rad
@radfounder.bsky.social
data-science masters⌛| founder: spatial x affective computing startup | ivy grad | Paris
My friend launched a crowdfunding campaign for her amazing startup that wants to save Museums and invigorate local communities in France!
Please think about contributing!!
www.kisskissbankbank.com/fr/projects/...
Please think about contributing!!
www.kisskissbankbank.com/fr/projects/...
CHARIT: Mécénat digital pour les Musées Régionaux et leurs Associations d’Amis par CHARIT
CHARIT, votre plateforme de mécénat digital pour soutenir les Musées Régionaux et leurs Associations d’Amis
www.kisskissbankbank.com
December 5, 2024 at 5:58 PM
My friend launched a crowdfunding campaign for her amazing startup that wants to save Museums and invigorate local communities in France!
Please think about contributing!!
www.kisskissbankbank.com/fr/projects/...
Please think about contributing!!
www.kisskissbankbank.com/fr/projects/...
Correct. There are no chambers on X, politics left the town square. All it is now is rage bait and slop that chase monetization.
I am not rooting for X to die, but I’m tired of the social engineered cycles of all accounts posting the same type of slop to get monetized.
I am not rooting for X to die, but I’m tired of the social engineered cycles of all accounts posting the same type of slop to get monetized.
I hate the way articles about BSky keep framing choice to move here in terms of left/right and echo chambers. Fundamentally, the bigger issues for me are wanting to see links/current news, wanting to see posts that aren't rage-baiting/content farming, and wanting actually diverse conversations.
November 27, 2024 at 2:56 PM
Correct. There are no chambers on X, politics left the town square. All it is now is rage bait and slop that chase monetization.
I am not rooting for X to die, but I’m tired of the social engineered cycles of all accounts posting the same type of slop to get monetized.
I am not rooting for X to die, but I’m tired of the social engineered cycles of all accounts posting the same type of slop to get monetized.
Reposted by rad
Supercharging Training using float8 and FSDP2
They demonstrate how they achieved up to 50% throughput speedup while achieving loss and evaluation benchmark parity in training over FSDP1 bf16 training.
pytorch.org/blog/trainin...
They demonstrate how they achieved up to 50% throughput speedup while achieving loss and evaluation benchmark parity in training over FSDP1 bf16 training.
pytorch.org/blog/trainin...
Supercharging Training using float8 and FSDP2
IBM: Tuan Hoang Trong, Alexei Karve, Yan Koyfman, Linsong Chu, Divya Kumari, Shweta Salaria, Robert Walkup, Praneet Adusumilli, Nirmit Desai, Raghu Ganti, Seetharami Seelam
Meta: Less Wright, Wei Feng...
pytorch.org
November 26, 2024 at 5:14 PM
Supercharging Training using float8 and FSDP2
They demonstrate how they achieved up to 50% throughput speedup while achieving loss and evaluation benchmark parity in training over FSDP1 bf16 training.
pytorch.org/blog/trainin...
They demonstrate how they achieved up to 50% throughput speedup while achieving loss and evaluation benchmark parity in training over FSDP1 bf16 training.
pytorch.org/blog/trainin...
Reposted by rad
This is the best pun title I've ever seen
Abducted by Bayesians?
This paper discusses the role of theoretical notions in making predictions and evaluating statistical models. The core idea of the paper is that such …
www.sciencedirect.com
November 25, 2024 at 9:36 PM
This is the best pun title I've ever seen
Reposted by rad
"La Fosse des Mariannes", a science fiction short story featuring AI-animated beings, has just been published in Solaris, the French-language magazine of fantasy, horror, and science fiction, celebrating its 50th anniversary. Its creation was inspired by this image, generated a year ago.
November 26, 2024 at 7:05 PM
"La Fosse des Mariannes", a science fiction short story featuring AI-animated beings, has just been published in Solaris, the French-language magazine of fantasy, horror, and science fiction, celebrating its 50th anniversary. Its creation was inspired by this image, generated a year ago.
