Tensorboard Installation Errors
1 attempt stage Dell 32-bit PC, original installation anaconda,
Conda create-n TensorFlow python=3.6
Activate TensorFlow
Pip install--ignore-installed--upgrade tensorflow results show that matching TensorFlow version is not found
2 Find data Find TensorFlow No 32-bit version Lenovo 64-bit
Conda create-n TensorFlow python=3.6
Activate TensorFlow
Pip Install--ignore-installed--upgrade tensorflow installation tensorflow success Tensorboard--logdir= "\path\to\log" Display html5lib error Cannot import name as Reparseexception
3 into the pit began to ask the successful installation of the classmate, may be a path problem, modify the path or not
According to the error, search, but no one has the same error, only Windows system has a reparse point to suspect that their computer installed this, in the process is not very understanding, first give up.
Since it is a problem in Html5lib, suspect html5lib version, remove Html5lib Toolkit, reinstall, or not
Since is reparse, the resolution direction question, installs the BF4, the result is not good.
Finally, reinstall Anaconda and use the python3.5 version instead
Conda create-n TensorFlow python=3.5
Activate TensorFlow
Pip Install--ignore-installed--upgrade tensorflow installation tensorflow successful Tensorboard--logdir= "Path\to\log" successfully turned on using code to regenerate the log, It's not possible. The problem may be that the Anaconda version is too high, re-download version 4.2.0, and finally found that the corresponding TensorFlow, or TensorFlow version is higher than Anaconda
4 finale Reinstall python3.5, which successfully opens the generated Tensorboard as before, but displays an error for the log that was just generated.
Finally accidentally opened Tensorboard--help
Convert Tensorboard open format.
Tensorboard--logdir Name:path run successfully.
Success Code:
Conda create-n TensorFlow python=3.6
Activate TensorFlow
Pip Install--ignore-installed--upgrade tensorflow installation tensorflow successful Tensorboard--logdir name: "\path\to\log"