Adam Roberts @ada_rob Twitter profile | Pikagi

Pikagi

Adam Roberts

@ada_rob

7,335

Followers

648

Following

124

Media

1,470

Statuses

ai researcher @ Google DeepMind :: ♫ (MusicVAE, NSynth, MusicLM, SingSong) & 📝 (T5, PaLM) & :: t5x & seqio // recovering comp biologist

Chapel Hill, NC

https://t.co/DLTHY7TDEh

Joined December 2011

Don't wanna be here? Send us removal request.

Pinned Tweet

@ada_rob

Adam Roberts

2 months

As promised, we're continuing to push our research 🔬 into #MusicFX DJ to make the experience more engaging and controllable--putting the human deeper in the infinite loop. First up, we've increased the responsiveness and musical diversity, and added a

10

8

88

Last Seen Profiles

@_wikkk

@AC_assettocorsa

@CHGO_Bears

@nix_eth

@zdanowicz_s

@L0veleeVita3

@KKatsis

@fernandukibro

@monicapalacios

@nancyrober43651

@OneStepFromEden

@APDasociacion

@ZineIncoming

@bowlds_rae57003

@PhilHs10

@mucc_tatsuro

@MrsWhitehouse74

@WithTheBest

@charlesofplay

@1JustHere4Facts

@Orln

@SaudiMobileShow

@mcnaughtonjenna

@Iraqimofa

@DeeAsiaa_

@Tathyung

@UWCL

@TheTenorGuy01

@leonpiasek

@donnamarti66947

@anthony_borre

@RosieDowell1

@Magnustalma16

@nancywalke71549

@ChillaKiana

@sanuzis

@ada_rob

Adam Roberts

6 years

We just launched Magenta Studio: a suite of free music-making tools using our machine learning models! They're available as an @ableton plugin or as standalone Electron apps. So excited to see what you make with them!! #MadeWithMagenta 👉 Download them at

24

284

1K

@ada_rob

Adam Roberts

2 years

That was fast. News sites are already using DALL-E (mini) to generate fake headline images. This move seems questionable to say the least @NextShark

Tweet media one

24

118

866

@ada_rob

Adam Roberts

4 years

We are releasing mT5: A massively-multilingual version of T5 that supports over 💯 languages! mT5 was pre-trained on a multilingual version of C4 and achieves SoTA on many cross-lingual NLP tasks. 📜Pre-print: 💾Code/models:

Tweet media one

4

124

507

@ada_rob

Adam Roberts

4 years

Check out our @GoogleAI blog post discussing the T5 paper from October and some fun follow-up work. Not only did T5 redefine SoTA for many NLP tasks 📈, it can also tell you how to make a peanut butter banana sandwich 🥪 even Elvis would be proud of.

9

112

479

@ada_rob

Adam Roberts

6 years

Introducing the MAESTRO dataset (172 hours of aligned MIDI and audio piano performances) and Wave2Midi2Wave: a factorized approach to modeling and generating piano performances. w/ @fjord41 Andriy Stasyuk @iansimon @huangcza @sedielem Read 👉 Listen 👇

8

144

404

@ada_rob

Adam Roberts

4 years

As promised, we have made the Text-To-Text Transfer Transformer (T5) models much easier to fine-tune for new tasks, and we just released a Colab notebook where you can try it yourself on a free TPU! 👇 (1/3)

6

109

404

@ada_rob

Adam Roberts

5 years

We just updated Magenta Studio for @Ableton !! 🎉 The big change is a 5th plugin called Drumify which generates 🥁 grooves to accompany melodies, chord progressions, & basslines using #ML . Check it out in action 🎬👇 Learn more and try it yourself at !

10

81

372

@ada_rob

Adam Roberts

6 years

Lots of exciting work to share today about #MusicVAE on the #Magenta blog: a full description, a JavaScript library built on @deeplearnjs , 3 browser-based apps to try it yourself, and examples galore!

Tweet card media

MusicVAE: Creating a palette for musical scores with machine learning.

When a painter creates a work of art, she first blends and explores color options on an artist’s palette before applying them to the canvas. This process is ...

magenta.tensorflow.org

7

130

357

@ada_rob

Adam Roberts

5 years

The birth of Fruit Genie is still my favorite performance (by @stevendrozd ). More behind-the-scenes at .

