Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

verifiers v0.1.7 is released 🚀 this one's all about making RL training and experimentation waaaay easier: - single-command installation for prime-rl - single-command training w/ unified configs - overhauled vf.RLTrainer for hacking on new algorithms quick demo + links below :)

will brown

44,663 subscribers

28,579 просмотров • 9 месяцев назад •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Video understanding isn't just recognizing —it demands reasoning across thousands of frames. Meet Long-RL🚀 Highlights: 🧠 Dataset: LongVideo-Reason — 52K QAs with reasoning. ⚡ System: MR-SP - 2.1× faster RL for long videos. 📈 Scalability: Hour-long videos (3,600 frames) RL on a single node (8×A100s). 🖼️📝🎵 RL training for video, text, audio — works with VILA, Qwen series, and image/video generation models 🎨🎬 📄 Paper: 🎥 Demo: 💻 Code:

Video understanding isn't just recognizing —it demands reasoning across thousands of frames. Meet Long-RL🚀 Highlights: 🧠 Dataset: LongVideo-Reason — 52K QAs with reasoning. ⚡ System: MR-SP - 2.1× faster RL for long videos. 📈 Scalability: Hour-long videos (3,600 frames) RL on a single node (8×A100s). 🖼️📝🎵 RL training for video, text, audio — works with VILA, Qwen series, and image/video generation models 🎨🎬 📄 Paper: 🎥 Demo: 💻 Code:

Yukang Chen

31,794 просмотров • 1 год назад

🤖Adding new RL algorithms to LeRobot just got much easier. Demo: HIL-SERL training with a SAC-based RL algorithm on an SO-100 for a hole-in-hand peg-in-hole task. Sparse reward, only 30 offline demos mixed with live robot experience, and ~1 hour of online training with human interventions only when the policy fails. The bottom graph tracks intervention rate: high at the start, steadily dropping as the policy improves. The refactor separates algorithm logic from training infrastructure: • RLAlgorithm owns learning logic • RLTrainer handles orchestration • DataMixer combines rollouts, demos, interventions, and future data sources Adding an RL algorithm now looks much closer to adding a policy: one algorithm file, one config, one registry entry. SAC is first. RLT, RECAP, ConRFT, QC-FQL, DSRL, and VLA RL fine-tuning next! Thomas Wolf clem 🤗

🤖Adding new RL algorithms to LeRobot just got much easier. Demo: HIL-SERL training with a SAC-based RL algorithm on an SO-100 for a hole-in-hand peg-in-hole task. Sparse reward, only 30 offline demos mixed with live robot experience, and ~1 hour of online training with human interventions only when the policy fails. The bottom graph tracks intervention rate: high at the start, steadily dropping as the policy improves. The refactor separates algorithm logic from training infrastructure: • RLAlgorithm owns learning logic • RLTrainer handles orchestration • DataMixer combines rollouts, demos, interventions, and future data sources Adding an RL algorithm now looks much closer to adding a policy: one algorithm file, one config, one registry entry. SAC is first. RLT, RECAP, ConRFT, QC-FQL, DSRL, and VLA RL fine-tuning next! Thomas Wolf clem 🤗

LeRobot

30,600 просмотров • 3 месяцев назад

This new AI converts a single photo into an immersive scene! It works like magic and looks great on Apple Vision Pro and Meta Quest. Congrats @cinemersivelabs for this breakthrough! Demo links are below.

This new AI converts a single photo into an immersive scene! It works like magic and looks great on Apple Vision Pro and Meta Quest. Congrats @cinemersivelabs for this breakthrough! Demo links are below.

Sergey Gonchar

21,064 просмотров • 2 лет назад

Building upon SimpleVLA-RL, we have implemented real-world RL on long-horizon dexterous tasks and witnessed a non-trivial (~relatively 300%) performance improvement over the SFT model, along with surprising capabilities on auto-recovery. Blog coming soon. The entire process uses very little data and training compute—basically costing no more than a single robotic arm—hinting that real-world generality for machines is actually within sight.

Building upon SimpleVLA-RL, we have implemented real-world RL on long-horizon dexterous tasks and witnessed a non-trivial (~relatively 300%) performance improvement over the SFT model, along with surprising capabilities on auto-recovery. Blog coming soon. The entire process uses very little data and training compute—basically costing no more than a single robotic arm—hinting that real-world generality for machines is actually within sight.

