Icon

02_​Basic_​Learner_​View_​Tutorial

How to use the Learner View

This workflow shows an example of the View of the DL4J Feedforward Leaner nodes.

How to use the Learner ViewThis workflow shows an example of the View shared by all DL4J Feedforward Learner nodes. During execution, the currentlearning status as well as the current error can be viewed in the Learner Node View. Additionally, the bottom of the viewshows a history of the error of the network after each epoch. The view updates automatically after each adjustment of thenetwork parameters. The view can be opened by Right Click > View: Learning Status. Additionally, the learning can bestopped manually by clicking the 'Stop Learning' button in the Learner Node View. This will stop learning after the currentbatch and save the model in the current status. This is especially useful if the error doesn't significantly decrease anymoreduring learning so we can stop the learning process without having to wait for all epochs to be finished. Note that after the'Stop Learning Button' was clicked the model will be directly ready to use.Workflow RequirementsKNIME Analytics Platform 3.4.0KNIME Deeplearning4J Integration This metanode creates a simple general purpose multi layer perceptron consisting oftwo fully connected (FC) layers. Parameters for all layers:Learning Rate: 0.01Activation Function: ReLU Right Click > View: Learning Status100 FCUnits100 FCUnitsData Generator DL4J Feedforward Learner(Classification) Dense Layer DL4J ModelInitializer Dense Layer How to use the Learner ViewThis workflow shows an example of the View shared by all DL4J Feedforward Learner nodes. During execution, the currentlearning status as well as the current error can be viewed in the Learner Node View. Additionally, the bottom of the viewshows a history of the error of the network after each epoch. The view updates automatically after each adjustment of thenetwork parameters. The view can be opened by Right Click > View: Learning Status. Additionally, the learning can bestopped manually by clicking the 'Stop Learning' button in the Learner Node View. This will stop learning after the currentbatch and save the model in the current status. This is especially useful if the error doesn't significantly decrease anymoreduring learning so we can stop the learning process without having to wait for all epochs to be finished. Note that after the'Stop Learning Button' was clicked the model will be directly ready to use.Workflow RequirementsKNIME Analytics Platform 3.4.0KNIME Deeplearning4J Integration This metanode creates a simple general purpose multi layer perceptron consisting oftwo fully connected (FC) layers. Parameters for all layers:Learning Rate: 0.01Activation Function: ReLU Right Click > View: Learning Status100 FCUnits100 FCUnitsData Generator DL4J Feedforward Learner(Classification) Dense Layer DL4J ModelInitializer Dense Layer

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: David Kolb
Created at: 2016-07-11
On NodePit since: 2019-07-12
Last update: 2024-04-18
Created with KNIME version: v4.1.0
Tags: deeplearning machine learning neural networks view

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!