Informers

:slightly_smiling_face: State-of-the-art natural language processing for Ruby

Supports:

Sentiment analysis
Question answering
Named-entity recognition
Text generation - in development
Summarization - in development
Translation - in development

Installation

Add this line to your application’s Gemfile:

gem 'informers'

On Mac, also install OpenMP:

brew install libomp

Sentiment Analysis

First, download the pretrained model.

Predict sentiment

model = Informers::SentimentAnalysis.new("sentiment-analysis.onnx")
model.predict("This is super cool")

This returns

{label: "positive", score: 0.999855186578301}

Predict multiple at once

model.predict(["This is super cool", "I didn't like it"])

Question Answering

First, download the pretrained model and add Numo to your application’s Gemfile:

gem 'numo-narray'

Ask a question with some context

model = Informers::QuestionAnswering.new("question-answering.onnx")
model.predict(
  question: "Who invented Ruby?",
  context: "Ruby is a programming language created by Matz"
)

This returns

{answer: "Matz", score: 0.9980658360049758, start: 42, end: 46}

Named-Entity Recognition

First, export the pretrained model.

Get entities

model = Informers::NER.new("ner.onnx")
model.predict("Nat works at GitHub in San Francisco")

This returns

[
  {text: "Nat",           tag: "person",   score: 0.9840519576513487, start: 0,  end: 3},
  {text: "GitHub",        tag: "org",      score: 0.9426134775785775, start: 13, end: 19},
  {text: "San Francisco", tag: "location", score: 0.9952414982243061, start: 23, end: 36}
]

Models

Task	Description	Contributor	License	Link
Sentiment analysis	DistilBERT fine-tuned on SST-2	Hugging Face	Apache-2.0	Link
Question answering	DistilBERT fine-tuned on SQuAD	Hugging Face	Apache-2.0	Link
Named-entity recognition	BERT fine-tuned on CoNLL03	Bayerische Staatsbibliothek	In-progress	Link

Models are quantized to make them faster and smaller.

Deployment

Check out Trove for deploying models.

trove push sentiment-analysis.onnx

Credits

This project uses many state-of-the-art technologies:

Transformers for transformer models
Bling Fire and BERT for high-performance text tokenization
ONNX Runtime for high-performance inference

Some code was ported from Transformers and is available under the same license.

History

View the changelog

Contributing

Everyone is encouraged to help improve this project. Here are a few ways you can help:

Report bugs
Fix bugs and submit pull requests
Write, clarify, or fix documentation
Suggest or add new features

To get started with development:

git clone https://github.com/ankane/informers.git
cd informers
bundle install

export MODELS_PATH=path/to/onnx/models
bundle exec rake test