openvino-tokenizers2024.4.0.0

Published 2 months ago

Convert tokenizers into OpenVINO models

pip install openvino-tokenizers

Package Downloads

Authors

OpenVINO Developers

Requires Python

>=3.8

Dependencies

OpenVINO Tokenizers

OpenVINO Tokenizers adds text processing operations to OpenVINO.

Features

Perform tokenization and detokenization without third-party dependencies
Convert a HuggingFace tokenizer into OpenVINO model tokenizer and detokenizer
Combine OpenVINO models into a single model
Add greedy decoding pipeline to text generation model

Installation

(Recommended) Create and activate virtual env:

python3 -m venv venv
source venv/bin/activate
 # or
conda create --name openvino_tokenizers
conda activate openvino_tokenizers

Minimal Installation

Use minimal installation when you have a converted OpenVINO tokenizer:

pip install openvino-tokenizers
 # or
conda install -c conda-forge openvino openvino-tokenizers

Convert Tokenizers Installation

If you want to convert HuggingFace tokenizers into OpenVINO tokenizers:

pip install openvino-tokenizers[transformers]
 # or
conda install -c conda-forge openvino openvino-tokenizers && pip install transformers[sentencepiece] tiktoken

Install Pre-release Version

Use openvino-tokenizers[transformers] to install tokenizers conversion dependencies.

pip install --pre -U openvino openvino-tokenizers --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly

Build and Install from Source

Using OpenVINO PyPI package

openvino-tokenizers build depends on openvino package which will be automatically installed from PyPI during the build process. To install unreleased versions, you would need to install openvino package from the nightly distribution channel using --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly

git clone https://github.com/openvinotoolkit/openvino_tokenizers.git
cd openvino_tokenizers
pip install . --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly

This command is the equivalent of minimal installation. Install tokenizers conversion dependencies if needed:

pip install transformers[sentencepiece] tiktoken

:warning: Latest commit of OpenVINO Tokenizers might rely on features that are not present in the release OpenVINO version. Use a nightly build of OpenVINO or build OpenVINO Tokenizers from a release branch if you have issues with the build process.

Using OpenVINO archive

Install OpenVINO archive distribution. Use --no-deps to avoid OpenVINO installation from PyPI into your current environment. --extra-index-url is needed to resolve build dependencies only.

source path/to/installed/openvino/setupvars.sh
git clone https://github.com/openvinotoolkit/openvino_tokenizers.git
cd openvino_tokenizers
pip install --no-deps . --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly

This command is the equivalent of minimal installation. Install tokenizers conversion dependencies if needed:

pip install transformers[sentencepiece] tiktoken

Build and install for development

Using OpenVINO PyPI package

git clone https://github.com/openvinotoolkit/openvino_tokenizers.git
cd openvino_tokenizers
pip install -e .[all] --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly
# verify installation by running tests
cd tests/
pytest .

Using OpenVINO archive

Install OpenVINO archive distribution. Use --no-deps to avoid OpenVINO installation from PyPI into your current environment. --extra-index-url is needed to resolve build dependencies only.

source path/to/installed/openvino/setupvars.sh
git clone https://github.com/openvinotoolkit/openvino_tokenizers.git
cd openvino_tokenizers
pip install -e .[all] --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly
# verify installation by running tests
cd tests/
pytest .

C++ Installation

You can use converted tokenizers in C++ pipelines with prebuild binaries.

Download OpenVINO archive distribution for your OS from here and extract the archive.
Download OpenVINO Tokenizers prebuild libraries from here. To ensure compatibility first three numbers of OpenVINO Tokenizers version should match OpenVINO version and OS.
Extract OpenVINO Tokenizers archive into OpenVINO installation directory. OpenVINO Tokenizers archive maintains the structure to be aligned with OpenVINO archive:
- Windows: <openvino_dir>\runtime\bin\intel64\Release\
- MacOS_x86: <openvino_dir>/runtime/lib/intel64/Release
- MacOS_arm64: <openvino_dir>/runtime/lib/arm64/Release/
- Linux_x86: <openvino_dir>/runtime/lib/intel64/
- Linux_arm64: <openvino_dir>/runtime/lib/aarch64/

After that you can add binary extension in the code with:

core.add_extension("openvino_tokenizers.dll") for Windows
core.add_extension("libopenvino_tokenizers.dylib") for MacOS
core.add_extension("libopenvino_tokenizers.so") for Linux

and read/compile converted (de)tokenizers models. If you use version 2023.3.0.0, the binary extension file is called (lib)user_ov_extension.(dll/dylib/so).

