Hugging Face实战-系列教程4:padding与attention_mask

🚩🚩🚩Hugging Face 实战系列 总目录

有任何问题欢迎在下面留言
本篇文章的代码运行界面均在notebook中进行
本篇文章配套的代码资源已经上传

上篇内容:
Hugging Face实战-系列教程3:文本2分类

下篇内容:
Hugging Face实战-系列教程5:NER上(命名实体识别/文本标注/Doccano工具使用/关键信息抽取/Token分类/源码解读/代码逐行解读/文本BIO处理/文本分类/序列标注)

1 padding的作用

sequence1_ids = [[200, 200, 200]]
sequence2_ids = [[200, 200]]

batched_ids = [
    [200, 200, 200],
    [200, 200, tokenizer.pad_token_id],
]

print(model(torch.tensor(sequence1_ids)).logits)
print(model(torch.tensor(sequence2_ids)).logits)

print(model(torch.tensor(batched_ids)).logits)

对应的输出:

tensor([[ 1.5694, -1.3895]], grad_fn=AddmmBackward0)
tensor([[ 0.5803, -0.4125]], grad_fn=AddmmBackward0)
tensor([[ 1.5694, -1.3895], [ 1.3374, -1.2163]], grad_fn=AddmmBackward0)

padding是很有用的,但是不需要了解的特别深入,看当前这一个例子就行了。

sequence1_ids、sequence1_ids分别表示两个样本,将两个样本做成batch。

输出实际上是两个对比,第一行和第二行是没有加入tokenizer.pad_token_id的输出,第三行是加入tokenizer.pad_token_id做成batch的输出。

sequence1_ids的结果是完全一样的,sequence2_ids的结果不同。很显然对于同一个样本加入来tokenizer.pad_token_id的输出是不相同的,这是因为默认的id和我们自己指定的id是不同的。

结论就是自己指定padding的时候,要同时指定attention_mask,不然就会把你指定的padding也计算来self-Attention。

怎么指定attention_mask呢?

2 attention_mask的作用

batched_ids = [
    [200, 200, 200],
    [200, 200, tokenizer.pad_token_id],
]

attention_mask = [
    [1, 1, 1],
    [1, 1, 0],
]

outputs = model(torch.tensor(batched_ids), attention_mask=torch.tensor(attention_mask))
print(outputs.logits)

输出结果:

tensor([[ 1.5694, -1.3895], [ 0.5803, -0.4125]], grad_fn=AddmmBackward0)

当需要自定义的时候,别忘记把Attention_mask指定上去,指定后,结果就能对的上了 。

3 不同padding方法

随便来组样本:

sequences = ["I've been waiting for a this course my whole life.", "So have I!", "I played basketball yesterday."]

在notebook中的输出:

{‘input_ids’: [[101, 1045, 1005, 2310, 2042, 3403, 2005, 1037, 2023, 2607, 2026, 2878, 2166, 1012, 102], [101, 2061, 2031, 1045, 999, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0], [101, 1045, 2209, 3455, 7483, 1012, 102, 0, 0, 0, 0, 0, 0, 0, 0]], ‘attention_mask’: [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]]}

一个三个句子,1最长,以1为标准,指定按照最长的为标准进行拼接填充:

# 按照最长的填充
model_inputs = tokenizer(sequences, padding="longest")
print(model_inputs)

当然也可以按照模型所能容忍的最大的长度来指定:

# BERT默认最大是512
model_inputs = tokenizer(sequences, padding="max_length")
print(model_inputs)

一般默认的最大的就是512,虽然openAI那个可以更大,但是一般那都hold不住了。

这是padding到512大小在notebook的输出:

{‘input_ids’: [[101, 1045, 1005, 2310, 2042, 3403, 2005, 1037, 2023, 2607, 2026, 2878, 2166, 1012, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [101, 2061, 2031, 1045, 999, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[101, 1045, 2209, 3455, 7483, 1012, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]],
‘attention_mask’: [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], [1, 1, 1, 1, 1, 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, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]}

这肯定是没有必要了,也可以自己指定一个确定的值:

填充到多少

model_inputs = tokenizer(sequences, padding="max_length", max_length=8)
print(model_inputs)

在notebook的输出:

{‘input_ids’: [[101, 1045, 1005, 2310, 2042, 3403, 2005, 1037, 2023,
2607, 2026, 2878, 2166, 1012, 102], [101, 2061, 2031, 1045, 999, 102,
0, 0], [101, 1045, 2209, 3455, 7483, 1012, 102, 0]], ‘attention_mask’:
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 0,
0], [1, 1, 1, 1, 1, 1, 1, 0]]}

在这个结果中,没有到8的padding到8,但是比8还长的没有处理,加入截断操作就可以进行处理了:

#到多少就截断

model_inputs = tokenizer(sequences, max_length=10, truncation=True)
print(model_inputs)

{‘input_ids’: [[101, 1045, 1005, 2310, 2042, 3403, 2005, 1037, 2023,
102], [101, 2061, 2031, 1045, 999, 102], [101, 1045, 2209, 3455, 7483,
1012, 102]], ‘attention_mask’: [[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1,
1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1]]}

4.4 返回数据格式
最后别忘记返回成PyTorch的tensor格式:

#最好返回tensor

model_inputs = tokenizer(sequences, padding=True, return_tensors="pt")
print(model_inputs)

{‘input_ids’: tensor([[ 101, 1045, 1005, 2310, 2042, 3403, 2005, 1037,
2023, 2607, 2026, 2878,
2166, 1012, 102],
[ 101, 2061, 2031, 1045, 999, 102, 0, 0, 0, 0, 0, 0,
0, 0, 0],
[ 101, 1045, 2209, 3455, 7483, 1012, 102, 0, 0, 0, 0, 0,
0, 0, 0]]), ‘attention_mask’: tensor([[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
[1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0]])}

现在你应该已经掌握了Hugging Face的工具包了,那我们应该怎样去训练我们的模型呢?

