Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

You can literally RL a 4B VLM to ace GeoGuesser 🌍 > Open-source code, RL environment, dataset, training setup, evals, and everything you need to reproduce it end-to-end > dropping soon!!

Adithya S K

14,044 subscribers

36,667 Aufrufe • vor 1 Tag •via X (Twitter)

Wissenschaft & Technologie Bildung

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

To solve AGI, we must first solve Geoguessr For that I built vlm-gym, a simple RL gym written in scratch, in JAX for Qwen3VL-4B (released yesterday) And added Geospot, a RL environment for geolocation and learned VLMs can learn how to geoguess. More:

To solve AGI, we must first solve Geoguessr For that I built vlm-gym, a simple RL gym written in scratch, in JAX for Qwen3VL-4B (released yesterday) And added Geospot, a RL environment for geolocation and learned VLMs can learn how to geoguess. More:

Surya Dantuluri

140,580 Aufrufe • vor 10 Monaten

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more

Introducing INTELLECT-3: Scaling RL to a 100B+ MoE model on our end-to-end stack Achieving state-of-the-art performance for its size across math, code and reasoning Built using the same tools we put in your hands, from environments & evals, RL frameworks, sandboxes & more

Prime Intellect

1,149,729 Aufrufe • vor 9 Monaten

Excited to present FastTD3: a simple, fast, and capable off-policy RL algorithm for humanoid control -- with an open-source code to run your own humanoid RL experiments in no time! Thread below 🧵

Excited to present FastTD3: a simple, fast, and capable off-policy RL algorithm for humanoid control -- with an open-source code to run your own humanoid RL experiments in no time! Thread below 🧵

Younggyo Seo

131,382 Aufrufe • vor 1 Jahr

I like to think of evals as something active, not passive -- it's a North Star that steers LLMs toward higher intelligence. Evals should drive your RL/SFT/post-training decisions. Internal evals at frontier labs make a huge difference -- and you can see it in how models behave differently (GPT seems better at one-shot tasks, Claude at multi-turn). If you want to learn more about building evals that actually improve your model in post-training, check out our AMD x DeeplearningAI course "Fine-tuning & RL for LLMs: Intro to Post-training" (content is free):

I like to think of evals as something active, not passive -- it's a North Star that steers LLMs toward higher intelligence. Evals should drive your RL/SFT/post-training decisions. Internal evals at frontier labs make a huge difference -- and you can see it in how models behave differently (GPT seems better at one-shot tasks, Claude at multi-turn). If you want to learn more about building evals that actually improve your model in post-training, check out our AMD x DeeplearningAI course "Fine-tuning & RL for LLMs: Intro to Post-training" (content is free):

Sharon Zhou

16,903 Aufrufe • vor 6 Monaten

"One of the very confusing things about the models right now: how to reconcile the fact that they are doing so well on evals. And you look at the evals and you go, 'Those are pretty hard evals.' But the economic impact seems to be dramatically behind. There is [a possible] explanation. Back when people were doing pre-training, the question of what data to train on was answered, because that answer was everything. So you don't have to think if it's going to be this data or that data. When people do RL training, they say, 'Okay, we want to have this kind of RL training for this thing and that kind of RL training for that thing.' You say, 'Hey, I would love our model to do really well when we release it. I want the evals to look great. What would be RL training that could help on this task?' If you combine this with generalization of the models actually being inadequate, that has the potential to explain a lot of what we are seeing, this disconnect between eval performance and actual real-world performance"

"One of the very confusing things about the models right now: how to reconcile the fact that they are doing so well on evals. And you look at the evals and you go, 'Those are pretty hard evals.' But the economic impact seems to be dramatically behind. There is [a possible] explanation. Back when people were doing pre-training, the question of what data to train on was answered, because that answer was everything. So you don't have to think if it's going to be this data or that data. When people do RL training, they say, 'Okay, we want to have this kind of RL training for this thing and that kind of RL training for that thing.' You say, 'Hey, I would love our model to do really well when we release it. I want the evals to look great. What would be RL training that could help on this task?' If you combine this with generalization of the models actually being inadequate, that has the potential to explain a lot of what we are seeing, this disconnect between eval performance and actual real-world performance"

Dwarkesh Patel

502,484 Aufrufe • vor 9 Monaten

This is a full end-to-end demo on how you can create your own product copilot using OpenCopilot: It's open source -> Chatting with products is truly an amazing experience instead of clicking around. Give it a try and let me know what you think!

