tensorflow / tensorflow/tensorboard

Feature Request: Visualizing Attention Weights of Attention Wrapper

Open
#160 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

type:feature (new plugin)
Dominant language
TypeScript
Stars
7.2k
Forks
1.7k
Avg merge
4d 22h
Merged PRs (30d)
1

Description

OS: macOS Sierra version 10.12.5
TensorFlow Version: v1.2.0-rc2-21-g12f033d 1.2.0

This is related to tensorflow.contrib.seq2seq. I would like the ability to visualize the attention weights of the AttentionWrapper. Would it be possible to add some tool that displays how the attention weights change as the decoder network produces outputs?

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

Start by reading tensorflow.contrib.seq2seq.AttentionWrapper and tracing how its attention weights could reach TensorBoard. No files or tests are identified in the issue, so first define the visualization entry point and data flow. Done means displaying how the weights change as decoder outputs are produced.

Written by the indexing model from the issue text.

Assessment

Domain
data-visualization, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.