Reducing the ICU Data Size

By default, all available ICU locales are supported, which significantly increases the package size. To reduce the size of the ICU libraries included in your final package, follow these steps:

Use the ICU Data Configuration File:
- This file specifies which features and locales to include in a custom data bundle. You can find more information here.
Set the ICU Data Filter File as an Environment Variable:
- On Unix-like systems (Linux, macOS): Set the ICU_DATA_FILTER_FILE environment variable to the path of your configuration file (filters.json):
```
export ICU_DATA_FILTER_FILE="filters.json"
```
- On Windows: Set the ICU_DATA_FILTER_FILE environment variable using the Command Prompt or PowerShell:
  
  Command Prompt:
```
set ICU_DATA_FILTER_FILE=filters.json
```
  PowerShell:
```
$env:ICU_DATA_FILTER_FILE="filters.json"
```

Create a Configuration File:

An example configuration file (filters.json) might look like this:

{
  "localeFilter": {
    "filterType": "language",
    "includelist": [
      "en"
    ]
  }
}

Configure OpenVINO Tokenizers:

When building OpenVINO tokenizers, set the following CMake option during the project configuration:

-DBUILD_FAST_TOKENIZERS=ON

Example for a pip installation path:

ICU_DATA_FILTER_FILE=</path/to/filters.json> pip install git+https://github.com/openvinotoolkit/openvino_tokenizers.git --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly --config-settings=override=cmake.options.BUILD_FAST_TOKENIZERS=ON

By following these instructions, you can effectively reduce the size of the ICU libraries in your final package.

Build OpenVINO Tokenizers without FastTokenizer Library

If a tokenizer doesn't use CaseFold, UnicodeNormalization or Wordpiece operations, you can drastically reduce package binary size by building OpenVINO Tokenizers without FastTokenizer dependency with this flag:

-DENABLE_FAST_TOKENIZERS=OFF

This option can also help with building for platform that is supported by FastTokenizer, for example Android x86_64.

Example for a pip installation path:


pip install git+https://github.com/openvinotoolkit/openvino_tokenizers.git --extra-index-url https://storage.openvinotoolkit.org/simple/wheels/nightly --config-settings=override=cmake.options.ENABLE_FAST_TOKENIZERS=OFF

Usage

:warning: OpenVINO Tokenizers can be inferred on a CPU device only.

Convert HuggingFace tokenizer

OpenVINO Tokenizers ships with CLI tool that can convert tokenizers from Huggingface Hub or Huggingface tokenizers saved on disk:

convert_tokenizer codellama/CodeLlama-7b-hf --with-detokenizer -o output_dir

There is also convert_tokenizer function that can convert tokenizer python object.

import numpy as np
from transformers import AutoTokenizer
from openvino import compile_model, save_model
from openvino_tokenizers import convert_tokenizer

hf_tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
ov_tokenizer = convert_tokenizer(hf_tokenizer)

compiled_tokenzier = compile_model(ov_tokenizer)
text_input = ["Test string"]

hf_output = hf_tokenizer(text_input, return_tensors="np")
ov_output = compiled_tokenzier(text_input)

for output_name in hf_output:
    print(f"OpenVINO {output_name} = {ov_output[output_name]}")
    print(f"HuggingFace {output_name} = {hf_output[output_name]}")
# OpenVINO input_ids = [[ 101 3231 5164  102]]
# HuggingFace input_ids = [[ 101 3231 5164  102]]
# OpenVINO token_type_ids = [[0 0 0 0]]
# HuggingFace token_type_ids = [[0 0 0 0]]
# OpenVINO attention_mask = [[1 1 1 1]]
# HuggingFace attention_mask = [[1 1 1 1]]

