\nIt seems the issue is related to Kaggle environment. I found there is no issue with Google colab and may refer to the [gist](https://colab.research.google.com/gist/SuryanarayanaY/153ff07cb2eb8c066c3987e526a5a72d/60784.ipynb).\r\n\r\nBy looking at the logs you might need to install setup tools separately in Kaggle.\r\n\r\nPlease try `pip install setuptools` first and then try to install `tflite-support` and let us know if still having issue after this.\r\n\r\nThanks!", "@SuryanarayanaY I tried that but got same error.", "@dsbyprateekg ,\r\n\r\nCould you please cross check the environment details using `pip list` and compare it with the Colab environment attached in gist above and let us know any differences. Also please confirm the setuptools version installed there.\r\n\r\nI am not sure,but maybe can you try uninstalling `setuptools_scm` and then install `setuptools` and let us know if it works.\r\n\r\nThanks!", "Hi @dsbyprateekg \r\n\r\nTo resolve this problem, you can try the following steps:\r\n\r\n1. Ensure that you have all the necessary dependencies and prerequisites installed on your system. Refer to the official documentation of tflite-support for any specific requirements.\r\n\r\n2. Update your pip package manager to the latest version by running the command:\r\n\r\n>> pip install --upgrade pip\r\n\r\n3. Check if you have the required development libraries and tools installed on your system. Some Python packages require certain system-level dependencies to build correctly. For example, on Ubuntu, you might need to install the build-essential package:\r\n\r\n>> sudo apt-get install build-essential\r\n\r\n4. Try installing the package again using the --no-binary flag to force a source build instead of using pre-compiled wheels:\r\n\r\n>> pip install --no-binary :all: tflite-support\r\n\r\n5. If the above steps don't resolve the issue, you can try cloning the tflite-support repository from GitHub and installing it directly from source:\r\n\r\n>> git clone https://github.com/tensorflow/tflite-support.git\r\n>> cd tflite-support\r\n>> pip install .", "@yasirulhadi I will follow mentioned steps and update you soon.", "@yasirulhadi after following your suggested steps, installation was done successfully. See below logs-\r\n\r\n\r\nbut now I am getting following error-\r\n`ModuleNotFoundError: No module named 'tflite_support.metadata_writers'`\r\n\r\nin this import `from tflite_support.metadata_writers import object_detector`", "Hi @dsbyprateekg \r\n\r\nThe error message ModuleNotFoundError: No module named 'tflite_support.metadata_writers' suggests that the tflite_support.metadata_writers module is not found. This could indicate a problem with the installation or the package itself.\r\n\r\n1. Double-check that the tflite-support package was installed correctly. You can verify this by running the following command and checking if tflite-support is listed:\r\n>> pip list\r\n\r\n2. Make sure you are using the correct import statement in your code. The tflite_support.metadata_writers module was introduced in a later version of the tflite-support package. If you are using an older version, it's possible that this module is not available. You can check the version of tflite-support installed by running:\r\n\r\n>> pip show tflite-support\r\n\r\nIf you have an older version installed, you can upgrade to the latest version using:\r\n\r\n>> pip install --upgrade tflite-support\r\n\r\n3. Finally, ensure that you have imported the necessary submodules correctly. For example, if you're trying to import the object_detector module, make sure you have the following import statements:\r\n\r\n>> from tflite_support.metadata_writers import object_detector\r\n>> from tflite_support.metadata import metadata_schema_py_generated as _metadata_fb\r\n\r\n4. If the issue persists, you might need to check the documentation or the source code of tflite-support.\r\n", "@yasirulhadi If you read older comments then you will come to know that !pip install tflite-support command is not working that's why I posted this issue.\r\n\r\nIf I use upgrade command with it then again I am facing following error which I posted as first post- `ERROR: Could not build wheels for tflite-support, which is required to install pyproject.toml-based projects`\r\n\r\nAlso as per your suggested command to install tflit-support- `!pip install --no-binary :all: tflite-support`, I am able to see installed version as `tflite-support-0.1.0a1`-\r\n\r\n\r\nI suggest you to please try yourself in kaggle environment and share me a working gist.", "@dsbyprateekg Let me try and work on it again. \r\n", "@dsbyprateekg \r\n\r\nWhat's the current python version in your kaggle environment?\r\n\r\nIf it is 3.10.10 then change it to 3.7. In some environments Tensorflow library doesn't run off with python version higher than 3.7.\r\n\r\nOnce it is changed to 3.7, try to run the above commands I share and let me know.\r\nHopefully it would work! ", "@yasirulhadi we cannot change Python version in kaggle environment.", "@dsbyprateekg \r\nhttps://www.kaggle.com/code/taylorsamarel/change-python-version-on-kaggle-taylor-amarel?scriptVersionId=115764070", "@yasirulhadi instead of changing Python version to older one, can this not be fixed by TF team to support Python 3.10 version?", "@dsbyprateekg \r\nIt is not the issue with tensorflow but with Kaggle environment's python version. Some libraries don't support the upgrade and works well with the downgraded version.", "@dsbyprateekg ,\r\n\r\nCan you report the issue to kaggle team. Its problem with Kaggle environment and we can't support issues with third party tools. Maybe you can refer the pip list as mentioned for checking the Colab environment packages with versions and try to install missing packages in kaggle environment.\r\n\r\n You can use Google Colab for free in case you want to train/test your model.\r\n\r\nThanks!", "@SuryanarayanaY Actually first I tried my TFLite model train/test/convert in Colab only but due to Python 3.10, there I faced issue while installing TFLite model maker library. So I have switched to Kaggle. Now in Kaggle I am able to train model but now issue is with TFLite-support library.", "@dsbyprateekg Were you able to change the python version?", "@yasirulhadi No, I have not.\r\nSo I did it on my personal laptop.", "Hi @dsbyprateekg \r\n\r\nThe tflite-support has has known issue of not running on python 3.10 because of which the model maker installation is hindered in the colab as well.\r\n\r\nYou can try installing `pip install tflite-support==0.4.2` and see if it works.\r\n\r\nThanks.", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60784\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60784\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 1097547538, "node_id": "MDU6TGFiZWwxMDk3NTQ3NTM4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:gpu", "name": "comp:gpu", "color": "0052cc", "default": false, "description": "GPU related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Hi @dark16sider ,\r\n\r\nThanks for reporting issue. Could you please submit a code snippet to replicate and understand the problem. Please confirm whether you are using prebuilt wheel or built from source? If built from source let us know the build command used.\r\n\r\nCould you please confirm whether you have cuDNN installed and what version was installed. For TF2.12 version the recommended configurations are listed below.\r\n\r\nVersion | Python version | Compiler | Build tools | cuDNN | CUDA\r\n-- | -- | -- | -- | -- | --\r\ntensorflow-2.13.0 | 3.8-3.11 | Clang 16.0.0 | Bazel 5.3.0 | 8.6 | 11.8\r\ntensorflow-2.12.0 | 3.8-3.11 | GCC 9.3.1 | Bazel 5.3.0 | 8.6 | 11.8\r\n\r\nAlso please confirm whether path configuration set for CUDA and cuDNN and confirm the commands for same.\r\n\r\nThanks!\r\n", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "It seemed like it was a OOM Error for me. Maybe try decreasing the batch size ?", "Hi @dark16sider ,\r\n\r\nCould you please confirm whether above comments worked for you. If you are getting this at the middle of training yeah there is possibility that it might be happening due to OOM problem also. Since you have not pasted any code snippet which may be custom code I would recommend to please check whether unwanted variables that are initialized are getting deleted when there is no more requirement of those ? Also ensure enough memory resources to remove OOM problem from your program.\r\n\r\nProviding more details like reproducible code snippet can help us to understand your problem better.\r\n\r\nThanks!", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60783\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60783\">No</a>\n" ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1097547147, "node_id": "MDU6TGFiZWwxMDk3NTQ3MTQ3", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:ops", "name": "comp:ops", "color": "0052cc", "default": false, "description": "OPs related issues" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 5508003926, "node_id": "LA_kwDOArmXAs8AAAABSE14Vg", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.13", "name": "TF 2.13", "color": "B13ACB", "default": false, "description": "For issues related to Tensorflow 2.13" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "Hi @15079044843 ,\r\n\r\nCould you please elaborate your issue along with commands you have used. Unfortunately provided logs not enough to understand the problem.\r\n\r\nThanks!", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further.", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60781\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60781\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" }, { "id": 3797168204, "node_id": "LA_kwDOArmXAs7iVDBM", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.8", "name": "TF 2.8", "color": "5DC9D0", "default": false, "description": "" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "Hi @puyiwen \r\n\r\nCould you please share a toy tflite model or code to reproduce the issue? \r\n\r\nThanks.\r\n\r\n", "> Hi @puyiwen\r\n> \r\n> Could you please share a toy tflite model or code to reproduce the issue?\r\n> \r\n> Thanks.\r\n\r\nOK, there is my tflite model.\r\n\r\n[qat_noquant_int8.zip](https://github.com/tensorflow/tensorflow/files/11660722/qat_noquant_int8.zip)\r\n\r\nand there is my onnx model \r\n[checkpoint_best.zip](https://github.com/tensorflow/tensorflow/files/11660732/checkpoint_best.zip)\r\n", "Hi @puyiwen \r\n\r\nThanks for sharing.\r\n\r\nI have observed that Batch MatMul operation is being converted into into Split - multi FullyConnected - Pack. \r\n\r\nTFlite converts model ops into simple [supported operations](https://www.tensorflow.org/lite/guide/ops_compatibility) that can be run in memory constrained devices like Edge devices and micro controllers. \r\n\r\nIf you are facing any issue regarding quantization or want to disable quantization for any node , you can refer [Quantization Debugger](https://www.tensorflow.org/lite/performance/quantization_debugger) and do [Selective Quantization](https://www.tensorflow.org/lite/performance/quantization_debugger#selective_quantization).\r\n\r\nThanks.", "> 你好@puyiwen\r\n> \r\n> 感谢分享。\r\n> \r\n> 我观察到 Batch MatMul 操作正在转换为 Split - multi FullyConnected - Pack。\r\n> \r\n> TFlite 将模型操作转换为简单的[支持操作](https://www.tensorflow.org/lite/guide/ops_compatibility)，这些操作可以在边缘设备和微控制器等内存受限设备中运行。\r\n> \r\n> 如果您遇到有关量化的任何问题或想要禁用任何节点的量化，您可以参考[量化调试器](https://www.tensorflow.org/lite/performance/quantization_debugger)并执行[选择性量化](https://www.tensorflow.org/lite/performance/quantization_debugger#selective_quantization)。\r\n> \r\n> 谢谢。\r\n\r\nThank you for your reply. The MatMul operator is 'nn.Linear' in Pytorch. I am wondering why **nn.Linear** is **Batch MatMul** in tflite. How can I convert the Batch MatMul to norm MatMul? Can you help me?", "Hi @puyiwen \r\n\r\nThe model can be analyzed using [Model Analyzer](https://www.tensorflow.org/lite/guide/model_analyzer) API and it can be observed that multiple MatMul operations are generated. Please find the [gist](https://colab.research.google.com/gist/pjpratik/404adf7abcaefd9e759ef10593d21167/60780.ipynb) here.\r\n\r\nThe TFlite converts model ops into [supported operations](https://www.tensorflow.org/lite/guide/ops_compatibility) that are natively supported by TensorFlow Lite.\r\n\r\nIs there any challenge in using the current conversion in your use case?\r\n\r\nThanks.", "> Hi @puyiwen\r\n> \r\n> The model can be analyzed using [Model Analyzer](https://www.tensorflow.org/lite/guide/model_analyzer) API and it can be observed that multiple MatMul operations are generated. Please find the [gist](https://colab.research.google.com/gist/pjpratik/404adf7abcaefd9e759ef10593d21167/60780.ipynb) here.\r\n> \r\n> The TFlite converts model ops into [supported operations](https://www.tensorflow.org/lite/guide/ops_compatibility) that are natively supported by TensorFlow Lite.\r\n> \r\n> Is there any challenge in using the current conversion in your use case?\r\n> \r\n> Thanks.\r\n\r\nThis kind of splitting will be very slow in reasoning on my deployment platform, and I wonder if it can be merged.\r\nThere is another problem. After my model is converted to tflite and statically quantized to in8, the model file size is larger than before it was not quantized. It is now about 5.6M. The file size of the model before quantization (I sent you earlier) It is about 3.6M. I compared the model structure before and after quantization, and tflite will automatically add a DeQuantize in front of the DIV operator, and then quantize it back after passing the DIV. I don't know if this part of the problem caused the model file to become larger after quantization. I don’t want to have this kind of operation. How should I do it?\r\nIt is my quantized tflite model.\r\n[qat_quant_static_int8.zip](https://github.com/tensorflow/tensorflow/files/11674934/qat_quant_static_int8.zip)", "Hi @puyiwen \r\n\r\nThanks for the information.\r\n\r\nThe converter adds `Quantize` and `DeQuantize` stubs for the ops that do not support int8 quantization, to give floating point view.\r\n\r\nAlso, can you try conversion in latest TF-Nightly and let us know your observation?\r\n\r\nThe converter adds unfolding matmul operation by default which is disabled in nightly version.\r\n\r\nThanks.", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 473184161, "node_id": "MDU6TGFiZWw0NzMxODQxNjE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:support", "name": "type:support", "color": "159b2e", "default": false, "description": "Support issues" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "pkgoogle", "id": 132095473, "node_id": "U_kgDOB9-d8Q", "avatar_url": "https://avatars.githubusercontent.com/u/132095473?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pkgoogle", "html_url": "https://github.com/pkgoogle", "followers_url": "https://api.github.com/users/pkgoogle/followers", "following_url": "https://api.github.com/users/pkgoogle/following{/other_user}", "gists_url": "https://api.github.com/users/pkgoogle/gists{/gist_id}", "starred_url": "https://api.github.com/users/pkgoogle/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pkgoogle/subscriptions", "organizations_url": "https://api.github.com/users/pkgoogle/orgs", "repos_url": "https://api.github.com/users/pkgoogle/repos", "events_url": "https://api.github.com/users/pkgoogle/events{/privacy}", "received_events_url": "https://api.github.com/users/pkgoogle/received_events", "type": "User", "site_admin": false } ]

