Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

I've gotten a mujoco sim RL training loop for a unitree robot at 200k SPS. I'm looking into the physics for friction, contact dynamics. My goal: can I reproduce & beat the mujoco playground RL baselines This is running in my web browser with raylib. Its the baseline

kache

402,818 subscribers

30,204 görüntüleme • 3 ay önce •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

System identification (sysid) is the process of finding the physical parameters that make a simulation match reality. If you're training an RL locomotion policy in simulation, the accuracy of your motor model directly affects how well the policy transfers to the real robot. A recent git commit by Kevin Zakka added a sysid toolbox to MuJoCo which automates this process: you provide recorded motor data and a MuJoCo model, and it optimizes the model parameters to minimize the difference between simulated and real trajectories. For my RobStride Dynamics RS02 QDD motors (17 Nm peak, 7.75:1 gear), I built a Rust tool that sends multi-sine torque excitation at 1 kHz and records position/velocity feedback. I then feed this data into MuJoCo's sysid optimizer.

System identification (sysid) is the process of finding the physical parameters that make a simulation match reality. If you're training an RL locomotion policy in simulation, the accuracy of your motor model directly affects how well the policy transfers to the real robot. A recent git commit by Kevin Zakka added a sysid toolbox to MuJoCo which automates this process: you provide recorded motor data and a MuJoCo model, and it optimizes the model parameters to minimize the difference between simulated and real trajectories. For my RobStride Dynamics RS02 QDD motors (17 Nm peak, 7.75:1 gear), I built a Rust tool that sends multi-sine torque excitation at 1 kHz and records position/velocity feedback. I then feed this data into MuJoCo's sysid optimizer.

David Bar

48,347 görüntüleme • 5 ay önce

GPT-6 Astra test 4/n Prompt: "Implement pen spinning with a dexterous hand. Use Isaac Lab for RL training, use the Sharpa hand, and create the pen mesh yourself. Give me a trained RL policy and a visualization video. You are free to search the web and download papers or anything else you need." After running autonomously for a day and a half (including training the policy), this is what we got: Credit: my student Chengyang Li

GPT-6 Astra test 4/n Prompt: "Implement pen spinning with a dexterous hand. Use Isaac Lab for RL training, use the Sharpa hand, and create the pen mesh yourself. Give me a trained RL policy and a visualization video. You are free to search the web and download papers or anything else you need." After running autonomously for a day and a half (including training the policy), this is what we got: Credit: my student Chengyang Li

Wentao Zhu

688,955 görüntüleme • 6 gün önce

This is running in my web browser on my laptop. It's deterministic and running using webgpu Tonight; I have a /goal going where it is going to set up one of my 4090s as the authority server, and then test all the rollback code. Goal is 50k boxes simmed per user, up to 512 users. Massively multiplayer massively rigid body physics online games inside your web browser. Why the hell not, we've got the tokens to spare!

This is running in my web browser on my laptop. It's deterministic and running using webgpu Tonight; I have a /goal going where it is going to set up one of my 4090s as the authority server, and then test all the rollback code. Goal is 50k boxes simmed per user, up to 512 users. Massively multiplayer massively rigid body physics online games inside your web browser. Why the hell not, we've got the tokens to spare!

kache

39,217 görüntüleme • 2 ay önce

I just released a first video deep dive into my LeHome Challenge solution (🥇 sim round, 🥈 real-robot final at ICRA 2026). I already shared what I did and the video series is about why: the reasoning, the alternatives I rejected, and the intuition behind each choice. I split the whole video explanation into 3 parts: Part 1 is about RL for VLAs 👇

I just released a first video deep dive into my LeHome Challenge solution (🥇 sim round, 🥈 real-robot final at ICRA 2026). I already shared what I did and the video series is about why: the reasoning, the alternatives I rejected, and the intuition behind each choice. I split the whole video explanation into 3 parts: Part 1 is about RL for VLAs 👇

