repository
stringclasses
156 values
issue title
stringlengths
1
1.01k
labels
stringclasses
8 values
body
stringlengths
1
270k
tensorflowtensorflow
performance bug in conv3d transpose cause it to be a factor 100 slow than equivalent computation
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow this only concern a nativ tensorflow function os platform and distribution e g linux ubuntu 16 04 test on arch linux os x and window 8 tensorflow instal from source or bi...
tensorflowtensorflow
use dataset api with estimator in mirroredstrategy non dma safe string tensor error
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow os platform and distribution e g linux ubuntu 16 04 cento tensorflow instal from source or binary pip install tensorflow gpu tensorflow version use command be...
tensorflowtensorflow
can not use adagradoptimizer with mirroredstrategy
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 ubuntu 18 04 tensorflow instal from source or binary binary tensorflow version use command below v...
tensorflowtensorflow
transform graph error when remove node op identity op checknumeric and fold old batch norm be use together
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary source tensorflow version use command below r1...
tensorflowtensorflow
tf datum dataset iterator be not clean when the loop end with a break
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 centos7 tensorflow instal from source or binary binary tensorflow version use command below 1 8 0 ...
tensorflowtensorflow
nhwc convolution sometimes incorrectly consider nchw
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command below...
tensorflowtensorflow
segfault with rpc op in eager mode
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary binary pip tensorflow version use command belo...
tensorflowtensorflow
manually place operation in eager execution raise failedpreconditionerror
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux tensorflow instal from source or binary binary tensorflow version use command below tensorfl...
tensorflowtensorflow
tf keras layers dropout have no be train argument
Bug
dropout layer be usually only use while training kera automatically enable dropout layer while training and disable they while predict thus tf keras layers dropout have no be train argument but it exist in tf layer dropout that can mislead kera and tf user when try to mix kera with tf layer inde...
tensorflowtensorflow
tf split s 1 support doesn t handle zero dimension
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 colab tensorflow instal from source or binary colab tensorflow version use command below unknow...
tensorflowtensorflow
java sbt can not register 2 metric with the same name on cross test
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 osx 10 11 6 tensorflow instal from source or binary binary tensorflow version use command below 1 ...
tensorflowtensorflow
tf gfile open have different semantic from open
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow n a os platform and distribution e g linux ubuntu 16 04 macos 10 13 tensorflow instal from source or binary binary tensorflow version use command below v1...
tensorflowtensorflow
tensor with gpubfcallocator generate segfault on tensorflow op
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary source tensorflow version use command below ...
tensorflowtensorflow
for loop not work in tf contrib autograph
Bug
the following example for autograph be not work see my system info below from tensorflow contrib import autograph def computation for k in range 4 print k return tf no op computation autographd autograph to graph computation verbose true if I remove the ...
tensorflowtensorflow
difference in output between cpu and gpu
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow custom code os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command belo...
tensorflowtensorflow
documentation of raw rnn be wrong
Bug
in the official documentation 1 of tf nn raw rnn we have emit structure as the third output of loop fn when the loop fn be run for the first time later on the emit structure be use to copy tf zero like emit structure to the minibatch entry that be finish by emit tf where finish tf zero like emit stru...
tensorflowtensorflow
tflite memory alignment error with some armv7 android device
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 build on mac os x 10 13 4 and linux ubuntu 17 10 android ndk revision 16 1 4479499 target various android d...
tensorflowtensorflow
tensorflow serve page miss on deploy page
Bug
from the tensorflow deploy page the link for tensorflow serve all give 404 page not find error for example try the installation page for example try the serve a tensorflow model
tensorflowtensorflow
code documetnantion probably misprint in function parameter description
Bug
l573 I believe there be a typo in the description in the penultimate word and which be probably worth replace with be example log every n step the frequency in term of global step that the loss and global step be log
tensorflowtensorflow
tensorflow lite interpreter allocatetensor fail with it size alloc size 15840 0
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 raspberry pi 3 tensorflow instal from source or binary source tensorflow version use command below ...
tensorflowtensorflow
faulty numpy randomness when use gpu
Bug
system information have I write custom code yes os platform and distribution test on slackware linux 14 2 and ubuntu 16 04 tensorflow instal from binary tensorflow version test on both 1 4 and 1 6 python version 3 6 cuda cudnn version 8 0 for tf 1 4 and 9 0 for tf 1 6 bazel ...
tensorflowtensorflow
invalidargumenterror for save restore of variable same version same os same directory
Bug
I get an invalidargumenterror with no further information when I try to save and then restore part of my model later to continue train it due to need my laptop for class initialization saver tf train saver embedding embedding weight nce weight bias nce bias save saver save sess ...