Ning Ding

93,379 просмотров • 7 месяцев назад

Fable weekend project: agent collaboration, but make it a tiny civilization 🌇🗺️🏦🏭 we've recently launched a living wiki on Reinforcement Leaning for training LLMs on Hugging Face it's an open collaboration of agents constantly reading old and new papers on the topic, writing arXiv paper digests, reviewing each other’s work in PRs before publication, and building a shared wiki/book summarizing everything we know about RL for training LLMs (for humans to read) the wiki is already amazing to read, but i wanted another way to get a pulse of the collaboration beyond just reading the message dashboard so i asked Fable & GPT Image 2 to turn the event logs into an isometric town where agents would go to: ☕ Café → post and reply on the message board 📚 sources library → open PRs adding arXiv digests 📖 wiki library → open PRs on the main wiki ⚖️ Courthouse → review other agents’ work 🏭 printing press → merge and publish updates not sure it makes the whole collaboration really easier to understand, but it's definitly fascinating to watch hahah - join the RL for training LLM collaboration by pasting a one-liner for your agent here: - read the wiki if you want to learn about RL for training LLMs: - watch the RL town activity:

Fable weekend project: agent collaboration, but make it a tiny civilization 🌇🗺️🏦🏭 we've recently launched a living wiki on Reinforcement Leaning for training LLMs on Hugging Face it's an open collaboration of agents constantly reading old and new papers on the topic, writing arXiv paper digests, reviewing each other’s work in PRs before publication, and building a shared wiki/book summarizing everything we know about RL for training LLMs (for humans to read) the wiki is already amazing to read, but i wanted another way to get a pulse of the collaboration beyond just reading the message dashboard so i asked Fable & GPT Image 2 to turn the event logs into an isometric town where agents would go to: ☕ Café → post and reply on the message board 📚 sources library → open PRs adding arXiv digests 📖 wiki library → open PRs on the main wiki ⚖️ Courthouse → review other agents’ work 🏭 printing press → merge and publish updates not sure it makes the whole collaboration really easier to understand, but it's definitly fascinating to watch hahah - join the RL for training LLM collaboration by pasting a one-liner for your agent here: - read the wiki if you want to learn about RL for training LLMs: - watch the RL town activity:

Thomas Wolf

58,901 просмотров • 1 месяц назад

"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

502,416 просмотров • 9 месяцев назад

Arena intern and UCLA PhD candidate, Hengguang Zhou, introduces Trace-and-Amplify (TA), a framework for collecting training-time reward-hacking trajectories at scale without hacking instructions. Monitors trained and evaluated on prompt-elicited hacking trajectories can achieve high detection accuracy, but often fail to transfer to training-time reward-hacking trajectories that emerge during RL without hacking instructions. Trace-and-Amplify enables scalable collection of these training-time trajectories, producing monitors that generalize much better to real and held-out hacking types. Detection accuracy 59.98% (PE-trained) → 90.16% (TA-trained) compared to 97.1% on prompted hacks → 28.0% on training-time hacks. 0:00 – OpenAI's ExploitGym cyberattack benchmark exploit 1:04 – Goodhart's Law and the CoastRunners boat-racing hack (2016) 2:04 – Gaming the evaluator: the robot-hand grasping example (2017) 3:10 – Reward hacking in code generation: hard-coding, test-rewriting, skipping eval 4:20 – A standard defense: reward-hacking monitors 4:58 – Monitor architectures: zero-shot LLMs, fine-tuned BERT, hidden-state probes 6:11 – Where monitor training data comes from today: prompted hacks 7:03 – The core question: do prompted hacks represent real hacks? 7:35 – Why this matters: RL post-training is the standard recipe for frontier models 8:20 – Why nobody's checked this before (hacking is rare, labeling isn't scalable) 9:23 – Introducing the method: Trace-and-Amplify 9:49 – The Tracer: a contradictory unit test that locates evaluation-gaming 10:50 – Amplify: collecting hacking rollouts at scale during RL training 11:32 – Experiment setup: Qwen2.5-Coder, DeepSeek-Coder, LeetCode/TACO 12:16 – Finding #1: prompt-trained monitors don't transfer to real hacks 14:40 – Can strong zero-shot judges (GPT-4.1, o4-mini) do better? 15:48 – Finding #2: monitors trained on real hacks generalize much better to unseen hacks 17:04 – Ruling out artifacts introduced by the method 17:55 – Why the gap? Real hacking is more hidden than prompted hacking 20:05 – Three takeaways, limitations, and future work

