tensorflow / tensorflow/tensorboard

Erratic behavior in accessing subdirs on Windows

Open
#1,353 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

os:windows
Dominant language
TypeScript
Stars
7.2k
Forks
1.7k
Avg merge
4d 22h
Merged PRs (30d)
1

Description

Consider Stack Overflow for getting support using TensorBoard - they have a larger community with better searchability:

https://stackoverflow.com/questions/tagged/tensorboard

For bug reports, please include the following:

  • TensorBoard version (from pip package, also printed out when running tensorboard)
    1.10.0
  • TensorFlow version if different from TensorBoard
    1.10.0
  • OS Platform and version (e.g., Linux Ubuntu 16.04)
    Windows 8
  • Python version (e.g. 2.7, 3.5)
    3.5.5
  • For browser-related issues:
    • Browser type and version (e.g. Chrome 64.0.3282.140)
      Chrome
    • Screenshot if it's a visual issue

image

Please describe the bug as clearly as possible, and if possible provide a minimal example (code, data, and/or command line) to reproduce the issue. Thanks!

  • I've multiple 3 subdirs containing .events file under a parent directory sh at C:\.
  • Launching, tensorboard in any way (from 'C:' or 'C:\sh') and pointing to sh parent dir (correctly) starts the tensorboard server at 6006 and in the console on the list of file names shows all the subdirs but never shows all three in the plots.
  • Sometimes, it shows one of them or sometimes two (and I've seen all the possible combinations of pairs out of three on multiple restarts of the server) but never all of them.
  • It works fine if I point it directly to a subdir, showing just one plot.

I've also tried switching to Firefox or downgrading to 0.x but no use. It'll be great if someone can help me out. A possible caveat might be that each of the three events file is pretty big ~90 MB.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Reproduce the report with TensorBoard 1.10.0 and TensorFlow 1.10.0 on Windows 8 using Python 3.5.5: place three roughly 90 MB event files in separate subdirectories under C:\sh and launch TensorBoard against the parent directory. Compare the parent-directory view with direct views of each subdirectory; done means all three plots load reliably together.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, tensorflow
Domain
data-visualization
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.