chatbot.py

# -*- coding: utf-8 -*-
"""
Created on Fri Jan  4 14:54:52 2019

@author: slacki
"""

#Using Python 3.5 which allows tensorflow work

#Building a Chatbot with deep NLP
import numpy as np
import tensorflow as tf
import re
import time 

### Part 1 - Data Preprocessing

#importing dataset
lines = open('movie_lines.txt', encoding = 'utf-8', errors = 'ignore').read().split('\n')
conversations = open('movie_conversations.txt', encoding = 'utf-8', errors = 'ignore').read().split('\n')

#Creating a dicitonary that maps each line and its id
id2line = {}
for line in lines:
    _line = line.split(' +++$+++ ') #_line means that this is a local variable
    id2line[_line[0]] = _line[len(_line)-1] #Take first and last string from the lines array and map to eachother
    
 #I'm not sure why he used this logic rather than the one I did above
 #   if len(_line) == 5: 
 #   id2line[_line[0]] = _line[4]
 
 #Creating a list of all of the conversations
conversations_ids = []
for conversation in conversations[:-1]:
    _conversation = conversation.split(' +++$+++ ')[-1][1:-1].replace("'","").replace(" ", "") #This is a weird way of just choosing the first and last element from the list 
    conversations_ids.append(_conversation.split(','))
    
# Getting the questions and answers separately
questions = []
answers = []
for conversation in conversations_ids:
    for i in range(len(conversation) - 1):
        questions.append(id2line[conversation[i]])
        answers.append(id2line[conversation[i + 1]])

#Doing a first cleaning of the texts. Consider not including . ? , 
def clean_text(text):
    text = text.lower()
    text = re.sub(r"i'm", "i am", text)
    text = re.sub(r"he's", "he is", text)
    text = re.sub(r"she's", "she is", text)
    text = re.sub(r"that's", "that is", text)
    text = re.sub(r"what's", "what is", text)
    text = re.sub(r"where's", "where is", text)
    text = re.sub(r"\'ll", " will", text)
    text = re.sub(r"\'ve", " have", text)
    text = re.sub(r"\'re", " are", text)
    text = re.sub(r"\'d", " would", text)
    text = re.sub(r"won't", "will not", text)
    text = re.sub(r"can't", "cannot", text)
    text = re.sub(r"[-()\"#/@;<>{}+=~|.?,)]", "", text)
    return text

#Cleaning the questions, did differently than in the video
clean_questions = []
for question in questions:
    _cleaned1 = clean_text(question)
    clean_questions.append(_cleaned1)

#Cleaning the answers
clean_answers = []
for answer in answers:
    _cleaned2 = clean_text(answer)
    clean_answers.append(_cleaned2)
    
#Creating a dictionary that maps each word to its number of occurences
#Mind that the numbers are different than in the video. If the bot doesn't work in the end, it will be worth reviewing my code 
word2count = {}
for question in clean_questions:
    for word in question.split():
        if word not in word2count:
            word2count[word] = 1
        else:
            word2count[word] += 1
for answer in clean_answers:
    for word in answer.split():
        if word not in word2count:
            word2count[word] = 1
        else:
            word2count[word] += 1

#Creating two dictionaries that map the questions words and the answers words to a unique integer. Genius. Every word has a number
treshold = 20
questionswords2int = {}
word_number = 0
for word, count in word2count.items():  #.items() is used to do things with dictionaries innit
    if count >= treshold:
        questionswords2int[word] = word_number
        word_number += 1
answerswords2int = {}
word_number = 0
for word, count in word2count.items():  #.items() is used to do things with dictionaries innit
    if count >= treshold:
        answerswords2int[word] = word_number
        word_number += 1

#Adding the last tokens to these two dictionaries
tokens = ['<PAD>','<EOS>', '<OUT>', '<SOS>']
for token in tokens: 
    questionswords2int[token] = len(questionswords2int) + 1
for token in tokens: 
    answerswords2int[token] = len(answerswords2int) + 1

#Creating the inverse dictionary of the answers2wordsint dictionary
answerints2word = {w_i: w for w, w_i in answerswords2int.items()}

#Adding the End of String (EOS) token to the end of every answer
for i in range(len(clean_answers)):
    clean_answers[i] += ' <EOS>'

#Translating all the questions and the answers into integers
# and replacing all words that were filtered out by <OUT>
questions_to_int = []
for question in clean_questions:
    ints = []
    for word in question.split():
        if word not in questionswords2int:
            ints.append(questionswords2int['<OUT>']) #takes from the inverse dictionary
        else:
            ints.append(questionswords2int[word])
    questions_to_int.append(ints)
        
answers_to_int = []
for answer in clean_answers:
    ints = []
    for word in answer.split():
        if word not in answerswords2int:
            ints.append(answerswords2int['<OUT>']) #takes from the inverse dictionary
        else:
            ints.append(answerswords2int[word])
    answers_to_int.append(ints)

#Sorting questions and answers by the length of questions because that reduces loss or something. And makes things more efficient
sorted_cleaned_questions = []
sorted_cleaned_answers = []
for length in range(1, 25 + 1):
    for i in enumerate(questions_to_int):
        if len(i[1]) == length:
            sorted_cleaned_questions.append(questions_to_int[i[0]])
            sorted_cleaned_answers.append(answers_to_int[i[0]])

########### PART 2 - BUILDING SEQ2SEQ MODEL #####################
            
# Creatin gplaceholders for the inputs and the targets
def model_inputs():
    inputs = tf.placeholder(tf.int32, [None, None], name = 'input')
    targets = tf.placeholder(tf.int32, [None, None], name = 'target')
    lr = tf.placeholder(tf.float32, name = 'learning_rate')
    keep_prob = tf.placeholder(tf.float32, name = 'keep_prob')
    return inputs, targets, lr, keep_prob

# Preprocessing the targets
def preprocess_targets(targets, word2int, batch_size):
    left_side = tf.fill([batch_size, 1], word2int['<SOS>'])
    right_side = tf.strided_slice(targets, [0,0], [batch_size, -1], [1,1])
    preprocessed_targets = tf.concat([left_side, right_side], 1)
    return preprocessed_targets
    
#Creating the Encoder RNN Layer
def encoder_rnn(rnn_inputs, rnn_size, num_layers, keep_prob, sequence_length):
    lstm = tf.contrib.rnn.BasicLSTMCell(rnn_size)
    lstm_dropout = tf.contrib.rnn.DropoutWrapper(lstm, input_keep_prob = keep_prob)
    encoder_cell = tf.contrib.rnn.MultiRNNCell([lstm_dropout] * num_layers)
    encoder_output, encoder_state = tf.nn.bidirectional_dynamic_rnn(cell_fw = encoder_cell,
                                                                    cell_bw = encoder_cell,
                                                                    sequence_length = sequence_length,
                                                                    inputs = rnn_inputs,
                                                                    dtype = tf.float32)
    return encoder_state

#Decoding the training set
def decode_training_set(encoder_state, decoder_cell, decoder_embedded_input, sequence_length, decoding_scope, output_function, keep_prob, batch_size):
    attention_states = tf.zeros([batch_size, 1, decoder_cell.output_size])
    attention_keys, attention_values, attention_score_function, attention_construct_function = tf.contrib.seq2seq.prepare_attention(attention_states, attention_option = 'bahdanau', num_units = decoder_cell.output_size)
    training_decoder_function = tf.contrib.seq2seq.attention_decoder_fn_train(encoder_state[0],
                                                                              attention_keys,
                                                                              attention_values,
                                                                              attention_score_function,
                                                                              attention_construct_function,
                                                                              name = 'attn_dec_train')
    decoder_output, decoder_final_state, decoder_final_context_state =  tf.contrib.seq2seq.dynamic_rnn_decoder(decoder_cell, 
                                                                                                               training_decoder_function,
                                                                                                               decoder_embedded_input,
                                                                                                               sequence_length,
                                                                                                               scope = decoding_scope)
    decoder_output_dropout = tf.nn.dropout(decoder_output, keep_prob)
    return output_function(decoder_output_dropout)

# Decoding the test/validation set
def decode_test_set(encoder_state, decoder_cell, decoder_embeddings_matrix, sos_id, eos_id, maximum_length, num_words, decoding_scope, output_function, keep_prob, batch_size):
    attention_states = tf.zeros([batch_size, 1, decoder_cell.output_size])
    attention_keys, attention_values, attention_score_function, attention_construct_function = tf.contrib.seq2seq.prepare_attention(attention_states, attention_option = "bahdanau", num_units = decoder_cell.output_size)
    test_decoder_function = tf.contrib.seq2seq.attention_decoder_fn_inference(output_function,
                                                                              encoder_state[0],
                                                                              attention_keys,
                                                                              attention_values,
                                                                              attention_score_function,
                                                                              attention_construct_function,
                                                                              decoder_embeddings_matrix,
                                                                              sos_id,
                                                                              eos_id,
                                                                              maximum_length,
                                                                              num_words,
                                                                              name = "attn_dec_inf")
    test_predictions, decoder_final_state, decoder_final_context_state = tf.contrib.seq2seq.dynamic_rnn_decoder(decoder_cell,
                                                                                                                test_decoder_function,
                                                                                                                scope = decoding_scope)
    return test_predictions

# Creating the Decoder RNN
def decoder_rnn(decoder_embedded_input, decoder_embeddings_matrix, encoder_state, num_words, sequence_length, rnn_size, num_layers, word2int, keep_prob, batch_size):
    with tf.variable_scope("decoding") as decoding_scope:
        lstm = tf.contrib.rnn.BasicLSTMCell(rnn_size)
        lstm_dropout = tf.contrib.rnn.DropoutWrapper(lstm, input_keep_prob = keep_prob)
        decoder_cell = tf.contrib.rnn.MultiRNNCell([lstm_dropout] * num_layers)
        weights = tf.truncated_normal_initializer(stddev = 0.1)
        biases = tf.zeros_initializer()
        output_function = lambda x: tf.contrib.layers.fully_connected(x,
                                                                      num_words,
                                                                      None,
                                                                      scope = decoding_scope,
                                                                      weights_initializer = weights,
                                                                      biases_initializer = biases)
        training_predictions = decode_training_set(encoder_state,
                                                   decoder_cell,
                                                   decoder_embedded_input,
                                                   sequence_length,
                                                   decoding_scope,
                                                   output_function,
                                                   keep_prob,
                                                   batch_size)
        decoding_scope.reuse_variables()
        test_predictions = decode_test_set(encoder_state,
                                           decoder_cell,
                                           decoder_embeddings_matrix,
                                           word2int['<SOS>'],
                                           word2int['<EOS>'],
                                           sequence_length - 1,
                                           num_words,
                                           decoding_scope,
                                           output_function,
                                           keep_prob,
                                           batch_size)
    return training_predictions, test_predictions
    
# Building the seq2seq model
def seq2seq_model(inputs, targets, keep_prob, batch_size, sequence_length, answers_num_words, questions_num_words, encoder_embedding_size, decoder_embedding_size, rnn_size, num_layers, questionswords2int):
    encoder_embedded_input = tf.contrib.layers.embed_sequence(inputs,
                                                              answers_num_words + 1,
                                                              encoder_embedding_size,
                                                              initializer = tf.random_uniform_initializer(0, 1))
    encoder_state = encoder_rnn(encoder_embedded_input, rnn_size, num_layers, keep_prob, sequence_length)
    preprocessed_targets = preprocess_targets(targets, questionswords2int, batch_size)
    decoder_embeddings_matrix = tf.Variable(tf.random_uniform([questions_num_words + 1, decoder_embedding_size], 0, 1))
    decoder_embedded_input = tf.nn.embedding_lookup(decoder_embeddings_matrix, preprocessed_targets)
    training_predictions, test_predictions = decoder_rnn(decoder_embedded_input,
                                                         decoder_embeddings_matrix,
                                                         encoder_state,
                                                         questions_num_words,
                                                         sequence_length,
                                                         rnn_size,
                                                         num_layers,
                                                         questionswords2int,
                                                         keep_prob,
                                                         batch_size)
    return training_predictions, test_predictions

################ Part 3 - training the seq2seq model #########

# Setting the Hyperparameters
epochs = 100 #setting the number of batches which affects speed of calculation
batch_size = 64
rnn_size = 512
num_layers = 3
encoding_embedding_size = 512
decoding_embedding_size = 512
learning_rate = 0.01
learning_rate_decay = 0.9
min_learning_rate = 0.0001
keep_probability = 0.5    

# Defining a session
tf.reset_default_graph()
session = tf.InteractiveSession()

# Loading the model inputs
inputs, targets, lr, keep_prob = model_inputs()

# Setting the sequence length
sequence_length = tf.placeholder_with_default(25, None, name = 'sequence_length')

# Getting the shape of the inputs tensor
input_shape = tf.shape(inputs)

# Getting the training and test predictions
training_predictions, test_predictions = seq2seq_model(tf.reverse(inputs, [-1]),
                                                       targets,
                                                       keep_prob,
                                                       batch_size,
                                                       sequence_length,
                                                       len(answerswords2int),
                                                       len(questionswords2int),
                                                       encoding_embedding_size,
                                                       decoding_embedding_size,
                                                       rnn_size,
                                                       num_layers,
                                                       questionswords2int)

# Setting up the loss error, the optimiser and gradient clipping
with tf.name_scope("optimization"):
    loss_error = tf.contrib.seq2seq.sequence_loss(training_predictions,
                                                  targets,
                                                  tf.ones([input_shape[0], sequence_length]))
    optimizer = tf.train.AdamOptimizer(learning_rate)
    gradients = optimizer.compute_gradients(loss_error)
    clipped_gradients = [(tf.clip_by_value(grad_tensor, -5., 5.), grad_variable) for grad_tensor, grad_variable in gradients if grad_tensor is not None]
    optimizer_gradient_clipping = optimizer.apply_gradients(clipped_gradients)
    
# Padding the sequences with the <PAD> token
# Question (who, are, you)
#Answer : (<SOS>, I, am, a, bot, ., <EOS>
#this requires the question to have adding such as (who, are, you, <PAD>, <PAD>, <PAD>)
# Now input has the same lenght as the output
def apply_padding(batch_of_sequences, word2int):
    max_sequence_length = max([len(sequence) for sequence in batch_of_sequences])
    return [sequence + [word2int['<PAD>']] * (max_sequence_length - len(sequence)) for sequence in batch_of_sequences]

# Splitting the data into batches of questions and answers
def split_into_batches(questions, answers, batch_size):
    for batch_index in range(0, len(questions) // batch_size):
        start_index = batch_index * batch_size
        questions_in_batch = questions[start_index : start_index + batch_size]
        answers_in_batch = answers[start_index : start_index + batch_size]
        padded_questions_in_batch = np.array(apply_padding(questions_in_batch, questionswords2int))
        padded_answers_in_batch = np.array(apply_padding(answers_in_batch, answerswords2int))
        yield padded_questions_in_batch, padded_answers_in_batch
        
# Splitting the questions and answers into training and validation sets
training_validation_split = int(len(sorted_cleaned_questions) * 0.15)
training_questions = sorted_cleaned_questions[training_validation_split:]
training_answers = sorted_cleaned_answers[training_validation_split:]
validation_questions = sorted_cleaned_questions[:training_validation_split]
validation_answers = sorted_cleaned_answers[:training_validation_split]

# TRAAAINING
batch_index_check_training_loss = 100
batch_index_check_validation_loss = ((len(training_questions)) // batch_size // 2) - 1
total_training_loss_error = 0 
list_validation_loss_error = []
early_stopping_check = 0
early_stopping_stop = 1000
checkpoint = "chatbot_weights.ckpt"
session.run(tf.global_variables_initializer())
for epoch in range(1, epochs + 1):
    for batch_index, (padded_questions_in_batch, padded_answers_in_batch) in enumerate(split_into_batches(training_questions, training_answers, batch_size)):
        starting_time = time.time()
        _, batch_training_loss_error = session.run([optimizer_gradient_clipping, loss_error], {inputs: padded_questions_in_batch,
                                                   targets: padded_answers_in_batch,
                                                   lr: learning_rate,
                                                   sequence_length: padded_answers_in_batch.shape[1],
                                                   keep_prob: keep_probability})
    total_training_loss_error += batch_training_loss_error
    ending_time = time.time()
    batch_time = ending_time - starting_time
    if batch_index % batch_index_check_training_loss == 0:
        print('Epoch: {:>3}/{}, Batch: {:>4}/{}, Training Loss Error: {:>6.3f}, Training Time on 100 Batches: {:d} seconds'.format(epoch,
                                                                                                                                   epochs,
                                                                                                                                   batch_index,
                                                                                                                                   len(training_questions) // batch_size,
                                                                                                                                   total_training_loss_error / batch_index_check_training_loss,
                                                                                                                                   int(batch_time * batch_index_check_training_loss)))
        total_training_loss_error = 0
    if batch_index % batch_index_check_validation_loss == 0 and batch_index > 0:
        total_validation_loss_error = 0
        starting_time = time.time()
        for batch_index_validation, (padded_questions_in_batch, padded_answers_in_batch) in enumerate(split_into_batches(validation_questions, validation_answers, batch_size)):
            batch_validation_loss_error = session.run(loss_error, {inputs: padded_questions_in_batch,
                                                       targets: padded_answers_in_batch,
                                                       lr: learning_rate,
                                                       sequence_length: padded_answers_in_batch.shape[1],
                                                       keep_prob: 1})
        total_validation_loss_error += batch_validation_loss_error
        ending_time = time.time()
        batch_time = ending_time - starting_time
        average_validation_loss_error = total_validation_loss_error / (len(validation_questions) / batch_size)
        print('Validation Loss Error: {:>6.3f}, Batch Validation Time: {:d} seconds'.format(average_validation_loss_error, int(batch_time)))
        learning_rate *= learning_rate_decay
        if learning_rate < min_learning_rate:
            learning_rate = min_learning_rate
        list_validation_loss_error.append(average_validation_loss_error)
        if average_validation_loss_error <= min(list_validation_loss_error):
            print('I speak better now!!')
            early_stopping_check = 0
            saver = tf.train.Saver()
            saver.save(session, checkpoint)
        else: 
            print("Sorry I do not speak better, I need to practice more")
            early_stopping_check += 1
            if early_stopping_check == early_stopping_stop:
                break
    if early_stopping_check == early_stopping_stop:
        print("My apologies, I cannot speak better anymore. This is the best I can do")
        break
print("Game over")