Ilia

32,610 görüntüleme • 2 ay önce

hi all, excited to join! i'm building an expressive mini "shoggoth" robot which will eventually be hooked up to gpt4o realtime voice. i'm currently working on the low-level policies, which are trained in a mujoco simulation with RL. to delay working on raw-pixels for now, i trained a pose-estimation model using deeplabcut and triangulate the position in 3d space using the stereo cameras. eventually, i'll use gpt4o's tool calling capabilities to activate several of these policies (closed and open loop) based on the dialog flow! captions: manual actuation of the tentacle / 3d pose estimation / target design

hi all, excited to join! i'm building an expressive mini "shoggoth" robot which will eventually be hooked up to gpt4o realtime voice. i'm currently working on the low-level policies, which are trained in a mujoco simulation with RL. to delay working on raw-pixels for now, i trained a pose-estimation model using deeplabcut and triangulate the position in 3d space using the stereo cameras. eventually, i'll use gpt4o's tool calling capabilities to activate several of these policies (closed and open loop) based on the dialog flow! captions: manual actuation of the tentacle / 3d pose estimation / target design

Matthieu LC

45,587 görüntüleme • 1 yıl önce

This is how Strike Robot turns Simulation into Reality! One of the biggest challenges in robotics is ensuring that behaviors validated in simulation work reliably in the real world. For this experiment, we reconstructed part of a real laboratory at Eastworlds inside SR Platform. The generated layout was then deployed into MuJoCo. Using SR Agentic, the robot was tasked with finding abnormal objects in a cluttered environment and sending a Telegram notification when detected. Before deployment, everything is validated in simulation.

This is how Strike Robot turns Simulation into Reality! One of the biggest challenges in robotics is ensuring that behaviors validated in simulation work reliably in the real world. For this experiment, we reconstructed part of a real laboratory at Eastworlds inside SR Platform. The generated layout was then deployed into MuJoCo. Using SR Agentic, the robot was tasked with finding abnormal objects in a cluttered environment and sending a Telegram notification when detected. Before deployment, everything is validated in simulation.

Strike Robot

15,350 görüntüleme • 3 ay önce

I spent some time last week vibe-coding "The Beginner's RL Playground" a simple web tool for exploring many basic concepts in reinforcement learning. Users can adjust all sorts of parameters of the algorithm and environment and see the effects of those changes in realtime.

I spent some time last week vibe-coding "The Beginner's RL Playground" a simple web tool for exploring many basic concepts in reinforcement learning. Users can adjust all sorts of parameters of the algorithm and environment and see the effects of those changes in realtime.

Arthur Juliani

45,743 görüntüleme • 1 yıl önce

What I love about Replit ⠕. Low floor to get started. Ultra high ceiling for what you can build. I told Agent to build a native C application that runs on my desktop without a web browser - AND compile it into JavaScript to run in the web browser. Here's a video showing both running

What I love about Replit ⠕. Low floor to get started. Ultra high ceiling for what you can build. I told Agent to build a native C application that runs on my desktop without a web browser - AND compile it into JavaScript to run in the web browser. Here's a video showing both running

jordwalke

43,773 görüntüleme • 7 ay önce

Bench is a constant struggle bus. Managed a single at 315 with the Slingshot this morning. My goal for 2025 was to hit this raw. Unlikely at this point. I've got pain in both shoulders now. Doing what I can...

Bench is a constant struggle bus. Managed a single at 315 with the Slingshot this morning. My goal for 2025 was to hit this raw. Unlikely at this point. I've got pain in both shoulders now. Doing what I can...

FLEXjs

81,729 görüntüleme • 9 ay önce

Chinese consumer electronics maker Honor will unveil its first humanoid robot on March 1 at MWC 2026 in Barcelona. The robot is designed for retail & home service. The company sold a whopping 71 million smartphones in 2025. Looks like a customized Unitree G1 with a 2DoF neck.

Chinese consumer electronics maker Honor will unveil its first humanoid robot on March 1 at MWC 2026 in Barcelona. The robot is designed for retail & home service. The company sold a whopping 71 million smartphones in 2025. Looks like a customized Unitree G1 with a 2DoF neck.

