ChunML

        # Combine the context vector and the LSTM output
        # Before combined, both have shape of (batch_size, 1, rnn_size),
        # so let's squeeze the axis 1 first
        # After combined, it will have shape of (batch_size, 2 * rnn_size)
        lstm_out = tf.concat([tf.squeeze(context, 1), tf.squeeze(lstm_out, 1)], 1)
        
        # lstm_out now has shape (batch_size, rnn_size)
        lstm_out = self.wc(lstm_out)
        
        # Finally, it is converted back to vocabulary space: (batch_size, vocab_size)

ChunML

for e in range(10):
    accuracy = []
    for batch, (text, label) in enumerate(train_data.take(-1)):
        with tf.GradientTape() as tape:
            logits = model(text)
            label = tf.expand_dims(label, 1)
            loss = loss_func(label, logits)

ChunML

def positional_embedding(pos, model_size):
    PE = np.zeros((1, model_size))
    for i in range(model_size):
        if i % 2 == 0:
            PE[:, i] = np.sin(pos / 10000 ** (i / model_size))
        else:
            PE[:, i] = np.cos(pos / 10000 ** ((i - 1) / model_size))
    return PE

max_length = max(len(data_en[0]), len(data_fr_in[0]))

ChunML

ChunML

    for e in range(50):
        batches = get_batches(in_text, out_text, flags.batch_size, flags.seq_size)
        state_h, state_c = net.zero_state(flags.batch_size)
        
        # Transfer data to GPU
        state_h = state_h.to(device)
        state_c = state_c.to(device)
        for x, y in batches:
            iteration += 1
            

ChunML

import * as tf from '@tensorflow/tfjs';
import React from 'react';
import ReactDOM from 'react-dom';

class CharacterTable {
  constructor(chars) {
    // chars must be a list of unique characters
    this.chars = chars;
    this.charIndices = {};
    this.indicesChar = {};

ChunML

What a ridiculous concept !
[[8, 5, 21, 22, 23]]
quel concept ridicule ! <end>
Your idea is not entirely crazy .
[[24, 25, 6, 26, 27, 28, 1]]
votre idee n est pas completement folle . <end>
A man s worth lies in what he is .
[[5, 29, 30, 31, 32, 9, 8, 7, 6, 1]]
la valeur d un homme reside dans ce qu il est . <end>
What he did is very wrong .

ChunML

Epoch 90 Loss 0.0023
Average elapsed time: 3.77s
A man s worth lies in what he is .
[[5, 29, 30, 31, 32, 9, 8, 7, 6, 1]]
la valeur d un homme reside dans ce qu il est . <end>

Epoch 100 Loss 0.0018
Average elapsed time: 3.76s
What a ridiculous concept !
[[8, 5, 21, 22, 23]]

ChunML

What a ridiculous concept !
[[8, 5, 21, 22, 23]]
ce concept ridicule ! <end>
Your idea is not entirely crazy .
[[24, 25, 6, 26, 27, 28, 1]]
votre idee n est pas completement pas . . . <end>
A man s worth lies in what he is .
[[5, 29, 30, 31, 32, 9, 8, 7, 6, 1]]
la valeur d un homme reside dans ce qu il est . <end>
What he did is very wrong .

ChunML

Epoch 50 Loss 0.0141
Average elapsed time: 4.84s
I can t believe you re giving up .
[[16, 11, 12, 95, 2, 96, 14, 97, 1]]
je n arrive pas a croire que vous abandonniez . <end>

Epoch 60 Loss 0.0063
Average elapsed time: 4.66s
What a ridiculous concept !
[[8, 5, 21, 22, 23]]