optimizing-deep-learning-models-for-arm-cortex-m-through-pruning's Introduction

Optimizing Deep Learning Models for ARM Cortex-M Through Pruning

I integrated two types of model pruning methods and ported them to the ARM Cortex-M using the CMSIS library.

Requirement

Pytorch(version: 1.5.0)
ARM-CMSIS_5
NuMaker-PFM-M487(development board)
Mbed OS(version: 5.15.1)
Ubuntu 16.04

Pruning methods

Convolutional layers pruning of this paper
Fully-connected layer pruning of this paper

System flow

Example command

Train Alexnet-Light on CIFAR10 and prune it.

python pruning.py --dataset cifar10 --model alexnet-light --lr 0.001 --epochs 80 --retrain-epochs 80 --prune-layers conv1 conv2 conv3 conv4 --prune-channels 0 0 10 10 --sensitivity 0.8

Load a well-trained model and prune it.

python pruning.py --dataset cifar10 --model alexnet-light --lr 0.001 --retrain-epochs 80 --prune-layers conv1 conv2 conv3 conv4 --prune-channels 0 0 10 10 --sensitivity 0.8 --no-train True --load-path .\saves\alexnet_nopruned_7722_7692.ptmodel

Quantization and generate `cortexm_weight.h` for `NuMaker-PFM-M487`

python inference_alexnet.py

Results

Inference time was measured on NuMaker-PFM-M487.
Accuracy was simulated on PC with CMSIS library.

LeNet-Light with MNIST

	Before	After
Inference time	30ms	22ms
Accuracy	99.11%	98.55%

Alexnet-Light with CIFAR10

	Before	After
Inference time	20ms	14ms
Accuracy	76.92%	75.15%

The inference time of Alexnet-Light(kernel size:3x3) is lower than LeNet-Light(kernel size:5x5) because of conv kernel size.

VGG-Light with CIFAR10

	Before	After
Inference time	47ms	32ms
Accuracy	78.98%	77.84%

optimizing-deep-learning-models-for-arm-cortex-m-through-pruning's People

Contributors

Stargazers

Watchers

optimizing-deep-learning-models-for-arm-cortex-m-through-pruning's Issues

Bias Shift bit

Hello. First of all, thank you for your nice code.
I wonder how you did the shift method for the bias during quantization.
Looking at your code, you filled the bias shift part with '0'. According to the ARM CMSIS Deploy Guide, bias left shift does the following.
https://developer.arm.com/documentation/102591/2208/Compute-the-layer-shifts
Have you ever used a method other than that?

Recommend Projects