Insights: tensorflow/tensorflow

23 Pull requests merged by 15 people

• Update Relese.md

  #60033 merged Mar 17, 2023

• [NVIDIA TF] CollectiveAllToAllV2 + NCCL all-to-all fixes

  #60001 merged Mar 17, 2023

• Clarifying language in the Breaking Changes section in RELEASE.md

  #60017 merged Mar 17, 2023

• [NVIDIA TF] Update cudnn frontend to v0.8

  #59675 merged Mar 17, 2023

• Generic Pattern Matching for FP8 GEMMs in XLA

  #59971 merged Mar 16, 2023

• Fix the endianness issue in v1 frozen graphs in python:lite_test on BE machines

  #58601 merged Mar 16, 2023

• Fix release notes

  #60016 merged Mar 16, 2023

• Disable a test that results in OOM+segfault

  #60014 merged Mar 16, 2023

• [NVIDIA XLA:GPU] Do not show the C stack trace when there is a too old cudnn version.

  #59915 merged Mar 16, 2023

• convert_image_dtype to handle complex dtype

  #59313 merged Mar 15, 2023

• Don't link tensorflow-lite against gemmlowp

  #58678 merged Mar 15, 2023

• Updating Release.md with Legal Language for Release Notes

  #60000 merged Mar 15, 2023

• Fix bad cherrypick

  #59998 merged Mar 15, 2023

• [NVIDIA TF] Add remaining GPU BF16 cwise ops

  #59972 merged Mar 15, 2023

• Add big-endian support to TFLite FlatBuffers

  #58494 merged Mar 15, 2023

• Fix 2.11 build

  #59992 merged Mar 14, 2023

• Use llvm-raw override for build w/ local LLVM repo

  #59942 merged Mar 14, 2023

• Remove references to the Bazel flag `--distinct_host_configuration`.

  #59955 merged Mar 14, 2023

• Update version numbers for TensorFlow 2.12.0

  #59988 merged Mar 14, 2023

• Adds Python 3.11 support to RELEASE.md

  #59977 merged Mar 14, 2023

• Update version numbers for TensorFlow 2.11.1

  #59978 merged Mar 14, 2023

• Update release notes for TensorFlow 2.11.1

  #59976 merged Mar 14, 2023

• [ROCm] add reserved gpu memory for rocm libraries with overflow check

  #59945 merged Mar 13, 2023

22 Pull requests opened by 19 people

• Update "Hello World" Example Doc and Fix Broken Links

  #59969 opened Mar 13, 2023

• Fix the broken links in overview.md and Fixing typo error in Calibrator.py

  #59974 opened Mar 13, 2023

• Update audio_classification/overview.md

  #59975 opened Mar 13, 2023

• Add big-endian conversion data files

  #59989 opened Mar 14, 2023

• Improved readability of TensorFlow description.

  #59991 opened Mar 14, 2023

• Add tf_gpu_tests_tags to gpu-related tests.

  #59993 opened Mar 15, 2023

• Add flag to select enabled CPU targets

  #59995 opened Mar 15, 2023

• [oneDNN] Fixing a failure in mkl_fused_batch_norm_op_test

  #59999 opened Mar 15, 2023

• [NVIDIA XLA:GPU] Fused MHA Support in XLA GPU: SE

  #60002 opened Mar 15, 2023

• build: use LINK_COMPONENTS for linking LLVMSupport

  #60003 opened Mar 15, 2023

• [tosa] Legalize tfl.Rsqrt to tosa.Rsqrt

  #60005 opened Mar 16, 2023

• Update label_image.py to be compatible with TF 2.x

  #60006 opened Mar 16, 2023

• Fix documentation of tflite interpreter's num_threads parameter

  #60010 opened Mar 16, 2023

• lite: Update CMakeLists.txt

  #60015 opened Mar 16, 2023

• [Intel oneDNN][Bug Fix] fix unit test failure with framework::function test

  #60018 opened Mar 16, 2023

• Fixes to build and run TFLite evaluation ios app

  #60019 opened Mar 16, 2023

• [NVIDIA XLA] Vector BiasAdd Epilogue Fusion for FP8 GEMMs

  #60020 opened Mar 16, 2023

• lite: Add config option to enable label_image example

  #60021 opened Mar 16, 2023

• Remove redundant `static_assert`s

  #60025 opened Mar 16, 2023

• Reduce MKL overheads on small shapes by not rewriting node to use MKL

  #60026 opened Mar 16, 2023

• Remove not needed build file

  #60038 opened Mar 19, 2023

• Improve readability of applied patch files

  #60039 opened Mar 19, 2023

55 Issues closed by 31 people

• Amount of RAM used

  #59771 closed Mar 20, 2023

• error with bazel - trying to build windows version of tensorflow for visual studio

