Skip to content

A TensorFlow 2.0 implementation of MnasNet: Platform-Aware Neural Architecture Search for Mobile.

Notifications You must be signed in to change notification settings

nsarang/MnasNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MnasNet Tensorflow 2 Implementation

Mingxing Tan, Bo Chen, Ruoming Pang, Vijay Vasudevan, Mark Sandler, Andrew Howard, Quoc V. Le. MnasNet: Platform-Aware Neural Architecture Search for Mobile. CVPR 2019. Arxiv link: https://arxiv.org/abs/1807.11626

Usage

Available implementations: a1, b1, small, d1, d1_320

from MnasNet_models import Build_MnasNet

# Standard model
model = Build_MnasNet('a1')


# Change default parameters:
model = Build_MnasNet('a1', dict(input_shape=(128, 128, 3), dropout_rate=0.5))

Pretrained models

Model Dataset Input Size Depth Multiplier Top-1 Accuracy Top-5 Accuracy Pixel 1 latency (ms) DownLoad Link
MnasNet-A1 ImageNet 224*224 1.0 75.2 95.2 78ms Google Drive

Reference

MnasNet - Official implementation for Cloud TPU

About

A TensorFlow 2.0 implementation of MnasNet: Platform-Aware Neural Architecture Search for Mobile.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages