From e0740395471e5f0b4335aea950930827f829f092 Mon Sep 17 00:00:00 2001 From: Kaoutar El Maghraoui Date: Sun, 20 Aug 2023 09:15:32 +0100 Subject: [PATCH] Update README.md Added AnalogNAS citation and awards --- README.md | 35 +++++++++++++++++++++++++++++++---- 1 file changed, 31 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 7ca89940..c282d353 100644 --- a/README.md +++ b/README.md @@ -1,11 +1,11 @@ # AnalogNAS ## Description -**AnalogNas** is a modular and flexible framework to facilitate implementation of Analog-aware Neural Architecture Search. It offers high-level classes to define: the search space, the accuracy evaluator, and the search strategy. It leverages [the aihwkit framework](https://github.com/IBM/aihwkit) to apply hardware-aware training with analog non-idealities and noise included. **AnalogNAS** obtained architectures are more robust during inference on Analog Hardware. We also include two evaluators trained to rank the architectures according to their analog training accuracy. +**AnalogNas** is a modular and flexible framework to facilitate the implementation of Analog-aware Neural Architecture Search. It offers high-level classes to define: the search space, the accuracy evaluator, and the search strategy. It leverages [the aihwkit framework](https://github.com/IBM/aihwkit) to apply hardware-aware training with analog non-idealities and noise included. **AnalogNAS** obtained architectures are more robust during inference on Analog Hardware. We also include two evaluators trained to rank the architectures according to their analog training accuracy. > :warning: This library is currently in beta and under active development. -> Please be mindful of potential issues and keep an eye for improvements, -> new features and bug fixes in upcoming versions. +> Please be mindful of potential issues and monitor for improvements, +> new features, and bug fixes in upcoming versions. [**Setup**](#setup) | [**Usage**](#usage) @@ -78,7 +78,7 @@ worker.result_summary() # print results ``` ## Usage -To get started, check out ```nas_search_demo.py``` and ```starter_notebook.ipynb``` to make sure that the installation went well. +To get started, check out ```nas_search_demo.py``` and ```starter_notebook.ipynb``` to ensure the installation went well. ## Authors AnalogNAS has been developed by IBM Research, @@ -89,6 +89,33 @@ Abu Sebastian, and Kaoutar El Maghraoui You can contact us by opening a new issue in the repository. +## How to cite? + +In case you are using the _AnalogNas_ toolkit for +your research, please cite the IEEE Edge 2023 paper that describes the toolkit: + +> Hadjer Benmeziane, Corey Lammie, Irem Boybat, Malte Rasch, Manuel Le Gallo, +> Hsinyu Tsai, Ramachandran Muralidhar, Smail Niar, Ouarnoughi Hamza, Vijay Narayanan, +> Abu Sebastian and Kaoutar El Maghraoui +> "AnalogNAS: A Neural Network Design Framework for Accurate Inference with Analog In-Memory Computing" (2023 IEEE INTERNATIONAL CONFERENCE ON EDGE +> COMPUTING & COMMUNICATIONS (IEEE Edge)) + +> https://arxiv.org/abs/2305.10459 + + +## Awards and Media Mentions + +* We are proud to share that AnalogNAS open source project the prestigious **IEEE OPEN SOURCE SCIENCE** in 2023 at the [IEEE 2023 Services Computing Congress](https://conferences.computer.org/services/2023/awards/). + + image + +* AnalogNAS paper received the **Best Paper Award** at [2023 IEEE EDGE (INTERNATIONAL CONFERENCE ON EDGE COMPUTING & COMMUNICATIONS)](https://conferences.computer.org/edge/2023/) + + image + + + + ## References * [Hardware-aware Neural Architecture Search: Survey and Taxonomy](https://www.ijcai.org/proceedings/2021/592) * [AIHWKit](https://ieeexplore.ieee.org/abstract/document/9458494)