[ "Hi @Bhuvan-1 \r\n\r\nCurrently, there is no straightforward way to extract all header files needed, so you must include all header files in tensorflow/lite/ from the TensorFlow repository. Additionally, you will need header files from FlatBuffers and Abseil.\r\n\r\nThanks", "@pjpratik , I don't have much idea on Flatbuffers and Abseil, I followed [this](https://www.youtube.com/watch?v=He2p2JLpYC0&ab_channel=EdgeAIGuy) as a reference. Currently I am including flatbuffers & tensorflow headers\r\n\r\nCan you please guide on how to include FlatBuffers & Abseil headers correctly.\r\n\r\nAlso is the command that I used to generate the library correct? or should it be built differently?\r\n\r\nThanks", "Hi @Bhuvan-1, did you follow your reference all the way through? including the CMakeLists.txt and CMake portion? While these steps are not technically required, it seems like this was started with that tool chain in mind, I believe if you follow those steps your includes, compilation, linking, and execution will be smoother.", "Thanks @pkgoogle , I used the CMakeLists.txt from the /lite/examples/minimal and its working smoothly now.", "Hi @Bhuvan-1,\r\n\r\nIf you are satisfied w/ your resolution and have no more open items for this issue, please feel free to close the issue. Thanks!", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60779\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60779\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 474725938, "node_id": "MDU6TGFiZWw0NzQ3MjU5Mzg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stale", "name": "stale", "color": "d4c5f9", "default": false, "description": "This label marks the issue/pr stale - to be closed automatically if no activity" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 1661751498, "node_id": "MDU6TGFiZWwxNjYxNzUxNDk4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TFLiteConverter", "name": "TFLiteConverter", "color": "bfdadc", "default": false, "description": "For issues related to TFLite converter" } ]

