Implementing Searching for MobileNetV3 paper using Pytorch

• The current model is a very early model. I will modify it as a general model as soon as possible.

Paper

• Searching for MobileNetV3 paper

• Author: Andrew Howard(Google Research), Mark Sandler(Google Research, Grace Chu(Google Research), Liang-Chieh Chen(Google Research), Bo Chen(Google Research), Mingxing Tan(Google Brain), Weijun Wang(Google Research), Yukun Zhu(Google Research), Ruoming Pang(Google Brain), Vijay Vasudevan(Google Brain), Quoc V. Le(Google Brain), Hartwig Adam(Google Research)

Todo

• Experimental need for ImageNet dataset.

• Code refactoring

MobileNetV3 Block

Experiments

• For CIFAR-100 data, I experimented with resize (224, 224).
  

Usage

Train

python main.py

• If you want to change hyper-parameters, you can check "python main.py --help"

Options:

• --dataset-mode (str) - which dataset you use, (example: CIFAR10, CIFAR100), (default: CIFAR100).

• --epochs (int) - number of epochs, (default: 100).

• --batch-size (int) - batch size, (default: 128).

• --learning-rate (float) - learning rate, (default: 1e-1).

• --dropout (float) - dropout rate, (default: 0.3).

• --model-mode (str) - which network you use, (example: LARGE, SMALL), (default: LARGE).

• --load-pretrained (bool) - (default: False).

• --evaluate (bool) - Used when testing. (default: False).

• --multiplier (float) - (default: 1.0).

Test

python main.py --evaluate True

• Put the saved model file in the checkpoint folder and saved graph file in the saved_graph folder and type "python main.py --evaluate True".

• If you want to change hyper-parameters, you can check "python test.py --help"

Options:

• --dataset-mode (str) - which dataset you use, (example: CIFAR10, CIFAR100), (default: CIFAR100).

• --epochs (int) - number of epochs, (default: 100).

• --batch-size (int) - batch size, (default: 128).

• --learning-rate (float) - learning rate, (default: 1e-1).

• --dropout (float) - dropout rate, (default: 0.3).

• --model-mode (str) - which network you use, (example: LARGE, SMALL), (default: LARGE).

• --load-pretrained (bool) - (default: False).

• --evaluate (bool) - Used when testing. (default: False).

• --multiplier (float) - (default: 1.0).

Number of Parameters

import torchfrom model import MobileNetV3def get_model_parameters(model):
    total_parameters = 0
    for layer in list(model.parameters()):
        layer_parameter = 1
        for l in list(layer.size()):
            layer_parameter *= l
        total_parameters += layer_parameter    return total_parameters

tmp = torch.randn((128, 3, 224, 224))
model = MobileNetV3(model_mode="LARGE", multiplier=1.0)print("Number of model parameters: ", get_model_parameters(model))

Requirements

• torch==1.0.1