  #59810 closed Mar 20, 2023

• Provide XLA as a standalone library

  #50681 closed Mar 19, 2023

• Saving Custom Model Not Working

  #60022 closed Mar 19, 2023

• Tensoflow Installation Problem

  #60037 closed Mar 19, 2023

• Library not loaded: @rpath/libgpr.20.dylib Rosetta2 conda env

  #59778 closed Mar 19, 2023

• assertAlmostEqual raise confusing error message when test fails

  #59789 closed Mar 19, 2023

• bug feedback

  #60028 closed Mar 18, 2023

• tf.data.Dataset is much slower than Python generator producing the same data

  #59744 closed Mar 18, 2023

• Failing to run hexagon delegates on android

  #59817 closed Mar 18, 2023

• mixed_float16 in Keras example stops the model to learn

  #59826 closed Mar 18, 2023

• one tflite model padding layer , the data is error(flatbuffers)

  #59845 closed Mar 18, 2023

• How to find accuracy of a pretrained tflite model

  #60023 closed Mar 17, 2023

• tf-nightly [2.13.0.dev20230316]: module 'tensorflow' has no attribute 'version'

  #60024 closed Mar 17, 2023

• RuntimeError: tensorflow/lite/kernels/elementwise.cc:88 Type INT16 is unsupported by op Rsqrt.Node number 34 (RSQRT) failed to prepare.Failed to apply the default TensorFlow Lite delegate indexed at 0.

  #59506 closed Mar 17, 2023

• Issue created for Rollback of PR #59233: Fix crash in stateless_random_binomial when shape type is int64

  #59300 closed Mar 17, 2023

• Extending tf.Tensor class

  #59472 closed Mar 17, 2023

• Overflow error when running depthwise_conv2d_native_backprop_filter

  #59059 closed Mar 17, 2023

• Process get killed when running tensorflow.python.ops.gen_math_ops.floor_div_eager_fallback

  #59376 closed Mar 17, 2023

• Cannot import tensorflow after latest tensorboard release: TypeError: Descriptors cannot not be created directly

  #59665 closed Mar 17, 2023

• How much FPS in iOS (Swift)?

  #58656 closed Mar 17, 2023

• Selectively build TFLite for IOS failure

  #59853 closed Mar 17, 2023

• Layer input_spec must be an instance of InputSpec!

  #59863 closed Mar 17, 2023

• can substitute boringssl with openssl?

  #60009 closed Mar 17, 2023

• Android digit_classifier not working at all

  #57064 closed Mar 17, 2023

• Add support of float in lite/kernels/kernel_util.cc#CalculateActivationRangeQuantized

  #39487 closed Mar 17, 2023

• why the scales in quantization are floats not in the format of m/2^n

  #44247 closed Mar 17, 2023

• Failed to convert QAT model to tflite when I use tf.lite.OpsSet.EXPERIMENTAL_TFLITE_BUILTINS_ACTIVATIONS_INT16_WEIGHTS_INT8

  #47746 closed Mar 17, 2023

• tflite esrgan super image resolution limited 50x50

  #49961 closed Mar 17, 2023

• Index out of bounds error

  #51070 closed Mar 17, 2023

• tensorflow build failed due to '*.so' file dlopen.

  #59769 closed Mar 17, 2023

• No wheels for tf-nightly Macos arm64

  #59801 closed Mar 17, 2023

• Encounter segment error when trying to deploy tflite model

  #59846 closed Mar 17, 2023

• ERROR: tensorflow/tensorflow/compiler/xla/service/gpu/runtime/BUILD:146:11: in deps attribute of cc_library rule //tensorflow/compiler/xla/service/gpu/runtime:gemm: target '//tensorflow/compiler/xla/service/gpu:non_atomically_upgradeable_rw_lock' does not exist

  #59878 closed Mar 17, 2023

• New

  #60027 closed Mar 17, 2023

• Getting error when calculating entropy for each images in the batch in the input tensor in a custom layer in tensorflow/Keras.

  #59899 closed Mar 16, 2023

• tr1

  #60008 closed Mar 16, 2023

• segment max does not output zero for empty segment

  #58123 closed Mar 16, 2023

• CUDA_ERROR_ILLEGAL_ADDRESS

  #59671 closed Mar 16, 2023

• Question about tf.image.convert_image_dtype parameter type?