8

111

330

@ada_rob

Adam Roberts

6 years

Excited to share my newest work #MusicVAE for interpolating and sampling melodies, beats, and three-part song segments from a VAE! Listen to samples and create your own in the #colab notebook (link in YT description) w/ @jesseengel @deck #magenta #nips2017

Tweet card media

From "MusicVAE: A Hierarchical Latent Vector Model for Learning Long-Term Structure in Music". https://g.co/magenta/music-vae

www.youtube.com

4

94

301

@ada_rob

Adam Roberts

3 years

For a year, the T5 team has collab'd with FLAX and JAX to build a successor to our research library, using it to train models at many scales 📈 on TPU... ...and now you can too! T5X is still in rapid development, but you can use it or find inspiration at !

Tweet card media

GitHub - google-research/t5x

Contribute to google-research/t5x development by creating an account on GitHub.

4

71

296

@ada_rob

Adam Roberts

5 years

Last summer, @GoogleMagenta intern @jon_gillick recorded MIDI of professional drummers to train MusicVAEs (GrooVAEs🥁). This awesome work is core Magenta Studio and was accepted to #icml2019 🎉 Today we release the dataset, code and colab on our blog: 👀

Tweet card media

GrooVAE: Generating and Controlling Expressive Drum Performances

In this post we introduce the GrooVAE class of models (pronounced “groovay”) for generating and controlling expressive drum performances, along with a new da...

magenta.tensorflow.org

9

82

292

@ada_rob

Adam Roberts

4 years

New preprint: How Much Knowledge Can You Pack into the Parameters of a Language Model? We show that T5 outperforms all previous open-domain QA systems *without using any external knowledge or context*. Joint work w/ @colinraffel & Noam Shazeer. (1/5)

Tweet media one

5

76

281

@ada_rob

Adam Roberts

5 years

Mission accomplished. I can now retire.

Tweet card media

AI Sheep Song (Using GPT-2, Synthesizer V, and Google Magenta)

This song was written, composed, and sung by several narrow AI technologies in about 4 minutes.LYRICS: Sample output from Open AI's GPT-2 when conditioned on...

www.youtube.com

9

65

274

@ada_rob

Adam Roberts

6 months

Does this mean we can go back to doing research again?

4

11

252

@ada_rob

Adam Roberts

5 months

T5 Reunion! ( @NoamShazeer was replaced by a sentinel token)

Tweet media one

3

4

255

@ada_rob

Adam Roberts

1 year

After a 2 year hiatus from music to work on LLMs, I'm psyched to be back at it with #MusicLM and #SingSong . #SingSong has been blowing my mind for weeks and I'm ecstatic to finally be able to share it with the world. Looking for ways to bring it closer to you...

@chrisdonahuey

Chris Donahue

1 year

Excited to share SingSong, a system which can generate instrumental accompaniments to pair with input vocals! 📄 🔊 Work co-led by myself, @antoine_caillon , and @ada_rob as part of @GoogleMagenta and the broader MusicLM project 🧵

147

764

3K

4

34

242

@ada_rob

Adam Roberts

4 years

I *love* these examples of doing *unsupervised* text style transfer with T5!

Tweet media one

1

39

230

@ada_rob

Adam Roberts

6 years

We just launched Magenta.js: an API for generating music with @GoogleMagenta models in the browser! Read about it and play with some demos at ! @iansimon @fjord41 @teropa

2

68

192

@ada_rob

Adam Roberts

6 years

The first convincing musical style transfer I've heard. Make sure you check out African drums and whistling to symphony! Nice work Mor et al! ( )

Tweet card media

A Universal Music Translation Network

1. String Quartet, Haydn (0:11)2. Cantata Opera, Bach (0:47)3. Classical Guitar (0:58)4. Symphony, Mozart (1:09)5. Music of Africa (1:27)6. Whistling (human)...

www.youtube.com

2

68

183

@ada_rob

Adam Roberts

2 years

Here is a real-world 🚲 example (not in the paper) for T5 Small (~60M params). The VeLO-trained model reaches the same loss as >5x steps of Adafactor and takes only ~1.5x as long per step in wall time 🕓...a >3x speedup⏫!! Pretty amazing for such an OOD task!

