tensorflow / tensorflow/tensorboard
SummaryWriter/TensorBoard should use a more efficient format
Open
Nobody has claimed this yet.
plugin:audio
plugin:images
plugin:scalars
theme:performance
- Dominant language
- TypeScript
- Stars
- 7.2k
- Forks
- 1.7k
- Avg merge
- 4d 22h
- Merged PRs (30d)
- 1
Description
The current format used becomes slow in many circumstances:
- Saving images/audio/...
- Saving many scalars
I suggest that the format is replaced by something that:
- Is compressed (and don't save tags for every event)
- Is columnar (don't load images if you only look at scalars)
- Easy to read and load from Python (efficiently)
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No files or tests are named. Start by locating the current SummaryWriter/TensorBoard event format and examine its behavior when saving images, audio, and many scalars. Done would require a compressed, columnar format that avoids repeated tags, supports selective loading, and remains efficient to read from Python.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data-visualization
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100