NLP——常见任务的批量加载2.0

目标：针对NLP子任务，如文本分类、命名实体识别、文本匹配、关系抽取等，如何使用keras批量加载训练集、验证集或测试集，来提升训练或预测效率？

参考博客：NLP——如何批量加载数据

1、NER任务的数据生成器

import numpy as np
from bert4keras.snippets import sequence_padding, DataGenerator
from bert4keras.tokenizers import Tokenizer
from loguru import loggerdef load_ner_data(filename):"""功能: 加载数据, 单条格式：[text, (start, end, label), (start, end, label), ...]，意味着text[start:end + 1]是类型为label的实体。:param filename: 数据路径:return:"""ner_data, label_map = [], set()with open(filename, "r", encoding="utf-8") as file:f = file.read()for l in f.split("\n\n"):if not l:continued = [""]for i, c in enumerate(l.split("\n")):char, flag = c.split(" ")d[0] += charif flag[0] == "B":d.append([i, i, flag[2:]])label_map.add(flag[2:])elif flag[0] == "I":d[-1][1] = iner_data.append(d)return ner_data, label_mapclass NerDataGenerator(DataGenerator):def __init__(self, max_length, dict_path, data, categories, batch_size, buffer_size):# 子类继承父类, 并进行初始化super(NerDataGenerator, self).__init__(data=data, batch_size=batch_size, buffer_size=buffer_size)self.max_length = max_lengthself.tokenizer = Tokenizer(dict_path, do_lower_case=True)self.categories = categoriesdef __iter__(self, random=False):batch_token_ids, batch_segment_ids, batch_labels = [], [], []for is_end, d in self.sample(random):tokens = self.tokenizer.tokenize(d[0], maxlen=self.max_length)mapping = self.tokenizer.rematch(d[0], tokens)start_mapping = {j[0]: i for i, j in enumerate(mapping) if j}end_mapping = {j[-1]: i for i, j in enumerate(mapping) if j}token_ids = self.tokenizer.tokens_to_ids(tokens)segment_ids = [0] * len(token_ids)labels = np.zeros(len(token_ids), dtype=int)for start, end, label in d[1:]:if start in start_mapping and end in end_mapping:start = start_mapping[start]end = end_mapping[end]labels[start] = self.categories.index(label) * 2 + 1labels[start + 1:end + 1] = self.categories.index(label) * 2 + 2batch_token_ids.append(token_ids)batch_segment_ids.append(segment_ids)batch_labels.append(labels)if len(batch_token_ids) == self.batch_size or is_end:batch_token_ids = sequence_padding(batch_token_ids)batch_segment_ids = sequence_padding(batch_segment_ids)batch_labels = sequence_padding(batch_labels)yield [batch_token_ids, batch_segment_ids], batch_labelsbatch_token_ids, batch_segment_ids, batch_labels = [], [], []if __name__ == "__main__":# 命名实体识别数据生成器train_data, categories = load_ner_data('./china-people-daily-ner-corpus/example.train')ner_data_generator = NerDataGenerator(max_length=128,dict_path="./vocab.txt",data=train_data,categories=list(sorted(categories)),batch_size=32,buffer_size=None)flag = -1for x_true, y_true in ner_data_generator:# 输出每个batch中的数据for idx in range(len(x_true[0])):flag += 1logger.info("第{0}条文本: \"{1}\"".format(flag, train_data[flag][0]))logger.info("word2id: {0}".format(str(list(x_true[0][idx]))))logger.info("mask: {0}".format(str(list(x_true[1][idx]))))logger.info("label: {0}\n\n".format(str(list(y_true[idx]))))break

