Tensorflow LSTM示例输入格式batches2string

import tensorflow as tf

# 假设有两个批次的字符串数据
batch1 = ['Hello', 'World']
batch2 = ['TensorFlow', 'LSTM']

# 构建词汇表
vocab = ['<pad>', '<start>', '<end>', 'H', 'e', 'l', 'o', 'W', 'r', 'd', 'T', 'n', 's', 'F', 'w', 'L', 'S', 'M']

# 将字符串转换为整数序列
def string_to_int(string, vocab):
    return [vocab.index(char) for char in string]

# 将批次的字符串数据转换为整数张量
def batches_to_tensor(batches, vocab):
    tensor = []
    for batch in batches:
        int_sequence = string_to_int(batch, vocab)
        tensor.append(int_sequence)
    return tf.constant(tensor)

# 将字符串批次转换为张量
input_batches = [batch1, batch2]
input_tensor = batches_to_tensor(input_batches, vocab)

print(input_tensor)