tensorflowtensorflow
documentation issue regard instal on ubuntu
Bug
while I be peruse the documentation for install linux I get to the section determine which tensorflow to install and then I continue on I be at determine how to install tensorflow and I settle on virtualenv so I go to that section and I follow the step but when I get to the sub section 2 st...
tensorflowtensorflow
tensorflow lite demo app crash when use inception v3 mobilenet v1 float instead of quantise mobilenet
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 ubuntu 14 04 tensorflow instal from source or binary git clone source tensorfl...
tensorflowtensorflow
request pre build support old cpu
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow na os platform and distribution e g linux ubuntu 16 04 debian gnu linux 9 4 stretch cpu model intel r xeon r cpu e7 8837 2 67ghz and any old cpu te...
tensorflowtensorflow
bug tf keras estimator model to estimator api give error when keras model contain lambda layer
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow custom code os platform and distribution e g linux ubuntu 16 04 window tensorflow instal from source or binary binary tensorflow vers...
tensorflowtensorflow
keras model fit method not converge with eager execution
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes see below os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary binary with pip3 tensorflow versio...
tensorflowtensorflow
tensorflow 1 8 0rc0 tf compat as str return bytes for python3 since 20180409
Bug
the issue appear first in tf nightly 1 8 0 dev20180409 but be now present in tensorflow 1 8 0rc0 reproduce step python3 c import tensorflow as tf print tf version type tf compat as str hello str be expect to always print true but get tensorflow ...
tensorflowtensorflow
xla implementation of fft on cpu pull in tf core framework
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 macos 10 11 tensorflow instal from source or binary source tensorflow version use command below co...
tensorflowtensorflow
tensorflow leak 1280 byte with each session open and close python api
Bug
have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 17 10 tensorflow instal from source or binary binary tensorflow version use command below 1 6 also test...
tensorflowtensorflow
eager mode fail with custom operator
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes I have write a custom tensorflow operator os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary ...
tensorflowtensorflow
in matmul be fp16 x fp16 accumulate to fp16 or fp32
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command bel...
tensorflowtensorflow
convert tensorflow frozen and inference model to lite fail with check fail array have shape
Bug
have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 mac os x high sierra tensorflow instal from source or binary source tensorflow version use command below v1 7 0 rc1 10...
tensorflowtensorflow
tensorflow mnist estimator do the training batch size affect the graph expect input
Bug
hello tensorflow team first of all let I thank you for your amazing job tensorflow be one of the great deep learn framework out there I be open this issue because I be experience something weird after follow the mnist tutorial with the new estimator high level interface the training and evaluation wo...
tensorflowtensorflow
can not return string from tf data map function
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command below ...
tensorflowtensorflow
bug 1 8 prefetch to device doesn t work with eager iterator
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes minimal non work example can be see below os platform and distribution e g linux ubuntu 16 04 window 10 tensorflow instal from source or binary pip install...
tensorflowtensorflow
documentation miss for tensorrt in python api
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 n a tensorflow instal from source or binary n a tensorflow version use command below 1 7 pyt...
tensorflowtensorflow
freeze graph py generate error message when run on sample model census wide deep
Bug
1 it must be a bug a feature request or a significant problem with documentation for small doc fix please send a pr instead 2 the form below must be fill out 3 it shouldn t be a tensorboard issue those go here here s why we have that policy tensorflow developer respond to issue we w...
tensorflowtensorflow
as per documentation tensorflow 1 5 with cuda 8
Bug
hi as per documentation in the below link installingvirtualenv url requirement for tensorflow 1 5 be cuda 8 I instal tensorflow use the follow tf binary url however still it be ask for cuda 9 be there any whl which support cuda 8 tensorflow 1 5
tensorflowtensorflow
build graph documentation
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command below 1 4 1 gpu python versio...
tensorflowtensorflow
eager tf size do not respect out type
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command belo...
tensorflowtensorflow
error in restore the tensorflow model and metagraph while use seq2seq and attention model
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary pip3 install tensorflow gpu 1 0 0 tensorflow vers...
tensorflowtensorflow
tf keras estimator create order io cast everything to floatx which break non floatx input
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 debian 3 16 36 tensorflow instal from source or binary instal via pip tensorflow version use command be...
tensorflowtensorflow
name scope problem of tensorflow python layer normalization py
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow os platform and distribution e g linux ubuntu 16 04 window 10 64bit tensorflow instal from source or binary binary tensorflow version use command below 1 7...
