smallbasic.plugins/mlpack/samples/data.bas at llama.cpp · smallbasic/smallbasic.plugins

28 lines (20 loc) · 838 Bytes

const RATIO = 0.1 ' 10%

import mlpack

i = image("chicken.png")

i.save(a)

dataset = mlpack.Load(a)

' Split the dataset into training and validation sets.

split_data = mlpack.split(dataset, RATIO)

trainData = split_data.train

validData = split_data.test

' The train and valid datasets contain both - the features as well as the

' prediction. Split these into separate matrices.

trainX = mlpack.submat(trainData, 1, 1, trainData.n_rows - 1, trainData.n_cols - 1)

validX = mlpack.submat(validData, 1, 1, validData.n_rows - 1, validData.n_cols - 1)

' Create prediction data for training and validatiion datasets.

trainY = mlpack.get_row(trainData, 0)

validY = mlpack.get_row(validData, 0)

' Scale all data into the range (0, 1) for increased numerical stability.

mlpack.scale(trainX, validX)

mlpack.scale(trainY, validY)

print "done!"

Provide feedback