[ { "login": "pjpratik", "id": 118897289, "node_id": "U_kgDOBxY6iQ", "avatar_url": "https://avatars.githubusercontent.com/u/118897289?v=4", "gravatar_id": "", "url": "https://api.github.com/users/pjpratik", "html_url": "https://github.com/pjpratik", "followers_url": "https://api.github.com/users/pjpratik/followers", "following_url": "https://api.github.com/users/pjpratik/following{/other_user}", "gists_url": "https://api.github.com/users/pjpratik/gists{/gist_id}", "starred_url": "https://api.github.com/users/pjpratik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/pjpratik/subscriptions", "organizations_url": "https://api.github.com/users/pjpratik/orgs", "repos_url": "https://api.github.com/users/pjpratik/repos", "events_url": "https://api.github.com/users/pjpratik/events{/privacy}", "received_events_url": "https://api.github.com/users/pjpratik/received_events", "type": "User", "site_admin": false } ]

[ "Hi @dsbyprateekg \r\n\r\nFrom the error log, it looks like the TF Model has select ops which are not available in TFLite builtins. \r\n\r\nTo be able to convert the model into TFLite, please add the below flags before conversion.\r\n\r\n```\r\nconverter.target_spec.supported_ops = [\r\n tf.lite.OpsSet.TFLITE_BUILTINS, # enable TensorFlow Lite ops.\r\n tf.lite.OpsSet.SELECT_TF_OPS # enable TensorFlow ops.\r\n]\r\n```\r\nThanks.", "@pjpratik thanks, with the above I am able to proceed.\r\nBut when I checked the input shape of the converted tflite model, it's array([1,1,1,3)] which I don't understand.\r\nCan you please share me Python inference code for TFLite? My original model is based on ssd_mobilenet_v2_fpnlite_640x640_coco17_tpu-8.config. ", "Hi @dsbyprateekg \r\n\r\n If your model has dynamic input shapes and you didn't specify a dimension size when building and converting your model, the converter will use 1 as the default dimension size.\r\n\r\n However, at inference time you can use Interpreter's [ResizeInputTensor](https://www.tensorflow.org/lite/api_docs/python/tf/lite/Interpreter#resize_tensor_input) API to change the input shape of the model\r\n\r\nYou can check this TF Hub example for the inference.\r\n\r\nhttps://hub.tensorflow.google.cn/google/lite-model/object_detection/mobile_object_localizer_v1/1/default/1\r\n\r\nThanks.", "This issue is stale because it has been open for 7 days with no activity. It will be closed if no further activity occurs. Thank you.", "This issue was closed because it has been inactive for 7 days since being marked as stale. Please reopen if you'd like to work on this further." ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 5508003926, "node_id": "LA_kwDOArmXAs8AAAABSE14Vg", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.13", "name": "TF 2.13", "color": "B13ACB", "default": false, "description": "For issues related to Tensorflow 2.13" } ]

[ { "login": "tilakrayal", "id": 81610181, "node_id": "MDQ6VXNlcjgxNjEwMTgx", "avatar_url": "https://avatars.githubusercontent.com/u/81610181?v=4", "gravatar_id": "", "url": "https://api.github.com/users/tilakrayal", "html_url": "https://github.com/tilakrayal", "followers_url": "https://api.github.com/users/tilakrayal/followers", "following_url": "https://api.github.com/users/tilakrayal/following{/other_user}", "gists_url": "https://api.github.com/users/tilakrayal/gists{/gist_id}", "starred_url": "https://api.github.com/users/tilakrayal/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/tilakrayal/subscriptions", "organizations_url": "https://api.github.com/users/tilakrayal/orgs", "repos_url": "https://api.github.com/users/tilakrayal/repos", "events_url": "https://api.github.com/users/tilakrayal/events{/privacy}", "received_events_url": "https://api.github.com/users/tilakrayal/received_events", "type": "User", "site_admin": false } ]