运行结果（前10行）：

2021-12-24 00:06:44.949 | INFO     | __main__:<module>:164 - 第0条文本: "海钓比赛地点在厦门与金门之间的海域。"
2021-12-24 00:06:44.950 | INFO     | __main__:<module>:165 - word2id: [101, 3862, 7157, 3683, 6612, 1765, 4157, 1762, 1336, 7305, 680, 7032, 7305, 722, 7313, 4638, 3862, 1818, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.950 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.950 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.950 | INFO     | __main__:<module>:164 - 第1条文本: "这座依山傍水的博物馆由国内一流的设计师主持设计，整个建筑群精美而恢宏。"
2021-12-24 00:06:44.950 | INFO     | __main__:<module>:165 - word2id: [101, 6821, 2429, 898, 2255, 988, 3717, 4638, 1300, 4289, 7667, 4507, 1744, 1079, 671, 3837, 4638, 6392, 6369, 2360, 712, 2898, 6392, 6369, 8024, 3146, 702, 2456, 5029, 5408, 5125, 5401, 5445, 2612, 2131, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.950 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.951 | INFO     | __main__:<module>:164 - 第2条文本:
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:165 - word2id: [101, 852, 868, 711, 671, 702, 1066, 772, 1054, 1447, 510, 782, 3696, 1062, 789, 8024, 2418, 2496, 5541, 2577, 2160, 7333, 8024, 4696, 3633, 976, 1168, 100, 1044, 1921, 678, 722, 2569, 5445, 2569, 8024, 1400, 1921, 678, 722, 727, 5445, 727, 100, 8024, 3909, 1265, 702, 782, 4638, 1399, 1164, 2533, 1927, 1469, 2143, 6802, 2650, 1599, 8024, 2828, 3121, 7484, 1920, 689, 3030, 1762, 7674, 855, 8024, 6821, 3416, 2798, 5543, 6631, 6632, 5632, 2769, 8024, 3030, 5564, 686, 921, 8024, 3300, 2792, 868, 711, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.951 | INFO     | __main__:<module>:164 - 第3条文本:
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:165 - word2id: [101, 1762, 1355, 6809, 1744, 2157, 8024, 2593, 3131, 924, 7372, 1282, 1146, 3249, 1350, 8024, 2347, 2768, 711, 4852, 833, 924, 7397, 860, 5143, 4638, 7028, 6206, 5299, 2768, 6956, 1146, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.951 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.952 | INFO     | __main__:<module>:164 - 第4条文本:
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:165 - word2id: [101, 3189, 915, 697, 1744, 1744, 1079, 3124, 2229, 6963, 1041, 4007, 1359, 3144, 8024, 2226, 5052, 3189, 915, 1068, 5143, 4680, 1184, 3221, 1325, 1380, 3297, 881, 3198, 3309, 8024, 852, 1071, 5546, 2483, 2595, 679, 6241, 5632, 3209, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:167 - label: [0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.952 | INFO     | __main__:<module>:164 - 第5条文本: "克马尔的女儿让娜今年读五年级，她所在的班上有30多名同学，该班的“家委会”由10名家长组成。"
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:165 - word2id: [101, 1046, 7716, 2209, 4638, 1957, 1036, 6375, 2025, 791, 2399, 6438, 758, 2399, 5277, 8024, 1961, 2792, 1762, 4638, 4408, 677, 3300, 8114, 1914, 1399, 1398, 2110, 8024, 6421, 4408, 4638, 100, 2157, 1999, 833, 100, 4507, 8108, 1399, 2157, 7270, 5299, 2768, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:167 - label: [0, 5, 6, 6, 0, 0, 0, 5, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.952 | INFO     | __main__:<module>:164 - 第6条文本: "参加步行的有男有女，有年轻人，也有中年人。"
2021-12-24 00:06:44.952 | INFO     | __main__:<module>:165 - word2id: [101, 1346, 1217, 3635, 6121, 4638, 3300, 4511, 3300, 1957, 8024, 3300, 2399, 6768, 782, 8024, 738, 3300, 704, 2399, 782, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.953 | INFO     | __main__:<module>:164 - 第7条文本: "沙特队教练佩雷拉：两支队都想胜，因此都作出了最大的努力。"
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:165 - word2id: [101, 3763, 4294, 7339, 3136, 5298, 877, 7440, 2861, 8038, 697, 3118, 7339, 6963, 2682, 5526, 8024, 1728, 3634, 6963, 868, 1139, 749, 3297, 1920, 4638, 1222, 1213, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:167 - label: [0, 3, 4, 4, 0, 0, 5, 6, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.953 | INFO     | __main__:<module>:164 - 第8条文本: "这种混乱局面导致有些海域使用者的合法权益难以得到维护。"
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:165 - word2id: [101, 6821, 4905, 3921, 744, 2229, 7481, 2193, 5636, 3300, 763, 3862, 1818, 886, 4500, 5442, 4638, 1394, 3791, 3326, 4660, 7410, 809, 2533, 1168, 5335, 2844, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.953 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.954 | INFO     | __main__:<module>:167 - label: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]2021-12-24 00:06:44.954 | INFO     | __main__:<module>:164 - 第9条文本:
2021-12-24 00:06:44.954 | INFO     | __main__:<module>:165 - word2id: [101, 7826, 2161, 3209, 4802, 2900, 1139, 8024, 2190, 3124, 2424, 4638, 6821, 4905, 2900, 2971, 2130, 1059, 3766, 3300, 752, 2141, 3418, 2945, 8024, 5401, 1744, 3124, 2424, 679, 2682, 738, 3766, 3300, 1403, 704, 1744, 6760, 6375, 3130, 2697, 2825, 3318, 8024, 752, 2141, 4696, 4685, 2600, 3300, 671, 1921, 833, 1920, 4635, 754, 1921, 678, 8039, 830, 6379, 7368, 4638, 6821, 4905, 976, 3791, 808, 782, 100, 7478, 2382, 1927, 3307, 100, 8024, 2199, 886, 5401, 1744, 4638, 1555, 689, 1310, 3215, 772, 689, 1358, 1168, 2014, 5516, 8024, 886, 5401, 1744, 4638, 4993, 751, 1213, 1358, 1168, 2938, 2154, 511, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.954 | INFO     | __main__:<module>:166 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2021-12-24 00:06:44.954 | INFO     | __main__:<module>:167 - label: [0, 5, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

