如何用Tensorflow实现RNN?

Question

如何用Tensorflow实现RNN

Answer 1

class TextLoader():
def __init__(self, data_dir, batch_size, seq_length, encoding='utf-8'):
self.data_dir = data_dir
self.batch_size = batch_size
self.seq_length = seq_length
self.encoding = encoding
#第一次运行程序时只有input.txt一个文件，剩下两个文件是运行之后产生的
input_file = os.path.join(data_dir, "input.txt")
vocab_file = os.path.join(data_dir, "vocab.pkl")
tensor_file = os.path.join(data_dir, "data.npy")
#如果是第一次执行则调用preprocess函数，否则调用load_preprocessed函数。
if not (os.path.exists(vocab_file) and os.path.exists(tensor_file)):
print("reading text file")
self.preprocess(input_file, vocab_file, tensor_file)
else:
print("loading preprocessed files")
self.load_preprocessed(vocab_file, tensor_file)
self.create_batches()
self.reset_batch_pointer()

def preprocess(self, input_file, vocab_file, tensor_file):
with codecs.open(input_file, "r", encoding=self.encoding) as f:
data = f.read()
#使用Counter函数对输入数据进行统计。counter保存data中每个字符出现的次数
counter = collections.Counter(data)
#对counter进行排序，出现次数最多的排在前面
count_pairs = sorted(counter.items(), key=lambda x: -x[1])
#将data中出现的所有字符保存，这里有65个，所以voacb_size=65
self.chars, _ = zip(*count_pairs)
self.vocab_size = len(self.chars)
#按照字符出现次数多少顺序将chars保存，vocab中存储的是char和顺序，这样方便将data转化为索引
self.vocab = dict(zip(self.chars, range(len(self.chars))))
with open(vocab_file, 'wb') as f:
#保存chars
cPickle.dump(self.chars, f)
#将data中每个字符转化为索引下标。
self.tensor = np.array(list(map(self.vocab.get, data)))
np.save(tensor_file, self.tensor)

def load_preprocessed(self, vocab_file, tensor_file):
#如果是第二次运行，则可以直接读取之前保存的chars和tensor
with open(vocab_file, 'rb') as f:
self.chars = cPickle.load(f)
self.vocab_size = len(self.chars)
self.vocab = dict(zip(self.chars, range(len(self.chars))))
self.tensor = np.load(tensor_file)
self.num_batches = int(self.tensor.size / (self.batch_size *
self.seq_length))

def create_batches(self):
#首先将数据按batch_size切割，然后每个batch_size在按照seq_length进行切割
self.num_batches = int(self.tensor.size / (self.batch_size *
self.seq_length))

if self.num_batches == 0:
assert False, "Not enough data. Make seq_length and batch_size small."

self.tensor = self.tensor[:self.num_batches * self.batch_size * self.seq_length]
xdata = self.tensor
#构造target，这里使用上一个词预测下一个词，所以直接将x向后一个字符即可
ydata = np.copy(self.tensor)
ydata[:-1] = xdata[1:]
ydata[-1] = xdata[0]
#将数据进行切分，这里我们假设数据总长度为10000，batch_size为100， seq_length为10.
# 所以num_batches=10，所以，xdata在reshape之后变成[100, 100],然后在第二个维度上切成10份，
# 所以最终得到[100, 10, 10]的数据
self.x_batches = np.split(xdata.reshape(self.batch_size, -1),
self.num_batches, 1)
self.y_batches = np.split(ydata.reshape(self.batch_size, -1),
self.num_batches, 1)

def next_batch(self):
x, y = self.x_batches[self.pointer], self.y_batches[self.pointer]
self.pointer += 1
return x, y

def reset_batch_pointer(self):
self.pointer = 0