Tweet media one

3

16

159

@ada_rob

Adam Roberts

6 years

Hear some AI generated music at our talk and poster today at #ICML2018 ! Talk is at 2:50 in Victoria. Poster is #175 and 👇 @jesseengel @colinraffel @fjord41 @douglas_eck

Tweet media one

1

30

136

@ada_rob

Adam Roberts

4 years

UPDATE: We have spent the past month “fine-tuning” our approach for Closed Book QA (CBQA, no access to external knowledge) w/ T5 and now our appendix is overflowing with interesting results and new SoTAs on open domain WebQuestions and TriviaQA! (1/7)

Tweet card media

How Much Knowledge Can You Pack Into the Parameters of a Language Model?

It has recently been observed that neural language models trained on unstructured text can implicitly store and retrieve knowledge using natural language queries. In this short paper, we measure...

@ada_rob

Adam Roberts

4 years

New preprint: How Much Knowledge Can You Pack into the Parameters of a Language Model? We show that T5 outperforms all previous open-domain QA systems *without using any external knowledge or context*. Joint work w/ @colinraffel & Noam Shazeer. (1/5)

Tweet media one

5

76

281

3

27

118

@ada_rob

Adam Roberts

6 years

The #beatblender AI Experiment built by @tBlankensmith and @hapticdata with #MusicVAE is now live! I am just having fun sitting here clicking "GENERATE NEW BEAT" over and over...

2

36

106

@ada_rob

Adam Roberts

6 years

Accidentally dropped a few layers in my WaveNet and it decided to smash my piano.

3

17

106

@ada_rob

Adam Roberts

5 years

More album liner notes should include paper citations.

Tweet media one

1

18

102

@ada_rob

Adam Roberts

2 years

I've found myself referring to this table in the appendix of our recent paper () A LOT so I thought I'd point it out for those who might benefit from it. Basic upshot is that ED:MLM works really well, especially for classification tasks, even in 0shot.

Tweet media one

5

13

95

@ada_rob

Adam Roberts

6 months

I'm really proud of the work our team continues to do to make generative AI a transformative tool for human creativity! *Music AI Tools* will bring unprecedented creative capability and control to more people, developed responsibly with artists throughout the process.

@demishassabis

Demis Hassabis

6 months

Thrilled to share #Lyria , the world's most sophisticated AI music generation system. From just a text prompt Lyria produces compelling music & vocals. Also: building new Music AI tools for artists to amplify creativity in partnership w/YT & music industry

111

537

3K

9

15

94

@ada_rob

Adam Roberts

2 years

This is paving the way for a very dangerous practice by media organizations. DALL-E mini is pretty obvious, but once models at the level of #dalle2 and #imagen are widely available, we are in trouble.

Tweet media one

7

3

85

@ada_rob

Adam Roberts

2 years

Very excited about this new work demonstrating the benefits of multitask finetuning / instruction tuning! Also happy to continue the trend of open-sourcing our T5 model variants via #t5x for everyone to use and build upon (even if you use @huggingface :) -- more to come!

@hwchung27

Hyung Won Chung

2 years

New paper + models! We extend instruction finetuning by 1. scaling to 540B model 2. scaling to 1.8K finetuning tasks 3. finetuning on chain-of-thought (CoT) data With these, our Flan-PaLM model achieves a new SoTA of 75.2% on MMLU.

Tweet media one

9

64

353

0

15

84

@ada_rob

Adam Roberts

4 years

Speaking totally from my own perspective, while I'm technically impressed with @OpenAI 's jukebox, I have many concerns from an ethical perspective. Why was the GPT-2 launch treated with such sensitivity but not this?

2

1

82

@ada_rob

Adam Roberts

4 years

I'm so excited about the program we've put together for Saturday's #NeurIPS2019 ML for Creativity and Design Workshop 3.0. Aside from the amazing accepted talks and posters, we have a diverse set of invited speakers I want to highlight in this thread.

NeurIPS2019 ML for Creativity & Design

NeurIPS 2019 Workshop on Machine Learning for Creativity and Design

neurips2019creativity.github.io

3

17

81

@ada_rob

Adam Roberts

3 years

TEXT PROMPT: crazy MAGA protestors staging a coup DALL-E:

Tweet media one

1

8

78

@ada_rob

Adam Roberts

2 years

After 2 years focused on infra, it was a joy to dive back into research with @colinraffel and the friendly folks at @BigscienceW while taking OSS T5X for a spin. Our systematic study fills some gaps in our knowledge of LLM design for zero-shot. Hope you find it useful!

