Gabriel Dulac-Arnold

@[email protected]
62 Followers
42 Following
7 Posts
AI researcher focused on practical RL @ Google Brain, previously DeepMind.
If not tinkering electronic brains, probably surfing, sailing or climbing.
Websitehttp://gabe.squirrelsoup.net
Gabriel Dulac-ArnoldJan 11, 2023
Olivier GriselJan 11, 2023

PEP 703 by Sam Gross and sponsored by @ambv on Making the Global Interpreter Lock Optional in CPython is now a public draft:

https://peps.python.org/pep-0703/

Others and I contributed some context on why this would help improve the efficiency and maintainability of #Python code in the context of #DataScience, #MachineLearning #DeepLearning and more generally in scientific programming / #scipy contexts.

Feel free to discuss this draft here: https://discuss.python.org/t/pep-703-making-the-global-interpreter-lock-optional/22606

#PyData

PEP 703 – Making the Global Interpreter Lock Optional in CPython | peps.python.org

CPython’s global interpreter lock (“GIL”) prevents multiple threads from executing Python code at the same time. The GIL is an obstacle to using multi-core CPUs from Python efficiently. This PEP proposes adding a build configuration (--disable-gil) to...

Python Enhancement Proposals (PEPs)
Gabriel Dulac-ArnoldDec 16, 2022
𝚛𝚊𝚠𝚡Dec 16, 2022

"a beautiful 20-step interpolation from typing to jazz" is especially juicy. #ai #aimusic
---
RT @_akhaliq
Riffusion, real-time music generation with stable diffusion

@huggingface model: https://huggingface.co/riffusion/riffusion-model-v1
project page: https://www.riffusion.com/about
https://twitter.com/_akhaliq/status/1603411250389700611

riffusion/riffusion-model-v1 · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

Gabriel Dulac-ArnoldDec 15, 2022
Marco CuturiDec 15, 2022
We still have an open research internship position @ Apple, in the Machine Learning Research, in Paris (onsite, not remote). Topics of interest include optimization and interplay with deep learning, diffusions/optimal transport, differentiable optimization, etc. We are looking for PhD students with a record of ML publications. Duration is between now and anytime before September 31st. Please reach out to [email protected] if interested.
Gabriel Dulac-ArnoldDec 15, 2022
Mark ReidDec 15, 2022

What if stable diffusion models are used to generate spectrograms and then they are converted to audio?

https://www.riffusion.com/about

The examples of interpolating loops are pretty neat (“typing to jazz” and “church bells to electronic beats”).

Early days, but I could imagine a nice UI being put in front of this and the results being used in a similar way to sample packs are used today in production, e.g., extra texture, fills, initial inspiration, etc.

#Riffusion #StableDiffusion #GenerativeMusic

Riffusion

Stable diffusion for real-time music generation

Riffusion
Gabriel Dulac-ArnoldNov 25, 2022
Pablo Samuel CastroNov 24, 2022

Really exciting to see yet another successful application of #ReinforcementLearning #RL in a meaningful, real-world problem.
Congrats, Cosmin and team!

https://twitter.com/CauseMean/status/1595821831805874176?t=MDpiMnS6whQlr7pbZjr7pA&s=19

Cosmin Paduraru on Twitter

“Excited to share the details of our work at @DeepMind on using reinforcement learning to help large-scale commercial cooling systems save energy and run more efficiently: https://t.co/i2M1cghANJ. Here’s what we found 🧵”

Twitter
Gabriel Dulac-ArnoldNov 24, 2022
Hi all, looking forward to following paper threads and sharing some cool work. Currently at Google Brain in France working on making RL useful and practical.