tokenizers

Provides an implementation of today's most used tokenizers, with a focus on performance and versatility.

Main features:

• Train new vocabularies and tokenize, using today's most used tokenizers.

• Extremely fast (both training and tokenization), thanks to the Rust implementation. Takes less than 20 seconds to tokenize a GB of text on a server's CPU.

• Easy to use, but also extremely versatile.

• Designed for research and production.

• Normalization comes with alignments tracking. It's always possible to get the part of the original sentence that corresponds to a given token.

• Does all the pre-processing: Truncate, Pad, add the special tokens your model needs.

Performances

Performances can vary depending on hardware, but running the ~/bindings/python/benches/test_tiktoken.py should give the following on a g6 aws instance:

Bindings

We provide bindings to the following languages (more to come!):

• Rust (Original implementation)

• Python

• Node.js

• Ruby (Contributed by @ankane, external repo)

Installation

You can install from source using:

pip install git+https://github.com/huggingface/tokenizers.git#subdirectory=bindings/python

our install the released versions with

pip install tokenizers

Quick example using Python:

Choose your model between Byte-Pair Encoding, WordPiece or Unigram and instantiate a tokenizer:

from tokenizers import Tokenizerfrom tokenizers.models import BPEtokenizer = Tokenizer(BPE())

You can customize how pre-tokenization (e.g., splitting into words) is done:

from tokenizers.pre_tokenizers import Whitespacetokenizer.pre_tokenizer = Whitespace()

Then training your tokenizer on a set of files just takes two lines of codes:

from tokenizers.trainers import BpeTrainertrainer = BpeTrainer(special_tokens=["[UNK]", "[CLS]", "[SEP]", "[PAD]", "[MASK]"])tokenizer.train(files=["wiki.train.raw", "wiki.valid.raw", "wiki.test.raw"], trainer=trainer)

Once your tokenizer is trained, encode any text with just one line:

output = tokenizer.encode("Hello, y'all! How are you ? ?")print(output.tokens)# ["Hello", ",", "y", "'", "all", "!", "How", "are", "you", "[UNK]", "?"]

Check the documentation or the quicktour to learn more!