@BigscienceW

BigScience Research Workshop

2 years

🧐 When targeting zero-shot use, should you train a T5, a PrefixLM, or a GPT? What if you plan to leverage multitask finetuning (à la T0)? 🤩 In , we explore how architectures & pretraining objectives impact zero-shot performance. ⬇️ Thread time!

Tweet media one

3

82

316

1

3

72

@ada_rob

Adam Roberts

2 years

We used T5X and SeqIO to train PaLM 🌴, and also released a short paper on this infrastructure alongside the model. Check it out at . Huge thanks to the many contributors and my co-lead authors @hwchung27 @anselmlevskaya @mishra_gaurav_ @jekbradbury !

Tweet card media

Scaling Up Models and Data with $\texttt{t5x}$ and $\texttt{seqio}$

Recent neural network-based language models have benefited greatly from scaling up the size of training datasets and the number of parameters in the models themselves. Scaling can be complicated...

@GoogleAI

Google AI

2 years

Introducing the 540 billion parameter Pathways Language Model. Trained on two Cloud #TPU v4 pods, it achieves state-of-the-art performance on benchmarks and shows exciting capabilities like mathematical reasoning, code writing, and even explaining jokes.

76

1K

4K

2

9

69

@ada_rob

Adam Roberts

5 years

Apparently we have been going after the wrong target audience. Kids love Magenta! #eastbaymakerfaire

Tweet media one

3

7

67

@ada_rob

Adam Roberts

1 year

Exciting new work! Biggest takeaway IMHO is that you should switch to finetuning the new👶Flan-T5 checkpoints over the OG🧓T5 ones for best results. Congrats to @ShayneRedford on leading this project and helping us continue the tradition of opensourcing our T5 work!

@ShayneRedford

Shayne Longpre

1 year

✨New Paper✨What’s the best completely public competitor to #ChatGPT ? Flan-T5 beats all public models we tested: Flan-T5 3B ▶️ T0++ 3B ▶️ OPT-IML 175B ▶️ GLM-130B ▶️ Flan 2021 3B ▶️ NIv2 3B We release the @GoogleAI 🌟Flan Collection🌟data + methods for Instruction Tuning! 1/

Tweet media one

Tweet media two

24

251

1K

1

11

66

@ada_rob

Adam Roberts

3 years

So happy to see @Fiery_Furnaces back together again. If you've never seen them live before, the window has now reopened for you to correct that error.

Tweet media one

2

9

63

@ada_rob

Adam Roberts

6 years

Check out the GrooVAE plugin based on a new variant of MusicVAE by Jon GIllick. Humanize your beats with machine learning! 🥁

2

9

63

@ada_rob

Adam Roberts

3 years

Happy to help make the fully-generated mC4 corpus available to the public!

@ai2_allennlp

AllenNLP

3 years

We just released the multilingual C4 dataset! 1.4TB of German? 1.6TB of Spanish? 3.6TB of Russian 😲? We got it all. Check out the announcement at !

2

100

435

0

9

63

@ada_rob

Adam Roberts

6 years

"It’s tempting to analogize implementations to scientific experiments, but usually they aren’t. It’s tempting to think of them as engineering solutions, but they usually aren’t. I think “implementations” are best understood as design solutions..." 👆👇

How should we evaluate progress in AI? | Better without AI

Improving artificial intelligence research with scientific testing, design practice, and meta-rational choice of methods and criteria

betterwithout.ai

1

15

56

@ada_rob

Adam Roberts

5 years

SOOOOO excited this album is finally out!! Musical parts were composed using MusicVAE and NSynth bleeps all over it! A truly groundbreaking human/AI collaboration involving many amazing artists, researchers, and ML models!

@YACHT

ʏ 𝗔 ⊂ ᴴ ᴛ̞

5 years

𝘊𝘩𝘢𝘪𝘯 𝘛𝘳𝘪𝘱𝘱𝘪𝘯𝘨 is out now planetwide on 𝘿 𝙁 𝘼 ⚡︎ @dfarecords ⚡︎ stream it anywhere:

5

15

73

1

17

55

@ada_rob

Adam Roberts

6 years

I never thought I'd say this, but I think I like JavaScript now.

