从代码学习深度学习 - LSTM PyTorch版

---

前言

深度学习中的循环神经网络（RNN）及其变种长短期记忆网络（LSTM）在处理序列数据（如文本、时间序列等）方面表现出色。本篇博客将通过一个完整的PyTorch实现，带你从零开始学习如何使用LSTM进行文本生成任务。我们将基于H.G. Wells的《时间机器》数据集，逐步展示数据预处理、模型定义、训练与预测的全过程。通过代码和文字的结合，帮助你深入理解LSTM的实现细节及其在自然语言处理中的应用。

本文的代码分为四个主要部分：

1. 数据加载与预处理（utils_for_data.py）
2. LSTM模型定义（Jupyter Notebook中的模型部分）
3. 训练与预测逻辑（utils_for_train.py）
4. 可视化工具（utils_for_huitu.py）

以下是详细的实现与解析。

---

一、数据加载与预处理

首先，我们需要加载《时间机器》数据集并进行预处理。以下是utils_for_data.py中的完整代码及其功能说明。

1.1 代码实现

import random
import re
import torch
from collections import Counter

def read_time_machine():
    """将时间机器数据集加载到文本行的列表中"""
    with open('timemachine.txt', 'r') as f:
        lines = f.readlines()
    return [re.sub('[^A-Za-z]+', ' ', line).strip().lower() for line in lines]

def tokenize(lines, token='word'):
    """将文本行拆分为单词或字符词元"""
    if token == 'word':
        return [line.split() for line in lines]
    elif token == 'char':
        return [list(line) for line in lines]
    else:
        print(f'错误：未知词元类型：{
     token}')

def count_corpus(tokens):
    """统计词元的频率"""
    if not tokens:
        return Counter()
    if isinstance(tokens[0], list):
        flattened_tokens = [token for sublist in tokens for token in sublist]
    else:
        flattened_tokens = tokens
    return Counter(flattened_tokens)

class Vocab:
    """文本词表类，用于管理词元及其索引的映射关系"""
    def __init__(self, tokens=None, min_freq=0, reserved_tokens=None):
        self.tokens = tokens if tokens is not None else []
        self.reserved_tokens = reserved_tokens if reserved_tokens is not None else []
        counter = self._count_corpus(self.tokens)
        self._token_freqs = sorted(counter.items(), key=lambda x: x[1], reverse=True)
        self.idx_to_token = ['<unk>'] + self.reserved_tokens
        self.token_to_idx = {
   token: idx for idx, token in enumerate(self.idx_to_token)}
        for token, freq in self._token_freqs:
            if freq < min_freq:
                break
            if token not in self.token_to_idx:
                self.idx_to_token.append(token)
                self.token_to_idx[token] = len(self.idx_to_token) - 1

    @staticmethod
    def _count_corpus(tokens):
        if not tokens:
            return Counter()
        if isinstance(tokens[0], list):
            tokens = [token for sublist in tokens for token in sublist]
        return Counter(tokens)

    def __len__(self):
        return len(self.idx_to_token)

    def __getitem__(self, tokens):
        if not isinstance(tokens, (list, tuple)):
            return self.token_to_idx.get(tokens, self.unk)
        return [self[token] for token in tokens]

    def to_tokens(self, indices):
        if not isinstance(indices, (list, tuple)):
            return self.idx_to_token[indices]
        return [self.idx_to_token[index] for index in indices]

    @property
    def unk(self):
        return 0

    @property
    def token_freqs(self):
        return self._token_freqs

def load_corpus_time_machine(max_tokens=-1):
    lines = read_time_machine()
    tokens = tokenize(lines, 'char')
    vocab = Vocab(tokens)
    corpus = [vocab[token] for line in tokens for token in line]
    if max_tokens > 0:
        corpus = corpus[:max_tokens]
    return corpus, vocab

def seq_data_iter_random(corpus, batch_size, num_steps):
    offset = random.randint(0, num_steps - 1)
    corpus = corpus[offset:]
    num_subseqs = (len(corpus) - 1) // num_steps
    initial_indices = list(range(0, num_subseqs * num_steps, num_steps))
    random.shuffle(initial_indices)

    def data(pos):
        return corpus[pos:pos + num_steps]

    num_batches = num_subseqs // batch_size
    for i in range(0, batch_size * num_batches, batch_size):
        initial_indices_per_batch = initial_indices[i:i + batch_size]
        X = [data(j) for j in initial_indices_per_batch]
        Y = [data(j + 1) for j in initial_indices_per_batch]
        yield torch.tensor(X), torch.tensor(Y)

def seq_data_iter_sequential(corpus, batch_size, num_steps):
    offset = random.randint(0, num_steps)
    num_tokens = ((len(corpus) - offset - 1) // batch_size) *