This is a full end-to-end demo on how you can create your own product copilot using OpenCopilot: It's open source -> Chatting with products is truly an amazing experience instead of clicking around. Give it a try and let me know what you think!

Mo Gharabat 🦇

22,331 Aufrufe • vor 2 Jahren

If you want a robot to do something well, you need to know how to talk to it. If you don't, you can learn, with Semantic Action RL! In our paper, Jagdeep Bhatia, Andrew Wagenmaker, Will Chen show how RL over VLA prompts enables new tasks and learns blazing fast in the real world!

If you want a robot to do something well, you need to know how to talk to it. If you don't, you can learn, with Semantic Action RL! In our paper, Jagdeep Bhatia, Andrew Wagenmaker, Will Chen show how RL over VLA prompts enables new tasks and learns blazing fast in the real world!

Sergey Levine

31,286 Aufrufe • vor 2 Monaten

I spent my summer building TinyTPU : An open source ML inference and training chip. it can do end to end inference + training ENTIRELY on chip. here's how I did it👇:

I spent my summer building TinyTPU : An open source ML inference and training chip. it can do end to end inference + training ENTIRELY on chip. here's how I did it👇:

surya

356,539 Aufrufe • vor 1 Jahr

introducing AlphaLab -an autonomous system that runs quantitative research end-to-end give it a dataset + goal, it explores, reads papers, builds evals, and runs parallel experiments across a GPU cluster, learning&evolving from the results all details, code, and paper below

introducing AlphaLab -an autonomous system that runs quantitative research end-to-end give it a dataset + goal, it explores, reads papers, builds evals, and runs parallel experiments across a GPU cluster, learning&evolving from the results all details, code, and paper below

Brendan Hogan

178,070 Aufrufe • vor 5 Monaten

I hand-wrote a 500-LoC RL stack to make hacking on RL research much easier. Most RL stacks are either massive and unhackable, or duct-taped research scripts. I am open-sourcing Mithrl, a modular RLVR stack. Next items on my checklist: adding more complex environment examples, supporting multi-gpu + async RL, and QoL fixes. I might scrap external runtime dependencies (Huggingface PEFT + vLLM) and write purpose-built, simpler versions from scratch if I feel the need. If you want to experiment with RL and are looking to own sovereign tools, I’d love to get on call, understand your requirements and help integrate for free.

I hand-wrote a 500-LoC RL stack to make hacking on RL research much easier. Most RL stacks are either massive and unhackable, or duct-taped research scripts. I am open-sourcing Mithrl, a modular RLVR stack. Next items on my checklist: adding more complex environment examples, supporting multi-gpu + async RL, and QoL fixes. I might scrap external runtime dependencies (Huggingface PEFT + vLLM) and write purpose-built, simpler versions from scratch if I feel the need. If you want to experiment with RL and are looking to own sovereign tools, I’d love to get on call, understand your requirements and help integrate for free.

omkaar

17,218 Aufrufe • vor 5 Monaten

There’s a way to learn trading without getting liquidated and qualify for a $24,000 funded account In this video, we'll go over: > What is AceTrader > How to Trade > Getting funded > Ace Trader 100 Everything you need to know, step by step 👇

There’s a way to learn trading without getting liquidated and qualify for a $24,000 funded account In this video, we'll go over: > What is AceTrader > How to Trade > Getting funded > Ace Trader 100 Everything you need to know, step by step 👇

Gorilla

12,669 Aufrufe • vor 7 Monaten

The open source X-Humanoid full stack robot body, motion control, VLM/VLA models, and the RoboMIND dataset. Runs ANY AI MODLES. It fully supports ROS2, MQTT, and TCP/IP, developers can customize. Clear “no call home” circuits. Garage builder prices dropping to below $10,000.

The open source X-Humanoid full stack robot body, motion control, VLM/VLA models, and the RoboMIND dataset. Runs ANY AI MODLES. It fully supports ROS2, MQTT, and TCP/IP, developers can customize. Clear “no call home” circuits. Garage builder prices dropping to below $10,000.

Brian Roemmele

28,728 Aufrufe • vor 4 Monaten

You should have the power to change anything and everything about your computer, your framework, and your server. End-to-end open source is your ticket to tell Apple — or anyone! — to go fuck themselves when they seek to dictate how you compute.