  #58699 closed Mar 15, 2023

• S3_USE_HTTPS 0 is being ignored - TF insists on HTTPS

  #59990 closed Mar 15, 2023

• Wrong metric calculations for masked out timeseries with sample weights

  #59934 closed Mar 15, 2023

• can't load saved Convnext model

  #59767 closed Mar 15, 2023

• //tensorflow/core/distributed_runtime/integration_test:c_api_session_coordination_test_cpu fails on manylinux2014_aarch64

  #56402 closed Mar 15, 2023

• Missing Operations for TfLite GPU Delegate

  #34679 closed Mar 15, 2023

• Process get killed tensorflow.python.ops.gen_stateless_random_ops.stateless_random_uniform_full_int

  #59351 closed Mar 15, 2023

• ParameterServerStrategy supporting GPU workers

  #47239 closed Mar 15, 2023

• Tensorflow optimizer.apply_gradients is very slow.

  #59961 closed Mar 15, 2023

• Fails to build with llvm-project repository override

  #59739 closed Mar 14, 2023

• [tensorflow-intel] Missing p39 package for tf-intel 2.10.1 on PyPi

  #59967 closed Mar 14, 2023

• Test issue no need to take any action.

  #59793 closed Mar 14, 2023

• Feature request: tf.gather which returns 0 on invalid indices

  #59724 closed Mar 13, 2023

• tensorflow-cpu-aws 2.12.0rc0 only has the Python 3.11 wheel

  #59731 closed Mar 13, 2023

• Outdated example

  #59900 closed Mar 13, 2023

• tf-nightly wheels missing

  #59489 closed Mar 13, 2023

25 Issues opened by 23 people

• Missing module

  #60040 opened Mar 19, 2023

• A CUDA_ERROR_MISALIGNED_ADDRESS/CUDA_ERROR_ILLEGAL_ADDRESS can be triggered in SparseSplit

  #60036 opened Mar 18, 2023

• A check fail can be triggered in MatrixSquareRoot

  #60035 opened Mar 18, 2023

• A check fail can be triggered in CholeskyGrad

  #60034 opened Mar 18, 2023

• failed to expand 'RETURN_IF_ROCTRACER_ERROR', it is invalid to use a preprocessor directive as macro parameter

  #60032 opened Mar 17, 2023

• Build failure: enumeration value ‘kFP8’ not handled in switch [-Werror=switch]

  #60031 opened Mar 17, 2023

• About Android calling TensorFlow Lite crash problem

  #60030 opened Mar 17, 2023

• pb file, StridedSlice op, attribute begin, Tensor data is empty

  #60029 opened Mar 17, 2023

• Not able to build TFLite GPU Delegate dynamic library with bazel for android arm64

  #60012 opened Mar 16, 2023

• Include tf.IsInf on Tensorflow Lite Operators

  #60011 opened Mar 16, 2023

• bug feedback

  #60007 opened Mar 16, 2023

• Cant assign trainable attribute to variables ->AttributeError: can't set attribute

  #60004 opened Mar 15, 2023

• How to benchmark TFLite object detection model

  #59997 opened Mar 15, 2023

• Strange Autograph Error

  #59996 opened Mar 15, 2023

• LSTM model on arduino nano ble 33 using TFLite

  #59994 opened Mar 15, 2023

• Using Teachable Machine to make a Audio model for Python

  #59985 opened Mar 14, 2023

• v2.11.0 build failure on linux/arm64

  #59984 opened Mar 14, 2023

• TFlite model has bad performance

  #59983 opened Mar 14, 2023

• `nan` while doing inference in C++ with half precision

  #59982 opened Mar 14, 2023

• TypeError: VariableMetaclass._variable_v1_call() got an unexpected keyword argument 'experimental_enable_variable_lifting'

  #59981 opened Mar 14, 2023

• Example android app show "GPU is not supportted".

  #59980 opened Mar 14, 2023

• tf.image.extract_patches fails to compute gradients

  #59979 opened Mar 14, 2023

• Tensorflow Lite Build issue - libtensorflow_framework.so.2: cannot open shared object file

  #59973 opened Mar 13, 2023

• armeabi-v7a assembler error

  #59970 opened Mar 13, 2023

• I don't get the label.txt when training the model.txt which causes me to fail to convert to tflite

  #59968 opened Mar 13, 2023

182 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• //tensorflow/tools/api/tests:api_compatibility_test fails with protobuf > 3.20.3

  #59932 commented on Mar 18, 2023 • 13 new comments

• [tosa] Add math.sign legalization

  #58785 commented on Mar 17, 2023 • 9 new comments

• third_party/icu/data missing big-endian conversion data files

  #59935 commented on Mar 16, 2023 • 7 new comments

• some errors

  #59883 commented on Mar 17, 2023 • 6 new comments

• BatchToSpaceND and SpaceToBatchND ERROR_GPU_NOT_COMPATIBLE

  #59870 commented on Mar 15, 2023 • 5 new comments

• Errors building Tensorflow from source

  #59943 commented on Mar 16, 2023 • 5 new comments

• Inexistant CUDA support for last version ?

