FilesExpand file tree

 master

/

ExampleLilDicky.java

Copy path

Blame

More file actions

Blame

More file actions

Latest commit

 

History

History

History

43 lines (34 loc) · 1.4 KB

 master

/

ExampleLilDicky.java

Copy path

Top

File metadata and controls

• Code
• Blame

43 lines (34 loc) · 1.4 KB

Raw

Copy raw file

Download raw file

Open symbols panel

Edit and raw actions

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

import java.util.Random;

import model.Model;

import trainer.Trainer;

import util.NeuralNetworkHelper;

import datasets.TextGenerationUnbroken;

import datastructs.DataSet;

public class ExampleLilDicky {

public static void main(String[] args) throws Exception {

Random rng = new Random();

int totalSequences = 2000;

int sequenceMinLength = 10;

int sequenceMaxLength = 100;

String textSource = "LilDicky";

DataSet data = new TextGenerationUnbroken("datasets/text/"+textSource+".txt", totalSequences, sequenceMinLength, sequenceMaxLength, rng);

String savePath = "saved_models/"+textSource+".ser";

boolean initFromSaved = true; //set this to false to start with a fresh model

boolean overwriteSaved = true;

TextGenerationUnbroken.reportSequenceLength = 500;

int bottleneckSize = 10; //one-hot input is squeezed through this

int hiddenDimension = 200;

int hiddenLayers = 1;

double learningRate = 0.001;

double initParamsStdDev = 0.08;

Model lstm = NeuralNetworkHelper.makeLstmWithInputBottleneck(

data.inputDimension, bottleneckSize,

hiddenDimension, hiddenLayers,

data.outputDimension, data.getModelOutputUnitToUse(),

initParamsStdDev, rng);

int reportEveryNthEpoch = 10;

int trainingEpochs = 1000;

Trainer.train(trainingEpochs, learningRate, lstm, data, reportEveryNthEpoch, initFromSaved, overwriteSaved, savePath, rng);

System.out.println("done.");

}

}