如何按令牌分割文本

语言模型有令牌限制。您不应超过令牌限制。因此，当您将文本分割成块时，计算令牌数量是个好主意。有许多令牌化工具。当您计算文本中的令牌时，应使用与语言模型中使用的相同的令牌化工具。

tiktoken

note

tiktoken 是由 OpenAI 创建的快速 BPE 令牌化工具。

我们可以使用 tiktoken 来估算使用的令牌。对于 OpenAI 模型，它可能会更准确。

1. 文本的分割方式：按传入的字符。
2. 块大小的测量方式：通过 tiktoken 令牌化工具。

CharacterTextSplitter、RecursiveCharacterTextSplitter 和 TokenTextSplitter 可以直接与 tiktoken 一起使用。

%pip install --upgrade --quiet langchain-text-splitters tiktoken

<!--IMPORTS:[{"imported": "CharacterTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/character/langchain_text_splitters.character.CharacterTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import CharacterTextSplitter

# This is a long document we can split up.
with open("state_of_the_union.txt") as f:
    state_of_the_union = f.read()

要使用 CharacterTextSplitter 进行分割，然后使用 tiktoken 合并块，请使用其 .from_tiktoken_encoder() 方法。请注意，此方法的分割可能大于 tiktoken 分词器测量的块大小。

.from_tiktoken_encoder() 方法接受 encoding_name 作为参数（例如 cl100k_base），或 model_name（例如 gpt-4）。所有其他参数如 chunk_size、chunk_overlap 和 separators 用于实例化 CharacterTextSplitter：

text_splitter = CharacterTextSplitter.from_tiktoken_encoder(
    encoding_name="cl100k_base", chunk_size=100, chunk_overlap=0
)
texts = text_splitter.split_text(state_of_the_union)

print(texts[0])

Madam Speaker, Madam Vice President, our First Lady and Second Gentleman. Members of Congress and the Cabinet. Justices of the Supreme Court. My fellow Americans.  

Last year COVID-19 kept us apart. This year we are finally together again. 

Tonight, we meet as Democrats Republicans and Independents. But most importantly as Americans. 

With a duty to one another to the American people to the Constitution.

要对块大小实施硬约束，我们可以使用 RecursiveCharacterTextSplitter.from_tiktoken_encoder，其中每个分割如果大小较大将被递归分割：

<!--IMPORTS:[{"imported": "RecursiveCharacterTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/character/langchain_text_splitters.character.RecursiveCharacterTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import RecursiveCharacterTextSplitter

text_splitter = RecursiveCharacterTextSplitter.from_tiktoken_encoder(
    model_name="gpt-4",
    chunk_size=100,
    chunk_overlap=0,
)

我们还可以加载 TokenTextSplitter 分割器，它直接与 tiktoken 一起工作，并确保每个分割小于块大小。

<!--IMPORTS:[{"imported": "TokenTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/base/langchain_text_splitters.base.TokenTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import TokenTextSplitter

text_splitter = TokenTextSplitter(chunk_size=10, chunk_overlap=0)

texts = text_splitter.split_text(state_of_the_union)
print(texts[0])

Madam Speaker, Madam Vice President, our

某些书写语言（例如中文和日文）中的字符编码为 2 个或更多的标记。直接使用 TokenTextSplitter 可能会在两个块之间分割字符的标记，导致格式错误的 Unicode 字符。请使用 RecursiveCharacterTextSplitter.from_tiktoken_encoder 或 CharacterTextSplitter.from_tiktoken_encoder 以确保块包含有效的 Unicode 字符串。

spaCy

note

spaCy 是一个用于高级自然语言处理的开源软件库，使用编程语言 Python 和 Cython 编写。

LangChain 实现了基于 spaCy 分词器 的分割器。

1. 文本如何分割：通过 spaCy 分词器。
2. 块大小如何测量：通过字符数量。

%pip install --upgrade --quiet  spacy

# This is a long document we can split up.
with open("state_of_the_union.txt") as f:
    state_of_the_union = f.read()

<!--IMPORTS:[{"imported": "SpacyTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/spacy/langchain_text_splitters.spacy.SpacyTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import SpacyTextSplitter

text_splitter = SpacyTextSplitter(chunk_size=1000)

texts = text_splitter.split_text(state_of_the_union)
print(texts[0])

Madam Speaker, Madam Vice President, our First Lady and Second Gentleman.

Members of Congress and the Cabinet.

Justices of the Supreme Court.

