200 lines
6.0 KiB
Markdown
200 lines
6.0 KiB
Markdown
|
# 使用 Keras TensorFlow 预测莎士比亚文本
|
||
|
|
|
||
|
|
> 原文：<https://www.askpython.com/python/examples/predict-shakespearean-text>
|
||
|
|
|
||
|
|
嘿伙计们！在本教程中，我们将了解如何使用 Python 中的 Keras TensorFlow API 创建递归神经网络模型来预测莎士比亚文本。
|
||
|
|
|
||
|
|
***也读作:[股价预测使用 Python](https://www.askpython.com/python/examples/stock-price-prediction-python)***
|
||
|
|
|
||
|
|
为了产生新的文本，我们将使用定制的 RNN 模型来训练 GitHub 莎士比亚文本数据集 。
|
||
|
|
|
||
|
|
* * *
|
||
|
|
|
||
|
|
## **第一步:导入库**
|
||
|
|
|
||
|
|
我们利用了一些最流行的深度学习库。Sweetviz 是一个新的软件包，可自动进行探索性数据分析，尤其有利于分析我们的训练数据集。
|
||
|
|
|
||
|
|
```py
|
||
|
|
pip install sweetviz
|
||
|
|
import numpy as np
|
||
|
|
import pandas as pd
|
||
|
|
import matplotlib.pyplot as plt
|
||
|
|
import tensorflow as tf
|
||
|
|
from tensorflow import keras
|
||
|
|
import sweetviz as sw
|
||
|
|
import seaborn as sns
|
||
|
|
sns.set()
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
## **第二步:加载数据集**
|
||
|
|
|
||
|
|
```py
|
||
|
|
shakespeare_url='https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt'
|
||
|
|
filepath=keras.utils.get_file('shakespeare.txt',shakespeare_url)
|
||
|
|
with open(filepath) as f:
|
||
|
|
shakespeare_text=f.read()
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
```py
|
||
|
|
Downloading data from https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt
|
||
|
|
1122304/1115394 [==============================] - 0s 0us/step
|
||
|
|
1130496/1115394 [==============================] - 0s 0us/step
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
既然我们已经将数据集下载到 Python 笔记本中，我们需要在利用它进行训练之前对其进行预处理。
|
||
|
|
|
||
|
|
## **第三步:预处理数据集**
|
||
|
|
|
||
|
|
标记化是将冗长的文本字符串分成较小部分或标记的过程。较大的文本块可以被标记成句子，然后变成单词。
|
||
|
|
|
||
|
|
预处理还包括从生成的标记中删除标点符号。
|
||
|
|
|
||
|
|
```py
|
||
|
|
tokenizer=keras.preprocessing.text.Tokenizer(char_level=True)
|
||
|
|
tokenizer.fit_on_texts(shakespeare_text)
|
||
|
|
|
||
|
|
max_id=len(tokenizer.word_index)
|
||
|
|
dataset_size=tokenizer.document_count
|
||
|
|
[encoded]=np.array(tokenizer.texts_to_sequences([shakespeare_text]))-1
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
## **步骤 4:准备数据集**
|
||
|
|
|
||
|
|
我们将使用`tf.data.Dataset`,它通常对大量的元素有用，比如大量的文本数据。
|
||
|
|
|
||
|
|
`Dataset.repeat()`遍历数据集并重复数据集指定的次数。`window()`就像一个滑动窗口，每次滑动指定数量的窗口，进行反复迭代。
|
||
|
|
|
||
|
|
```py
|
||
|
|
train_size=dataset_size*90//100
|
||
|
|
dataset=tf.data.Dataset.from_tensor_slices(encoded[:train_size])
|
||
|
|
|
||
|
|
n_steps=100
|
||
|
|
window_length=n_steps+1
|
||
|
|
dataset=dataset.repeat().window(window_length,shift=1,drop_remainder=True)
|
||
|
|
|
||
|
|
dataset=dataset.flat_map(lambda window: window.batch(window_length))
|
||
|
|
|
||
|
|
batch_size=32
|
||
|
|
dataset=dataset.shuffle(10000).batch(batch_size)
|
||
|
|
dataset=dataset.map(lambda windows: (windows[:,:-1],windows[:,1:]))
|
||
|
|
dataset=dataset.map(lambda X_batch,Y_batch: (tf.one_hot(X_batch,depth=max_id),Y_batch))
|
||
|
|
dataset=dataset.prefetch(1)
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
## **第五步:建立模型**
|
||
|
|
|
||
|
|
模型构建非常简单。我们将创建一个顺序模型，并向该模型添加具有某些特征的层。
|
||
|
|
|
||
|
|
```py
|
||
|
|
model=keras.models.Sequential()
|
||
|
|
model.add(keras.layers.GRU(128,return_sequences=True,input_shape=[None,max_id]))
|
||
|
|