  #59905 commented on Mar 18, 2023 • 5 new comments

• Fix LSTMBlockCell Check Fail

  #59534 commented on Mar 19, 2023 • 5 new comments

• [tosa][fix] align scale_n of resize to the spec for input size of 1 case

  #59914 commented on Mar 17, 2023 • 5 new comments

• Bad CRC while building TensorFlowLiteSelectTfOps_framework

  #59721 commented on Mar 16, 2023 • 4 new comments

• Tensorflow Does Not Seem to Work With Latest cuDNN 8.8 Version

  #59966 commented on Mar 16, 2023 • 4 new comments

• step 4 error

  #59815 commented on Mar 17, 2023 • 4 new comments

• Allow selecting enabled targets

  #59595 commented on Mar 17, 2023 • 4 new comments

• Description of Build Order and bootstrapped builds

  #53582 commented on Mar 14, 2023 • 3 new comments

• Can't convert openimages_v4/ssd/mobilenet_v2

  #59965 commented on Mar 15, 2023 • 3 new comments

• To use the argument `sampling_rate`, you should install tensorflow_io. You can install it via `pip install tensorflow-io`

  #58762 commented on Mar 16, 2023 • 3 new comments

• Bfloat16 on GPU for sigmoid/swish activations, dropout and LSTM layers

  #59743 commented on Mar 16, 2023 • 3 new comments

• A check fail can be triggered in LSTMBlockCell

  #58270 commented on Mar 17, 2023 • 3 new comments

• Tensorflow Import error

  #59814 commented on Mar 17, 2023 • 3 new comments

• tf.io.FixedLenSequenceFeature shows error when padding value for unformant serialized features

  #59950 commented on Mar 13, 2023 • 2 new comments

• unable to use my classifiers.tflite and label.txt. Please help

  #59807 commented on Mar 14, 2023 • 2 new comments

• Tensorflow2.11 build error /root/tensorflow/tensorflow/lite/python/BUILD:68:10 Middleman _middlemen/tensorflow_Slite_Spython_Stflite_Uconvert-runfiles failed:

  #59940 commented on Mar 14, 2023 • 2 new comments

• Tensorflow lite build error for Aarch64

  #59824 commented on Mar 14, 2023 • 2 new comments

• TensorFlow 2.7 does not detect CUDA installed through conda

  #52988 commented on Mar 14, 2023 • 2 new comments

• tf.nn.learned_unigram_candidate_sampler function causes silent breakdown

  #59752 commented on Mar 14, 2023 • 2 new comments

• Cannot save model when not giving a name in Layer.add_weight call in "build"

  #59954 commented on Mar 14, 2023 • 2 new comments

• disable linkage to nativewindow when API_LEVEL<26

  #59861 commented on Mar 14, 2023 • 2 new comments

• Resize OP doesn't work in Metal delegate with dynamic shape input in TFLite

  #59911 commented on Mar 14, 2023 • 2 new comments

• crosstool_wrapper_driver_is_not_gcc error while building TF v2.11

  #59924 commented on Mar 16, 2023 • 2 new comments

• Support Python 3.11

  #58032 commented on Mar 16, 2023 • 2 new comments

• Problems with converted 8-bit TFLite models of CycleGAN and running inference (specially allocating tensors)

  #59922 commented on Mar 17, 2023 • 2 new comments

• Error to train or load data on RTX 4080

  #59897 commented on Mar 17, 2023 • 2 new comments

• Post-training quantization same model get different result for tf.concat

  #45330 commented on Mar 17, 2023 • 2 new comments

• ValueError: setting an array element with a sequence.

  #59912 commented on Mar 17, 2023 • 2 new comments

• Building libtensorflowlite_gpu_delegate.so with bazel raise an error.

  #59794 commented on Mar 17, 2023 • 2 new comments

• //tensorflow/python/kernel_tests/proto:decode_proto_op_test fails with protobuf > 3.20.3

  #59930 commented on Mar 18, 2023 • 2 new comments

• "unknown rank error" in tensorflow.keras.layers.Layer with tensorflow.py_function output

  #59726 commented on Mar 18, 2023 • 2 new comments

• quantize_and_dequantize_v2 crashes when computing gradients in forward mode

  #59960 commented on Mar 19, 2023 • 2 new comments

• different output value in pytorch->onnx->tflite(int8 quantization)

  #52357 commented on Mar 20, 2023 • 2 new comments

• About fine-tune

  #59908 commented on Mar 20, 2023 • 2 new comments

• [TFLite] Add int16x8 support for SPACE_TO_BATCH_ND and BATCH_TO_SPACE operators

  #52576 commented on Mar 15, 2023 • 2 new comments

• [NVIDIA XLA] Improve allreduce reassociation

  #59424 commented on Mar 14, 2023 • 2 new comments

• [NVIDIA XLA] Support Conv-Bias-Relu6/LeakyRelu fusion in XLA using cuDNN runtime fusion

