algorithmiahq

algorithmiahq / pytorchDemo / 0.2.0

README.md

Overview

This is the simplest possible Convolutional Neural Net (one convolutional layer, one pooling layer) trained on the CIFAR-10 dataset.

Applicable Scenarios and Problems

Classifying images that belong to one of the CIFAR-10 classes: ['plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']

Usage

Data

This demo only supports loading images from Algorithmia's Data Collections. For more information about how to upload your data into a Collection, refer to the docs here.

Input

ParameterDescription
inputsquare image

The input image can be any size, but it must be square.

Output

ParameterDescription
outputclass label