Marc Romeyn

Senior ML engineer @nvidia. Building open-source LLM tooling (NeMo & NeMo-Run). Deep Learning, ML-infra & Recsys. ex @spotify, @efLDN alum. 🇳🇱

[Bryan Catanzaro @ctnzr]

NVIDIA Nemotron 3 Ultra is now live!

Frontier accuracy, 5X greater speed, 30% lower cost.

Deploy however you need - on-premise, on the cloud, or at the edge.

Model is live on HuggingFace under the OpenMDW 1.1 license.

https://t.co/IOfAwv3jB6

[Oleksii Kuchaiev @kuchaev]

Nemotron 3 Super is here — 120B total / 12B active, Hybrid SSM Latent MoE, designed for Blackwell.
Truly open: permissive license, open data, open training infra. See analysis on @ArtificialAnlys

Details in thread 🧵below:

[The Kobeissi Letter @KobeissiLetter]

BREAKING: Warren Buffett's Berkshire Hathaway, $BRK.A, officially becomes the first US company outside of tech to hit $1 trillion in market cap.

[Marc Romeyn @MarcRomeyn]

Cool post from my old team at Spotify!

[TensorFlow @TensorFlow]

Crafting listening experiences for users through reinforcement learning. 🧠💡🎧

Explore how @Spotify leveraged the TensorFlow ecosystem to solve recommendation problems, design an extendable offline simulator, and train RL Agents to generate playlists. → https://t.co/IdbWzDzdH5

[Marc Romeyn @MarcRomeyn]

Very excited this is out! Showcases some new APIs of Merlin Models we have been working on last summer.

[Gabriel Moreira @gspmoreira]

In #recsys you might want to predict different types of user events (click, like, share, purchase).
In this post we introduce multi-task learning (MTL) for ranking models and how to easily build and train such models using #Merlin and #tensorflow.
https://t.co/pMMNRt3Ig6

[Marc Romeyn @MarcRomeyn]

@Quaxi_ Learning Dutch?!

[Marc Romeyn @MarcRomeyn]

Good thread on some of the new features of Merlin that I have been working on for the last year. Stay tuned, much more to come!

[Shoya Yoshida @decisiontree7]

I've been playing around with NVIDIA's Merlin Models library for the last couple days for the H&M Kaggle competition and here are my thoughts on the library! 🧵

[Marc Romeyn @MarcRomeyn]

I had fun time talking to @anders_arpteg on the AIAW podcast. We spoke about a wide variety of different topics, spanning from @nvidia, @Spotify, @join_ef, Recsys, ML-ops & Web-3. It can be found as video (https://t.co/87bIfZ6nxZ) + podcast platforms like https://t.co/kV6ivVhFvg

[Marc Romeyn @MarcRomeyn]

@karlhigley Do you happen to know a recsys modeling library that makes this mix&matching extremely easy?;)

[Marc Romeyn @MarcRomeyn]

@eugeneyan @nvidia Nice, that makes it easier since our TF examples are a bit behind our PyTorch ones. Just lemme know if you get stuck somewhere in our tutorials

[Marc Romeyn @MarcRomeyn]

@rwstephe Wrong choice dude…

[Marc Romeyn @MarcRomeyn]

@gspmoreira @eggie5 It would be interesting to see if this happens more often on larger recsys datasets as well. We made it very easy in Transformers4Rec to combine various masking-schemes with transformer-architectures from Huggingface. Docs: https://t.co/SOdZQiEyAO

[Marc Romeyn @MarcRomeyn]

I am a bit embarrassed that I never really knew what ASML did as a Dutch person. I learned a lot from this thread

[Trung Phan @TrungTPhan]

ASML is the most important company you've never heard of.

The $300B+ Dutch firm makes the machines that make semiconductors. Each one costs $150m and access to them are a huge geopolitical flashpoint.

Here's a breakdown 🧵

[Marc Romeyn @MarcRomeyn]

@mattnworb Penalties are anti-Dutch

[Marc Romeyn @MarcRomeyn]

I am very excited about this! We're currently working hard on getting the library ready for open-sourcing, so stay tuned for that

[Even Oldridge @Even_Oldridge]

Incredibly excited to share that our @NVIDIAAI paper "Transformers4Rec: Bridging the Gap between NLP and Sequential/Session-Based Recommendation" which wraps @huggingface transformers for use in RecSys has been accepted to @ACMRecSys. See you in Amsterdam #RecSys2021!

[Marc Romeyn @MarcRomeyn]

4/ I can't wait to help trying to raise the bar of this tooling with the goal of advancing and democratizing Recsys for everyone.

[Marc Romeyn @MarcRomeyn]

1/ Personal update: after 4 amazing years at @spotify, I am very excited to start a new chapter today at @nvidia. I will be working on https://t.co/VamhiLXIZy, an open source end-to-end GPU-accelerated recsys platform.

[Marc Romeyn @MarcRomeyn]

3/ Building large-scale recommender systems is notoriously hard, and open-source tooling is arguably behind compared to more mature ML-fields like Computer Vision & NLP.