  #59614 commented on Mar 18, 2023 • 2 new comments

• [NVIDIA XLA:GPU] Fused MHA - Infrastructure + BMM-BMM fusion

  #59832 commented on Mar 14, 2023 • 2 new comments

• [NextPluggableDevice]fix NextPluggableDevice allocator(use pjrt allocator) and Sync()

  #59898 commented on Mar 16, 2023 • 2 new comments

• Movie recommender model doesn't converge

  #59944 commented on Mar 13, 2023 • 1 new comment

• Problems of tflite conversion，about the file of tflite(.h5 to .tflite)

  #59921 commented on Mar 13, 2023 • 1 new comment

• tf.gather Converting sparse IndexedSlices warning

  #38168 commented on Mar 13, 2023 • 1 new comment

• error using the Dispatch

  #59933 commented on Mar 13, 2023 • 1 new comment

• gpu/cpu time performance

  #59963 commented on Mar 13, 2023 • 1 new comment

• unable to build Tf lite .whl file for raspberry pi 0

  #59037 commented on Mar 13, 2023 • 1 new comment

• Upgrade TF to CUDA 11.4 and cuDNN 8.2

  #51659 commented on Mar 13, 2023 • 1 new comment

• “Cloud TPU” console spam on every TensorFlow import

  #50423 commented on Mar 13, 2023 • 1 new comment

• Build with tpu support fails with protobuf error during api generation

  #50539 commented on Mar 13, 2023 • 1 new comment

• Tensorflow lite for microcontrollers documentation has broken links to .cc files

  #59959 commented on Mar 13, 2023 • 1 new comment

• The results of tf.image.convert_image_dtype running on CPU and GPU are very different.

  #58749 commented on Mar 13, 2023 • 1 new comment

• Freeze zero weights during training

  #59868 commented on Mar 14, 2023 • 1 new comment

• [TF:TRT] Tensorflow Docker 2.7.0-gpu has incompatible TRT version 8.0.0

  #53697 commented on Mar 14, 2023 • 1 new comment

• not gpu registered kernels for complex numbers ( tf.complex64 and tf.complex128) for the gather_nd method of tf.Variable

  #59754 commented on Mar 14, 2023 • 1 new comment

• tflite model has bad accuracy

  #59684 commented on Mar 14, 2023 • 1 new comment

• Windows build fails with "pywrap_tensorflow_import_lib_file: variable '$<' : no input file"

  #59938 commented on Mar 14, 2023 • 1 new comment

• Install tensorboard as an extra

  #59605 commented on Mar 14, 2023 • 1 new comment

• Resource exhausted error during training

  #59906 commented on Mar 15, 2023 • 1 new comment

• int8 quantization with int16 activations support for TFMicro and Cortex-M

  #45759 commented on Mar 15, 2023 • 1 new comment

• Allow using raw values for Projector?

  #6219 commented on Mar 15, 2023 • 1 new comment

• The problem with the application performing speech command recognition on the mobile phone

  #45792 commented on Mar 15, 2023 • 1 new comment

• Efficient Ragged Tensor support on TFlite

  #44776 commented on Mar 15, 2023 • 1 new comment

• Remove CMSIS Core include from all Makefiles

  #43696 commented on Mar 15, 2023 • 1 new comment

• tensorflowlite v2.3 + XNNPACK run into error {ModifyGraphWithDelegate is disallowed}

  #42757 commented on Mar 15, 2023 • 1 new comment

• Potential use-after-free bug in function `EagerTensor_shape_tuple` by using an object after all its references released (a static analyzer report)

  #52085 commented on Mar 16, 2023 • 1 new comment

• Misspelling of the word "argument"

  #58383 commented on Mar 16, 2023 • 1 new comment

• LinearOperatorInversion gives weird result on GPU for linear operators

  #59884 commented on Mar 16, 2023 • 1 new comment

• [RNN] LSTM after conversion to TFLite not run on GPU.