[ "@nVietUK,\r\n`nvidia-docker v2` uses --runtime=nvidia instead of --gpus all. nvidia-docker v1 uses the nvidia-docker alias, rather than the **--runtime=nvidia or --gpus** all command line flags.\r\n\r\nDownload and run a GPU-enabled TensorFlow image:\r\n`docker run --gpus all -it --rm tensorflow/tensorflow:latest-gpu \\`\r\n\r\nUse the latest TensorFlow GPU image to start a bash shell session in the container:\r\n`docker run --gpus all -it tensorflow/tensorflow:latest-gpu bash`\r\n\r\nhttps://www.tensorflow.org/install/docker\r\nThank you!", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60777\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60777\">No</a>\n" ]

[ { "id": 386191887, "node_id": "MDU6TGFiZWwzODYxOTE4ODc=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20response", "name": "stat:awaiting response", "color": "f4b400", "default": false, "description": "Status - Awaiting response from author" }, { "id": 473173272, "node_id": "MDU6TGFiZWw0NzMxNzMyNzI=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:feature", "name": "type:feature", "color": "159b2e", "default": false, "description": "Feature requests" }, { "id": 473184161, "node_id": "MDU6TGFiZWw0NzMxODQxNjE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:support", "name": "type:support", "color": "159b2e", "default": false, "description": "Support issues" }, { "id": 4829271983, "node_id": "LA_kwDOArmXAs8AAAABH9jXrw", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.11", "name": "TF 2.11", "color": "46B4D7", "default": false, "description": "Issues related to TF 2.11" } ]

[ { "login": "SuryanarayanaY", "id": 116063290, "node_id": "U_kgDOBur8Og", "avatar_url": "https://avatars.githubusercontent.com/u/116063290?v=4", "gravatar_id": "", "url": "https://api.github.com/users/SuryanarayanaY", "html_url": "https://github.com/SuryanarayanaY", "followers_url": "https://api.github.com/users/SuryanarayanaY/followers", "following_url": "https://api.github.com/users/SuryanarayanaY/following{/other_user}", "gists_url": "https://api.github.com/users/SuryanarayanaY/gists{/gist_id}", "starred_url": "https://api.github.com/users/SuryanarayanaY/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/SuryanarayanaY/subscriptions", "organizations_url": "https://api.github.com/users/SuryanarayanaY/orgs", "repos_url": "https://api.github.com/users/SuryanarayanaY/repos", "events_url": "https://api.github.com/users/SuryanarayanaY/events{/privacy}", "received_events_url": "https://api.github.com/users/SuryanarayanaY/received_events", "type": "User", "site_admin": false } ]

