AttributeError: module 'tensorboard.util' has no attribute 'PersistentOpEvaluator'

[smelfungus]

Getting next error on tensorboard launch attempt:

PS C:\Users\user> tensorboard
Traceback (most recent call last):
  File "c:\users\user\appdata\local\programs\python\python36\lib\runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "c:\users\user\appdata\local\programs\python\python36\lib\runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "c:\users\user\appdata\local\programs\python\python36\Scripts\tensorboard.exe\__main__.py", line 5, in <module>
  File "c:\users\user\appdata\local\programs\python\python36\lib\site-packages\tensorboard\main.py", line 45, in <module>
    from tensorboard import default
  File "c:\users\user\appdata\local\programs\python\python36\lib\site-packages\tensorboard\default.py", line 38, in <module>
    from tensorboard.plugins.beholder import beholder_plugin
  File "c:\users\user\appdata\local\programs\python\python36\lib\site-packages\tensorboard\plugins\beholder\__init__.py", line 15, in <module>
    from tensorboard.plugins.beholder.beholder import Beholder
  File "c:\users\user\appdata\local\programs\python\python36\lib\site-packages\tensorboard\plugins\beholder\beholder.py", line 25, in <module>
    from tensorboard.plugins.beholder import im_util
  File "c:\users\user\appdata\local\programs\python\python36\lib\site-packages\tensorboard\plugins\beholder\im_util.py", line 89, in <module>
    class PNGDecoder(util.PersistentOpEvaluator):
AttributeError: module 'tensorboard.util' has no attribute 'PersistentOpEvaluator'

Version:

PS C:\Users\user> pip show tensorboard
Name: tensorboard
Version: 1.12.1
Summary: TensorBoard lets you watch Tensors Flow
Home-page: https://github.com/tensorflow/tensorboard
Author: Google Inc.
Author-email: opensource@google.com
License: Apache 2.0
Location: c:\users\user\appdata\local\programs\python\python36\lib\site-packages
Requires: werkzeug, markdown, protobuf, six, numpy, grpcio, wheel
Required-by: tensorflow-gpu

• Windows 10, 64
• Python version 3.6

Thank you!

[npv14]

I solved this problem by trying to merge my tensorboard\plugins folder with https://github.com/tensorflow/tensorboard

[RiteshBM]

Same Issue , @npv14 's solution didn't work for me.

Would it be correct if I said that there should be a class PersistentOpEvaluator in tensorboard/util.py file?

and there is a class by that name over there, but still it doesn't work for some reason.
Seems like there is something wrong with tensorboard.util. I commented out the parts that called util.PersistentOpEvaluator and now the error is repeated with util.Retrier:-

C:\Users\User>tensorboard
Traceback (most recent call last):
  File "C:\Users\User\Anaconda3\Scripts\tensorboard-script.py", line 10, in <module>
    sys.exit(run_main())
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\main.py", line 48, in run_main
    program.setup_environment()
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\lazy.py", line 49, in __getattr__
    module = self._load()
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\lazy.py", line 38, in _load
    module = importlib.import_module(self.__name__)
  File "C:\Users\User\Anaconda3\lib\importlib\__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 985, in _gcd_import
  File "<frozen importlib._bootstrap>", line 968, in _find_and_load
  File "<frozen importlib._bootstrap>", line 957, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 673, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 697, in exec_module
  File "<frozen importlib._bootstrap>", line 222, in _call_with_frames_removed
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\program.py", line 44, in <module>
    from tensorboard.backend import application
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\backend\application.py", line 37, in <module>
    from tensorboard import db
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\db.py", line 42, in <module>
    class TensorBase(object):
  File "C:\Users\User\Anaconda3\lib\site-packages\tensorboard\db.py", line 48, in TensorBase
    def __init__(self, db_connection_provider, retrier_factory=util.Retrier):
AttributeError: module 'tensorboard.util' has no attribute 'Retrier'

[eelcor]

Same issue here. Have been compiling Tensorflow 1.12 myself to work on a Xeon without AVX instructions (but with Cuda). Both compiled Tensorboard and pipped Tensorboard show the same behavior.

