tensorflow / tensorflow/privacy

A Question about Research mi_lira_2021: why set training=True at Inference

Open
#427 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
2k
Forks
477
Avg merge
22h 12m
Merged PRs (30d)
1

Description

Hi, I have a question about the code of Research mi_lira_2021.
It set training=True at inference:
https://github.com/tensorflow/privacy/blob/4dd8d0ffde4ddb1575d5c2fc02e0693e08f4f4a1/research/mi_lira_2021/inference.py#L72

But since the model contains batchnorm layers, it should be training=True to avoid updating the parameters in the inference.
So why training=True here?

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 at research/mi_lira_2021/inference.py around line 72 and inspect how the model is called during inference, especially its batch-normalization behavior. Determine why training=True is used and document the rationale or clarify the code if the current behavior is incorrect; the issue is done when the question has a maintainer-backed answer.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, tensorflow
Domain
machine-learning
Issue type
Documentation
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.