# save tokenizer for later use
save_model(ov_tokenizer, "openvino_tokenizer.xml")

loaded_tokenizer = compile_model("openvino_tokenizer.xml")
loaded_ov_output = loaded_tokenizer(text_input)
for output_name in hf_output:
    assert np.all(loaded_ov_output[output_name] == ov_output[output_name])

Connect Tokenizer to a Model

To infer and convert the original model, install torch or torch-cpu to the virtual environment.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
from openvino import compile_model, convert_model
from openvino_tokenizers import convert_tokenizer, connect_models

checkpoint = "mrm8488/bert-tiny-finetuned-sms-spam-detection"
hf_tokenizer = AutoTokenizer.from_pretrained(checkpoint)
hf_model = AutoModelForSequenceClassification.from_pretrained(checkpoint)

text_input = ["Free money!!!"]
hf_input = hf_tokenizer(text_input, return_tensors="pt")
hf_output = hf_model(**hf_input)

ov_tokenizer = convert_tokenizer(hf_tokenizer)
ov_model = convert_model(hf_model, example_input=hf_input.data)
combined_model = connect_models(ov_tokenizer, ov_model)
compiled_combined_model = compile_model(combined_model)

openvino_output = compiled_combined_model(text_input)

print(f"OpenVINO logits: {openvino_output['logits']}")
# OpenVINO logits: [[ 1.2007061 -1.4698029]]
print(f"HuggingFace logits {hf_output.logits}")
# HuggingFace logits tensor([[ 1.2007, -1.4698]], grad_fn=<AddmmBackward0>)

Use Extension With Converted (De)Tokenizer or Model With (De)Tokenizer

Import openvino_tokenizers will add all tokenizer-related operations to OpenVINO, after which you can work with saved tokenizers and detokenizers.

import numpy as np
import openvino_tokenizers
from openvino import Core

core = Core()

# detokenizer from codellama sentencepiece model
compiled_detokenizer = core.compile_model("detokenizer.xml")

token_ids = np.random.randint(100, 1000, size=(3, 5))
openvino_output = compiled_detokenizer(token_ids)

print(openvino_output["string_output"])
# ['sc�ouition�', 'intvenord hasient', 'g shouldwer M more']

Text generation pipeline

import numpy as np
from openvino import compile_model, convert_model
from openvino_tokenizers import add_greedy_decoding, convert_tokenizer
from transformers import AutoModelForCausalLM, AutoTokenizer


model_checkpoint = "JackFram/llama-68m"
hf_tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
hf_model = AutoModelForCausalLM.from_pretrained(model_checkpoint, use_cache=False)

# convert hf tokenizer
text_input = ["Quick brown fox jumped "]
ov_tokenizer, ov_detokenizer = convert_tokenizer(hf_tokenizer, with_detokenizer=True)
compiled_tokenizer = compile_model(ov_tokenizer)

# transform input text into tokens
ov_input = compiled_tokenizer(text_input)
hf_input = hf_tokenizer(text_input, return_tensors="pt")

# convert Pytorch model to OpenVINO IR and add greedy decoding pipeline to it
ov_model = convert_model(hf_model, example_input=hf_input.data)
ov_model_with_greedy_decoding = add_greedy_decoding(ov_model)
compiled_model = compile_model(ov_model_with_greedy_decoding)

# generate new tokens
new_tokens_size = 10
prompt_size = ov_input["input_ids"].shape[-1]
input_dict = {
    output.any_name: np.hstack([tensor, np.zeros(shape=(1, new_tokens_size), dtype=np.int_)])
    for output, tensor in ov_input.items()
}
for idx in range(prompt_size, prompt_size + new_tokens_size):
    output = compiled_model(input_dict)["token_ids"]
    input_dict["input_ids"][:, idx] = output[:, idx - 1]
    input_dict["attention_mask"][:, idx] = 1
ov_token_ids = input_dict["input_ids"]

hf_token_ids = hf_model.generate(
    **hf_input,
    min_new_tokens=new_tokens_size,
    max_new_tokens=new_tokens_size,
    temperature=0,  # greedy decoding
)