3

5

46

@ada_rob

Adam Roberts

3 years

my @zzznah #deepdream commission is 🍄

Tweet media one

5

4

48

@ada_rob

Adam Roberts

4 years

@boredyannlecun and @ylecun 's shared talk at #NeurIPS4Creativity is a must-see. @MirowskiPiotr @korymath

Tweet card media

Yann LeCun · Machine Learning For Creativity and Designs - Posters

1

13

41

@ada_rob

Adam Roberts

6 years

Check out the updated MusicVAE paper for #icml2018 featuring new attribute vector results w/ Hinton diagrams courtesy of @colinraffel and an example page thanks to @fjord41 ! paper: examples:

Tweet media one

0

15

41

@ada_rob

Adam Roberts

6 years

come walk on our latent space @sonarplusd

Tweet media one

1

9

41

@ada_rob

Adam Roberts

6 years

Absolutely brilliant work.

Tweet card media

Everybody Dance Now

Full paper - https://arxiv.org/pdf/1808.07371.pdfWebsite - https://carolineec.github.io/everybody_dance_now/

www.youtube.com

1

20

40

@ada_rob

Adam Roberts

5 years

Very happy that the Creativity Workshop was accepted for a third year at NeurIPS, but even happier to see this wonderful breakdown of the selection process by the selection committee!

0

6

40

@ada_rob

Adam Roberts

1 year

While things may be heating up in the competitive space of AI and LLMs, it's important that we share resources and artifacts for the benefit of science and progress! Also important to have talented students like @ShayneRedford remind us of this fact.

@GoogleAI

Google AI

1 year

Today we’re releasing a new collection of tasks, templates and methods for instruction tuning of #ML models. Training on this collection can enable language models to reason more competently over arbitrary, unseen tasks. Learn all about it at:

Tweet media one

9

72

393

2

5

39

@ada_rob

Adam Roberts

5 years

Now available for Windows as standalone apps!

Tweet card media

Magenta Studio - Standalone

Magenta Studio is a collection of music plugins built on Magenta's open source tools and models.

magenta.tensorflow.org

@ada_rob

Adam Roberts

5 years

We just updated Magenta Studio for @Ableton !! 🎉 The big change is a 5th plugin called Drumify which generates 🥁 grooves to accompany melodies, chord progressions, & basslines using #ML . Check it out in action 🎬👇 Learn more and try it yourself at !

10

81

372

1

5

38

@ada_rob

Adam Roberts

5 years

@yacht is making some truly groundbreaking art with collaborators across the world of machine learning and art as @TheUniverse discussed at #io19 yesterday (). Lots more exciting stuff to come from them so stay tuned!!!

Tweet media one

1

8

37

@ada_rob

Adam Roberts

2 years

Kudos to @OpenAI for making this type of generation impossible with their model and API. That will slow things down a bit... In the meantime it is important to solidify journalistic standards on this and distribution mechanisms (like Yahoo and Google News) to enforce them.

5

2

33

@ada_rob

Adam Roberts

5 years

First demo using GANSynth in Magenta.js!

@teropa

Tero Parviainen

5 years

Circular interpolation in the GAN-generated timbre space. Kind of like a 2-dimensional sinusoidal LFO I suppose. 〜

1

7

18

0

6

34

@ada_rob

Adam Roberts

1 year

This is awesome, especially the interpolations. Feels like MusicVAE for audio!!!

@_akhaliq

AK

1 year

Riffusion, real-time music generation with stable diffusion @huggingface model: project page:

Tweet media one

64

627

3K

0

2

33

@ada_rob

Adam Roberts

4 years

Artistic and cultural appropriation? Deep faking specific artists voices and style? Use of sensitive language outside of personal context.

2

1

30

@ada_rob

Adam Roberts

5 years

I recently gave a talk at @GrayAreaorg for the #AlgorithmicArtAssembly covering the rapid progress in generative models in recent years and how @GoogleMagenta is exploring their use as creative tools. ...and it's now available on YouTube! Take a 👀:

Tweet card media

Adam Roberts - Exploring the Creative Potential of Deep Learning...

https://ai.google/research/people/104881Exploring the Creative Potential of Deep Learning through the Magenta ProjectOver the past 3 years, the increase in t...

www.youtube.com

4

6

33

@ada_rob

Adam Roberts

6 years

If you're at #abletonLoop today, come hear @jesseengel and I talk about how you can use now use Magenta's machine learning models as part of your @ableton workflow at 11:45!