Reposted by rad
What's the difference between explainability and interpretability? Does the machine learning community have an agreed-upon definition?
No.
There's a great overview, which is from this paper: arxiv.org/abs/2211.08943
My take: I prefer interpretability since the term explainability is too strong.
No.
There's a great overview, which is from this paper: arxiv.org/abs/2211.08943
My take: I prefer interpretability since the term explainability is too strong.
November 26, 2024 at 3:41 PM
What's the difference between explainability and interpretability? Does the machine learning community have an agreed-upon definition?
No.
There's a great overview, which is from this paper: arxiv.org/abs/2211.08943
My take: I prefer interpretability since the term explainability is too strong.
No.
There's a great overview, which is from this paper: arxiv.org/abs/2211.08943
My take: I prefer interpretability since the term explainability is too strong.
Reposted by rad
While everybody has been distracted with the opensourcing of the Model Context Protocol - and that’s awesome too - there was a library released that supports a unified interface for different GenAI models.
Such a time savior.
By @andrewyng.bsky.social himself.
github.com/andrewyng/ai...
Such a time savior.
By @andrewyng.bsky.social himself.
github.com/andrewyng/ai...
GitHub - andrewyng/aisuite: Simple, unified interface to multiple Generative AI providers
Simple, unified interface to multiple Generative AI providers - GitHub - andrewyng/aisuite: Simple, unified interface to multiple Generative AI providers
github.com
November 26, 2024 at 7:00 PM
While everybody has been distracted with the opensourcing of the Model Context Protocol - and that’s awesome too - there was a library released that supports a unified interface for different GenAI models.
Such a time savior.
By @andrewyng.bsky.social himself.
github.com/andrewyng/ai...
Such a time savior.
By @andrewyng.bsky.social himself.
github.com/andrewyng/ai...
Reposted by rad
November 25, 2024 at 4:02 PM
Day 2 of using X and Bluesky simultaneously:
- 0 ML/AI tools, resources or papers encountered on X.
- 18 to 19 papers, resources and datasets pasted into my personal discord from Bluesky.
I don't want to compare. I prefer having everyone in the same space. But monetization on X was a mistake.
- 0 ML/AI tools, resources or papers encountered on X.
- 18 to 19 papers, resources and datasets pasted into my personal discord from Bluesky.
I don't want to compare. I prefer having everyone in the same space. But monetization on X was a mistake.
November 26, 2024 at 6:56 PM
Day 2 of using X and Bluesky simultaneously:
- 0 ML/AI tools, resources or papers encountered on X.
- 18 to 19 papers, resources and datasets pasted into my personal discord from Bluesky.
I don't want to compare. I prefer having everyone in the same space. But monetization on X was a mistake.
- 0 ML/AI tools, resources or papers encountered on X.
- 18 to 19 papers, resources and datasets pasted into my personal discord from Bluesky.
I don't want to compare. I prefer having everyone in the same space. But monetization on X was a mistake.
Reposted by rad
I'm really enjoying this AI papers feed - thanks for making it @sethlazar.org !
And what a cool feature of this place. 🦋
bsky.app/profile/did:...
And what a cool feature of this place. 🦋
bsky.app/profile/did:...
November 26, 2024 at 3:06 PM
I'm really enjoying this AI papers feed - thanks for making it @sethlazar.org !
And what a cool feature of this place. 🦋
bsky.app/profile/did:...
And what a cool feature of this place. 🦋
bsky.app/profile/did:...
Reposted by rad
What AI podcast do you listen to❓
18 ideas here
Tx to the podcasters! You're appreciated!
on Bluesky
@paulroetzer.bsky.social, @andreykurenkov.bsky.social, @gavinpurcell.bsky.social, @dwarkesh.bsky.social, @eladgil.bsky.social, @swyx.io, @reidhoffman.bsky.social
READ medium.com/@robert_1489...
