From 78f76fac31a99552bcd3d9700fd531f081eb4eb1 Mon Sep 17 00:00:00 2001 From: Yurii Sushko Date: Wed, 16 Sep 2020 12:04:07 -0700 Subject: [PATCH] Add a note about API update to the docs. PiperOrigin-RevId: 332057406 --- .../privacy/membership_inference_attack/README.md | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/tensorflow_privacy/privacy/membership_inference_attack/README.md b/tensorflow_privacy/privacy/membership_inference_attack/README.md index f9dc2f6..654207b 100644 --- a/tensorflow_privacy/privacy/membership_inference_attack/README.md +++ b/tensorflow_privacy/privacy/membership_inference_attack/README.md @@ -18,6 +18,20 @@ the model are used (e.g., losses, logits, predictions). Neither model internals ## How to use + +### API revamp note +We're **revamping our attacks API to make it more structured, modular and +extensible**. The docs below refers to the legacy experimental API and will be +updated soon. Stay tuned! + +For a quick preview, you can take a look at [data_structures.py] +(https://github.com/tensorflow/privacy/blob/master/tensorflow_privacy/privacy/membership_inference_attack/data_structures.py) +and [membership_inference_attack_new.py] +(https://github.com/tensorflow/privacy/blob/master/tensorflow_privacy/privacy/membership_inference_attack/membership_inference_attack_new.py). + +For now, here's a reference to the legacy API. + + ### Codelab The easiest way to get started is to go through [the introductory codelab](https://github.com/tensorflow/privacy/blob/master/tensorflow_privacy/privacy/membership_inference_attack/codelab.ipynb).