CheeseShop
tokenizers
Our great sponsors
CheeseShop | tokenizers | |
---|---|---|
2 | 8 | |
1 | 8,424 | |
- | 3.1% | |
3.8 | 8.5 | |
7 months ago | 2 days ago | |
Rust | Rust | |
MIT License | Apache License 2.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
CheeseShop
-
Apache Spark UDFs in Rust
By comparison, PyO3 handles virtually all that boilerplate, so your Rust functions can accept and return many native Rust types and everything just works (for example). Or maybe I'm missing some fundamental difference with how JVM data are handled versus Python.
-
PyO3: Rust Bindings for the Python Interpreter
At work, I'm using PyO3 for a project that churns through a lot of data (step 1) and does some pattern mining (step 2). This is the second generation of the project and is on-demand compared with the large, batch project in Spark that it is replacing. The Rust+Python project has really good performance, and using Rust for the core logic is such a joy compared with Scala or Python that a lot of other pieces are written in.
Learning PyO3, I cobbled together a sample project[0] to demonstrate how some functionality works. It's a little outdated (uses PyO3 0.11.0 compared with the current 0.13.1) and doesn't show everything, but I think it's reasonably clear.
One thing I noticed is that passing very large data from Rust and into Python's memory space is a bit of a challenge. I haven't quite grokked who owns what when and how memory gets correctly dropped, but I think the issues I've had are with the amount of RAM used at any moment and not with any memory leaks.
[0] https://github.com/aeshirey/CheeseShop
tokenizers
-
HF Transfer: Speed up file transfers
Hugging Face seems to like Rust. They also wrote Tokenizers in Rust.
-
LLM custom dictionary
Your intuition is right. There are two ways (in increasing order of result performance) : 1. You can simply extend vocab file of the tokenizer and test the predictions 2. You can extend the vocab file and re-train your model on custom data which has these new tokens. Check the following issue on GitHub : https://github.com/huggingface/tokenizers/issues/247
-
[D] SentencePiece, WordPiece, BPE... Which tokenizer is the best one?
SentencePiece -> implementation of some algorithms (there are several others, https://github.com/microsoft/BlingFire https://github.com/glample/fastBPE https://github.com/huggingface/tokenizers )
-
Portability of Rust in 2021
In sum I would like the idea to go with Rust as I more or less got to rewrite the whole thing anyway, but I am a bit skeptical if I will be able to interface with everything that might come up at some point. Or probably end up in a wrapper hell if I got to use more C++ libraries. On the other hand there are definitely a few Rust projects out there that might come in handy (for example https://github.com/huggingface/tokenizers). And the build process is pretty awful right now (CMake it is but with lots of hacks).
-
[D] What's going to be the dominant language for machine learning in 5 years?
A full machine learning pipeline usually comprises far more than just the model, and this is the area where Rust may shine (the recent work by HuggingFace and their https://github.com/huggingface/tokenizers library is a good example)
-
substitute for tokenizer in torchtext
As for other tokenizers, you can take a look at - Huggingface tokenizers library: https://github.com/huggingface/tokenizers - NLTK tokenize: https://www.nltk.org/api/nltk.tokenize.html - Polygot: https://pypi.org/project/polyglot/
-
PyO3: Rust Bindings for the Python Interpreter
Huggingface Tokenizers (https://github.com/huggingface/tokenizers), which are now used by default in their Transformers Python library, use pyO3 and became popular due to the pitch that it encoded text an order of magnitude faster with zero config changes.
It lives up to that claim. (I had issues with return object typing when going between Python/Rust at first but those are more consistent now)
-
Rusticles #19 - Wed Nov 11 2020
huggingface/tokenizers (Rust): 💥Fast State-of-the-Art Tokenizers optimized for Research and Production
What are some alternatives?
ffi-overhead - comparing the c ffi (foreign function interface) overhead on various programming languages
onnx-tensorflow - Tensorflow Backend for ONNX
whatlang-pyo3 - Python Binding for Rust WhatLang, a language detection library
onnxruntime - ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
dtparse - Fast datetime parser for Python written in Rust
setuptools-rust - Setuptools plugin for Rust support
rust-numpy - PyO3-based Rust bindings of the NumPy C-API
BlingFire - A lightning fast Finite State machine and REgular expression manipulation library.
pythran - Ahead of Time compiler for numeric kernels
rayon - Rayon: A data parallelism library for Rust
tch-rs - Rust bindings for the C++ api of PyTorch.