1. X
  2. Adam Roberts
Log inSign up
Adam Roberts
1,669 posts
Image
user avatar
Adam Roberts
@ada_rob
Principal Scientist @ Google DeepMind :: 🌈psychedelic ai :: 📝 Foundational LLM work: TL for T5 & C4, core PaLM :: 👨‍💻 TL for T5x & SeqIO
Alameda, CA
g.co/magenta
Joined December 2011
732
Following
8,409
Followers
RepliesRepliesMediaMedia
  • Pinned
    user avatar
    Adam Roberts
    @ada_rob
    Jun 20, 2025
    Introducing Magenta RealTime, our open weights live music model! Like Lyria RealTime, this model provides the user continuous control of the musical flow 🌊. We will be sharing some of our favorite ways to extend these models over the next few weeks. Can't wait to see what you
    user avatar
    Chris Donahue
    @chrisdonahuey
    Jun 20, 2025
    Excited to announce 🎵Magenta RealTime, the first open weights music generation model capable of real-time audio generation with real-time control. 👋 **Try Magenta RT on Colab TPUs**: colab.research.google.com/github/magenta… 👀 Blog post: g.co/magenta/rt 🧵 below
    Image
    00:00
  • user avatar
    Adam Roberts
    @ada_rob
    Jun 12
    Just got carded for Mythos in Greece. This is getting out of hand @AnthropicAI.
    Image
  • user avatar
    Adam Roberts
    @ada_rob
    Jun 4
    Finally, you can truly *play* the model as an instrument, letting you decide how much control you want and how much you want it to riff along with you. AND open weights AND runs on your machine for ultra-low latency. Hope you enjoy it!
    user avatar
    Google Magenta Project
    @GoogleMagenta
    Jun 4
    Introducing Magenta RealTime 2 (MRT2): the live music model you can play as an instrument. MRT2 offers MIDI and prompt controls, and runs natively on a MacBook with <200ms latency. Open weights. Open source inference engine. Suite of apps and plugins. Hear what it can do and
    Image
    00:00
  • user avatar
    Adam Roberts
    @ada_rob
    Dec 19, 2025
    Long live the enc-dec! The reports of its demise have been greatly exaggerated. In fact, has there been a report that doesn't show it excelling?
    user avatar
    Omar Sanseviero
    Google AI Studio
    @osanseviero
    Dec 18, 2025
    Introducing T5Gemma 2, the next generation of encoder-decoder models 🚀 Built on top of Gemma 3, we were able to build compact models at sizes of 270m-270m, 1B-1B, and 4B-4B sizes. While most models today are decoder-only, T5Gemma 2 is the first (I'm aware of) multimodal,
    Image
  • user avatar
    Adam Roberts
    @ada_rob
    Nov 24, 2025
    AGI
    Image
    Image

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Advertisement
Advertisement