# decode model output
compiled_detokenizer = compile_model(ov_detokenizer)
ov_output = compiled_detokenizer(ov_token_ids)["string_output"]
hf_output = hf_tokenizer.batch_decode(hf_token_ids, skip_special_tokens=True)
print(f"OpenVINO output string: `{ov_output}`")
# OpenVINO output string: `['Quick brown fox was walking through the forest. He was looking for something']`
print(f"HuggingFace output string: `{hf_output}`")
# HuggingFace output string: `['Quick brown fox was walking through the forest. He was looking for something']`

TensorFlow Text Integration

OpenVINO Tokenizers include converters for certain TensorFlow Text operations. Currently, only the MUSE model is supported. Here is an example of model conversion and inference:

import numpy as np
import tensorflow_hub as hub
import tensorflow_text  # register tf text ops
from openvino import convert_model, compile_model
import openvino_tokenizers  # register ov tokenizer ops and translators


sentences = ["dog",  "I cuccioli sono carini.", "私は犬と一緒にビーチを散歩するのが好きです"]
tf_embed = hub.load(
    "https://www.kaggle.com/models/google/universal-sentence-encoder/frameworks/"
    "TensorFlow2/variations/multilingual/versions/2"
)
# convert model that uses Sentencepiece tokenizer op from TF Text
ov_model = convert_model(tf_embed)
ov_embed = compile_model(ov_model, "CPU")

ov_result = ov_embed(sentences)[ov_embed.output()]
tf_result = tf_embed(sentences)

assert np.all(np.isclose(ov_result, tf_result, atol=1e-4))

RWKV Tokenizer

from urllib.request import urlopen

from openvino import compile_model
from openvino_tokenizers import build_rwkv_tokenizer


rwkv_vocab_url = (
    "https://raw.githubusercontent.com/BlinkDL/ChatRWKV/main/tokenizer/rwkv_vocab_v20230424.txt"
)

with urlopen(rwkv_vocab_url) as vocab_file:
    vocab = map(bytes.decode, vocab_file)
    tokenizer, detokenizer = build_rwkv_tokenizer(vocab)

tokenizer, detokenizer = compile_model(tokenizer), compile_model(detokenizer)

print(tokenized := tokenizer(["Test string"])["input_ids"])  # [[24235 47429]]
print(detokenizer(tokenized)["string_output"])  # ['Test string']

Supported Tokenizer Types

Huggingface Tokenizer Type	Tokenizer Model Type	Tokenizer	Detokenizer
Fast	WordPiece	✅	❌
	BPE	✅	✅
	Unigram	❌	❌
Legacy	SentencePiece .model	✅	✅
Custom	tiktoken	✅	✅
RWKV	Trie	✅	✅

Test Results

This report is autogenerated and includes tokenizers and detokenizers tests. The Output Matched, % column shows the percent of test strings for which the results of OpenVINO and Huggingface Tokenizers are the same. To update the report run pytest --update_readme tokenizers_test.py in tests directory.

Output Match by Tokenizer Type

Tokenizer Type	Output Matched, %	Number of Tests
BPE	95.57	5932
SentencePiece	88.23	6534
Tiktoken	99.19	494
WordPiece	99.10	1327