tensorflowtensorflow
incorrect description of num sample parameter in tf nn nce loss function num sample be number of negative example per 1 positive example not per batch
Bug
hi all look like parameter num sample in tf nn nce loss function define the number of negative example per a positive example but not per a batch as describe in tensorflow documentation see the next code import tensorflow as tf import numpy as np compute sample logit be invoke ...
tensorflowtensorflow
format issue in tf debugger documentation
Bug
image in the faq section frequently ask question there seem to be some markdown formatting problem
tensorflowtensorflow
tf contrib data bucket by sequence length fail for nest dataset element
Bug
hello everyone I just try the new function to group variable length input for the dataset api namely tf contrib data bucket by sequence length for a small estimator model I implement the input fn such that it return a dataset where each element be a tuple feature dict label create input...
tensorflowtensorflow
maskedautoregressiveflow example tf contrib distribution raise valueerror
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 google colab also try on a windows 10 machine with tf 1 6 tensorflow instal from source or binary default fro...
tensorflowtensorflow
contrib nccl broadcast raise unimplementederror
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary docker image tensorflow tensorflow nightly devel gpu...
tensorflowtensorflow
tf1 6 1 7 ps worker distribute run fail with unavailableerror os error when job be not run on current machine
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary official 1 6 0 release binary or build from master branc...
tensorflowtensorflow
beam search terminology
Bug
this issue be with terminology use in documentation and code system information not applicable the documentation and code in tf contrib seq2seq beamsearchdecoder and friend seem to use beam to mean search state whereas beam conventionally mean a collection of search state this non standard usage...
tensorflowtensorflow
mkl dnn be build during non mkl build
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 ubuntu 14 04 tensorflow instal from source or binary source tensorflow version use command below ...
tensorflowtensorflow
import a meta graph which contain a summarywriter doesn t work
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 17 10 tensorflow instal from source or binary binary via pip tensorflow version use comman...
tensorflowtensorflow
unimplementederror broadcast between and be not support yet
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 window 10 pro 64 bit 10 0 build 16299 tensorflow instal from source or binary binary tensorflow v...
tensorflowtensorflow
map and batch tensor shape do not match value of tensor in the same way that call map and batch individually do
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow happen with stock code os platform and distribution e g linux ubuntu 16 04 linux ubuntu server 17 10 1 tensorflow instal from source or binary source tensorfl...
tensorflowtensorflow
tf matrix solve ls documentation corrupt
Bug
hi I just find tf matrix solve ls and read its documentation at grafik seem like there be some markdown error have I write custom code no os platform and distribution n a tensorflow instal from n a tensorflow version n a bazel version n a cuda cudnn version n a gpu model and...
tensorflowtensorflow
error command in installation guild
Bug
please go to stack overflow for help and support there be an small error in the installation guild determine which tensorflow to install which be 7th step under the instal with virtualenv section the site give an example command of instal tensorflow in the active virtualenv for macos pytho...
tensorflowtensorflow
convert result mismatch between command line and python api
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary by pip3 install upgrade tensorflow gp...
tensorflowtensorflow
possible issue of tensorflow kera not handle shape correctly
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 win10 1709 x64 tensorflow instal from source or binary binary from pypi tensorflow version use command b...
tensorflowtensorflow
full fledge java api roadmap
Bug
be there any estimate plan for java api development in term of functionality the only absolute choice for java language nowadays be deeplearning4j any plan to occupy a niche for java language by the term full fledge I mean construct a graph from scratch include most common math function operation res...
tensorflowtensorflow
documentation link for version 1 5 silently redirect to 1 6
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command below ...
tensorflowtensorflow
break qualcomm link on tensorflow org
Bug
on tensorflow org there be a logo with a break link to err time out look like it should be sorry if create an issue be wrong but I do not find any contact possibility
tensorflowtensorflow
performance drop in cifar10 in tf 1 6 0 compile with mkl
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 3 lts tensorflow instal from source or binary source tensorflow version use command below ...
tensorflowtensorflow
dead link on version page
Bug
on there s a link to this link be 404
tensorflowtensorflow
potential resource leak cause by unclear java example
Bug
system information java example at for tensorflow 1 6 0 describe the problem org tensorflow session runner run return list of closable javadoc clearly state that the caller be responsible to free all of they none of the java example I find at highlight that I realize it by happy acc...
tensorflowtensorflow
miss documentation for tf train saver
Bug
the documentation for tf train saver be not explain how to actually use the saver the crucial missing part saver tf train saver need to be execute right before finalize the graph
tensorflowtensorflow
lack of documentation about save restore graph other language api about this
Bug
I think that one of the main feature of tensorflow be its portability I want to exploit it in a huge code write in c where a small but computational heavy part can be compute use gpu through tf so I start to study the problem about save a graph load it to another code possibly in another programming language ...