说明：

• 结果中部分文本未显示，笔者已删除。原因：最近CSDN政治审核较为严格。
• example.train的部分数据格式如下：

海 O
钓 O
比 O
赛 O
地 O
点 O
在 O
厦 B-LOC
门 I-LOC
与 O
金 B-LOC
门 I-LOC
之 O
间 O
的 O
海 O
域 O
。 O这 O
座 O
依 O
山 O
傍 O
水 O
的 O
博 O
物 O
馆 O
由 O
国 O
内 O
一 O
流 O
的 O
设 O
计 O
师 O
主 O
持 O
设 O
计 O
， O
整 O
个 O
建 O
筑 O
群 O
精 O
美 O
而 O
恢 O
宏 O
。 O

• vocab.txt文件：大规模预训练模型的词典文件

2、基于R-drop的文本分类数据生成器

import numpy as np
from bert4keras.snippets import sequence_padding, DataGenerator
from bert4keras.tokenizers import Tokenizer
from loguru import loggerdef load_cls_data(filename):data = []with open(filename, "r", encoding="utf-8") as file:for line in file:# text1, text2, label = line.strip().split("\t")# util.append((text1, text2, int(label)))text, label = line.strip().split("\t")data.append((text, int(label)))return dataclass RDropDataGenerator(DataGenerator):def __init__(self, max_length, dict_path, data, batch_size, buffer_size):super(RDropDataGenerator, self).__init__(data=data, batch_size=batch_size, buffer_size=buffer_size)self.max_length = max_lengthself.tokenizer = Tokenizer(dict_path, do_lower_case=True)def __iter__(self, random=False):batch_token_ids, batch_segment_ids, batch_labels = [], [], []for is_end, (text, label) in self.sample(random):token_ids, segment_ids = self.tokenizer.encode(text, maxlen=self.max_length)for i in range(2):batch_token_ids.append(token_ids)batch_segment_ids.append(segment_ids)batch_labels.append([label])if len(batch_token_ids) == self.batch_size * 2 or is_end:batch_token_ids = sequence_padding(batch_token_ids)batch_segment_ids = sequence_padding(batch_segment_ids)batch_labels = sequence_padding(batch_labels)yield [batch_token_ids, batch_segment_ids], batch_labelsbatch_token_ids, batch_segment_ids, batch_labels = [], [], []if __name__ == "__main__":# 1.分类数据生成器data_generator = RDropDataGenerator(max_length=128,dict_path="../vocab.txt",data=load_cls_data("../r_drop.txt"),batch_size=32,buffer_size=None)flag = -1for x_true, y_true in data_generator:# 输出每个batch中的数据for idx in range(len(x_true[0])//2):flag += 1logger.info("第{0}条文本: \"{1}\"".format(flag, train_data[flag][0]))logger.info("word2id: {0}".format(str(list(x_true[0][idx]))))logger.info("mask: {0}".format(str(list(x_true[1][idx]))))logger.info("label: {0}\n\n".format(str(list(y_true[idx]))))