Output Match by Model

Tokenizer Type	Model	Output Matched, %	Number of Tests
BPE	EleutherAI/gpt-j-6b	95.29	255
BPE	EleutherAI/gpt-neo-125m	95.29	255
BPE	EleutherAI/gpt-neox-20b	95.82	239
BPE	EleutherAI/pythia-12b-deduped	95.82	239
BPE	KoboldAI/fairseq-dense-13B	96.65	239
BPE	NousResearch/Meta-Llama-3-8B-Instruct	100.00	241
BPE	Salesforce/codegen-16B-multi	96.08	255
BPE	Xenova/gpt-4o	100.00	255
BPE	ai-forever/rugpt3large_based_on_gpt2	94.51	255
BPE	bigscience/bloom	97.49	239
BPE	databricks/dolly-v2-3b	95.82	239
BPE	deepseek-ai/deepseek-coder-6.7b-instruct	100.00	257
BPE	facebook/bart-large-mnli	95.29	255
BPE	facebook/galactica-120b	95.82	239
BPE	facebook/opt-66b	96.65	239
BPE	gpt2	95.29	255
BPE	laion/CLIP-ViT-bigG-14-laion2B-39B-b160k	75.29	255
BPE	microsoft/deberta-base	96.65	239
BPE	roberta-base	95.29	255
BPE	sentence-transformers/all-roberta-large-v1	95.29	255
BPE	stabilityai/stablecode-completion-alpha-3b-4k	95.82	239
BPE	stabilityai/stablelm-2-1_6b	100.00	239
BPE	stabilityai/stablelm-tuned-alpha-7b	95.82	239
BPE	tiiuae/falcon-7b	94.51	255
SentencePiece	NousResearch/Llama-2-13b-hf	96.65	239
SentencePiece	NousResearch/Llama-2-13b-hf_legacy	100.00	239
SentencePiece	NousResearch/Llama-2-13b-hf_sp_backend	100.00	239
SentencePiece	THUDM/chatglm2-6b_legacy	100.00	153
SentencePiece	THUDM/chatglm3-6b_legacy	50.97	155
SentencePiece	camembert-base	52.30	239
SentencePiece	camembert-base_legacy	76.15	239
SentencePiece	codellama/CodeLlama-7b-hf	96.65	239
SentencePiece	codellama/CodeLlama-7b-hf_legacy	96.65	239
SentencePiece	codellama/CodeLlama-7b-hf_sp_backend	94.98	239
SentencePiece	facebook/musicgen-small	84.52	239
SentencePiece	facebook/musicgen-small_legacy	79.92	239
SentencePiece	microsoft/Phi-3-mini-128k-instruct	95.85	241
SentencePiece	microsoft/Phi-3-mini-128k-instruct_legacy	95.85	241
SentencePiece	microsoft/Phi-3-mini-128k-instruct_sp_backend	94.19	241
SentencePiece	microsoft/deberta-v3-base	96.65	239
SentencePiece	microsoft/deberta-v3-base_legacy	100.00	239
SentencePiece	mlx-community/quantized-gemma-7b-it	99.17	241
SentencePiece	mlx-community/quantized-gemma-7b-it_legacy	99.17	241
SentencePiece	mlx-community/quantized-gemma-7b-it_sp_backend	100.00	241
SentencePiece	rinna/bilingual-gpt-neox-4b	80.75	239
SentencePiece	rinna/bilingual-gpt-neox-4b_legacy	86.61	239
SentencePiece	t5-base	85.77	239
SentencePiece	t5-base_legacy	81.17	239
SentencePiece	xlm-roberta-base	96.23	239
SentencePiece	xlm-roberta-base_legacy	96.23	239
SentencePiece	xlnet-base-cased	65.27	239
SentencePiece	xlnet-base-cased_legacy	59.41	239
Tiktoken	Qwen/Qwen-14B-Chat	100.00	255
Tiktoken	THUDM/glm-4-9b	98.33	239
WordPiece	ProsusAI/finbert	100.00	107
WordPiece	bert-base-multilingual-cased	100.00	107
WordPiece	bert-base-uncased	100.00	107
WordPiece	cointegrated/rubert-tiny2	100.00	107
WordPiece	distilbert-base-uncased-finetuned-sst-2-english	100.00	107
WordPiece	google/electra-base-discriminator	100.00	107
WordPiece	google/mobilebert-uncased	100.00	91
WordPiece	jhgan/ko-sbert-sts	100.00	107
WordPiece	prajjwal1/bert-mini	100.00	91
WordPiece	rajiv003/ernie-finetuned-qqp	100.00	91
WordPiece	rasa/LaBSE	88.79	107
WordPiece	sentence-transformers/all-MiniLM-L6-v2	100.00	107
WordPiece	squeezebert/squeezebert-uncased	100.00	91

Recreating Tokenizers From Tests

In some tokenizers, you need to select certain settings so that their output is closer to the Huggingface tokenizers:

THUDM/chatglm2-6b detokenizer always skips special tokens. Use skip_special_tokens=True during conversion
THUDM/chatglm3-6b detokenizer don't skips special tokens. Use skip_special_tokens=False during conversion
All tested tiktoken based detokenizers leave extra spaces. Use clean_up_tokenization_spaces=False during conversion