Q1 ‘22 highlights and achievements

Posted by Nari Yoon, Hee Jung, DevRel Group Supervisor / Soonson Kwon, DevRel Program Supervisor

Let’s discover highlights and accomplishments of huge Google Machine Studying communities over the primary quarter of the yr! We’re enthusiastic and grateful about all of the actions that the communities throughout the globe do. Listed here are the highlights!

ML Ecosystem Marketing campaign Highlights

ML Olympiad is an related Kaggle Group Competitions hosted by Machine Studying Google Builders Consultants (ML GDEs) or TensorFlow Person Teams (TFUGs) sponsored by Google. The primary spherical was hosted from January to March, suggesting fixing important issues of our time. Competitors highlights embody Autism Prediction Problem, Arabic_Poems, Hausa Sentiment Evaluation, High quality Training, Good Well being and Effectively Being. Thanks TFUG Saudi, New York, Guatemala, São Paulo, Pune, Mysuru, Chennai, Bauchi, Casablanca, Agadir, Ibadan, Abidjan, Malaysia and ML GDE Ruqiya Bin Safi, Vinicius Fernandes Caridá, Yogesh Kulkarni, Mohammed buallay, Sayed Ali Alkamel, Yannick Serge Obam, Elyes Manai, Thierno Ibrahima DIOP, Poo Kuan Hoong for internet hosting ML Olympiad!

Highlights and Achievements of ML Communities

TFUG organizer Ali Mustufa Shaikh (TFUG Mumbai) and Rishit Dagli gained the TensorFlow Group Highlight award (paper and code). This venture was supported by supplied Google Cloud credit score.

ML GDE Sachin Kumar (Qatar) posted Construct a retail digital agent from scratch with Dialogflow CX – Final Chatbot Tutorials. On this tutorial, you’ll learn to construct a chatbot and voice bot from scratch utilizing Dialogflow CX, a Conversational AI Platform (CAIP) for constructing conversational UIs.

ML GDE Ngoc Ba (Vietnam) posted MTet: Multi-domain Translation for English and Vietnamese. This venture is about the way to accumulate prime quality information and prepare a state-of-the-art neural machine translation mannequin for Vietnamese. And it utilized Google Cloud TPU, Cloud Storage and associated GCP merchandise for sooner coaching.

Kaggle introduced the Google Open Supply Prize early this yr (Winners announcement web page). In January, ML GDE Aakash Kumar Nain (India)’s Constructing fashions in JAX – Part1 (Stax) was awarded.

In February, ML GDE Victor Dibia (USA)’s pocket book Signature Picture Cleansing with Tensorflow 2.0 and ML GDE Sayak Paul (India) & Soumik Rakshit’s pocket book gaugan-keras had been awarded.

TFUG organizer Usha Rengaraju posted Variable Choice Networks (AI for Local weather Change) and Probabilistic Bayesian Neural Networks utilizing TensorFlow Chance notebooks on Kaggle. They each bought gold medals, and she or he has grow to be a Triple GrandMaster!

TFUG Chennai hosted the 2 occasions, Transformers – A Journey into consideration and Intro to Deep Reinforcement Studying. These occasions had been deliberate for inexperienced persons. Occasions embody introductory classes explaining the transformers analysis papers and the essential idea of reinforcement studying.

ML GDE Margaret Maynard-Reid (USA), Nived P A, and Joel Shor posted Our Summer time of Code Undertaking on TF-GAN. This text describes enhancements made to the TensorFlow GAN library (TF-GAN) of the final summer season.

ML GDE Aakash Nain (India) launched a collection of tutorials about constructing fashions in JAX. Within the second tutorial, Aakash makes use of probably the most well-known and most generally used high-level libraries for Jax to construct a classifier. Within the pocket book, you’ll be taking a deep dive into Flax, too.

ML GDE Bhavesh Bhatt (India) constructed a mannequin for braille to audio with 95% accuracy. He created a mannequin that interprets braille to textual content and audio, lending a serving to hand to folks with visible disabilities.

ML GDE Sayak Paul (India) lately wrote Publishing ConvNeXt Fashions on TensorFlow Hub. This can be a contribution from the 30 variations of the mannequin, prepared for inference and switch studying, with documentation and pattern code. And he additionally posted First Steps in GSoC to encourage the guy ML GDEs’ participation in Google Summer time of Code (GSoC).

ML GDE Merve Noyan (Turkey) skilled 40 fashions on keras.io/examples; constructed demos for them with Streamlit and Gradio. And people are at the moment being hosted right here. She additionally held workshops entitled NLP workshop with TensorFlow for TFUG Delhi, TFUG Chennai, TFUG Hyderabad and TFUG Casablanca. It lined the essential to superior subjects in NLP proper from Transformers until mannequin internet hosting in Hugging Face, utilizing TFX and TF Serve.