Skip to content

Instantly share code, notes, and snippets.

@riverajefer

riverajefer/train.js

Created May 30, 2019 04:14
Show Gist options
  • Save riverajefer/bef718ca0f23c2ab6eb0e3853eed1393 to your computer and use it in GitHub Desktop.
Save riverajefer/bef718ca0f23c2ab6eb0e3853eed1393 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
train.js
// import * as tf from '@tensorflow/tfjs';
// Solve for XOR
const LEARNING_RATE = 0.1;
const EPOCHS = 500;
// Define the training data
const xs = [
[
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
1,
1,
1,
1,
1,
1,
0,
1,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
0,
1,
1,
1,
0,
1,
1,
0,
1,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
],
[
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0
]
];
const ys = [1, 1, 1, 1, 0, 0, 0, 0, 0];
// Instantiate the training tensors
let xTrain = tf.tensor2d(xs, [9, 51]);
let yTrain = tf.oneHot(tf.tensor1d(ys).toInt(), 2);
// Define the model.
const model = tf.sequential();
// Set up the network layers
model.add(tf.layers.dense({
units: 25,
activation: 'sigmoid',
inputShape: [51]
}));
model.add(tf.layers.dense({
units: 2,
activation: 'softmax',
outputShape: [2]
}));
// Define the optimizer
const optimizer = tf.train.adam(LEARNING_RATE);
// Init the model
model.compile({
optimizer: optimizer,
loss: 'categoricalCrossentropy',
metrics: ['accuracy'],
});
console.log('Entrenando... ');
// Train the model
const history = model.fit(xTrain, yTrain, {
epochs: EPOCHS,
validationData: [xTrain, yTrain],
}).then(() => {
console.log('Fin entrenamiento ! ');
const vectorPrueba1 = [
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
0,
1,
0,
1,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0
];
const vectorPruebaZero = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0
];
const vectorPruebaZero2 = [
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
];
const vectorPrueba1_2 = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
1,
1,
1,
0,
1,
0,
1,
1,
1,
1,
1,
1,
0,
1,
1,
0,
0
];
const vectorPruebaZero3 = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
];
const algo = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
1,
0,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0
];
const unoraro = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0,
0
];
const zeroRaro = [
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
1,
0,
0,
0
];
// Try the model on a value
const input = tf.tensor2d(zeroRaro, [1, 51]);
const predictOut = model.predict(input);
const logits = Array.from(predictOut.dataSync());
console.log('prediction: ', logits, 'SALIDA : ' + predictOut.argMax(-1).dataSync()[0]);
});
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment