How to quantize a Sequence classification network in Matlab using calibrate

1 次查看（过去 30 天）

显示更早的评论

Kevin 2022-12-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1872357-how-to-quantize-a-sequence-classification-network-in-matlab-using-calibrate

评论： Dhananjay Kumar 2023-7-17

I used this network as a basic for my own Sequence classification network:

https://de.mathworks.com/help/deeplearning/ug/sequence-classification-using-1-d-convolutions.html

I used dlquantizer to quantize my network

quantObj = dlquantizer(net)

Now i want to use calibrate

calResults = calibrate(quantObj,XValidation);

but it gives me the error "scalare structure required" If i use

calResults = calibrate(quantObj,TValidation);

it outputs me the error "Datastore must be either an imageDatastore, augmentedImageDatastore, TransformedDatastore, PixelLabelImageDatastore or CombinedDatastore."

How can I load the Calibrationdata?

TValidation and XValidation are 370x1 cell with Validation Data

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Dhananjay Kumar 2023-7-17

Hi Kevin , calibrate API is enhanced in R2023b to support a lot more types of input data. You should be able to do that in R2023b.

https://www.mathworks.com/downloads/prerelease/

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

请先登录，再回答此问题。

类别

AI, Data Science, and Statistics Deep Learning Toolbox Code Generation Quantization, Projection, and Pruning

在 Help Center 和 File Exchange 中查找有关 Quantization, Projection, and Pruning 的更多信息

产品

Deep Learning Toolbox

版本

R2022b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by