正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

World models are moving beyond offline generation towards interactive, real-time experiences. Introducing ⚡FlashDreams⚡: an open-source high-performance inference and serving library built for autoregressive world models: 🔥 Up to 3.10× faster LingBot-World inference 🔥 Up to 2.12× faster Self-Forcing inference 🔥 Up to 1.40× faster Wan2.1 inference 🔥 8 integrated... models 🔥 Multi-GPU, streaming, low-latency serving 🔥 Agentic skills that teach you how to use it FlashDreams is designed for a new generation of AI systems that continuously evolve over time while responding to user interactions. It powers applications across robotics, autonomous vehicle simulation, gaming, and virtual worlds. Github: Docs: Research page: Join the #flashdreams Discord channel at FlashDreams is also the runtime backbone behind NVIDIA OmniDreams ( 1/n #AI #WorldModels #FastInference #PhysicalAI #OpenSource #NVIDIAshow more

Ruilong Li

5,311 subscribers

99,435 次观看 • 3 个月前 •via X (Twitter)

教育科学技术

Anya Rossi• Live Now

Private livecam show

24 条评论

Ruilong Li 的头像

Ruilong Li3 个月前

World models are quickly advancing, but for most users the experience is still passive: watching impressive promo videos, not running the models, probing them, or building applications on top of them. FlashDreams aims to change that. Spin up a server, select a model, and interact with it live. 2/n

Ruilong Li 的头像

Ruilong Li3 个月前

A world model learns to generate and evolve an environment over time: video, actions, state, audio, sensors, control signals, and more. Serving it means keeping a live session running while input, model state, GPU inference, and output evolve together. 3/n

Ruilong Li 的头像

Ruilong Li3 个月前

Serving a world model is different from offline video generation. The goal is not just prompt → wait → video. A live world-model runtime must stay responsive while the world keeps advancing: low latency, high throughput, steady streaming, and rolling state. 4/n

Ruilong Li 的头像

Ruilong Li3 个月前

Efficient serving infra like vLLM and SGLang helped move LLMs from research checkpoints into real apps. World models are reaching a similar moment, but with a different interactive pattern than LLM: initialize once, then keep evolving the world forward at a constant pace. 5/n

Ruilong Li 的头像

Ruilong Li3 个月前

Similar to vLLM and SGLang, FlashDreams is built from the ground up for high-performance serving. FlashDreams delivers up to 2.12× speedup on Self-Forcing, 3.10× on LingBot-World, and 1.40× on Wan2.1. The system is designed specifically for autoregressive world models, where generation unfolds continuously over time. At the same time, the architecture naturally extends to bidirectional models by treating each generation task as a single autoregressive rollout. 6/n

Ruilong Li 的头像

Ruilong Li3 个月前

FlashDreams is more than an inference engine. It is a serving stack purpose-built for interactive world-model applications, supporting persistent sessions, streaming inputs and outputs, rolling model state, and multi-GPU execution. FlashDreams already integrates eight world models, with many more on the way. 7/n

Ruilong Li 的头像

Ruilong Li3 个月前

Ending this thread with a fun walk on w/ FlashDreams: Try it out and tag us with your exploration. #flashdreams

Ruilong Li 的头像

Ruilong Li3 个月前

The team behind this is extremely talented. I can't be more happy to work with them! Listing in random order: @jatentaki @wilson_over @jmartinezesturo @JunchenLiu77 @csyonghe, Gangzheng Tong, Jesse Archer, @mccaffrey_j_b, William Lew, @heuristic42, Aidan Foster, Wenqing Wang, @AdityaMaha47495, @FidlerSanja @ZGojcic, @natdoesgraphics, Matt Cragun, @wang_jialiang, Anis Ahmad, @killerooo, Deep Shekhar, Michael Chock, @ATMGuille9

Xun Huang 的头像

Xun Huang3 个月前

Congrats! Fantastic work

Ruilong Li 的头像

Ruilong Li3 个月前

Congrats to you too 😄

Andrew Carr 🤸 的头像

Andrew Carr 🤸3 个月前

oh this is super nice, well done!

Boyang Deng 的头像

Boyang Deng3 个月前

This is so cool and gonna be so useful! Congrats ... and thanks for the contribution! 🙂

Jiawei Yang 的头像

Jiawei Yang3 个月前

Legendary as always!

Jathushan Rajasegaran 的头像

Jathushan Rajasegaran3 个月前

Congratulations! Looks amazing!

Yossi Gandelsman 的头像

Yossi Gandelsman3 个月前

congrats!

Ruilong Li 的头像

Ruilong Li3 个月前

to you as well! big day for both of us : )

Richard Palethorpe 的头像

Richard Palethorpe3 个月前

Amazing stuff, does RTX 6000 or Spark have enough compute to handle these models or is a data center GPU needed?