[kenpachiii]

I got it working with the following steps:

1. pip uninstall tensorflow
2. pip uninstall tensorboard
3. go to site-packages and remove tensorflow/tensorboard folders
4. Reinstalled tensorflow - pip install [whatever wheel package]

You can find the wheel packages here for specific install types:
https://www.tensorflow.org/install/pip

Note: I only stated the specific wheel package because that's what I did.

I know this doesn't explain the 'why' of the issue but it got things going for me. Hope it helps!

[RiteshBM]

@kenpachiii I reinstalled tensorflow and tensorboard too (using conda however) but it still refuses to work(tried tensorflow-gpu 1.12.0 with tensorboard 1.12.1/1.12.0).

[eelcor]

Wat version of python do you use? I need to compile it manually due to the lack of avx instructions. Kind regards, Eelco Op ma 31 dec. 2018 03:22 schreef RiteshBM <notifications@github.com:

…

@kenpachiii <https://github.com/kenpachiii> I reinstalled tensorflow and tensorboard too (using conda however) but it still refuses to work(tried tensorflow-gpu 1.12.0 with tensorboard 1.12.1/1.12.0). — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1724 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AC_RQ-_m77Ei3YhHwYaeQo99V3UD_Adoks5u-XTlgaJpZM4Zf-r9> .

[stephanwlee]

Hello everyone, thanks for reporting the issue. Which version of TensorFlow have you installed?
This can happen if you have tb-nightly installed on your virtualenv then install tensorboard==1.12.x.

Like @kenpachiii put, can you try to delete tensorboad directory in site-packages (Location in pip show tensorboard) then pip install tensorboard==1.12?

[komcat]

I got same issue. Here is how I fix it by delete tensorboard folder & tensorboard.dist.info and reinstall tensorboard with respective version according to tensorflow.

mine is running
tf-gpu 1.10
win10
python 3.6.4

[NingMingHao]

Hello everyone, thanks for reporting the issue. Which version of TensorFlow have you installed?
This can happen if you have tb-nightly installed on your virtualenv then install tensorboard==1.12.x.

Like @kenpachiii put, can you try to delete tensorboad directory in site-packages (Location in pip show tensorboard) then pip install tensorboard==1.12?

Thanks, it seems the problem from tb-nightly. I uninstalled tb-nightly and reinstalled the tensorboard.
Finally, it works!!!

[lucventurini]

Another hack to solve it:

First edit your $PYTHONPATH/tensorboard/util/init.py (normally empty) to contain the following:

from .util import *

Second, you should modify the $PYTHONPATH/tensorboard/plugins/beholder/im_util.py file, and change it on line 25 like this:

from

from tensorboard import util

to

from tensorboard import util
from tensorboard.util import op_evaluator

Change then the two classes that inherit from util.PersistentOpEvaluator so that they inherit from op_evaluator:

class PNGDecoder(op_evaluator.PersistentOpEvaluator):
[...]
class Resizer(op_evaluator.PersistentOpEvaluator):

This will take care of the incompatibilities.

[RengXian]

@lucventurini your solution pop out another error.

AttributeError: module ‘tensorboard' has no attribute 'lazy'

[wookayin]

@RengXian Make sure that you don't have tb-nightly installed in your env.

[RengXian]

@wookayin Thanks a lot! it works just by uninstall the tb-nightly.

[Hayashi-Yudai]

@kenpachiii, @stephanwlee Thanks a lot. I use anaconda (Python 3.6.7), and installed tensorflow with conda command. I found two versions of tensorboard (1.12 and 1.13) were installed. So I delete both and reinstalled 1.12. As a result, the problem was solved.

[Harshini-Gadige]

@DummyCo Hi, did you happen to resolve this ? Please keep us posted.

[smelfungus]

@hgadig yep seems like clean reinstallation of tf and tb resolved the problem, Thanks everyone for sharing 'n attempts!

[Nfreshbird]

thanks,it do work! use pip show tensorboard and then delete all directories about tensorboard ,next reinstall a lower version of tensorboard (mine is 1.12).

[Kushagra3911]

Just type pip install tensorboard in the terminal.