Tokenizers

A fast and easy to use implementation of today's most used tokenizers.

High Level design: master

This API is currently in the process of being stabilized. We might introduce breaking changes really often in the coming days/weeks, so use at your own risks.

Installation

With pip:

pip install tokenizers

From sources:

To use this method, you need to have the Rust nightly toolchain installed.

# Install with:
curl https://sh.rustup.rs -sSf | sh -s -- -default-toolchain nightly-2019-11-01 -y
export PATH="$HOME/.cargo/bin:$PATH"

# Or select the right toolchain:
rustup default nightly-2019-11-01

Once Rust is installed and using the right toolchain you can do the following.

git clone https://github.com/huggingface/tokenizers
cd tokenizers/bindings/python

# Create a virtual env (you can use yours as well)
python -m venv .env
source .env/bin/activate

# Install `tokenizers` in the current virtual env
pip install maturin
maturin develop --release

Usage

Use a pre-trained tokenizer

from tokenizers import Tokenizer, models, pre_tokenizers, decoders

# Load a BPE Model
vocab = "./path/to/vocab.json"
merges = "./path/to/merges.txt"
bpe = models.BPE.from_files(vocab, merges)

# Initialize a tokenizer
tokenizer = Tokenizer(bpe)

# Customize pre-tokenization and decoding
tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel.new(add_prefix_space=True)
tokenizer.decoder = decoders.ByteLevel.new()

# And then encode:
encoded = tokenizer.encode("I can feel the magic, can you?")
print(encoded)

# Or tokenize multiple sentences at once:
encoded = tokenizer.encode_batch([
	"I can feel the magic, can you?",
	"The quick brown fox jumps over the lazy dog"
])
print(encoded)

Train a new tokenizer

from tokenizers import Tokenizer, models, pre_tokenizers, decoders, trainers

# Initialize a tokenizer
tokenizer = Tokenizer(models.BPE.empty())

# Customize pre-tokenization and decoding
tokenizer.pre_tokenizer = pre_tokenizers.ByteLevel.new(add_prefix_space=True)
tokenizer.decoder = decoders.ByteLevel.new())

# And then train
trainer = trainers.BpeTrainer.new(vocab_size=20000, min_frequency=2)
tokenizer.train(trainer, [
	"./path/to/dataset/1.txt",
	"./path/to/dataset/2.txt",
	"./path/to/dataset/3.txt"
])

# Now we can encode
encoded = tokenizer.encode("I can feel the magic, can you?")
print(encoded)

2.5 KiB Raw Blame History