3

7

33

@ada_rob

Adam Roberts

5 years

My thoughts on @YACHT 's fantastic new album.

YACHT's new album is powered by ML + Artists

The LA-based dance-pop trio YACHT just released their new album, a collaboration with Magenta and other ML researchers and artists! We hope to be able to fea...

magenta.tensorflow.org

1

17

32

@ada_rob

Adam Roberts

4 years

Just added much improved support for exporting SavedModels to the T5 codebase. This makes it much easier and faster to do inference on your fine-tuned model! Check out the new example at the bottom of the colab:

0

5

29

@ada_rob

Adam Roberts

5 years

A few of you at #ismir2019 asked me what I've been up to. Well, I took a sorta-but-not-really-hiatus from Magenta to work on NLP! The result was T5, which has been a very rewarding experience. Now, I'm looking forward to bringing some new insights back to music generation.

@colinraffel

Colin Raffel

5 years

New paper! We perform a systematic study of transfer learning for NLP using a unified text-to-text model, then push the limits to achieve SoTA on GLUE, SuperGLUE, CNN/DM, and SQuAD. Paper: Code/models/data/etc: Summary ⬇️ (1/14)

Tweet media one

9

370

1K

2

5

31

@ada_rob

Adam Roberts

2 years

@hardmaru tells me y'all like bears. "TMZ photo of a Teddy bear leaving LA's hippest new club with her entourage." #dalle

Tweet media one

1

3

30

@ada_rob

Adam Roberts

5 years

That new GAN is cool and all, but don't sleep on this!

@huangcza

Anna Huang

5 years

Check out our new blog post on music transformer that generates minute-long pieces that I find myself listening to again! It also generates compelling harmonizations and realizes chord progressions. Which is your favorite? Ours is the Jazz 🎷 👂

39

414

1K

0

5

30

@ada_rob

Adam Roberts

4 years

13 years ago yesterday I asked awrness to marry me and 13 years ago today we went to the courthouse in durhamnc to make it official! A few weeks later we moved to the Bay Area and have been there ever since. Today,…

7

0

28

@ada_rob

Adam Roberts

5 years

@ylecun and it uses CNNs!

2

3

28

@ada_rob

Adam Roberts

6 months

I've said it once and I'll say it again: I 'd love to do a second PhD and have Colin be my advisor. Actually, maybe I'll apply...but you should too!

@colinraffel

Colin Raffel

6 months

Also, I am 1000% hiring PhD students this round! If you want to work on - open models - collaborative/decentralized training - building models like OSS - coordinating model ecosystems - mitigating risks you should definitely apply! Deadline is Friday 😬

12

75

458

0

1

29

@ada_rob

Adam Roberts

6 months

Seems plausible that this comes down to limited compute resources. Ilya wants to scale up, Sam wants to sell more subscriptions.

4

0

28

@ada_rob

Adam Roberts

2 years

Finally, thanks to 🌴, my family will be able to understand my brand of humor. Credit to Jacob Devlin for this one.

Tweet media one

1

2

27

@ada_rob

Adam Roberts

6 years

First year CS students should take a "history of computing" course and @TheUniverse 's "Broad Band" should be required reading. I deputize @justinesherry to help make this a reality :)

2

4

26

@ada_rob

Adam Roberts

6 years

Also read about how @teropa has used Magenta.js to build some fantastic music-making apps!

Tweet card media

Connecting with Music Through Magenta.js

Editorial Note: Over the past few months, we have noticed many impressive apps built by Tero Parviainen ofcreative.ai popping up on Twitter, many of which us...

magenta.tensorflow.org

1

4

26

@ada_rob

Adam Roberts

2 years

Very excited to have been a small part of this amazing project. More work to be done to make this optimizer the go-to for giant (1B+) models, but it's in sight!