The Humanoid Hub

58,290 görüntüleme • 7 ay önce

A simple MPC controller for path tracking implemented in python. Few resources actually teach it without assuming a PhD first. This open-source project does exactly that: An iterative MPC tracking controller (built with CVXPY and MuJoCo) that shows how to make a robot (or autonomous vehicle) follow a path and avoid obstacles in real time. → Autonomous logistics, robot navigation, vehicle control; there's a path-following algorithm behind all of it. → Most implementations are black boxes. This one shows you what's actually happening. → It uses convex optimization (no non-linear solver), and still handles real vehicle physics through iterative linearization. What makes it practical is that you can run it with the mushr car model in MuJoCo, or headless without any physics simulation if you just want to understand the logic. If you're making the jump from classical control to real-time optimisation, this is a very good starting point. 📌 Repository: ——- Weekly robotics and AI insights. Subscribe free:

A simple MPC controller for path tracking implemented in python. Few resources actually teach it without assuming a PhD first. This open-source project does exactly that: An iterative MPC tracking controller (built with CVXPY and MuJoCo) that shows how to make a robot (or autonomous vehicle) follow a path and avoid obstacles in real time. → Autonomous logistics, robot navigation, vehicle control; there's a path-following algorithm behind all of it. → Most implementations are black boxes. This one shows you what's actually happening. → It uses convex optimization (no non-linear solver), and still handles real vehicle physics through iterative linearization. What makes it practical is that you can run it with the mushr car model in MuJoCo, or headless without any physics simulation if you just want to understand the logic. If you're making the jump from classical control to real-time optimisation, this is a very good starting point. 📌 Repository: ——- Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

16,064 görüntüleme • 11 gün önce

Introducing RL Environment Creator Skill Now any one can create RL environments $ npx skills add adithya-s-k/RL_Envs_101 > You can create environments across multiple frameworks like OpenEnv, OpenReward, Verifiers, NemoGym ... > the repo has live working examples of environments that your coding agent can reference > The skill is design to first understand what type of model you are training and create an environment while keeping that in mind ps. There’s a lot more to building RL environments that can be used for training. One major aspect is the data, which this skill can’t directly solve. However, the skill will help with implementing tools, rewards, and other components of an RL environment, making it easier to go from idea to implementation quickly across different frameworks. Let me know if you’d be interested in a detailed, end-to-end blog/tutorial on building an environment and actually training a model for a useful use case.

Introducing RL Environment Creator Skill Now any one can create RL environments $ npx skills add adithya-s-k/RL_Envs_101 > You can create environments across multiple frameworks like OpenEnv, OpenReward, Verifiers, NemoGym ... > the repo has live working examples of environments that your coding agent can reference > The skill is design to first understand what type of model you are training and create an environment while keeping that in mind ps. There’s a lot more to building RL environments that can be used for training. One major aspect is the data, which this skill can’t directly solve. However, the skill will help with implementing tools, rewards, and other components of an RL environment, making it easier to go from idea to implementation quickly across different frameworks. Let me know if you’d be interested in a detailed, end-to-end blog/tutorial on building an environment and actually training a model for a useful use case.

Adithya S K

47,002 görüntüleme • 4 ay önce

GPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size. My assessment of each: - Low: weird slop - Medium: kinda cooked - High: sort of tried but ultimately incoherent - Extra High: elite - really nice idea and well executed Obviously this is just one shot, but worth trying different reasoning levels for the new models, medium seems to be pretty good for GPT-5.5 and it was really bad for many previous GPT models.

GPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size. My assessment of each: - Low: weird slop - Medium: kinda cooked - High: sort of tried but ultimately incoherent - Extra High: elite - really nice idea and well executed Obviously this is just one shot, but worth trying different reasoning levels for the new models, medium seems to be pretty good for GPT-5.5 and it was really bad for many previous GPT models.

