Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

btw we're streaming our RL training runs. Inspired by Fuli Luo.

kickingkeys's profile picture

Surya

6,643 subscribers

60,081 görüntüleme • 3 gün önce •via X (Twitter)

Bilim & Teknoloji
LIVE

Anya Rossi• Live Now

Private livecam show

18 Yorum

Surya profil fotoğrafı
Surya3 gün önce

Training done with @Primeintellect Prime RL on a @sfcompute node.

pradheep profil fotoğrafı
pradheep3 gün önce

@_LuoFuli love this 🤌🏻

Sriraam profil fotoğrafı
Sriraam3 gün önce

@_LuoFuli Let’s goo

Surya profil fotoğrafı
Surya3 gün önce

@_LuoFuli 🫶🫶

Avi profil fotoğrafı
Avi3 gün önce

@_LuoFuli how can I get half as cool as you

Surya profil fotoğrafı
Surya3 gün önce

@_LuoFuli I'm learning from @hollowaya_x

Vincent Weisser profil fotoğrafı
Vincent Weisser3 gün önce

@_LuoFuli awesome work!!

Surya profil fotoğrafı
Surya3 gün önce

@_LuoFuli Thank you Vincent!

Karan Ganesan profil fotoğrafı
Karan Ganesan3 gün önce

@_LuoFuli slick!

Oscar Moxon profil fotoğrafı
Oscar Moxon3 gün önce

@_LuoFuli by far the best thing to hit the internet this week

Surya profil fotoğrafı
Surya3 gün önce

@_LuoFuli Ty king 👑

Karthik Uppu profil fotoğrafı
Karthik Uppu3 gün önce

@_LuoFuli this is incredible!

Tensor Fiend profil fotoğrafı
Tensor Fiend3 gün önce

@_LuoFuli This is awesome🔥

Alek profil fotoğrafı
Alek3 gün önce

@_LuoFuli Surya - what changed between steps 141 and 180?

Surya profil fotoğrafı
Surya3 gün önce

@_LuoFuli It 3x code length but we penalise for failed code so it adjusted p quickly

Dheeraj Vislawath profil fotoğrafı
Dheeraj Vislawath3 gün önce

@_LuoFuli this is really cool

AI Psychosis profil fotoğrafı
AI Psychosis3 gün önce

@_LuoFuli No one cares stop chasing a trend and do your own thing retard

Cupsey profil fotoğrafı
Cupsey3 gün önce

@_LuoFuli watching the policy improve in public is gonna be oddly addictive

Benzer Videolar

We asked Sholto Douglas from Anthropic about the costs of RL (Reinforcement Learning) runs. "In Dario Amodei's essay, he said that RL runs cost only $1M back in December." "RL is a more naively parallelizable and scalable than pre-training." "With pre-training, you need everything in one big data center ideally. For RL, in theory, you could scale all over the world."
0:37

We asked Sholto Douglas from Anthropic about the costs of RL (Reinforcement Learning) runs. "In Dario Amodei's essay, he said that RL runs cost only $1M back in December." "RL is a more naively parallelizable and scalable than pre-training." "With pre-training, you need everything in one big data center ideally. For RL, in theory, you could scale all over the world."

TBPN

76,696 görüntüleme • 1 yıl önce

Introducing Prime Sandboxes: MicroVM sandboxes purpose-built for RL training. Model training requires running tens of thousands of concurrent sandboxes, leading to complex and costly configuration. We built Prime Sandboxes for our own team. Today we're releasing them publicly.
0:26

Introducing Prime Sandboxes: MicroVM sandboxes purpose-built for RL training. Model training requires running tens of thousands of concurrent sandboxes, leading to complex and costly configuration. We built Prime Sandboxes for our own team. Today we're releasing them publicly.

Prime Intellect

109,687 görüntüleme • 6 gün önce

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.
2:15

We developed an RL method for fine-tuning our models for precise tasks in just a few hours or even minutes. Instead of training the whole model, we add an “RL token” output to π-0.6, our latest model, which is used by a tiny actor and critic to learn quickly with RL.

Physical Intelligence

444,783 görüntüleme • 6 ay önce

From Wrexham to Patagonia, our heritage runs deep. Introducing the 25/26 Third Kit: inspired by the Welsh diaspora and united by tradition. 🔴⚪ #WxmAFC
1:27

From Wrexham to Patagonia, our heritage runs deep. Introducing the 25/26 Third Kit: inspired by the Welsh diaspora and united by tradition. 🔴⚪ #WxmAFC

Wrexham AFC

140,617 görüntüleme • 1 yıl önce

Inspired by OW Zach Wallace goal for @hcbloemendaal last week, our U12 hockey players had some fun after training last night. #WhitgiftHockey
0:45

Inspired by OW Zach Wallace goal for @hcbloemendaal last week, our U12 hockey players had some fun after training last night. #WhitgiftHockey