tensorflowtensorflow
suggestion svd gpu op be not efficient than cpu
Bug
when I use tf 1 2 version there be no problem with use svd but few day ago I update my tensorflow version to 1 5 and my code get extremely slow so I test all possibility and check issue finally I find that someone commit svd gpu op make by cudasolver in june 2017 he say that it contain memcopy ...
tensorflowtensorflow
example and tutorial miss from pip tensorflow package
Bug
after search for and eventually post a question on stackoverflow I be recommend to report this as a bug when instal tensorflow e g tensorflow 1 5 0 cp36 cp36 m manylinux1 x86 64 whl from from pip there be only the tutorial mnist example in tensorflow example it would be nice to have the example availa...
tensorflowtensorflow
still see avx warning on tf 1 6 docker image
Bug
system information cat etc issue linux node014 jupyter 20170708 132328 4 4 0 1022 aw 31 ubuntu smp tue jun 27 11 27 55 utc 2017 x86 64 x86 64 x86 64 gnu linux version 16 04 2 lts xenial xerus version i d 16 04 version codename xenial ...
tensorflowtensorflow
fetch value of variable unnecessarily slow
Bug
do sess run var be about 5x slow than sess run var 1 python variable fetch bug report py 100 mb variable fetch cpu variable 2 5 gb sec min 40 74 median 41 33 mean 42 08 fetch cpu variable add 12 6 gb sec min 7 96 median 8 54 mean 8 71 fetch cpu variable concat 14 0...
tensorflowtensorflow
4x slowdown in feed dict in tf nightly gpu
Bug
upgrade to pip install tf nightly gpu from pip install tensorflow 1 5 slow down feed about 4x feed 100 mb array use to take 15ms and it take 60ms after the change I think this be due to change in alignment requirement for avx compile binary I want to start the thread to figure out how to regain the...
tensorflowtensorflow
htod take 2 5x long than d2h
Bug
I m notice that htod copy be take significantly long than dtoh ie do sess run gpu var vs sess run assign gpu var feed dict gpu var initial value arr 100 mb tensor take 8ms on v100 to fetch 12 5 gb sec but 21 67ms to feed 4 5 gb sec benchmark be there a way to make it fast ...
tensorflowtensorflow
can you document tf metric use in tf keras
Bug
there be no official documentation on how to use tf metric in tf keras two stackoverflow reference but I don t know if it be a api stable solution
tensorflowtensorflow
window installation page list wrong cudnn version
Bug
the windows installation page specifically ask to use cudnn 6 link then when run tensorflow it look specifically for cudnn7 build info cudnn dll name build info cudnn version number importerror could not find cudnn64 7 dll tensorflow require that this dll be instal in a directory that ...
tensorflowtensorflow
dnnconversioncreate f32 fail when run tf with optimize mkl
Bug
system information have I write custom code yes os platform and distribution linux cento 7 tensorflow instal from compile from source from tensorflow version 1 6 0 rc0 python version 2 7 bazel version 0 10 0 gcc compiler version stock cento...
tensorflowtensorflow
iterator get next documentation improvement request
Bug
system information n a describe the problem recently I ve write some code use dataset api and I would like to request a problem with documentation imo instead of hardcoding comment here l31 please move mrry s annotation about iterator get next and get next call warning threshold...
tensorflowtensorflow
window installation tutorial have wrong cudnn version requirement 7 require for late version
Bug
inaccuracy in the documentation in the installation documentation tensorflow in window say that the system must be run a version of cuda 9 0 correctly and cudnn 6 0 not correct because on the nvidia website we can download and install only cudnn v7 0 5 for cuda 8 0 9 0 9 1 c...
tensorflowtensorflow
tf queuebase dequeue many return list instead of tuple
Bug
the dequeue many operation return a list of tensor while the documentation state that it should be a tuple which be more sensible return the tuple of concatenate tensor that be dequeue version 1 5 0 from pip python3 on os x
tensorflowtensorflow
tensorflow 1 5 tf dynamic partition on gpu calculate wrong gradient shape
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow none os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary binary pip install tensorflow gpu 1 5 0 ...
tensorflowtensorflow
tf contrib estimator replicate model fn fail when a trainable variable doesn t have gradient
Bug
tf contrib estimator replicate model fn fail when the gradient of a trainable variable be none the error message be file usr local lib python2 7 dist package tensorflow python estimator estimator py line 302 in train loss self train model input fn hook save listener file usr...