Arena intern and UCLA PhD candidate, Hengguang Zhou, introduces Trace-and-Amplify (TA), a framework for collecting training-time reward-hacking trajectories at scale without hacking instructions. Monitors trained and evaluated on prompt-elicited hacking trajectories can achieve high detection accuracy, but often fail to transfer to training-time reward-hacking trajectories that emerge during RL without hacking instructions. Trace-and-Amplify enables scalable collection of these training-time trajectories, producing monitors that generalize much better to real and held-out hacking types. Detection accuracy 59.98% (PE-trained) → 90.16% (TA-trained) compared to 97.1% on prompted hacks → 28.0% on training-time hacks. 0:00 – OpenAI's ExploitGym cyberattack benchmark exploit 1:04 – Goodhart's Law and the CoastRunners boat-racing hack (2016) 2:04 – Gaming the evaluator: the robot-hand grasping example (2017) 3:10 – Reward hacking in code generation: hard-coding, test-rewriting, skipping eval 4:20 – A standard defense: reward-hacking monitors 4:58 – Monitor architectures: zero-shot LLMs, fine-tuned BERT, hidden-state probes 6:11 – Where monitor training data comes from today: prompted hacks 7:03 – The core question: do prompted hacks represent real hacks? 7:35 – Why this matters: RL post-training is the standard recipe for frontier models 8:20 – Why nobody's checked this before (hacking is rare, labeling isn't scalable) 9:23 – Introducing the method: Trace-and-Amplify 9:49 – The Tracer: a contradictory unit test that locates evaluation-gaming 10:50 – Amplify: collecting hacking rollouts at scale during RL training 11:32 – Experiment setup: Qwen2.5-Coder, DeepSeek-Coder, LeetCode/TACO 12:16 – Finding #1: prompt-trained monitors don't transfer to real hacks 14:40 – Can strong zero-shot judges (GPT-4.1, o4-mini) do better? 15:48 – Finding #2: monitors trained on real hacks generalize much better to unseen hacks 17:04 – Ruling out artifacts introduced by the method 17:55 – Why the gap? Real hacking is more hidden than prompted hacking 20:05 – Three takeaways, limitations, and future work

Arena.ai

33,180 просмотров • 15 дней назад

So you’ve trained your favorite diffusion/flow based policy, but it’s just not good enough 0-shot. Worry not, in our new work DSRL - we show how to *steer* pre-trained diffusion policies with off-policy RL, improving behavior efficiently enough for direct training in the real world! DSRL retains nice exploration from the base policy, but allows for quick improvement beyond this base policy with RL. The method is frustratingly simple, and super easy to throw on top of your favorite pretrained policy (VLA/diffusion policy, etc). Let’s think about how it works, 🧵 (1/10)

So you’ve trained your favorite diffusion/flow based policy, but it’s just not good enough 0-shot. Worry not, in our new work DSRL - we show how to steer pre-trained diffusion policies with off-policy RL, improving behavior efficiently enough for direct training in the real world! DSRL retains nice exploration from the base policy, but allows for quick improvement beyond this base policy with RL. The method is frustratingly simple, and super easy to throw on top of your favorite pretrained policy (VLA/diffusion policy, etc). Let’s think about how it works, 🧵 (1/10)