18 ideas here
Tx to the podcasters! You're appreciated!
on Bluesky
@paulroetzer.bsky.social, @andreykurenkov.bsky.social, @gavinpurcell.bsky.social, @dwarkesh.bsky.social, @eladgil.bsky.social, @swyx.io, @reidhoffman.bsky.social
READ medium.com/@robert_1489...
10 AI Podcasts to Listen To + Bonuses 🎧🤖
I listen to about 10 AI podcasts daily. I recommend finding at least one that you listen to weekly — or more if you have the time. Staying…
medium.com
November 26, 2024 at 6:38 PM
What AI podcast do you listen to❓
18 ideas here
Tx to the podcasters! You're appreciated!
on Bluesky
@paulroetzer.bsky.social, @andreykurenkov.bsky.social, @gavinpurcell.bsky.social, @dwarkesh.bsky.social, @eladgil.bsky.social, @swyx.io, @reidhoffman.bsky.social
READ medium.com/@robert_1489...
18 ideas here
Tx to the podcasters! You're appreciated!
on Bluesky
@paulroetzer.bsky.social, @andreykurenkov.bsky.social, @gavinpurcell.bsky.social, @dwarkesh.bsky.social, @eladgil.bsky.social, @swyx.io, @reidhoffman.bsky.social
READ medium.com/@robert_1489...
Reposted by rad
🎬 Want AI to be your video editor? This tool lets you remix videos with just words - drop in your clips, images & audio, then let AI do the FFMPEG magic! Natural language → custom video content 🪄
Check it out: huggingface.co/spaces/huggi...
Check it out: huggingface.co/spaces/huggi...
November 25, 2024 at 8:07 PM
🎬 Want AI to be your video editor? This tool lets you remix videos with just words - drop in your clips, images & audio, then let AI do the FFMPEG magic! Natural language → custom video content 🪄
Check it out: huggingface.co/spaces/huggi...
Check it out: huggingface.co/spaces/huggi...
Reposted by rad
“We are very clear that our primary user is the end user that we’re serving, and so we’re building features for them, ” Wang said. “If brands come, great, but you should be authentic; that’s what we tell them. It’s not going to work if you try to sound like a brand.” I did say this 👍
My colleague @katiehicks.bsky.social at Marketing Brew interviewed Bluesky COO @rose.bsky.team. "Bluesky has more than 21 million users. Where does it go from here?"
Bluesky has more than 21 million users. Where does it go from here?
The platform won’t touch ads for now, and is instead focused on paid subscriptions, payments, and getting more people signed up, its COO said.
www.marketingbrew.com
November 25, 2024 at 6:35 PM
“We are very clear that our primary user is the end user that we’re serving, and so we’re building features for them, ” Wang said. “If brands come, great, but you should be authentic; that’s what we tell them. It’s not going to work if you try to sound like a brand.” I did say this 👍
Reposted by rad
At @huggingface.bsky.social 🤗 we're preparing a collaborative annotation effort to build an open-source multilingual dataset.
If you'd like to get high-quality open data for your language, check if yours is listed in this form and sign up!
forms.gle/DHJdtvoSNxAA...
If you'd like to get high-quality open data for your language, check if yours is listed in this form and sign up!
forms.gle/DHJdtvoSNxAA...
Language Lead sign-up
At Hugging Face 🤗, we're launching a big community initiative to improve LLM training for many languages. We're looking for Language Leads to help us cultivate specific languages during this initiativ...
forms.gle
November 26, 2024 at 1:16 PM
At @huggingface.bsky.social 🤗 we're preparing a collaborative annotation effort to build an open-source multilingual dataset.
If you'd like to get high-quality open data for your language, check if yours is listed in this form and sign up!
forms.gle/DHJdtvoSNxAA...
If you'd like to get high-quality open data for your language, check if yours is listed in this form and sign up!
forms.gle/DHJdtvoSNxAA...
Reposted by rad
NuminaMath dataset now under Apache 2.0 license.
AI-MO/NuminaMath-CoT · Datasets at Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
November 25, 2024 at 9:31 AM
NuminaMath dataset now under Apache 2.0 license.