Whitgift Sport

13,529 görüntüleme • 3 yıl önce

Seungyoon wrote Air inspired from a diffuser gifted by a fan btw 💙
2:18

Seungyoon wrote Air inspired from a diffuser gifted by a fan btw 💙

◟♯ . / pea . ! ˚₊‧꒰ა ✦ ໒꒱ ‧₊˚

23,361 görüntüleme • 6 ay önce

Boxing runs deep. 🥊 Inspired by the greatest. Ali, Tyson, Holmes, Durán. Legends who shaped the sport and our passion for it. #BoxingLegends #RiyadhSeason #GOAT
0:36

Boxing runs deep. 🥊 Inspired by the greatest. Ali, Tyson, Holmes, Durán. Legends who shaped the sport and our passion for it. #BoxingLegends #RiyadhSeason #GOAT

TURKI ALALSHIKH

97,564 görüntüleme • 1 yıl önce

During Training Camp, we were honored to celebrate one of our favorite days of the summer with patients and families from UH Rainbow Babies. We're inspired by their resilience and we always look forward to seeing their smiling faces to brighten up our practice! University Hospitals | #DawgPound
0:45

During Training Camp, we were honored to celebrate one of our favorite days of the summer with patients and families from UH Rainbow Babies. We're inspired by their resilience and we always look forward to seeing their smiling faces to brighten up our practice! University Hospitals | #DawgPound

Cleveland Browns

32,438 görüntüleme • 27 gün önce

Inspired by real events. Defined by real courage. #TheLostBus — Streaming October 3 on Apple TV+
1:57

Inspired by real events. Defined by real courage. #TheLostBus — Streaming October 3 on Apple TV+

Apple TV

1,335,278 görüntüleme • 1 yıl önce

Today we're announcing cua-bench: a framework for benchmarking, training data, and RL environments for computer-use AI agents. Why? Current agents show 10x variance across minor UI changes. Here's how we're fixing it.
2:02

Today we're announcing cua-bench: a framework for benchmarking, training data, and RL environments for computer-use AI agents. Why? Current agents show 10x variance across minor UI changes. Here's how we're fixing it.

Cua

189,823 görüntüleme • 9 ay önce

Meet Kofi Abatey, a football prodigy whose love for the game runs deep. Inspired by his idol, Lamine Yamal, he spends countless nights training under the lights as he relentlessly pursues his dream.
2:04

Meet Kofi Abatey, a football prodigy whose love for the game runs deep. Inspired by his idol, Lamine Yamal, he spends countless nights training under the lights as he relentlessly pursues his dream.

𝐀𝐒𝐊

82,824 görüntüleme • 3 ay önce

We didn't have $1 million for a Game Awards Trailer so we're posting our indie game inspired by JRPG classics here instead for free!
1:40

We didn't have $1 million for a Game Awards Trailer so we're posting our indie game inspired by JRPG classics here instead for free!

❖ Runa ❖ Wishlist Now on Steam!

180,638 görüntüleme • 9 ay önce

Inspired by our past, our home and our identity. 💙
1:11

Inspired by our past, our home and our identity. 💙

Everton

585,478 görüntüleme • 2 yıl önce

RL in the real world presents some big challenges, but also some really big opportunities. In our new work, HIL-SERL, Charles Xu, Jeffrey Wu, Jianlan Luo show that real-world RL can learn a huge range of precise and robust tasks, and perform them much faster than imitation.
0:37

RL in the real world presents some big challenges, but also some really big opportunities. In our new work, HIL-SERL, Charles Xu, Jeffrey Wu, Jianlan Luo show that real-world RL can learn a huge range of precise and robust tasks, and perform them much faster than imitation.

Sergey Levine

36,544 görüntüleme • 1 yıl önce

our princess diana… icespice was inspired by our queen
0:59

our princess diana… icespice was inspired by our queen

malaytwt

163,073 görüntüleme • 3 yıl önce

inspired by meet, greet & bye. suob, out now on all streaming platforms ♥️
0:23

inspired by meet, greet & bye. suob, out now on all streaming platforms ♥️

juan karlos labajo

17,203 görüntüleme • 10 ay önce

you can create complex agentic environments and launch RL training runs with a single prompt. deploy trained inference endpoints with a single click. no GPUs, no SSH, no vLLM. just `prime`. guide:
9:59

you can create complex agentic environments and launch RL training runs with a single prompt. deploy trained inference endpoints with a single click. no GPUs, no SSH, no vLLM. just `prime`. guide:

will brown

139,747 görüntüleme • 6 ay önce

We're 10 devs who left AAA to make our dream RPG inspired by OSR tabletop games. Your support means the world to us. THANK YOU. ❤️
0:27

We're 10 devs who left AAA to make our dream RPG inspired by OSR tabletop games. Your support means the world to us. THANK YOU. ❤️

Wyldheart

33,370 görüntüleme • 5 ay önce

Inspired by our players. Worn by our fans. Designed by Nike. Be the first to represent:
0:38

Inspired by our players. Worn by our fans. Designed by Nike. Be the first to represent:

U.S. Soccer

383,399 görüntüleme • 1 yıl önce

btw this is AI and we're soooooo fucked vid by showtoolsai on insta
1:28

btw this is AI and we're soooooo fucked vid by showtoolsai on insta

WARIO HIMSELF

81,114 görüntüleme • 10 ay önce