[ "You can always save a model before, during, and after training using the same `model.save()` API. Once loading, you can always continue training if you so desire. There is no need for a separate API (`save_pretrained`).\r\n\r\nRegarding sharding (`max_shard_size`), there is no such parameter, TF does automatic sharding.\r\n\r\nSee https://www.tensorflow.org/api_docs/python/tf/saved_model/save, https://www.tensorflow.org/api_docs/python/tf/saved_model/SaveOptions and https://www.tensorflow.org/api_docs/python/tf/keras/Model#save", "Hi @SmileTM ,\r\n\r\n\r\nApart from resources mentioned in above, you can also refer to attached tutorial links for end to end examples of creating,training, saving and loading the models.\r\nPlease refer [serialization_and_saving](https://www.tensorflow.org/guide/keras/serialization_and_saving) and [save_and_load](https://www.tensorflow.org/tutorials/keras/save_and_load) also for a demo.\r\n\r\nTensorflow has options to [save](https://www.tensorflow.org/api_docs/python/tf/keras/Model#save) complete model architecture including optimizer state or only [weights](https://www.tensorflow.org/api_docs/python/tf/keras/Model#save_weights) of the layers also.\r\n\r\nIf you still looking for a particular use case please let us know.\r\n\r\nThanks!", "Thanks for the quick response.\r\n\r\n", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60776\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60776\">No</a>\n" ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473173351, "node_id": "MDU6TGFiZWw0NzMxNzMzNTE=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:build/install", "name": "type:build/install", "color": "159b2e", "default": false, "description": "Build and install issues" }, { "id": 1205615612, "node_id": "MDU6TGFiZWwxMjA1NjE1NjEy", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/subtype:%20ubuntu/linux", "name": "subtype: ubuntu/linux", "color": "b619ea", "default": false, "description": "Ubuntu/Linux Build/Installation Issues" }, { "id": 4829271983, "node_id": "LA_kwDOArmXAs8AAAABH9jXrw", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.11", "name": "TF 2.11", "color": "46B4D7", "default": false, "description": "Issues related to TF 2.11" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" }, { "id": 5508003926, "node_id": "LA_kwDOArmXAs8AAAABSE14Vg", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.13", "name": "TF 2.13", "color": "B13ACB", "default": false, "description": "For issues related to Tensorflow 2.13" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "Can you run post the output of `python -m pip list` and `python -c 'import tensorflow'` run from the same environment?", "Pip list:\r\n\r\n\r\nimport tensorflow:\r\n\r\n", "```\r\n_libgcc_mutex 0.1 main \r\n_openmp_mutex 5.1 1_gnu \r\nabsl-py 1.4.0 pypi_0 pypi\r\nastunparse 1.6.3 pypi_0 pypi\r\nca-certificates 2023.5.7 hbcca054_0 conda-forge\r\ncachetools 5.3.1 pypi_0 pypi\r\ncertifi 2023.5.7 pypi_0 pypi\r\ncharset-normalizer 3.1.0 pypi_0 pypi\r\ncudatoolkit 11.8.0 h6a678d5_0 \r\nflatbuffers 23.5.26 pypi_0 pypi\r\ngast 0.4.0 pypi_0 pypi\r\ngoogle-auth 2.19.1 pypi_0 pypi\r\ngoogle-auth-oauthlib 1.0.0 pypi_0 pypi\r\ngoogle-pasta 0.2.0 pypi_0 pypi\r\ngrpcio 1.54.2 pypi_0 pypi\r\nh5py 3.8.0 pypi_0 pypi\r\nidna 3.4 pypi_0 pypi\r\nimportlib-metadata 6.6.0 pypi_0 pypi\r\nkeras-nightly 2.14.0.dev2023060107 pypi_0 pypi\r\nld_impl_linux-64 2.38 h1181459_1 \r\nlibclang 16.0.0 pypi_0 pypi\r\nlibffi 3.4.4 h6a678d5_0 \r\nlibgcc-ng 11.2.0 h1234567_1 \r\nlibgomp 11.2.0 h1234567_1 \r\nlibstdcxx-ng 11.2.0 h1234567_1 \r\nmarkdown 3.4.3 pypi_0 pypi\r\nmarkupsafe 2.1.3 pypi_0 pypi\r\nncurses 6.4 h6a678d5_0 \r\nnumpy 1.24.3 pypi_0 pypi\r\noauthlib 3.2.2 pypi_0 pypi\r\nopenssl 1.1.1t h7f8727e_0 \r\nopt-einsum 3.3.0 pypi_0 pypi\r\npackaging 23.1 pypi_0 pypi\r\npip 23.0.1 py39h06a4308_0 \r\nprotobuf 4.23.2 pypi_0 pypi\r\npyasn1 0.5.0 pypi_0 pypi\r\npyasn1-modules 0.3.0 pypi_0 pypi\r\npython 3.9.16 h7a1cb2a_2 \r\nreadline 8.2 h5eee18b_0 \r\nrequests 2.31.0 pypi_0 pypi\r\nrequests-oauthlib 1.3.1 pypi_0 pypi\r\nrsa 4.9 pypi_0 pypi\r\nsetuptools 67.8.0 py39h06a4308_0 \r\nsix 1.16.0 pypi_0 pypi\r\nsqlite 3.41.2 h5eee18b_0 \r\ntb-nightly 2.14.0a20230603 pypi_0 pypi\r\ntensorboard-data-server 0.7.0 pypi_0 pypi\r\ntensorflow-io-gcs-filesystem 0.32.0 pypi_0 pypi\r\ntermcolor 2.3.0 pypi_0 pypi\r\ntf-estimator-nightly 2.14.0.dev2023060108 pypi_0 pypi\r\ntf-nightly 2.14.0.dev20230603 pypi_0 pypi\r\ntk 8.6.12 h1ccaba5_0 \r\ntyping-extensions 4.5.0 pypi_0 pypi\r\ntzdata 2023c h04d1e81_0 \r\nurllib3 1.26.16 pypi_0 pypi\r\nwerkzeug 2.3.4 pypi_0 pypi\r\nwheel 0.38.4 py39h06a4308_0 \r\nwrapt 1.15.0 pypi_0 pypi\r\nxz 5.4.2 h5eee18b_0 \r\nzipp 3.15.0 pypi_0 pypi\r\nzlib 1.2.13 h5eee18b_0 \r\n\r\n```\r\nConda list just in case", "So, this is using latest `tf-nightly`. I think there is an actual issue in these builds, but they are offered on a best effort basis, there is a rotation that looks at the build and tries to solve issues such as this before a release.\r\n\r\nCan you run the same on an environment where you use 2.13.0?", "@mihaimaruseac \r\nAs i said, I've met this problem using 2.11 and 2.12. Anyway, here is the output of those two commands. I've ran them on tensorflow 2.12 though, not 2.13.0. Pip tells me that there are only 2.13.0rc0, 2.13.0rc1are available. Do you want me to use them?\r\n\r\n\r\n\r\n\r\n\r\n", "Yes, please try `rc1`. Your initial screenshots were showing nightly, that's why I asked for a released version.", "> Your initial screenshots were showing nightly\r\n\r\nThat is because firstly I've tested 2.12 and 2.11, then I've seen that issue template requires testing on tf-nightly and installed it instead of regular tf.\r\nAnyway, here is the outputs with 2.13.0rc1\r\n\r\n\r\n\r\n\r\n", "The missing symbols is `tensorflow::OpKernel::TraceString[abi:cxx11](tensorflow::OpKernelcontext const&, bool) const`\r\n\r\nThis makes me think you are combining wheels with the old C++ ABI and with the new one. Can you try recreating an empty environment and installing there?", "@mihaimaruseac \r\n\r\nThat is already recreated environment. I've deleted it and created again. It just has the same name.", "Does the issue reproduce if you use `venv` instead of *conda?", "Done. Same problem. tf is the venv's name.\r\n\r\n\r\n\r\n\r\n\r\n", "There must be something wrong in your environment, as I cannot reproduce:\r\n\r\n```console\r\nmihaimaruseac@ankh:/tmp$ rm -rf venv && python3 -m venv venv && source venv/bin/activate && pip install -q tensorflow==2.13.0rc1 && python -c \"import tensorflow as tf\"\r\n2023-06-07 10:44:45.071830: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\r\nTo enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\r\n2023-06-07 10:44:46.494478: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\r\n(venv) mihaimaruseac@ankh:/tmp$ \r\n```\r\n\r\n\r\n```console\r\n(venv) mihaimaruseac@ankh:/tmp$ pip list\r\nPackage Version\r\n---------------------------- ---------\r\nabsl-py 1.4.0\r\nastunparse 1.6.3\r\ncachetools 5.3.1\r\ncertifi 2023.5.7\r\ncharset-normalizer 3.1.0\r\nflatbuffers 23.5.26\r\ngast 0.4.0\r\ngoogle-auth 2.19.1\r\ngoogle-auth-oauthlib 1.0.0\r\ngoogle-pasta 0.2.0\r\ngrpcio 1.54.2\r\nh5py 3.8.0\r\nidna 3.4\r\nkeras 2.13.1rc0\r\nlibclang 16.0.0\r\nMarkdown 3.4.3\r\nMarkupSafe 2.1.3\r\nnumpy 1.24.3\r\noauthlib 3.2.2\r\nopt-einsum 3.3.0\r\npackaging 23.1\r\npip 23.0.1\r\nprotobuf 4.23.2\r\npyasn1 0.5.0\r\npyasn1-modules 0.3.0\r\nrequests 2.31.0\r\nrequests-oauthlib 1.3.1\r\nrsa 4.9\r\nsetuptools 66.1.1\r\nsix 1.16.0\r\ntensorboard 2.13.0\r\ntensorboard-data-server 0.7.0\r\ntensorflow 2.13.0rc1\r\ntensorflow-estimator 2.13.0rc0\r\ntensorflow-io-gcs-filesystem 0.32.0\r\ntermcolor 2.3.0\r\ntyping_extensions 4.6.3\r\nurllib3 1.26.16\r\nWerkzeug 2.3.4\r\nwheel 0.40.0\r\nwrapt 1.15.0\r\n```\r\n\r\nYou do have some packages which don't exist on the fresh environment (`zipp` for example). You have 2 `nvidia-*` wheels which also don't exist but could be causing issues. And you have an old version of `setuptools` (try `python -m pip install --upgrade pip setuptools` and try again?)\r\n\r\nMy experiments are with python3.11.2:\r\n\r\n```console\r\n(venv) mihaimaruseac@ankh:/tmp$ python\r\nPython 3.11.2 (main, Feb 12 2023, 00:48:52) [GCC 12.2.0] on linux\r\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\r\n>>> \r\n```", "I've used python 3.9 on conda env, on venv i'm using python 3.8. I've tried what you asked (upgrade setup tools), here are results.\r\n\r\n\r\n\r\n\r\n\r\n", "And preventing move, I've installed python 3.11 and created venv using it. Same.\r\n\r\n\r\n\r\n\r\n\r\n", "~~Note that you skipped over the error in activating the virtual environment.~~ Edit: I reread the screenshot. In general, I'd recommend posting code/errors as formatted text, not images. Makes it easier to quote, cross-reference, search\r\n\r\nComing back to the issue, your last environment seems at a quick glance to be the same as mine. Will probably need to dig deeper into the binary, as it seems to be an issue specifically with your environment", "In the last virtualenvironment, if you do `cd tf/lib/python3.11/site-packages/tensorflow`, what is the output of `nm libtensorflow_cc.so.2 | grep _ZNK10tensorflow8OpKernel`?\r\n\r\nMine is\r\n\r\n```console\r\n...