  #56482 commented on Mar 16, 2023 • 1 new comment

• TFLite Converter for BiLSTM based architecture

  #45361 commented on Mar 16, 2023 • 1 new comment

• ValueError: Failed to parse the model: pybind11::init(): factory function returned nullptr. when convert and quantize tf model

  #46821 commented on Mar 16, 2023 • 1 new comment

• Feature Request: GPUOptions for Go binding

  #22926 commented on Mar 16, 2023 • 1 new comment

• tflite with CoreML delegate errors with mediapipe models

  #54375 commented on Mar 16, 2023 • 1 new comment

• Support for convolutional ops for complex variable inputs

  #48454 commented on Mar 17, 2023 • 1 new comment

• compile tensorflow_cc.so with c++17 failed with cuda support

  #59887 commented on Mar 17, 2023 • 1 new comment

• Cannot place 2 pre-trained models on separate GPUs

  #59865 commented on Mar 17, 2023 • 1 new comment

• Exporting 🤗 TF BERT model with multiple dynamic axes to MLIR fails with invalid ops generated

  #58835 commented on Mar 17, 2023 • 1 new comment

• OperatorNotAllowedInGraphError: Iterating over a symbolic `tf.Tensor` is not allowed when using a dataset with tuples

  #59510 commented on Mar 17, 2023 • 1 new comment

• '_UserObject' object has no attribute 'call_and_return_conditional_losses'

  #59797 commented on Mar 17, 2023 • 1 new comment

• `Unsupported object type numpy.ndarray` on multi-input `Dataset`

  #59848 commented on Mar 17, 2023 • 1 new comment

• //tensorflow/python/util/protobuf:protobuf_compare_test fails with protobuf > 3.20.3

  #59931 commented on Mar 17, 2023 • 1 new comment

• Tensorflow int8 conversion error for segmentation U-Net

  #59879 commented on Mar 17, 2023 • 1 new comment

• TF Lite issue when loading a saved TF Lite model on platforms with different endianness

  #45009 commented on Mar 17, 2023 • 1 new comment

• atrous_conv2d does not support NCHW format

  #9705 commented on Mar 17, 2023 • 1 new comment

• locals() not working in tensorflow.function decorator

  #59725 commented on Mar 17, 2023 • 1 new comment

• GPU:0 is joining a group with incompatible device typeCPU

  #59888 commented on Mar 18, 2023 • 1 new comment

• Running tensorflow distributed on Multiple workers

  #59799 commented on Mar 18, 2023 • 1 new comment

• Fail to configure on Windows(CMake3.11.0-rc2, VS2015, python3.5.3, swigwin-3.0.12)

  #23472 commented on Mar 18, 2023 • 1 new comment

• Java Tensorflow save & restore without using filesystem

  #26133 commented on Mar 18, 2023 • 1 new comment

• Why we cannot define two networks in the same graph when they are trained independently

  #28421 commented on Mar 18, 2023 • 1 new comment

• STM32F7 Hello World example fails for arm_cmplx_mag_squared_q10p6.c

  #35889 commented on Mar 18, 2023 • 1 new comment

• failed to build //tensorflow/compiler/tf2tensorrt:segment_test on mac osx 12.3.1

  #56021 commented on Mar 18, 2023 • 1 new comment

• precompile model executable binaries

  #49740 commented on Mar 18, 2023 • 1 new comment

• java.lang.UnsatisfiedLinkError:Failed to load native TensorFlow Lite methods. Check that the correct native libraries are present, and, if using a custom native library, have been properly loaded via System.loadLibrary():

  #59929 commented on Mar 18, 2023 • 1 new comment

• Inconsistent Log Level in cuda_driver InternalInit()

  #59849 commented on Mar 18, 2023 • 1 new comment

• Tensorflow Lite conversion output vs. Tensorflow output

  #58206 commented on Mar 18, 2023 • 1 new comment

• Input Layer can't be defined in tensorflow 2.11.0

  #59920 commented on Mar 20, 2023 • 1 new comment

• [TF-TRT] Variable converters follow-up: INT8 support and edge cases

  #56050 commented on Mar 15, 2023 • 1 new comment

• [iOS] Add initial xcframework bazel support

  #56789 commented on Mar 13, 2023 • 1 new comment

• Add tensorflow io support in Docker container

  #58742 commented on Mar 15, 2023 • 1 new comment

• [NVIDIA TF-TRT] Improvements of the TRTConverterV2 summary - Golden APIs Updates

  #59298 commented on Mar 15, 2023 • 1 new comment

• enable a oneDNN ITT feature

  #59320 commented on Mar 15, 2023 • 1 new comment

• [oneDNN]: Adding a build flag to conditionally compile with oneDNN v3.x

  #59599 commented on Mar 15, 2023 • 1 new comment

• Adding support for dynamic shapes to the Unpack converter.

  #59632 commented on Mar 15, 2023 • 1 new comment

• Refactor BitcastOp_Create and BitcastOp_Compute

  #59683 commented on Mar 15, 2023 • 1 new comment

• [tosa] Slice the input of kernel based ops to the actual used size

  #59821 commented on Mar 15, 2023 • 1 new comment

• Added option for Arm NN delegate

  #59952 commented on Mar 17, 2023 • 1 new comment

• include cstdint

  #59964 commented on Mar 16, 2023 • 1 new comment

• Can't install tflite-model-maker from pip or github

  #59855 commented on Mar 13, 2023 • 0 new comments

• Segmentation fault when running gen_nn_ops.fractional_avg_pool

  #59163 commented on Mar 13, 2023 • 0 new comments

• Please bring back native Windows CUDA support!

