tensorflow / tensorflow/tensorboard

[Feature Request]: Table view

Open
#3,074 1 comment 0 reactions 1 assignee View on GitHub

@bileschi is already working on this.

Since Dec 30, 2019.

core:frontend stat:awaiting tensorflower type:feature
Dominant language
TypeScript
Stars
7.2k
Forks
1.7k
Avg merge
4d 22h
Merged PRs (30d)
1

Description

I have been using Tensorboard to track training progress for my experiments and I would also like to store test results in Tensorboard.
I currently store my test results in a simple table, something like:

Exp name Metric 1 Metric 2
Exp 1 12 123
Exp 2 23 234

Unfortunately, I can't find a way to create a table in Tensorboard.

I would like something like writer.add_table_cell(row_name, column_name, result).

I am currently abusing HParams plugin table view, but it lacks some features:

  • I can't select subset of experiments to compare. I frequently have more than 20 experiments, so this is quite important for my workflow.
  • It has hparams and metrics separation, which is confusing and either not required for simple setup or not enough for complex one.
  • Sorting doesn't seem to work.
  • It has different design to other views.

Some additional functionality which would be nice:

  • Sorting using one of the columns
  • Creation of additional columns using values from other ones.

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.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.