@jmes_harrison

James Harrison

2 years

Tired of tuning your neural network optimizer? Wish there was an optimizer that just worked? We’re excited to release VeLO 🚲, the first hyperparameter-free learned optimizer that outperforms hand-designed optimizers on real-world problems: 🧵

Tweet media one

12

171

971

0

2

26

@ada_rob

Adam Roberts

4 years

🤔😉

Open source multilingual C4 (mC4). by tfds-copybara · Pull Request #2565 · tensorflow/datasets

Open source multilingual C4 (mC4).

2

3

23

@ada_rob

Adam Roberts

5 years

Come see me, @yotammann , @notwaldorf , @adamflorin , etc talk about algorithms & art & ...assembly language...? #nerddadjoke

@GrayAreaorg

Gray Area

5 years

Algorithmic Art Assembly is a brand new conference and music festival, focused on algorithmic tools and processes. Join us for two days of talks and two nights of live music featuring artists using algorithms for aesthetics. March 22 – 23rd Tickets here:

Tweet media one

1

8

30

4

5

25

@ada_rob

Adam Roberts

5 years

Looking forward to @TheUniverse (of @YACHT ) speaking on the band's recent work involving machine learning and participating in a discussion with her, @jesseengel , and another (surprise!) artist on Wednesday. You can watch live on YouTube!

@artwithMI

Artists + Machine Intelligence

5 years

(2) Claire Evans AKA @YACHT talks about deep learning as a tool in the creative process of YACHT's new album – with @Google Magenta's @ada_rob and @jesseengel Tune in live 5/8 12:30pm PT or visit us in-person on Stage 4. Learn more ↓

1

7

16

1

6

25

@ada_rob

Adam Roberts

6 years

So happy that the Magenta.js piano transcription found its way into one of these accessibility projects thanks to @yotammann . Really amazing and inspiring work by everyone involved!

2

3

25

@ada_rob

Adam Roberts

5 years

Whoa! I think this is one of the coolest / most usable things I've seen at #ismir2019 .

@ismir2019

ISMIR2019

5 years

"Unmixer is a web interface where users can upload music, extract loops, remix them, and mash-up loops from different songs. To extract loops with source separation, we use a nonnegative tensor factorization method improved with a sparsity constraint."

0

2

10

2

3

24

@ada_rob

Adam Roberts

6 years

One great thing about having a kid is you occasionally rediscover amazing artifacts like a @philipglass scored animation about the geometry of circles from @sesamestreet circa 79

1

3

24

@ada_rob

Adam Roberts

4 years

The EfficientQA @NeurIPSConf Competition () is offering GCP grants and office hours to help expand participation amongst underrepresented groups. Please apply (even if you're new to QA) at by July 30 to be considered for a grant!

1

12

24

@ada_rob

Adam Roberts

4 years

In concurrent work (also posted to arXiv today) our friends at Google Research released the incredibly cool REALM, an open-book QA system that is trained end-to-end to retrieve documents and answer questions. They outperform us on NQ and WQ.

2

3

23

@ada_rob

Adam Roberts

5 years

My first trip to Belgium!

@pvanimpe

GrandZaddy

5 years

Lots of questions for Adam Roberts from @Google here at the @dsbrussels neetup about arts & music. Check the magenta project.

0

3

7

2

1

24

@ada_rob

Adam Roberts

4 years

To see how to train T5 on a simplified variant of this problem with a free TPU, check out our colab notebook: Questions, comments, and feedback welcome! (5/5)

2

4

20

@ada_rob

Adam Roberts

4 years

emergent cheetos

1

0

21

@ada_rob

Adam Roberts

4 years

I think my biggest concern is conditioning on specific artists without their permission, which is something that has rubbed me the wrong way in past work as well. It seems even more problematic now that it's this good.

1

0

19

@ada_rob

Adam Roberts

6 years

I just reheard this very impressive WaveNet-based singing synthesis from Merlijn Blaauw and @JordiBonada during @emiliagogu 's excellent keynote at the #ICML2018 music workshop, and I was once again blown away. Take a listen if you haven't already.

1

10

21

@ada_rob

Adam Roberts

4 years

GrooVAE lives on! @jon_gillick

@notwaldorf

Monica Dinculescu

4 years

We built a really cool machine learning drummer that accompanies your piano in real-time!! This is DrumBot! It’s a web app, so it can come along to your next gig. 🎉 🥁 📝 Blog post: 🎧 Live action:

20

269

1K

1

2

20

@ada_rob

Adam Roberts

6 years

Looking forward to presenting some exciting new Magenta work with @jesseengel and @yotammann at @Ableton Loop in November!