My fellow Americans.  



Last year COVID-19 kept us apart.

This year we are finally together again. 



Tonight, we meet as Democrats Republicans and Independents.

But most importantly as Americans. 



With a duty to one another to the American people to the Constitution. 



And with an unwavering resolve that freedom will always triumph over tyranny. 



Six days ago, Russia’s Vladimir Putin sought to shake the foundations of the free world thinking he could make it bend to his menacing ways.

But he badly miscalculated. 



He thought he could roll into Ukraine and the world would roll over.

Instead he met a wall of strength he never imagined. 



He met the Ukrainian people. 



From President Zelenskyy to every Ukrainian, their fearlessness, their courage, their determination, inspires the world.

句子变换器

【SentenceTransformersTokenTextSplitter】(https://python.langchain.com/api_reference/text_splitters/sentence_transformers/langchain_text_splitters.sentence_transformers.SentenceTransformersTokenTextSplitter.html) 是一个专门用于句子变换器模型的文本分割器。默认行为是将文本分割成适合您想要使用的句子变换器模型的令牌窗口的块。

要根据句子变换器的分词器分割文本并限制令牌数量，请实例化一个 SentenceTransformersTokenTextSplitter。您可以选择性地指定：

• chunk_overlap: 令牌重叠的整数计数；
• model_name: 句子变换器模型名称，默认为 "sentence-transformers/all-mpnet-base-v2"；
• tokens_per_chunk: 每个块所需的令牌数量。

<!--IMPORTS:[{"imported": "SentenceTransformersTokenTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/sentence_transformers/langchain_text_splitters.sentence_transformers.SentenceTransformersTokenTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import SentenceTransformersTokenTextSplitter

splitter = SentenceTransformersTokenTextSplitter(chunk_overlap=0)
text = "Lorem "

count_start_and_stop_tokens = 2
text_token_count = splitter.count_tokens(text=text) - count_start_and_stop_tokens
print(text_token_count)

2

token_multiplier = splitter.maximum_tokens_per_chunk // text_token_count + 1

# `text_to_split` does not fit in a single chunk
text_to_split = text * token_multiplier

print(f"tokens in text to split: {splitter.count_tokens(text=text_to_split)}")

tokens in text to split: 514

text_chunks = splitter.split_text(text=text_to_split)

print(text_chunks[1])

lorem

NLTK

note

自然语言工具包，或更常见的 NLTK，是一个用于英语的符号和统计自然语言处理 (NLP) 的库和程序套件，使用 Python 编程语言编写。

我们可以使用 NLTK 基于 NLTK 分词器 来分割，而不仅仅是基于 "\n\n"。

1. 文本的分割方式：通过 NLTK 分词器。
2. 块大小的测量方式：按字符数。

# pip install nltk

# This is a long document we can split up.
with open("state_of_the_union.txt") as f:
    state_of_the_union = f.read()

<!--IMPORTS:[{"imported": "NLTKTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/nltk/langchain_text_splitters.nltk.NLTKTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import NLTKTextSplitter

text_splitter = NLTKTextSplitter(chunk_size=1000)

texts = text_splitter.split_text(state_of_the_union)
print(texts[0])

Madam Speaker, Madam Vice President, our First Lady and Second Gentleman.

Members of Congress and the Cabinet.

Justices of the Supreme Court.

My fellow Americans.

Last year COVID-19 kept us apart.

This year we are finally together again.

Tonight, we meet as Democrats Republicans and Independents.

But most importantly as Americans.

With a duty to one another to the American people to the Constitution.

And with an unwavering resolve that freedom will always triumph over tyranny.

Six days ago, Russia’s Vladimir Putin sought to shake the foundations of the free world thinking he could make it bend to his menacing ways.

But he badly miscalculated.

He thought he could roll into Ukraine and the world would roll over.

Instead he met a wall of strength he never imagined.

He met the Ukrainian people.

From President Zelenskyy to every Ukrainian, their fearlessness, their courage, their determination, inspires the world.

Groups of citizens blocking tanks with their bodies.