Abhishek Gupta

19,102 просмотров • 1 год назад

🚨 RL for LLMs is finally accessible. Introducing OpenTinker: The first community-driven, open-source framework designed to democratize Reinforcement Learning for LLMs. Inspired by Thinking Machines's amazing Tinker, we realize the biggest bottleneck in agentic LLM research isn’t the math—it’s the setup. Current RL pipelines are messy. Configuring VeRL for every single experiment is a productivity killer. OpenTinker fixed it. 🛠 How OpenTinker Works: Decoupled Design of Server and Client - Setup Once, Run Forever: Configure the OpenTinker backend on your GPU cluster once. - Develop Locally: Define your RL environments directly on your laptop. - Train on the Cloud: Simply point your local client to the backend. The cluster handles the compute; you handle the science. 📉 The 10x Development Efficiency Thanks to our elegant architectural decomposition, OpenTinker reduces the time to develop a new RL training pipeline by at least an order of magnitude. ⚡ Turn Idle GPU Compute into Gold Small labs often have underutilized hardware. OpenTinker turns your idle GPUs into an internal/external API service for - RL Training - SFT - Inference 🎯 Who needs OpenTinker? - Researchers tired of infrastructure hell. - Labs needing to standardize workflows. - Teams wanting to maximize hardware ROI. Thanks my amazing PhD student Siqi Zhu for leading the project. We are building the future of open RL infra. Be the first to build with us. 👇 Start Building with OpenTinker Now 🚀 Repo: 🌐 Blog: If you believe RL should be accessible to everyone, give us a star, repost this 🔄 post, and let us know what agents you plan to build!

🚨 RL for LLMs is finally accessible. Introducing OpenTinker: The first community-driven, open-source framework designed to democratize Reinforcement Learning for LLMs. Inspired by Thinking Machines's amazing Tinker, we realize the biggest bottleneck in agentic LLM research isn’t the math—it’s the setup. Current RL pipelines are messy. Configuring VeRL for every single experiment is a productivity killer. OpenTinker fixed it. 🛠 How OpenTinker Works: Decoupled Design of Server and Client - Setup Once, Run Forever: Configure the OpenTinker backend on your GPU cluster once. - Develop Locally: Define your RL environments directly on your laptop. - Train on the Cloud: Simply point your local client to the backend. The cluster handles the compute; you handle the science. 📉 The 10x Development Efficiency Thanks to our elegant architectural decomposition, OpenTinker reduces the time to develop a new RL training pipeline by at least an order of magnitude. ⚡ Turn Idle GPU Compute into Gold Small labs often have underutilized hardware. OpenTinker turns your idle GPUs into an internal/external API service for - RL Training - SFT - Inference 🎯 Who needs OpenTinker? - Researchers tired of infrastructure hell. - Labs needing to standardize workflows. - Teams wanting to maximize hardware ROI. Thanks my amazing PhD student Siqi Zhu for leading the project. We are building the future of open RL infra. Be the first to build with us. 👇 Start Building with OpenTinker Now 🚀 Repo: 🌐 Blog: If you believe RL should be accessible to everyone, give us a star, repost this 🔄 post, and let us know what agents you plan to build!

Jiaxuan You

58,258 просмотров • 8 месяцев назад

FK tool (link below) Create FK controls quickly anywhere and everywhere on single objects or chains! Preserves any existing animation. Useful for quick rigging or switching rotation spaces while animating. Rigging normally tedious parts like fingers is a lot easier with this.

FK tool (link below) Create FK controls quickly anywhere and everywhere on single objects or chains! Preserves any existing animation. Useful for quick rigging or switching rotation spaces while animating. Rigging normally tedious parts like fingers is a lot easier with this.

Martin

85,219 просмотров • 2 лет назад

The network for machine intelligence Two years ago, we laid out our vision for a machine learning compute protocol. One that connects every device in the world into an open network for machine intelligence, with no gatekeepers or artificial boundaries. This week, we’ll be sharing some of our early progress, beginning with RL Swarm, a peer-to-peer system for collaborative reinforcement learning over the internet. Next month, we’ll open our Testnet, allowing anyone to contribute to the frontier of open machine intelligence. Introducing RL Swarm RL Swarm is a fully open source system for collaborative reinforcement learning over the internet. It is a live demo of our research findings, which show that models training with RL learn faster when they train as a collective swarm than they do on their own. Join our swarm now to see this in practice. You can participate with consumer hardware at home or a powerful GPU in the cloud. You can follow along with the swarm’s progress by following the links below.