1

4

21

@ada_rob

Adam Roberts

2 years

These models have some journalistic value in making news article illustrations, but not photorealistic images that could be mistaken for photographic evidence.

2

1

19

@ada_rob

Adam Roberts

2 years

@sarahookr was the first AI Resident I mentored, but I am confident in saying that over the years I've learned much more from her than she ever did from me! I'm very excited to see what incredible things For AI will do under her leadership!

@sarahookr

Sara Hooker

2 years

I'm excited to finally share what I have been working on. Today we are officially launching Cohere For AI @forai_ml a non-profit research lab that aims to reimagine how, where, and by whom research is done.

Tweet media one

63

118

825

0

0

20

@ada_rob

Adam Roberts

1 year

@lpachter awww, shucks

1

0

19

@ada_rob

Adam Roberts

4 years

It may not be as flashy, but I am going to continue focusing my work in the music space on interactive tools to enhance creativity for professionals and amateurs alike!

1

0

18

@ada_rob

Adam Roberts

6 years

Morning #ICML2018 . Come check out the awesome set of talks at the workshop on Machine Learning for Music! I'll be presenting the recently launched Magenta.js library and talking about how it is engaging artists, musicians, and creative coders.

0

4

17

@ada_rob

Adam Roberts

5 years

Just wow.

@ctpt_studio

Counterpoint

5 years

Announcing new project: GANHarp This is an experimental musical instrument based on Magenta's GANSynth and Magenta.js. It synthesizes timbres from latent space, and then lets you morph between them as you play. #MadeWithMagenta 🎹

6

80

310

0

5

19

@ada_rob

Adam Roberts

6 years

Also big ups to @yotammann for putting the Magenta.js into @ableton !

2

2

19

@ada_rob

Adam Roberts

5 years

Read/watch a bit more about @GoogleMagenta 's I/O collaboration with @theflaminglips on Fruit Genie and long-term work with @YACHT on their forthcoming album. Still lots more to come! 🔨🍓 Also, learn how to build your own Fruit Genie at !

Tweet card media

Magenta + Deeplocal + The Flaming Lips = Fruit Genie

Editorial Note: At I/O 2019, we announced collaborations with two musical artists during our talk session: YACHT and The Flaming Lips. You can read about the...

magenta.tensorflow.org

@Google

Google

5 years

On the second day of #io19 , two bands took the stage—with a little help from machine learning. Go behind Magenta, the tech that rocked I/O 🎶 →

52

78

502

0

2

18

@ada_rob

Adam Roberts

5 years

Some fun work by our team in the Google Doodle today to celebrate Bach's birthday! Super proud of @huangcza who did a lot of this work as my intern 3 years ago (w/ very little help from me) and continues to do amazing things!!

@huangcza

Anna Huang

5 years

Check out our new blog post on Coconet 🥥, the #ml behind the Bach Doodle that’s live now! It’s a flexible infilling model that generates counterpoint through rewriting. @koning_robot , @notwaldorf , @ada_rob , @fjord41 , @AaronCourville , @douglas_eck

13

165

391

0

2

18

@ada_rob

Adam Roberts

1 year

Forget about the LLM hype, go see @kinggizzard in a cave.

@ada_rob

Adam Roberts

1 year

Forget about the LLM hype, go see @kinggizzard .

0

0

12

2

3

17

@ada_rob

Adam Roberts

5 years

Highly recommended that anyone interested in starting/continuing deep learning research (especially those from underrepresented groups and non-traditional backgrounds) APPLY!

@GoogleAI

Google AI

5 years

Applications for the 2020 Google AI Residency program are now open! Visit for application information. To learn more about the research accomplishments of the 2019 alumni, check out the post below!

9

212

571

0

2

18

@ada_rob

Adam Roberts

5 years

So psyched to hear such awesome music being made with #musicvae and #nsynth ! Such a great example of human/"AI" collaboration. #madewithmagenta

@YACHT

ʏ 𝗔 ⊂ ᴴ ᴛ̞

5 years

“(Downtown) Dancing” ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧ A transmission from deep latent space. Now streaming everywhere: ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧ ↧

1

5

15

0

0

17