  #59918 commented on Mar 13, 2023 • 0 new comments

• Crash due to CUDA_ERROR_ILLEGAL_ADDRESS can be triggered in SparseReorder, SparseFillEmptyRowsGrad, SparseSlice and SparseTensorToCSRSparseMatrix

  #59928 commented on Mar 13, 2023 • 0 new comments

• You must feed a value for placeholder tensor ... with dtype int32

  #59779 commented on Mar 13, 2023 • 0 new comments

• The shared library of C++ API lacks of operation symbols on windows.

  #30246 commented on Mar 13, 2023 • 0 new comments

• Jetson Nano build fails with missing "cudnn_version" string (default=7)

  #40784 commented on Mar 13, 2023 • 0 new comments

• Golang Tensorflow v2.3.0 installation fails

  #41808 commented on Mar 13, 2023 • 0 new comments

• Crash when running tensorflow.python.ops.gen_array_ops.upper_bound

  #59415 commented on Mar 13, 2023 • 0 new comments

• [TFlite dynamic range quantization]How can I disable dynamic quantization of activations? in inference?

  #38781 commented on Mar 14, 2023 • 0 new comments

• learned_unigram_candidate_sampler may enter infinite loop when range_max is a very large number

  #40263 commented on Mar 14, 2023 • 0 new comments

• Functionality of writing back Tflite model with multiple subgraphs that is currently loaded in the Tflite Interpreter.

  #44606 commented on Mar 14, 2023 • 0 new comments

• clip_by_value handles python float and numpy float differently

  #38658 commented on Mar 14, 2023 • 0 new comments

• EfficientNet loading slowly from SavedModel format

  #44773 commented on Mar 14, 2023 • 0 new comments

• Adding a feature of "Hard" attention could be very useful

  #50474 commented on Mar 14, 2023 • 0 new comments

• Specifying resource subtypes when converting graphdef to MLIR

  #50521 commented on Mar 14, 2023 • 0 new comments

• Bert ModelSpec Export to TFLITE does not work

  #56110 commented on Mar 14, 2023 • 0 new comments

• Check failure when running tensorflow.python.ops.gen_ragged_conversion_ops.ragged_tensor_to_variant

  #59084 commented on Mar 14, 2023 • 0 new comments

• Add API to convert variable to constant in Keras saved model to improve inference performance

  #59736 commented on Mar 14, 2023 • 0 new comments

• can't install throught `pip install -q git+https://github.com/tensorflow/docs`

  #59747 commented on Mar 14, 2023 • 0 new comments

• ideal development mode discussion

  #59770 commented on Mar 14, 2023 • 0 new comments

• How to use exported automl tensorflow model for tabular data in Python

  #59637 commented on Mar 14, 2023 • 0 new comments

• Loading and stitching TF graphs has tf.Variable conflict even if using the name scope in loading

  #46745 commented on Mar 14, 2023 • 0 new comments

• `PluggableGraphOptimizer failed: NOT_FOUND: Op type not registered '_CopyFromGpuToHost'`

  #58995 commented on Mar 14, 2023 • 0 new comments

• A crash due to check-fail can be triggered in QuantizedConv2D

  #59927 commented on Mar 14, 2023 • 0 new comments

• Build issue by way of third party benchmark from head

  #59958 commented on Mar 14, 2023 • 0 new comments

• Reproducible sparse-dense matrix multiplication

  #56547 commented on Mar 14, 2023 • 0 new comments

• Crash due to check-fail in DenseBincount

  #58178 commented on Mar 14, 2023 • 0 new comments

• libtpu.so 1.5.0 or tpu-vm-tf-2.11.0-pod runtime bug

  #59809 commented on Mar 15, 2023 • 0 new comments

• Optional xnnpack feature relies on weak symbols which are unavailable on Windows

  #59913 commented on Mar 15, 2023 • 0 new comments

• Inconsistent Runtime of XLA Compiled Model Inference

  #59719 commented on Mar 15, 2023 • 0 new comments

• Is it possible to set the random seed for tf.feature_column.categorical_column_with_hash_bucket on operation-level？

  #26574 commented on Mar 15, 2023 • 0 new comments

• [Enhancement] Automatically chose either memory or storage to cache in tf.data API