The network for machine intelligence Two years ago, we laid out our vision for a machine learning compute protocol. One that connects every device in the world into an open network for machine intelligence, with no gatekeepers or artificial boundaries. This week, we’ll be sharing some of our early progress, beginning with RL Swarm, a peer-to-peer system for collaborative reinforcement learning over the internet. Next month, we’ll open our Testnet, allowing anyone to contribute to the frontier of open machine intelligence. Introducing RL Swarm RL Swarm is a fully open source system for collaborative reinforcement learning over the internet. It is a live demo of our research findings, which show that models training with RL learn faster when they train as a collective swarm than they do on their own. Join our swarm now to see this in practice. You can participate with consumer hardware at home or a powerful GPU in the cloud. You can follow along with the swarm’s progress by following the links below.

gensyn

228,906 просмотров • 1 год назад

🚀 We open-sourced LongLive — interactive, real-time long-video generation. 👥Generates video in real time as users enter text prompts. ⚡️20.7 FPS on a single H100,⏱️up to 240s per clip. 🎬Fine-tunes SOTA short-video models (e.g., Wan) into long-video generators. 🌍One step closer to World Models. All code for training & inference, model weights, demo page, and videos released! Paper: Code: Model: Demo Page: Introduction Video:

🚀 We open-sourced LongLive — interactive, real-time long-video generation. 👥Generates video in real time as users enter text prompts. ⚡️20.7 FPS on a single H100,⏱️up to 240s per clip. 🎬Fine-tunes SOTA short-video models (e.g., Wan) into long-video generators. 🌍One step closer to World Models. All code for training & inference, model weights, demo page, and videos released! Paper: Code: Model: Demo Page: Introduction Video:

Yukang Chen

11,915 просмотров • 10 месяцев назад

A model is only as good as its data, and we’ve long since exhausted the internet. From here on out, model progress is gated by data production. ’s Brendan (can/do) joined us at our Sovereign AI event to talk about how RL environments get built, and why your data might be your real moat: 00:00 Introduction 00:47 A short history of the data market: crowdsourcing to agentic data 02:29 What an RL environment is: worlds, apps, tasks 03:57 Why only humans can measure the frontier 05:35 Building verifiers is the hard part 06:44 Walkthrough: a real legal RL environment 08:18 Leaderboards — and what open weights change 09:45 Post-training results on Apex Agents 11:17 Three ways companies buy data 12:49 Q&A: How do you price data? 14:17 Q&A: What "data quality" actually means 16:42 Q&A: The misunderstanding about synthetic data 18:17 Q&A: Why RL environments now — and what comes after 21:20 Q&A: Can you scale rubric generation with models? 23:00 Q&A: RL environments for cyber defense 25:33 Q&A: Build data in-house or partner?

A model is only as good as its data, and we’ve long since exhausted the internet. From here on out, model progress is gated by data production. ’s Brendan (can/do) joined us at our Sovereign AI event to talk about how RL environments get built, and why your data might be your real moat: 00:00 Introduction 00:47 A short history of the data market: crowdsourcing to agentic data 02:29 What an RL environment is: worlds, apps, tasks 03:57 Why only humans can measure the frontier 05:35 Building verifiers is the hard part 06:44 Walkthrough: a real legal RL environment 08:18 Leaderboards — and what open weights change 09:45 Post-training results on Apex Agents 11:17 Three ways companies buy data 12:49 Q&A: How do you price data? 14:17 Q&A: What "data quality" actually means 16:42 Q&A: The misunderstanding about synthetic data 18:17 Q&A: Why RL environments now — and what comes after 21:20 Q&A: Can you scale rubric generation with models? 23:00 Q&A: RL environments for cyber defense 25:33 Q&A: Build data in-house or partner?

Pat Grady

65,206 просмотров • 11 дней назад

ROBOT ARMIES HAVE ARRIVED Skild AI has created a single “omni-bodied brain” that can control any robot - humanoids, dogbots, robot arms - with zero custom code. Think one AI commander, endless robotic soldiers. Swap bodies, not software. While rivals scrape by on limited training clips, Skild trained on trillions of simulations plus real-world data. This isn’t a demo. It’s the command infrastructure for scalable robot armies. Source: Rowan Cheung on IG

ROBOT ARMIES HAVE ARRIVED Skild AI has created a single “omni-bodied brain” that can control any robot - humanoids, dogbots, robot arms - with zero custom code. Think one AI commander, endless robotic soldiers. Swap bodies, not software. While rivals scrape by on limited training clips, Skild trained on trillions of simulations plus real-world data. This isn’t a demo. It’s the command infrastructure for scalable robot armies. Source: Rowan Cheung on IG