KoNLPY

note

KoNLPy: Python中的韩语自然语言处理 是一个用于韩语自然语言处理（NLP）的Python包。

令牌分割涉及将文本分割成更小、更易管理的单位，称为令牌。这些令牌通常是单词、短语、符号或其他对进一步处理和分析至关重要的有意义元素。在英语等语言中，令牌分割通常涉及通过空格和标点符号分隔单词。令牌分割的有效性在很大程度上取决于分词器对语言结构的理解，以确保生成有意义的令牌。由于为英语设计的分词器无法理解其他语言（如韩语）的独特语义结构，因此无法有效用于韩语处理。

使用KoNLPY的Kkma分析器进行韩语令牌分割

在韩语文本的情况下，KoNLPY包含一个形态分析器，称为Kkma（韩语知识语素分析器）。Kkma提供韩语文本的详细形态分析。它将句子分解为单词，并将单词分解为各自的语素，识别每个令牌的词性。它可以将一段文本分割成单独的句子，这对于处理长文本特别有用。

使用注意事项

虽然Kkma以其详细分析而闻名，但需要注意的是，这种精确性可能会影响处理速度。因此，Kkma最适合那些优先考虑分析深度而非快速文本处理的应用。

# pip install konlpy

# This is a long Korean document that we want to split up into its component sentences.
with open("./your_korean_doc.txt") as f:
    korean_document = f.read()

<!--IMPORTS:[{"imported": "KonlpyTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/konlpy/langchain_text_splitters.konlpy.KonlpyTextSplitter.html", "title": "How to split text by tokens "}]-->
from langchain_text_splitters import KonlpyTextSplitter

text_splitter = KonlpyTextSplitter()

texts = text_splitter.split_text(korean_document)
# The sentences are split with "\n\n" characters.
print(texts[0])

춘향전 옛날에 남원에 이 도령이라는 벼슬아치 아들이 있었다.

그의 외모는 빛나는 달처럼 잘생겼고, 그의 학식과 기예는 남보다 뛰어났다.

한편, 이 마을에는 춘향이라는 절세 가인이 살고 있었다.

춘 향의 아름다움은 꽃과 같아 마을 사람들 로부터 많은 사랑을 받았다.

어느 봄날, 도령은 친구들과 놀러 나갔다가 춘 향을 만 나 첫 눈에 반하고 말았다.

두 사람은 서로 사랑하게 되었고, 이내 비밀스러운 사랑의 맹세를 나누었다.

하지만 좋은 날들은 오래가지 않았다.

도령의 아버지가 다른 곳으로 전근을 가게 되어 도령도 떠나 야만 했다.

이별의 아픔 속에서도, 두 사람은 재회를 기약하며 서로를 믿고 기다리기로 했다.

그러나 새로 부임한 관아의 사또가 춘 향의 아름다움에 욕심을 내 어 그녀에게 강요를 시작했다.

춘 향 은 도령에 대한 자신의 사랑을 지키기 위해, 사또의 요구를 단호히 거절했다.

이에 분노한 사또는 춘 향을 감옥에 가두고 혹독한 형벌을 내렸다.

이야기는 이 도령이 고위 관직에 오른 후, 춘 향을 구해 내는 것으로 끝난다.

두 사람은 오랜 시련 끝에 다시 만나게 되고, 그들의 사랑은 온 세상에 전해 지며 후세에까지 이어진다.

- 춘향전 (The Tale of Chunhyang)

Hugging Face分词器

Hugging Face 有许多分词器。

我们使用 Hugging Face 的分词器 GPT2TokenizerFast 来计算文本长度的标记数。

1. 文本的分割方式：按传入的字符进行分割。
2. 块大小的测量方式：通过 Hugging Face 分词器计算的标记数量。

from transformers import GPT2TokenizerFast

tokenizer = GPT2TokenizerFast.from_pretrained("gpt2")

<!--IMPORTS:[{"imported": "CharacterTextSplitter", "source": "langchain_text_splitters", "docs": "https://python.langchain.com/api_reference/text_splitters/character/langchain_text_splitters.character.CharacterTextSplitter.html", "title": "How to split text by tokens "}]-->
# This is a long document we can split up.
with open("state_of_the_union.txt") as f:
    state_of_the_union = f.read()
from langchain_text_splitters import CharacterTextSplitter

text_splitter = CharacterTextSplitter.from_huggingface_tokenizer(
    tokenizer, chunk_size=100, chunk_overlap=0
)
texts = text_splitter.split_text(state_of_the_union)

print(texts[0])

Madam Speaker, Madam Vice President, our First Lady and Second Gentleman. Members of Congress and the Cabinet. Justices of the Supreme Court. My fellow Americans.  

Last year COVID-19 kept us apart. This year we are finally together again. 

Tonight, we meet as Democrats Republicans and Independents. But most importantly as Americans. 

With a duty to one another to the American people to the Constitution.