$ nm libtensorflow_cc.so.2 | grep _ZNK10tensorflow8OpKernel\r\n U _ZNK10tensorflow8OpKernel10InputRangeESt17basic_string_viewIcSt11char_traitsIcEEPiS5_\r\n U _ZNK10tensorflow8OpKernel11TraceStringB5cxx11ERKNS_15OpKernelContextEb\r\n000000000476d340 W _ZNK10tensorflow8OpKernel12const_tensorEv\r\n U _ZNK10tensorflow8OpKernel16ShapeTraceStringB5cxx11ERKNS_15OpKernelContextE\r\n```", "Same\r\n\r\n```\r\n U _ZNK10tensorflow8OpKernel10InputRangeESt17basic_string_viewIcSt11char_traitsIcEEPiS5_\r\n U _ZNK10tensorflow8OpKernel11TraceStringB5cxx11ERKNS_15OpKernelContextEb\r\n000000000476d340 W _ZNK10tensorflow8OpKernel12const_tensorEv\r\n U _ZNK10tensorflow8OpKernel16ShapeTraceStringB5cxx11ERKNS_15OpKernelContextE\r\n```\r\n\r\n", "I think I have an idea on what happens, but just to test, can you try `LD_DEBUG=libs python -c \"import tensorflow as tf\" 2>file` and then attach the file? It is a large file, estimated to be several hundreds lines, but should give an idea on what libraries are attempted to be loaded.", "Done\r\n[file.txt](https://github.com/tensorflow/tensorflow/files/11708280/file.txt)\r\n", "Hi, I am experiencing the same issue: I am running inside pyenv virtualenv with venv name tensorflow.\r\n```\r\npython\r\nPython 3.11.4 (main, Jul 22 2023, 12:30:49) [GCC 11.3.0] on linux\r\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\r\n>>> import tensorflow as tf\r\nTraceback (most recent call last):\r\n File \"<stdin>\", line 1, in <module>\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/__init__.py\", line 38, in <module>\r\n from tensorflow.python.tools import module_util as _module_util\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/__init__.py\", line 36, in <module>\r\n from tensorflow.python import pywrap_tensorflow as _pywrap_tensorflow\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/pywrap_tensorflow.py\", line 26, in <module>\r\n self_check.preload_check()\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/platform/self_check.py\", line 63, in preload_check\r\n from tensorflow.python.platform import _pywrap_cpu_feature_guard\r\nImportError: /home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/platform/../../libtensorflow_cc.so.2: undefined symbol: _ZNK10tensorflow8OpKernel11TraceStringB5cxx11ERKNS_15OpKernelContextEb\r\n```\r\n\r\nHere is my environment:\r\n\r\n```\r\npython -m pip list\r\nPackage Version\r\n---------------------------- ---------\r\nabsl-py 1.4.0\r\nastunparse 1.6.3\r\ncachetools 5.3.1\r\ncertifi 2023.7.22\r\ncharset-normalizer 3.2.0\r\nflatbuffers 23.5.26\r\ngast 0.4.0\r\ngoogle-auth 2.22.0\r\ngoogle-auth-oauthlib 1.0.0\r\ngoogle-pasta 0.2.0\r\ngrpcio 1.56.2\r\nh5py 3.9.0\r\nidna 3.4\r\nkeras 2.13.1\r\nlibclang 16.0.6\r\nMarkdown 3.4.3\r\nMarkupSafe 2.1.3\r\nnumpy 1.24.3\r\noauthlib 3.2.2\r\nopt-einsum 3.3.0\r\npackaging 23.1\r\npip 23.2.1\r\nprotobuf 4.23.4\r\npyasn1 0.5.0\r\npyasn1-modules 0.3.0\r\nrequests 2.31.0\r\nrequests-oauthlib 1.3.1\r\nrsa 4.9\r\nsetuptools 65.5.0\r\nsix 1.16.0\r\ntensorboard 2.13.0\r\ntensorboard-data-server 0.7.1\r\ntensorflow 2.13.0\r\ntensorflow-estimator 2.13.0\r\ntensorflow-io-gcs-filesystem 0.32.0\r\ntermcolor 2.3.0\r\ntyping_extensions 4.5.0\r\nurllib3 1.26.16\r\nWerkzeug 2.3.6\r\nwheel 0.41.0\r\nwrapt 1.15.0\r\n```\r\nMachine:\r\n```\r\nsudo uname -a\r\nLinux johannes-guest-room 5.19.0-46-generic #47~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jun 21 15:35:31 UTC 2 x86_64 x86_64 x86_64 GNU/Linux\r\n```\r\nOutput of LD_DEBUG=libs python -c \"import tensorflow as tf\" 2>file.txt\r\n[file.txt](https://github.com/tensorflow/tensorflow/files/12136444/file.txt)\r\n\r\nHip", "Hi, I am experiencing the same issue: I am running inside pyenv virtualenv with venv name tensorflow.\r\n```\r\npython\r\nPython 3.11.4 (main, Jul 22 2023, 12:30:49) [GCC 11.3.0] on linux\r\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\r\n>>> import tensorflow as tf\r\nTraceback (most recent call last):\r\n File \"<stdin>\", line 1, in <module>\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/__init__.py\", line 38, in <module>\r\n from tensorflow.python.tools import module_util as _module_util\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/__init__.py\", line 36, in <module>\r\n from tensorflow.python import pywrap_tensorflow as _pywrap_tensorflow\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/pywrap_tensorflow.py\", line 26, in <module>\r\n self_check.preload_check()\r\n File \"/home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/platform/self_check.py\", line 63, in preload_check\r\n from tensorflow.python.platform import _pywrap_cpu_feature_guard\r\nImportError: /home/johannes/.pyenv/versions/tensorflow/lib/python3.11/site-packages/tensorflow/python/platform/../../libtensorflow_cc.so.2: undefined symbol: _ZNK10tensorflow8OpKernel11TraceStringB5cxx11ERKNS_15OpKernelContextEb\r\n```\r\n\r\nHere is my environment:\r\n\r\n```\r\npython -m pip list\r\nPackage Version\r\n---------------------------- ---------\r\nabsl-py 1.4.0\r\nastunparse 1.6.3\r\ncachetools 5.3.1\r\ncertifi 2023.7.22\r\ncharset-normalizer 3.2.0\r\nflatbuffers 23.5.26\r\ngast 0.4.0\r\ngoogle-auth 2.22.0\r\ngoogle-auth-oauthlib 1.0.0\r\ngoogle-pasta 0.2.0\r\ngrpcio 1.56.2\r\nh5py 3.9.0\r\nidna 3.4\r\nkeras 2.13.1\r\nlibclang 16.0.6\r\nMarkdown 3.4.3\r\nMarkupSafe 2.1.3\r\nnumpy 1.24.3\r\noauthlib 3.2.2\r\nopt-einsum 3.3.0\r\npackaging 23.1\r\npip 23.2.1\r\nprotobuf 4.23.4\r\npyasn1 0.5.0\r\npyasn1-modules 0.3.0\r\nrequests 2.31.0\r\nrequests-oauthlib 1.3.1\r\nrsa 4.9\r\nsetuptools 65.5.0\r\nsix 1.16.0\r\ntensorboard 2.13.0\r\ntensorboard-data-server 0.7.1\r\ntensorflow 2.13.0\r\ntensorflow-estimator 2.13.0\r\ntensorflow-io-gcs-filesystem 0.32.0\r\ntermcolor 2.3.0\r\ntyping_extensions 4.5.0\r\nurllib3 1.26.16\r\nWerkzeug 2.3.6\r\nwheel 0.41.0\r\nwrapt 1.15.0\r\n```\r\nMachine:\r\n```\r\nsudo uname -a\r\nLinux johannes-guest-room 5.19.0-46-generic #47~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jun 21 15:35:31 UTC 2 x86_64 x86_64 x86_64 GNU/Linux\r\n```\r\nOutput of LD_DEBUG=libs python -c \"import tensorflow as tf\" 2>file.txt\r\n[file.txt](https://github.com/tensorflow/tensorflow/files/12136444/file.txt)\r\n\r\nHip\r\n\r\nAdditionally I tested a few other things but with no success:\r\n\r\nI noticed I had the ubuntu cuda toolkit installed which was 11 but the driver said 12. So I installed the latest 12.2 toolkit, but to no avail. I also Upgraded to a recent mainline kernel version 6.3.13, this also did not help.\r\nI am running out of ideas.", "Ha, found it. Another installation had `tensorflow_framework.so` put in the library path. To find this take the path after `Import Error:` and put `ldd` in front of it. It will show you the offending library. Remove it in any way you deem appropriate.\r\nIn my case it was wasmedge which was the culprit. @DaddyWesker This is likely a problem with your installation, too. Hih.", "@jmfrank63 Thanks, I will try.", "@jmfrank63 \r\nOkay, so which one of those libs i need to delete exactly? \r\n\r\n", "Hi DaddyWesker,\r\n\r\n For me it was libtensorflow_framework. Check if\r\n/lib/x86_64-linux-gnu/libtensorflow_framework.so.2 exists. Instead of\r\ndeleting, renaming is better, as you can undo it.\r\n\r\nhih Johannes\r\n\r\nDipl.-Ing. (FH) Johannes-Maria Frank\r\nMaster CS UCL Birkbeck\r\n30 Craster Square\r\nNewcastle upon Tyne\r\nNE3 3PL\r\nUnited Kingdom\r\nphone: +447400987848\r\ne-mail: ***@***.***\r\n\r\n\r\n\r\nOn Mon, Jul 24, 2023 at 11:19 AM DaddyWesker ***@***.***>\r\nwrote:\r\n\r\n> @jmfrank63 <https://github.com/jmfrank63>\r\n> Okay, so which one of those libs i need to delete exactly?\r\n> [image: image]\r\n> <https://user-images.githubusercontent.com/10774222/255560808-426f8dc5-cfe1-415c-ab88-6876396b3876.png>\r\n>\r\n> —\r\n> Reply to this email directly, view it on GitHub\r\n> <https://github.com/tensorflow/tensorflow/issues/60775#issuecomment-1647629818>,\r\n> or unsubscribe\r\n> <https://github.com/notifications/unsubscribe-auth/AAXSQHQOWIELOKD5GDSET6TXRZD3HANCNFSM6AAAAAAYZZZW7E>\r\n> .\r\n> You are receiving this because you were mentioned.Message ID:\r\n> ***@***.***>\r\n>\r\n", "Well, for some reason renaming \r\n/lib/libtensorflow_framework.so.2\r\ndoes made a trick. At least tf is importing now and examples like \r\n```\r\npython3 -c \"import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))\"\r\npython3 -c \"import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))\"\r\n```\r\nworks well. I need to check it on some sort of NN. If everything is fine, then I'll close the issue. But still, why this so file is generating then when installing TF at the first place?..", "Thanks for debugging the issue! It seems this is a new side effect that arises when you mix global TF installations with installs in a virtual environment. We recommend all TF installs should be done in a venv to prevent these contagions.\r\n\r\nTF has a large C++ component that needs to be shipped with the wheel in order for users to use it. We could have localized the shared lib (`.so` file) to within the wheel (like smaller `.so`s are), but `libtensorflow_framework` is needed also if you want to do TF from other languages. So, for example, if you want to use TF from, say, Rust, you'd install the wheel, then tell your build process to link against this `.so` (and a few others).\r\n\r\nThere is no standard channel for distributing C++ artifacts, so it was decided to distribute them via the wheel and a GCS bucket. Recently, to reduce the number of builds, the GCS bucket was discontinued, so now users can get the shared objects only via the wheel.\r\n\r\nOverall, this issue points to a new debugging item for similar issues: we should ask users to confirm whether they have global TF installations or everything is in a venv. ", "Are you satisfied with the resolution of your issue?\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=Yes&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60775\">Yes</a>\n<a href=\"https://docs.google.com/forms/d/e/1FAIpQLSfaP12TRhd9xSxjXZjcZFNXPGk4kc1-qMdv3gc6bEP90vY1ew/viewform?entry.85265664=No&entry.2137816233=https://github.com/tensorflow/tensorflow/issues/60775\">No</a>\n" ]

[ { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).\n\nView this [failed invocation](https://github.com/tensorflow/tensorflow/pull/60774/checks?check_run_id=13985733590) of the CLA check for more information.\n\nFor the most up to date status, view the checks section at the bottom of the pull request." ]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 473172988, "node_id": "MDU6TGFiZWw0NzMxNzI5ODg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:bug", "name": "type:bug", "color": "159b2e", "default": false, "description": "Bug" }, { "id": 1478826728, "node_id": "MDU6TGFiZWwxNDc4ODI2NzI4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:core", "name": "comp:core", "color": "024391", "default": false, "description": "issues related to core part of tensorflow" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[]

[ { "id": 404586594, "node_id": "MDU6TGFiZWw0MDQ1ODY1OTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/stat:awaiting%20tensorflower", "name": "stat:awaiting tensorflower", "color": "f4b400", "default": false, "description": "Status - Awaiting response from tensorflower" }, { "id": 1097547147, "node_id": "MDU6TGFiZWwxMDk3NTQ3MTQ3", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:ops", "name": "comp:ops", "color": "0052cc", "default": false, "description": "OPs related issues" }, { "id": 1463677878, "node_id": "MDU6TGFiZWwxNDYzNjc3ODc4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/type:performance", "name": "type:performance", "color": "159b2e", "default": false, "description": "Performance Issue" }, { "id": 5206407904, "node_id": "LA_kwDOArmXAs8AAAABNlN64A", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/TF%202.12", "name": "TF 2.12", "color": "c5def5", "default": false, "description": "For issues related to Tensorflow 2.12" } ]