tensorflowtensorflow
python3 support of doc generation
Bug
1 replace codegen by astor 2 use tuple asdict to replace tuple dict fix 9437
tensorflowtensorflow
tensorflow 1 5 0 import error on cuda 9 0
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no I use the stock version os platform and distribution e g linux ubuntu 16 04 window 10 tensorflow instal from source or binary pip binary for window with gpu supp...
tensorflowtensorflow
no documentation on the order of eigenvalue return by tf self adjoint eig
Bug
system information os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary binary tensorflow version use command below v1 4 0 rc1 11 g130a514 1 4 0 python version 2 7 14 describe the problem from t...
tensorflowtensorflow
the link for the tutorial on google s tensorflow syntaxnet page give 404 error
Bug
go tot the page and click the tutorial link it get a 404 error the target of the link be installation
tensorflowtensorflow
fail install on window
Bug
python 3 6 4 there be a strange error when I install tensorflow 1 5 collect tensorflow use cache requirement already up to date wheel 0 26 in d python python36 lib site package from tensorflow requirement already up to date numpy 1 12 1 in d python python36 lib site package from ...
tensorflowtensorflow
keras multi input and estimator
Bug
if I ve interpret it correctly seem that there be some strange behavior with keras multi inputs and the estimator why input layer be rename with 1 suffix why tb display a 2 suffix parallel sub graph I ve attach a snippet runnable on colab and the tb render image import tensorf...
tensorflowtensorflow
assignaddvariableop have no output
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 arch linux tensorflow instal from source or binary source tensorflow versio...
tensorflowtensorflow
request for update kera dataset file to r1 5
Bug
system information execute keras sample code imdb fasttext py window 7 tensorflow instal from binary tensorflow version 1 5 0rc0 python version 3 5 1 describe the problem keras sample program do not work there be a bug for numpy arange method wrong usage ...
tensorflowtensorflow
change release md to specify cuda version
Bug
the release md state that prebuilt binary be now build against cuda 9 and cudnn 7 say that cuda 9 1 be not support could we change the release md so that it say prebuilt binary be now build against cuda 9 0 and cudnn 7 until later version be support
tensorflowtensorflow
compute gradient of loop variable return none
Bug
system information window 10 x64 instal from binary tensorflow 1 3 0 python 3 6 3 cudnn 6 4 6 cuda 8 0 nvidia geforce 940 m bug description issue 15403 be close since it be label as not a bug or feature request I believe it be indeed a bug the fundamental probl...
tensorflowtensorflow
lack of clarity in tf while loop documentation
Bug
I believe that the documentation for tf while loop be lack usage clarity and actually provide contradictory statement specifically it seem that many people be use the tf while loop as a for loop see stackoverflow however the tf while loop while loop docs state for correct program wh...
tensorflowtensorflow
loss softmax cross entropy documentation on tensor rank
Bug
documentation issue it seem loss softmax cross entropy work just fine with any tensor rank shape as long as the last dimension be class and not just shape of batch size num class as the documentation indicate the documentation also indicate weight should be rank 0 or 1 also not true i...
tensorflowtensorflow
tensorflow debug tfdbg valueerror with combine loss function
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes include os platform and distribution e g linux ubuntu 16 04 window 10 home x64 1709 tensorflow instal from source or binary binary pip3 tensorflow ver...
tensorflowtensorflow
doc feature request graphical representation of operation
Bug
it would be cool if the documentation of tf operation would contain graphical example e g the tf dynamic partition operation contain such a visualization image this would be especially helpful to understand the different slicing join operation
tensorflowtensorflow
tensorflow contrib gan loss impl test fail with assertionerror
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 s390x tensorflow instal from source or binary source tensorflow version use command...
tensorflowtensorflow
performance issue with tf1 5 on cpu
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 linux ubuntu 16 04 tensorflow instal from source or binary source tensorflow version use command below ...
tensorflowtensorflow
documentation method template improvement
Bug
system information n a describe the problem the method class template in documentation should include a full function path to the method instead of just truncating to the method s name I e this be what we have at present bad this be a more practical and copy paste friendly version ...
tensorflowtensorflow
dataset from generator doesn t release memory after recreate the session
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow yes os platform and distribution e g linux ubuntu 16 04 linux ubuntu 17 10 tensorflow instal from source or binary binary tensorflow version use command below ...
tensorflowtensorflow
inference on v100 with tf1 5 be extremely slow
Bug
system information have I write custom code as oppose to use a stock example script provide in tensorflow no os platform and distribution e g linux ubuntu 16 04 ubuntu 16 04 tensorflow instal from source or binary source and virtual env problem doesn t change tenso...