tensorflow: ModuleNotFoundError: No module named 'tensorflow.tensorboard.tensorboard'

Tensorboard not working on Tensorflow built from sources

System information

Linux Ubuntu 16.04
TensorFlow installed from sources using Bazel
TensorFlow version v1.2.0-1126-gb7acb6a
Bazel version 0.51
CUDA/cuDNN version 8.0/6.0
Found device 0 with properties: name: GeForce GTX 1070 major: 6 minor: 1 memoryClockRate (GHz) 1.683 pciBusID 0000:01:00.0 Total memory: 7.92GiB Free memory: 5.15GiB
Exact command to reproduce: $ tensorboard Installed into fresh anaconda3 environment ‘tensorflow’, environment is activated when performing command.

Description: The TensorBoard visualization doesn’t work

Source code / logs

$ tensorboard
Traceback (most recent call last):
  File "/home/gpu/anaconda3/envs/tensorflow/bin/tensorboard", line 7, in <module>
    from tensorflow.tensorboard.tensorboard import main
ModuleNotFoundError: No module named 'tensorflow.tensorboard.tensorboard'

About this issue

Original URL
State: closed
Created 7 years ago
Comments: 17 (4 by maintainers)

Most upvoted comments

For anybody else finding this issue, since tensorboard was split out from tensorflow the command to run is just tensorboard --logdir=$MODEL_DIR instead of python -m tensorflow.tensorboard --logdir=$MODEL_DIR as it was previously.

jcomfort4 on Aug 18, 2017

Thanks @jcomfort4. After Cloud SDK update 171.0.0 I had the same issue. Solved with tensorboard --logdir=output as replacement for: python -m tensorflow.tensorboard --logdir=output Please update: https://cloud.google.com/ml-engine/docs/how-tos/getting-started-training-prediction

GalMoore on Sep 17, 2017

Windows users might want to check out these two batch lines:

cd C:/Users/$USERNAME/AppData/Local/Programs/Python/Python35/Scripts/
tensorboard.exe --logdir=$MODEL_DIR

vossthor on Sep 7, 2017