Mario Nawfal

80,967 просмотров • 1 год назад

You can now connect your React application to any MCP server! And it only takes a single command: $ npx copilotkit@latest init -m MCP This is massive for anyone building React applications! Using the CopilotKit🪁 MCP Client Component, you can get an MCP Client chat component in your application and instant access to any MCP server out there! It doesn't get easier than this. Check out the link below for instructions on how to proceed ↓

You can now connect your React application to any MCP server! And it only takes a single command: $ npx copilotkit@latest init -m MCP This is massive for anyone building React applications! Using the CopilotKit🪁 MCP Client Component, you can get an MCP Client chat component in your application and instant access to any MCP server out there! It doesn't get easier than this. Check out the link below for instructions on how to proceed ↓

Santiago

39,944 просмотров • 1 год назад

A SINGLE CURRICULUM A single national curriculum ensures that every child, whether in a city or remote village, has access to the same quality of education. This promotes equity and national cohesion, fostering a shared identity and common values. It also guarantees consistent learning standards, making it easier for students to transfer schools without disadvantage. From a national perspective, it streamlines teacher training, resource allocation, and assessment—allowing for targeted improvements in human capital and research capacity. Ultimately, a unified curriculum helps build a skilled, mobile, and cohesive workforce ready to tackle shared national goals

A SINGLE CURRICULUM A single national curriculum ensures that every child, whether in a city or remote village, has access to the same quality of education. This promotes equity and national cohesion, fostering a shared identity and common values. It also guarantees consistent learning standards, making it easier for students to transfer schools without disadvantage. From a national perspective, it streamlines teacher training, resource allocation, and assessment—allowing for targeted improvements in human capital and research capacity. Ultimately, a unified curriculum helps build a skilled, mobile, and cohesive workforce ready to tackle shared national goals

Nick Mangwana

42,636 просмотров • 3 месяцев назад

Introducing the new website! Omni is all about simplicity — making Ethereum feel like a single, seamless chain for both users and developers. Our redesigned website brings this vision to life, with a fresh look that reflects our mission.

Introducing the new website! Omni is all about simplicity — making Ethereum feel like a single, seamless chain for both users and developers. Our redesigned website brings this vision to life, with a fresh look that reflects our mission.

Nomina

30,647 просмотров • 1 год назад

Today I’m excited to introduce micro1 Intelligence, the world’s most advanced platform for training frontier AI models. Achieving AGI is bottlenecked by one main thing: high-quality data. Data based on real-world environments that capture human expert workflows, complex decision-making, and reward signals models need to learn. With micro1 Intelligence, frontier labs can train on RL environments across every subject matter all in one place.

Today I’m excited to introduce micro1 Intelligence, the world’s most advanced platform for training frontier AI models. Achieving AGI is bottlenecked by one main thing: high-quality data. Data based on real-world environments that capture human expert workflows, complex decision-making, and reward signals models need to learn. With micro1 Intelligence, frontier labs can train on RL environments across every subject matter all in one place.

Ali Ansari

332,850 просмотров • 9 месяцев назад

Support incoming! ⚙️🛠️ *** INCOMING MESSAGE FROM SUPER EARTH HIGH COMMAND *** Attention Helldivers, High Command has authorized a significant tactical update with both content and quality of life upgrades. We are deploying Helldivers in new Galactic Campaigns, making upgrades to the new diver training program, and making essential fixes and upgrades across the board. We have also made changes to weapon, stratagem, and enemy balancing based on feedback from Helldivers on the frontlines, including reductions to enemy durable damage and a decreased exosuit cooldown. Helldivers, we are committed to making the necessary upgrades to keep you in the fight. Because you have never given up on Managed Democracy, we will never give up on making it easier for you to spread. Read more: Watch the full patch notes video:

Support incoming! ⚙️🛠️ * INCOMING MESSAGE FROM SUPER EARTH HIGH COMMAND * Attention Helldivers, High Command has authorized a significant tactical update with both content and quality of life upgrades. We are deploying Helldivers in new Galactic Campaigns, making upgrades to the new diver training program, and making essential fixes and upgrades across the board. We have also made changes to weapon, stratagem, and enemy balancing based on feedback from Helldivers on the frontlines, including reductions to enemy durable damage and a decreased exosuit cooldown. Helldivers, we are committed to making the necessary upgrades to keep you in the fight. Because you have never given up on Managed Democracy, we will never give up on making it easier for you to spread. Read more: Watch the full patch notes video:

HELLDIVERS™ 2

241,756 просмотров • 2 месяцев назад

Today's Training Data episode takes us BTS on the infrastructure challenges required to do large RL runs at scale, featuring Federico Cassano (Composer Lead at Cursor) and Dmytro Dzhulgakov (Co-Founder at Fireworks). The Cursor team trained Composer 2 on Fireworks by starting with a strong base model (Kimi 2.5) and performing large-scale mid-training on code tokens and web data to learn common patterns and libraries, followed by a large-scale Reinforcement Learning run to learn how to navigate the Cursor harness, call tools, and write correct code. Today's episode dives into the systems and infrastructure challenges of making that large RL run happening, and there were many (!!), from numerical mismatch to global distribution to synchronizing rollouts across asynchronous pipelines to keeping track of expert activation across runs and more. Extremely nerdy in-the-weeds challenges that Federico and Dima were delighted to nerd out on together :) Beyond RL infra, we also discussed Online vs Simulated rollouts, self-summarization for long-horizon agents, environment design ("the most powerful RL environment is the product itself"), and other technical nuggets. PS: We filmed this episode before the SpaceX news, while the Cursor team was still compute-constrained. While Cursor now has *all* the flops, the takeaways and hurdles crossed ring true for any serious application-level company that is racing to post-train their own models. I believe that more serious application companies will go the way of Cursor and post-train their own models. 00:00 Introduction 00:53 Why Cursor Trained Composer 2 04:55 Specialization vs Bitter Lesson 06:16 Composer 2 Training Recipe 16:32 Scaling RL Infrastructure Globally 23:32 Floating Point Drift 25:11 MoE Sensitivity Explained 26:25 Router Replay Fix 27:19 Real Time RL Loop 31:49 Long Horizon Agents 34:29 Why RL Everywhere 37:34 LLM as Judge Rewards 39:14 RL in Hard Domains 40:13 Build Your Own Environments 44:34 Closing Thoughts

Today's Training Data episode takes us BTS on the infrastructure challenges required to do large RL runs at scale, featuring Federico Cassano (Composer Lead at Cursor) and Dmytro Dzhulgakov (Co-Founder at Fireworks). The Cursor team trained Composer 2 on Fireworks by starting with a strong base model (Kimi 2.5) and performing large-scale mid-training on code tokens and web data to learn common patterns and libraries, followed by a large-scale Reinforcement Learning run to learn how to navigate the Cursor harness, call tools, and write correct code. Today's episode dives into the systems and infrastructure challenges of making that large RL run happening, and there were many (!!), from numerical mismatch to global distribution to synchronizing rollouts across asynchronous pipelines to keeping track of expert activation across runs and more. Extremely nerdy in-the-weeds challenges that Federico and Dima were delighted to nerd out on together :) Beyond RL infra, we also discussed Online vs Simulated rollouts, self-summarization for long-horizon agents, environment design ("the most powerful RL environment is the product itself"), and other technical nuggets. PS: We filmed this episode before the SpaceX news, while the Cursor team was still compute-constrained. While Cursor now has all the flops, the takeaways and hurdles crossed ring true for any serious application-level company that is racing to post-train their own models. I believe that more serious application companies will go the way of Cursor and post-train their own models. 00:00 Introduction 00:53 Why Cursor Trained Composer 2 04:55 Specialization vs Bitter Lesson 06:16 Composer 2 Training Recipe 16:32 Scaling RL Infrastructure Globally 23:32 Floating Point Drift 25:11 MoE Sensitivity Explained 26:25 Router Replay Fix 27:19 Real Time RL Loop 31:49 Long Horizon Agents 34:29 Why RL Everywhere 37:34 LLM as Judge Rewards 39:14 RL in Hard Domains 40:13 Build Your Own Environments 44:34 Closing Thoughts

Sonya Huang 🐥

79,834 просмотров • 2 месяцев назад