Peter Gostev (SF: 22-26 June)

209,258 görüntüleme • 4 ay önce

So this was my problem: Keeping the physics running while on the inventory menu (game is paused). Paused game means paused physics. No jiggling. UNLESS... I find out that you can create a separate PhysicsScene on the fly for the avatar. #gamedev #indiedev #madewithunity #RPG

Sérgio Rodrigues 🇵🇹 Working on Troubled Passage

37,948 görüntüleme • 2 yıl önce

Training clips from a couple sessions this week. Looking into a polygamy agreement so I can add the Roger’s squat into my marriage. Just have to get the state & my wife on board. Them damn Mormons may be onto something This is my dose of get the fuck after it for the timeline today. One fact remains: I have to get better 🦍

Training clips from a couple sessions this week. Looking into a polygamy agreement so I can add the Roger’s squat into my marriage. Just have to get the state & my wife on board. Them damn Mormons may be onto something This is my dose of get the fuck after it for the timeline today. One fact remains: I have to get better 🦍

🦍

58,531 görüntüleme • 10 ay önce

NVIDIA DROPPED A MOTION DIFFUSION MODEL FOR HUMANOID ROBOTS trained on 700 hours of mocap data kimodo generates high-quality 3D human and robot motions from text prompts you control it with: → full-body pose keyframes → end-effector positions/rotations → 2D paths and waypoints works on human skeletons and unitree G1 robot plug the outputs directly into mujoco or retarget to other robots using GMR has a web-based interactive demo with a timeline editor. runs locally needs ~17GB VRAM to run inference open source under apache 2.0

NVIDIA DROPPED A MOTION DIFFUSION MODEL FOR HUMANOID ROBOTS trained on 700 hours of mocap data kimodo generates high-quality 3D human and robot motions from text prompts you control it with: → full-body pose keyframes → end-effector positions/rotations → 2D paths and waypoints works on human skeletons and unitree G1 robot plug the outputs directly into mujoco or retarget to other robots using GMR has a web-based interactive demo with a timeline editor. runs locally needs ~17GB VRAM to run inference open source under apache 2.0

Vaishnavi

17,572 görüntüleme • 5 ay önce

Tij Iginla says he’s ready to make the team this year: “I think I’m ready. I’m looking forward to a big summer of training & pushing it hard in the gym. I feel ready.” “My goal is definitely to make the team this year.”

Tij Iginla says he’s ready to make the team this year: “I think I’m ready. I’m looking forward to a big summer of training & pushing it hard in the gym. I feel ready.” “My goal is definitely to make the team this year.”

Cole Bagley

16,725 görüntüleme • 2 ay önce

we just released a new blog "Training a coding agent using the OpenCode harness in remote HF sandboxes with TRL and OpenEnv" you can take a real coding agent (OpenCode), let it run its own tool loop against real coding problems, and train it with RL on the exact tokens it produced and every rollout runs in its own remote HF sandbox, so rollouts scale out beyond one machine the loop: - OpenCode owns its tool loop inside an OpenEnv sandbox - an in-sandbox proxy records the real token ids + logprobs, per turn - a hidden-test verifier scores the result, and that is the reward - TRL trains with AsyncGRPO, weights sync back to vLLM over NCCL blog + runnable example:

we just released a new blog "Training a coding agent using the OpenCode harness in remote HF sandboxes with TRL and OpenEnv" you can take a real coding agent (OpenCode), let it run its own tool loop against real coding problems, and train it with RL on the exact tokens it produced and every rollout runs in its own remote HF sandbox, so rollouts scale out beyond one machine the loop: - OpenCode owns its tool loop inside an OpenEnv sandbox - an in-sandbox proxy records the real token ids + logprobs, per turn - a hidden-test verifier scores the result, and that is the reward - TRL trains with AsyncGRPO, weights sync back to vLLM over NCCL blog + runnable example:

Sergio Paniego

37,936 görüntüleme • 1 ay önce

Bringing new meaning to the word NeRD at #CoRL2025. 🤓 Neural Robot Dynamics (NeRD) bridges the sim-to-real gap with a hybrid of neural modeling and analytical simulation, unlocking zero-shot transfer and continuous learning loops for optimal robotics training. Read more & try it in Newton 🤖

Bringing new meaning to the word NeRD at #CoRL2025. 🤓 Neural Robot Dynamics (NeRD) bridges the sim-to-real gap with a hybrid of neural modeling and analytical simulation, unlocking zero-shot transfer and continuous learning loops for optimal robotics training. Read more & try it in Newton 🤖

NVIDIA Robotics

17,556 görüntüleme • 11 ay önce