model.add(keras.layers.GRU(128,return_sequences=True))
|
||
|
|
model.add(keras.layers.TimeDistributed(keras.layers.Dense(max_id,activation='softmax')))
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
接下来，我们将编译模型并在数据集上拟合模型。我们将使用`Adam`优化器，但是你也可以根据你的喜好使用其他可用的优化器。
|
||
|
|
|
||
|
|
```py
|
||
|
|
model.compile(loss='sparse_categorical_crossentropy',optimizer='adam')
|
||
|
|
history=model.fit(dataset,steps_per_epoch=train_size // batch_size,epochs=1)
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
```py
|
||
|
|
31370/31370 [==============================] - 1598s 51ms/step - loss: 0.9528
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
## **第六步:测试模型**
|
||
|
|
|
||
|
|
我们在下面提到的代码片段中定义了一些函数。这些函数将根据我们定义的模型预处理和准备输入数据，并预测下一个字符，直到指定的字符数。
|
||
|
|
|
||
|
|
```py
|
||
|
|
def preprocess(texts):
|
||
|
|
X=np.array(tokenizer.texts_to_sequences(texts))-1
|
||
|
|
return tf.one_hot(X,max_id)
|
||
|
|
|
||
|
|
def next_char(text,temperature=1):
|
||
|
|
X_new=preprocess([text])
|
||
|
|
y_proba=model.predict(X_new)[0,-1:,:]
|
||
|
|
rescaled_logits=tf.math.log(y_proba)/temperature
|
||
|
|
char_id=tf.random.categorical(rescaled_logits,num_samples=1)+1
|
||
|
|
return tokenizer.sequences_to_texts(char_id.numpy())[0]
|
||
|
|
|
||
|
|
def complete_text(text,n_chars=50,temperature=1):
|
||
|
|
for _ in range(n_chars):
|
||
|
|
text+=next_char(text,temperature)
|
||
|
|
return text
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
让我们使用下面提到的代码来预测某个字母或单词的文本。
|
||
|
|
|
||
|
|
```py
|
||
|
|
print("Some predicted texts for letter 'D' are as follows:\n ")
|
||
|
|
for i in range(3):
|
||
|
|
print(complete_text('d'))
|
||
|
|
print()
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
```py
|
||
|
|
Some predicted texts for letter 'D' are as follows:
|
||
|
|
|
||
|
|
d, swalld tell you in mine,
|
||
|
|
the remeiviss if i shou
|
||
|
|
|
||
|
|
dima's for me, sir, to comes what this roguty.
|
||
|
|
|
||
|
|
dening to girl, ne'er i was deckong?
|
||
|
|
which never be
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
```py
|
||
|
|
print("Some predicted texts for word 'SHINE' are as follows:\n ")
|
||
|
|
for i in range(3):
|
||
|
|
print(complete_text('shine'))
|
||
|
|
print()
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
**输出:**
|
||
|
|
|
||
|
|
```py
|
||
|
|
Some predicted texts for word 'SHINE' are as follows:
|
||
|
|
|
||
|
|
shine on here is your viririno penaite the cursue,
|
||
|
|
i'll
|
||
|
|
|
||
|
|
shine yet it the become done to-k
|
||
|
|
make you his ocrowing
|
||
|
|
|
||
|
|
shine dises'-leck a word or my head
|
||
|
|
not oning,
|
||
|
|
so long
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
* * *
|
||
|
|
|
||
|
|
## **结论**
|
||
|
|
|
||
|
|
恭喜你！您刚刚学习了如何使用 RNN 构建一个莎士比亚文本预测器。希望你喜欢它！😇
|
||
|
|
|
||
|
|
喜欢这个教程吗？无论如何，我建议你看一下下面提到的教程:
|
||
|
|
|
||
|
|
1. [利用 Python 预测股价](https://www.askpython.com/python/examples/stock-price-prediction-python)
|
||
|
|
2. [用 Python 进行加密价格预测](https://www.askpython.com/python/examples/crypto-price-prediction)
|
||
|
|
3. [利用 Python 进行股票价格预测](https://www.askpython.com/python/examples/stock-price-prediction-python)
|
||
|
|
4. [Python 中的票房收入预测——简单易行](https://www.askpython.com/python/examples/box-office-revenue-prediction)
|
||
|
|
|
||
|
|
感谢您抽出时间！希望你学到了新的东西！！😄
|
||
|
|
|
||
|
|
* * *
|