[ { "login": "sachinprasadhs", "id": 73069040, "node_id": "MDQ6VXNlcjczMDY5MDQw", "avatar_url": "https://avatars.githubusercontent.com/u/73069040?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sachinprasadhs", "html_url": "https://github.com/sachinprasadhs", "followers_url": "https://api.github.com/users/sachinprasadhs/followers", "following_url": "https://api.github.com/users/sachinprasadhs/following{/other_user}", "gists_url": "https://api.github.com/users/sachinprasadhs/gists{/gist_id}", "starred_url": "https://api.github.com/users/sachinprasadhs/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sachinprasadhs/subscriptions", "organizations_url": "https://api.github.com/users/sachinprasadhs/orgs", "repos_url": "https://api.github.com/users/sachinprasadhs/repos", "events_url": "https://api.github.com/users/sachinprasadhs/events{/privacy}", "received_events_url": "https://api.github.com/users/sachinprasadhs/received_events", "type": "User", "site_admin": false } ]

[ "Hi @ngc92 ,\r\n\r\nI have replicated the reported behaviour on CPU runtime as per attached [cpu-gist](https://colab.research.google.com/gist/SuryanarayanaY/39d9263b1b84cd59ddad059140e29ae6/60772_cpu.ipynb).\r\n\r\nBut with GPU runtime the case is totally different. With `uint32` as dtype actually it is taking much more time than `int64` since some Ops with `unit32` as dtype are not supported on GPU . It seems this API implementation supports `int64` for best performance. I tried with `int32` also where performance is not good at all compared to `int64` even though all Ops ran on GPU only. I have attached all the results in the attached [gpu-gist](https://colab.research.google.com/gist/SuryanarayanaY/2d9a40aeb2c35f2ae8363b8d58c8a16e/60772_gpu.ipynb).\r\n\r\nPlease go through the results and shave your views. Thanks!\r\n", "@SuryanarayanaY \r\nThanks for the replication. \r\n\r\nFirst, your results seem to confirm that skipping the redundant comparison with zero leads to a measureable improvement in performance in both cases, so that should be a relatively straightforward improvement.\r\n\r\nThe missing GPU kernels make the other improvements more tricky, though.\r\n\r\nI've now run the current `count_nonzeros` implementation also on GPU under the tf profiler, with the following results:\r\n```\r\nKernel Name Op Name Total Avg Min Max\r\nCast_GPU_DT_BOOL_DT_INT64_kernel count_nonzero/Cast 12,488 4,162 4,018 4,303\r\ncub::DeviceSegmentedReduceKernel count_nonzero/Sum 9,973 3,324 3,249 3,470\r\nNotEqual_GPU_DT_BOOL_DT_BOOL_kernel count_nonzero/NotEqual 7,768 2,589 2,429 2,907\r\n```\r\nAs you can see, the conversion to int64 actually takes *more* time than the reduction itself." ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "I am able to compile locally w/ these changes" ]

[ { "id": 750616506, "node_id": "MDU6TGFiZWw3NTA2MTY1MDY=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/comp:lite", "name": "comp:lite", "color": "0052cc", "default": false, "description": "TF Lite related issues" }, { "id": 987666414, "node_id": "MDU6TGFiZWw5ODc2NjY0MTQ=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/ready%20to%20pull", "name": "ready to pull", "color": "2cd643", "default": false, "description": "PR ready for merge process" }, { "id": 1169364259, "node_id": "MDU6TGFiZWwxMTY5MzY0MjU5", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:XS", "name": "size:XS", "color": "adafea", "default": false, "description": "CL Change Size: Extra Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[ "Hi @terryheo Can you please review this PR ? Thank you!" ]

[ { "id": 390482148, "node_id": "MDU6TGFiZWwzOTA0ODIxNDg=", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/awaiting%20review", "name": "awaiting review", "color": "bc3869", "default": false, "description": "Pull request awaiting review" }, { "id": 1169364458, "node_id": "MDU6TGFiZWwxMTY5MzY0NDU4", "url": "https://api.github.com/repos/tensorflow/tensorflow/labels/size:S", "name": "size:S", "color": "adafea", "default": false, "description": "CL Change Size: Small" } ]

[ { "login": "gbaned", "id": 48215717, "node_id": "MDQ6VXNlcjQ4MjE1NzE3", "avatar_url": "https://avatars.githubusercontent.com/u/48215717?v=4", "gravatar_id": "", "url": "https://api.github.com/users/gbaned", "html_url": "https://github.com/gbaned", "followers_url": "https://api.github.com/users/gbaned/followers", "following_url": "https://api.github.com/users/gbaned/following{/other_user}", "gists_url": "https://api.github.com/users/gbaned/gists{/gist_id}", "starred_url": "https://api.github.com/users/gbaned/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/gbaned/subscriptions", "organizations_url": "https://api.github.com/users/gbaned/orgs", "repos_url": "https://api.github.com/users/gbaned/repos", "events_url": "https://api.github.com/users/gbaned/events{/privacy}", "received_events_url": "https://api.github.com/users/gbaned/received_events", "type": "User", "site_admin": false } ]

[]