运行结果：

2022-03-07 23:23:14.683 | INFO     | __main__:<module>:149 - 第0条文本: "这张高清图，谁有"
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:150 - word2id: [101, 6821, 2476, 7770, 3926, 1745, 8024, 6443, 3300, 102, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:152 - label: [0]2022-03-07 23:23:14.683 | INFO     | __main__:<module>:149 - 第1条文本: "英雄联盟最好英雄是什么"
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:150 - word2id: [101, 6821, 2476, 7770, 3926, 1745, 8024, 6443, 3300, 102, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.683 | INFO     | __main__:<module>:152 - label: [0]2022-03-07 23:23:14.684 | INFO     | __main__:<module>:149 - 第2条文本: "我也是醉了，这是什么意思"
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:150 - word2id: [101, 5739, 7413, 5468, 4673, 3297, 1962, 5739, 7413, 3221, 784, 720, 102, 0, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:152 - label: [1]2022-03-07 23:23:14.684 | INFO     | __main__:<module>:149 - 第3条文本: "现在有什么好看的动画片吗？"
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:150 - word2id: [101, 5739, 7413, 5468, 4673, 3297, 1962, 5739, 7413, 3221, 784, 720, 102, 0, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:152 - label: [1]2022-03-07 23:23:14.684 | INFO     | __main__:<module>:149 - 第4条文本: "三星电子厂工资待遇怎么样啊"
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:150 - word2id: [101, 2769, 738, 3221, 7004, 749, 8024, 6821, 3221, 784, 720, 2692, 2590, 102, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.684 | INFO     | __main__:<module>:152 - label: [0]2022-03-07 23:23:14.685 | INFO     | __main__:<module>:149 - 第5条文本: "送闺蜜什么生日礼物好"
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:150 - word2id: [101, 4385, 1762, 3300, 784, 720, 1962, 4692, 4638, 1220, 4514, 4275, 1408, 8043, 102]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:152 - label: [1]2022-03-07 23:23:14.685 | INFO     | __main__:<module>:149 - 第6条文本: "近期上映的电影有哪些"
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:150 - word2id: [101, 4385, 1762, 3300, 784, 720, 1962, 4692, 4638, 1220, 4514, 4275, 1408, 8043, 102]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:152 - label: [1]2022-03-07 23:23:14.685 | INFO     | __main__:<module>:149 - 第7条文本: "英雄联盟，求大神带~"
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:150 - word2id: [101, 676, 3215, 4510, 2094, 1322, 2339, 6598, 2521, 6878, 2582, 720, 3416, 1557, 102]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:152 - label: [0]2022-03-07 23:23:14.685 | INFO     | __main__:<module>:149 - 第8条文本: "给东加上部首是什么字？"
2022-03-07 23:23:14.685 | INFO     | __main__:<module>:150 - word2id: [101, 676, 3215, 4510, 2094, 1322, 2339, 6598, 2521, 6878, 2582, 720, 3416, 1557, 102]
2022-03-07 23:23:14.686 | INFO     | __main__:<module>:151 - mask: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
2022-03-07 23:23:14.686 | INFO     | __main__:<module>:152 - label: [0]

说明：

• R_drop后的x_true维度为train_data维度的2倍，请特别注意，所以len(x_true[0]) // 2
• r_drop.txt的部分数据格式如下：

这张高清图，谁有  0
英雄联盟最好英雄是什么 1
我也是醉了，这是什么意思 0
现在有什么好看的动画片吗？    1
三星电子厂工资待遇怎么样啊   0
送闺蜜什么生日礼物好  1
近期上映的电影有哪些  1
英雄联盟，求大神带~   1
给东加上部首是什么字？  0

​​​​​​​

NLP——常见任务的批量加载2.0相关推荐

1. 2021年大数据HBase（十五）：HBase的Bulk Load批量加载操作

   全网最详细的大数据HBase文章系列,强烈建议收藏加关注! 新文章都已经列出历史文章目录,帮助大家回顾前面的知识重点. 目录 系列历史文章 HBase的Bulk Load批量加载操作 一.Bulk L ...