You should have the power to change anything and everything about your computer, your framework, and your server. End-to-end open source is your ticket to tell Apple — or anyone! — to go fuck themselves when they seek to dictate how you compute.

DHH

141,124 Aufrufe • vor 1 Jahr

This is the Claude Code moment for Reinforcement Learning. Every frontier lab knows the secret: post-training is where the magic happens. If you can eval a task, you can use RL to benchmax your model. But building and scaling the RL loop has remained a dark art reserved for the top 1% of AI Labs. Until today. Introducing Optimus — an autonomous RL Agent Harness, and the first of many we are about to ship Iacon Autonomics. We are removing the manual grind of RL and adding taste and relentlessness to the process. Just hand Optimus a codebase, an API key, and a goal. It spins up a fleet of sub-agents to run hundreds of parallel experiments. It learns from its own history, autonomously building an evolutionary tree of runs where every failure sets the next batch up to win. Train LLMs, agents, robots in sim, game npcs and much more. If you can score it, Optimus can train it. Stop doing manual RL. Let the agents cook.

This is the Claude Code moment for Reinforcement Learning. Every frontier lab knows the secret: post-training is where the magic happens. If you can eval a task, you can use RL to benchmax your model. But building and scaling the RL loop has remained a dark art reserved for the top 1% of AI Labs. Until today. Introducing Optimus — an autonomous RL Agent Harness, and the first of many we are about to ship Iacon Autonomics. We are removing the manual grind of RL and adding taste and relentlessness to the process. Just hand Optimus a codebase, an API key, and a goal. It spins up a fleet of sub-agents to run hundreds of parallel experiments. It learns from its own history, autonomously building an evolutionary tree of runs where every failure sets the next batch up to win. Train LLMs, agents, robots in sim, game npcs and much more. If you can score it, Optimus can train it. Stop doing manual RL. Let the agents cook.

Dev Shah

13,309 Aufrufe • vor 16 Tagen

ELON: THE FUNDAMENTAL RL TEST IS AGAINST REALITY "You must know what is physically real for things to physically work. The really fundamental RL test is really gonna be, you are all against reality. You can't fool physics. But you can fool our ability to tell what it did with reality." Source: Cheeky Pint Podcast, Elon Musk

ELON: THE FUNDAMENTAL RL TEST IS AGAINST REALITY "You must know what is physically real for things to physically work. The really fundamental RL test is really gonna be, you are all against reality. You can't fool physics. But you can fool our ability to tell what it did with reality." Source: Cheeky Pint Podcast, Elon Musk

Mario Nawfal

47,876 Aufrufe • vor 7 Monaten

This is the ONLY Rama Tut clip you need to see... Wong -> Mystique -> Rama Tut -> Arnim Zola Believe me... the end is worth it... MARVEL SNAP #MarvelSnap

This is the ONLY Rama Tut clip you need to see... Wong -> Mystique -> Rama Tut -> Arnim Zola Believe me... the end is worth it... MARVEL SNAP #MarvelSnap

Guest

29,661 Aufrufe • vor 5 Monaten

😈 Today, we introduce WebGym, the largest-to-date open-source RL environment for web agent training that contains 300k tasks and a rollout framework optimized specifically for web environments' rollout speed. We reveal the effects of essential scaling directions we observe with WebGym. 1/n

😈 Today, we introduce WebGym, the largest-to-date open-source RL environment for web agent training that contains 300k tasks and a rollout framework optimized specifically for web environments' rollout speed. We reveal the effects of essential scaling directions we observe with WebGym. 1/n

Jack Bai

45,147 Aufrufe • vor 7 Monaten

Turn any open-source LLM into reasoning powerhouse! Using reinforcement finetuning you can add reasoning abilities to any LLM, even without a labelled dataset. Step-by-step explanation with code:

Turn any open-source LLM into reasoning powerhouse! Using reinforcement finetuning you can add reasoning abilities to any LLM, even without a labelled dataset. Step-by-step explanation with code:

Akshay 🚀

50,423 Aufrufe • vor 1 Jahr

> bored at home > open higgsfield > you can start making an AI film > no setup, no studio > turn an idea into something real > submit it to their original series > win $ 150k in an AI contest > book a flight and disappear > live your best life for a while

> bored at home > open higgsfield > you can start making an AI film > no setup, no studio > turn an idea into something real > submit it to their original series > win $ 150k in an AI contest > book a flight and disappear > live your best life for a while

CG

80,454 Aufrufe • vor 5 Monaten