Name
Description
Project mode
Project state
Collaborators
Select a dataset:
Dataset description
Dataset type
Dataset training size
Dataset testing size
Dataset input dimensions
Dataset input shape
Dataset output neurons
Dataset mode
Choose how you will load your dataset
Detect JSON:
Please choose a suitable dataset type above before uploading your JSON to ensure it is rendered on the right:
Need help choosing?
JSON format:
Rendering method
Editing method
Testing method
The dataset "Addition" is recommended for begginers
"MNIST", "CIFAR10", and "QuickDraw30" are recommended for people with a little more experience
"Custom" is recommended for people with much more experience that want to build neural networks for their own projects
Dataset viewer:
Architecture:
Properties:
No layer selectedBATCH SIZE
EPOCHS
LEARNING RATE
OPTIMIZER
LOSS FUNCTION
Trained models
Selected model
Model options
Test:
Output: X
JS Demo
You can use this trained model in your own code by copying it over.
Input variable: Change the input variable to the input that you desire. "input" is a list of the values that the neurons will take, and in cases where your input layer has several dimensions, the list will also have several dimensions. The example below is one in which all the neurons will take the value of 0
Structure variable: The variable structure must have the value seen below. This value may also be obtained by clicking the "Export" button below the "JS Demo" button. This indicated to the program what you Neural Network looks like, and hence how to go from the input to the output.
Output variable: The "output" variable will take the value returned from the "runNN" function. This will be a list of the values of all the output neurons.
let input = ;
let output = runNN(structure, input)
console.log(output)
Delete project
Deleting this project will also delete all its associated trained models. It will also delete its datasets, including the rendering, editing and testing functions. Furthermore, the structure will be deleted, including all its layers. Finally, your shared trained models will no longer appear publicly, and all their associated data will be removed.