2. ARM的批量加载/存储指令

   批量加载/存储指令格式: LDMXX|STMXX{条件符}{寻址模式}Rb{!},{寄存器列表} 其中XX表示IA/IB/DA/DB. LDMIA/STMIA    访问/存储后地址递增(Increm ...

3. Oracle通过OCI批量加载需要注意的问题

   ORACLE调用接口(Oracle Call Interface简称OCI)提供了一组可对ORACLE数据库进行存取的接口子例程(函数),通过在第三代程序设计语言(如C语言)中进行调用可达到存取ORA ...

4. Metal之MTLBuffer批量加载顶点数量较多的图形渲染

   渲染原理 本文是基于"Metal渲染绘制三角形"这样顶点较少图形基础之上的延伸, 在渲染三角形的时候, 顶点数据的存储使用的是数组,当顶点传递时通过setVertexBytes(_ ...

5. 【机器视觉】 Halcon批量加载图像

   00. 目录 文章目录 00. 目录 01. 概述 02. 开发环境 03. 加载图像文件方式一 04. 加载图像文件方式二 05. 下载 06. 附录 01. 概述 halcon是一款非常不错的视觉 ...

6. Android实现图片的高效批量加载

   前言:图片的加载,图片的处理,是Android开发程序员在开发中经常遇到的问题,比如,图片的压缩,图片批量的网络获取等.但是往往处理不当,就会报oom,那么如何解决这一类问题了,在分析了弘扬大牛的图片 ...

7. vue项目批量加载url文件并打包到zip下载

   vue项目批量加载url文件并打包到zip下载 项目里之前是遇到这样一个需求哈,需要根据选择的不同动态批量生成并下载图片, 而且这个图片不是一张一张生成下载,而是要等他选好条件之后, 把对应的图片动态 ...

8. insert into select 批量加载出错解决方案

   当使用insert into select 批量加载数据的时候,可能会碰到因为某些数据不符合加载条件,而导致整个insert 语句无法执行,全部rollback.这时可以使用DML 错误日志的特性,解 ...

9. 【ARM汇编-LDM】如何向寄存器批量加载数据？

   大家好,我是汤姆凯特. 文章目录 如何从存储器中批量加载/存储数据? 传送门: 给大家先详细介绍一下LDM.STM这两个批量加载.存储指令: 格式: 功能: 有两个常用的用法 一.如何用STM指令进行 ...

最新文章

1. PHP5.3.8连接Sql Server SQLSRV30
2. 使用MarkDown编辑器做出有意思的柱状图（完整代码）
3. 从“人肉扩缩容”到云原生容量，90 后程序员的进化
4. DPI — 业务识别技术
5. 蓝桥杯-8-1因式分解（java）
6. hutool读取和导出excel_Java编程第44讲——非常好用的hutool工具介绍
7. 通过Blazor使用C#开发SPA单页面应用程序(4) - Ant Design
8. 算法题复习（栈与队列、二叉树）
9. python实现五大基本算法_算法基础：五大排序算法Python实战教程
10. TensorFlow实战——深度学习训练个性化推荐系统
11. Windows Phone 8初学者开发—第6部分：设置应用程序的样式
12. 再延期！曝华为P50六月发布
13. 在vmware中创建共享磁盘
14. 企业级监控软件使用zabbix key监控nginx status各种状态
15. [Erlang 0020]网页游戏分线到不分线
16. html推箱子过关检测函数,HTML5推箱子实现
17. ajax java servlet_$.ajax(); 对应后台java servlet?
18. 三个简单的C语言程序
19. Qtum量子链发布QIP-19支持隐私资产技术提案
20. 微信红包和转账有哪些区别？原来还有这些不同！涨知识了

热门文章

1. 威科达科技—资深工程师关于防治干扰的秘籍分享
2. x86汇编_移位和循环移位指令简介_笔记46
3. vnpy入门操作1——安装
4. vnpy文件结构说明
5. 0801-3 VNPY安装
6. Numpy库学习—squeeze()函数
7. PDU配电单元推荐——同为科技（TOWE）自接线工程安全机柜PDU
8. 一篇弄懂 scrollWidth、scrollHeight、scrollLeft和scrollTop的区别！（2）
9. 输入一行字符，分别统计出其中英文字母、数字、空格和其他字符的个数。
10. 实验六 Java异常处理与集合的使用