Reposted by rad
Lots of folks talking about scaling LLM inference over this last year
Internally, I’ve been developing and using a library that makes this extremely easy, and I decided to open-source it
Meet the decoding library: github.com/benlipkin/de...
1/7
Internally, I’ve been developing and using a library that makes this extremely easy, and I decided to open-source it
Meet the decoding library: github.com/benlipkin/de...
1/7
GitHub - benlipkin/decoding: Composable inference algorithms with LLMs and programmable logic
Composable inference algorithms with LLMs and programmable logic - benlipkin/decoding
github.com
November 25, 2024 at 4:19 PM
Lots of folks talking about scaling LLM inference over this last year
Internally, I’ve been developing and using a library that makes this extremely easy, and I decided to open-source it
Meet the decoding library: github.com/benlipkin/de...
1/7
Internally, I’ve been developing and using a library that makes this extremely easy, and I decided to open-source it
Meet the decoding library: github.com/benlipkin/de...
1/7
Reposted by rad
We're going to see some very cool open multilingual efforts shared before the end of the year...
Find the OpenGPT-X models here: huggingface.co/openGPT-X
Find the OpenGPT-X models here: huggingface.co/openGPT-X
November 26, 2024 at 10:33 AM
We're going to see some very cool open multilingual efforts shared before the end of the year...
Find the OpenGPT-X models here: huggingface.co/openGPT-X
Find the OpenGPT-X models here: huggingface.co/openGPT-X
Reposted by rad
What are the brain’s “real” tuning curves?
Our new preprint "SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” argues that existing techniques for latent variable discovery are lacking.
We suggest a much simpl-er way to do things.
1/21🧵
Our new preprint "SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” argues that existing techniques for latent variable discovery are lacking.
We suggest a much simpl-er way to do things.
1/21🧵
November 25, 2024 at 1:39 PM
What are the brain’s “real” tuning curves?
Our new preprint "SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” argues that existing techniques for latent variable discovery are lacking.
We suggest a much simpl-er way to do things.
1/21🧵
Our new preprint "SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” argues that existing techniques for latent variable discovery are lacking.
We suggest a much simpl-er way to do things.
1/21🧵
Reposted by rad
We've had sunny weather during the last two days and I suddenly feel 5 years younger, more energized and not as mentally strained 😌
November 26, 2024 at 11:52 AM
We've had sunny weather during the last two days and I suddenly feel 5 years younger, more energized and not as mentally strained 😌
Reposted by rad
My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.
fleuret.org/dlc/
And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)
fleuret.org/lbdl/
fleuret.org/dlc/
And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)
fleuret.org/lbdl/
November 26, 2024 at 6:15 AM
My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.
fleuret.org/dlc/
And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)
fleuret.org/lbdl/
fleuret.org/dlc/
And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)
fleuret.org/lbdl/
Reposted by rad
How will AI affect the way we treat each other?
In "hybrid systems" of humans and machines, how will AI (whether simple or complex, like LLMs) affect not just human-machine interactions, but human-human interactions in the presence of machines?
Might AI change human social and ethical behavior? 1/
In "hybrid systems" of humans and machines, how will AI (whether simple or complex, like LLMs) affect not just human-machine interactions, but human-human interactions in the presence of machines?
Might AI change human social and ethical behavior? 1/
November 25, 2024 at 2:38 PM
How will AI affect the way we treat each other?
In "hybrid systems" of humans and machines, how will AI (whether simple or complex, like LLMs) affect not just human-machine interactions, but human-human interactions in the presence of machines?
Might AI change human social and ethical behavior? 1/
In "hybrid systems" of humans and machines, how will AI (whether simple or complex, like LLMs) affect not just human-machine interactions, but human-human interactions in the presence of machines?
Might AI change human social and ethical behavior? 1/
Reposted by rad
November 24, 2024 at 6:43 PM
😂
Hey, @rynelson.bsky.social, literally my first thought was about you 😁
November 25, 2024 at 6:28 PM
😂