  #27449 commented on Mar 15, 2023 • 0 new comments

• TensorFlow Lite conversion of frozen graph error

  #28438 commented on Mar 15, 2023 • 0 new comments

• Bug in LSTMBlockCellBpropWithCUDA when using peephole connections

  #34171 commented on Mar 15, 2023 • 0 new comments

• Gradient of matmul in while_loop works when run eagerly but not as tf.function

  #35012 commented on Mar 15, 2023 • 0 new comments

• Difference between android and iOS inference result

  #48170 commented on Mar 15, 2023 • 0 new comments

• custom op on gpu/dsp

  #47448 commented on Mar 15, 2023 • 0 new comments

• Testing Operator "raw_ops.ApplyAdam"

  #42215 commented on Mar 15, 2023 • 0 new comments

• Process killed on tf.dynamic_partition

  #40905 commented on Mar 15, 2023 • 0 new comments

• Causal padding for tfp.layers.Convolution1DFlipout

  #39479 commented on Mar 15, 2023 • 0 new comments

• Translate use of Adam Optimizer from TF 1 to TF2 causing TF2 to coredump

  #50727 commented on Mar 15, 2023 • 0 new comments

• Tensorflow grad

  #59790 commented on Mar 16, 2023 • 0 new comments

• How to "tf.data.experimental.AutoShardPolicy.OFF" if tf.data is not used? But gets this message while using tf.keras.model.fit.

  #50202 commented on Mar 16, 2023 • 0 new comments

• Pass epoch number as arg to `tf.data.Dataset.from_generator`

  #49935 commented on Mar 16, 2023 • 0 new comments

• Segmentation fault when running tensorflow.python.ops.gen_sparse_ops.sparse_fill_empty_rows

  #59393 commented on Mar 16, 2023 • 0 new comments

• free(): invalid pointer when running tensorflow.python.ops.gen_ragged_array_ops.ragged_cross

  #59368 commented on Mar 16, 2023 • 0 new comments

• Integer dtype not supported for tensorflow.math.tan

  #57192 commented on Mar 16, 2023 • 0 new comments

• tf.nn.local_response_normalization lack validation for depth_radius on cpu

  #58122 commented on Mar 16, 2023 • 0 new comments

• TFLite undefined symbol when cross-compiling for android on linux

  #59692 commented on Mar 16, 2023 • 0 new comments

• Memory leak problem due to multi-line comments

  #58856 commented on Mar 16, 2023 • 0 new comments

• [StableHLO/MHLO] `tf-lower-to-mlprogram-and-hlo`

  #59857 commented on Mar 17, 2023 • 0 new comments

• XLA inference fails complaning about branch shape mismatches

  #59818 commented on Mar 17, 2023 • 0 new comments

• tape.gradient computes none after model in model training

  #59652 commented on Mar 17, 2023 • 0 new comments

• SavedModel in C++ with Multiple Signatures gives "You must feed a value for placeholder tensor 'infer_float_board_state' with dtype float and shape [?,19,19,7]"

  #59956 commented on Mar 17, 2023 • 0 new comments

• NCCLAllReduce fails with CUDA 12.0 and CuDNN 8.8.0 when using MirroredStrategy

  #59890 commented on Mar 17, 2023 • 0 new comments

• SVD Function has (very) different GPU and CPU outputs

  #45756 commented on Mar 17, 2023 • 0 new comments

• Add support for kwargs for tf.recompute_grad API for graph mode

  #56692 commented on Mar 18, 2023 • 0 new comments

• Retrieving exactly the same recommendations for all users

  #52831 commented on Mar 18, 2023 • 0 new comments

• LSTM & BiLSTM can't run correct results while batch processing on Mobile

  #51561 commented on Mar 18, 2023 • 0 new comments

• TF 2.11.0: tf.image.ssim return_index_map=True outputs wrong shape

  #59067 commented on Mar 19, 2023 • 0 new comments

• [TFLite] Add int16x8 support for RELU_N1_TO_1 and PRELU operators

  #52913 commented on Mar 17, 2023 • 0 new comments

• Support GPUCompatiblePaddingFIFOQueue

  #57107 commented on Mar 15, 2023 • 0 new comments

• Fix some symbols error for cmake.

  #57168 commented on Mar 15, 2023 • 0 new comments

• [oneDNN] Fusing _FusedConv2D/3D with {LeakyRelu, Mish}

  #58080 commented on Mar 15, 2023 • 0 new comments

• adding tests for binary ops over large tensors

  #59295 commented on Mar 15, 2023 • 0 new comments

• Update ndk version to NDK 25b

  #59422 commented on Mar 17, 2023 • 0 new comments

• Fix crash when invalid input was provided to ragged_tensor_to_variant

  #59660 commented on Mar 15, 2023 • 0 new comments

• Get CUDA dependencies from PyPI packages

  #59825 commented on Mar 15, 2023 • 0 new comments

• [DO NOT REVIEW] Introduce multiple streams execution in TensorFlow.

  #59843 commented on Mar 16, 2023 • 0 new comments