上篇内容:
Hugging Face实战-系列教程3:文本2分类

下篇内容:
Hugging Face实战-系列教程5:NER上(命名实体识别/文本标注/Doccano工具使用/关键信息抽取/Token分类/源码解读/代码逐行解读/文本BIO处理/文本分类/序列标注)

  • 1
    点赞
  • 2
    收藏
    觉得还不错? 一键收藏
  • 打赏
    打赏
  • 0
    评论
Hugging Face Transformers 是一个基于 PyTorch 和 TensorFlow 的自然语言处理(NLP)库,它提供了用于训练、微调和使用最先进的预训练模型的工具和接口。以下是使用 Hugging Face Transformers 进行实战的一些示例。 1. 文本分类 文本分类是将文本分为不同的类别或标签的任务。在这个示例中,我们将使用 Hugging Face Transformers 中的 DistilBERT 模型来训练一个情感分析分类器,以将电影评论分为正面或负面。 ```python from transformers import DistilBertTokenizer, DistilBertForSequenceClassification import torch tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased') model = DistilBertForSequenceClassification.from_pretrained('distilbert-base-uncased') # 训练数据 train_texts = ["I really liked this movie", "The plot was boring and predictable"] train_labels = [1, 0] # 将文本编码为输入张量 train_encodings = tokenizer(train_texts, truncation=True, padding=True) # 将标签编码为张量 train_labels = torch.tensor(train_labels) # 训练模型 model.train() optimizer = torch.optim.AdamW(model.parameters(), lr=5e-5) for epoch in range(3): optimizer.zero_grad() outputs = model(**train_encodings, labels=train_labels) loss = outputs.loss loss.backward() optimizer.step() # 预测新的评论 texts = ["This is a great movie", "I hated this movie"] encodings = tokenizer(texts, truncation=True, padding=True) model.eval() with torch.no_grad(): outputs = model(**encodings) predictions = torch.argmax(outputs.logits, dim=1) print(predictions) ``` 2. 问答系统 问答系统是回答用户提出的问题的模型。在这个示例中,我们将使用 Hugging Face Transformers 中的 DistilBERT 模型和 SQuAD 数据集来训练一个简单的问答系统。 ```python from transformers import DistilBertTokenizer, DistilBertForQuestionAnswering import torch tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased') model = DistilBertForQuestionAnswering.from_pretrained('distilbert-base-uncased') # 加载 SQuAD 数据集 from transformers import squad_convert_examples_to_features, SquadExample, SquadFeatures, squad_processors processor = squad_processors['squad'] examples = processor.get_train_examples('data') features = squad_convert_examples_to_features(examples=examples, tokenizer=tokenizer, max_seq_length=384, doc_stride=128, max_query_length=64, is_training=True) # 训练模型 model.train() optimizer = torch.optim.AdamW(model.parameters(), lr=5e-5) for epoch in range(3): for feature in features: optimizer.zero_grad() outputs = model(input_ids=torch.tensor([feature.input_ids]), attention_mask=torch.tensor([feature.attention_mask]), start_positions=torch.tensor([feature.start_position]), end_positions=torch.tensor([feature.end_position])) loss = outputs.loss loss.backward() optimizer.step() # 预测新的问题 text = "What is the capital of France?" question = "What country's capital is Paris?" inputs = tokenizer.encode_plus(question, text, add_special_tokens=True, return_tensors='pt') model.eval() with torch.no_grad(): start_scores, end_scores = model(**inputs) start_index = torch.argmax(start_scores) end_index = torch.argmax(end_scores) answer = tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(inputs['input_ids'][0][start_index:end_index+1])) print(answer) ``` 3. 文本生成 文本生成是使用预训练模型生成自然语言文本的任务。在这个示例中,我们将使用 Hugging Face Transformers 中的 GPT-2 模型生成一些小说的开头。 ```python from transformers import GPT2Tokenizer, GPT2LMHeadModel import torch tokenizer = GPT2Tokenizer.from_pretrained('gpt2') model = GPT2LMHeadModel.from_pretrained('gpt2') # 生成新的文本 seed_text = "In a hole in the ground there lived a hobbit." encoded = tokenizer.encode(seed_text, return_tensors='pt') model.eval() with torch.no_grad(): output = model.generate(encoded, max_length=100, do_sample=True) generated = tokenizer.decode(output[0], skip_special_tokens=True) print(generated) ``` 这些示例只是 Hugging Face Transformers 库的一部分功能。您可以通过访问 Hugging Face Transformers 官方文档来了解更多信息。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

机器学习杨卓越

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值