Ruilong Li 的头像

Ruilong Li3 个月前

This is running 30 fps live on a single RTX 6000 pro

Richard Palethorpe 的头像

Richard Palethorpe3 个月前

Wow

Prithvi Jadwani | AI SEO | GEO | REDDIT SEO | GMB 的头像

Prithvi Jadwani | AI SEO | GEO | REDDIT SEO | GMB3 个月前

That's an impressive benchmark, but how does FlashDreams handle model drift from continuous evolution?

Ruilong Li 的头像

Ruilong Li3 个月前

There are some kv cache tricks can alleviate that during inference time. You can check out this configuration if interested: self-forcing-wan2.1-t2v-1.3b-sink5-window7-rerope Doc:

Maya Srinivasan 的头像

Maya Srinivasan2 个月前

real-time world models

Neha Patel 的头像

Neha Patel2 个月前

real-time world models finally

Mathias Heide 的头像

Mathias Heide3 个月前

Oh, i'd love to introduce this into Summer Engine for automated playtesting of games - possible?

相关视频

NVIDIA just released a new open source transcription model, Nemotron Speech ASR, designed from the ground up for low-latency use cases like voice agents. Here's a voice agent built with this new model. 24ms transcription finalization and total voice-to-voice inference time under 500ms. This agent actually uses *three* NVIDIA open source models: - Nemotron Speech ASR - Nemotron 3 Nano 30GB in a 4-bit quant (released in December) - A preview checkpoint of the upcoming Magpie text-to-speech model These models are all truly open source: weights, training data, training code, and inference code. This is a big deal! Jensen said in the CES keynote yesterday that he expects open source models to catch up to proprietary models this year in a number of categories. NVIDIA is putting their weight behind making this happen. (As Alan Kay said, the best way to predict the future is to invent it.) The code for this agent is open source too, of course. You can deploy it to production with Modal and Pipecat AI cloud, or run locally on an NVIDIA DGX Spark or RTX 5090.

NVIDIA just released a new open source transcription model, Nemotron Speech ASR, designed from the ground up for low-latency use cases like voice agents. Here's a voice agent built with this new model. 24ms transcription finalization and total voice-to-voice inference time under 500ms. This agent actually uses three NVIDIA open source models: - Nemotron Speech ASR - Nemotron 3 Nano 30GB in a 4-bit quant (released in December) - A preview checkpoint of the upcoming Magpie text-to-speech model These models are all truly open source: weights, training data, training code, and inference code. This is a big deal! Jensen said in the CES keynote yesterday that he expects open source models to catch up to proprietary models this year in a number of categories. NVIDIA is putting their weight behind making this happen. (As Alan Kay said, the best way to predict the future is to invent it.) The code for this agent is open source too, of course. You can deploy it to production with Modal and Pipecat AI cloud, or run locally on an NVIDIA DGX Spark or RTX 5090.

kwindla

274,801 次观看 • 8 个月前

Cyber defense is entering the superintelligence era. ⚡ Introducing the CrowdStrike Cyber Superintelligence Lab, the first frontier AI research organization built for cyberdefense and AI safety. The Lab delivers CrowdStrike SafeMind, a family of purpose-built security models and harnesses. SafeMind runs as one system designed to deliver AI safety: an offensive model that finds the attack path, a defensive model that closes it, and the harnesses that operate both in the same loop. Built using NVIDIA Nemotron open models and using CoreWeave’s AI Cloud for training and inference, SafeMind goes beyond AI that simply identifies threats, it defeats them. Learn more:

Cyber defense is entering the superintelligence era. ⚡ Introducing the CrowdStrike Cyber Superintelligence Lab, the first frontier AI research organization built for cyberdefense and AI safety. The Lab delivers CrowdStrike SafeMind, a family of purpose-built security models and harnesses. SafeMind runs as one system designed to deliver AI safety: an offensive model that finds the attack path, a defensive model that closes it, and the harnesses that operate both in the same loop. Built using NVIDIA Nemotron open models and using CoreWeave’s AI Cloud for training and inference, SafeMind goes beyond AI that simply identifies threats, it defeats them. Learn more:

CrowdStrike

57,939 次观看 • 15 天前

🔥🔥🔥We’ve been listening to your feedback! Our latest world model HY-World 1.5 just got a major upgrade to make world generation more accessible than ever: 🛠️ Open Training Code: Fully customizable code for building and training your own models. ⚡ Accelerated Inference: Turbocharged speed and optimized VRAM for real-time interaction. 📉 Lite 5B Model: A new lightweight model that fits into small-VRAM GPUs. 🙌 Zero Waitlist: Our online app is now fully open to everyone—no application required. This is just the beginning. HY-World is building the future of spatial intelligence—open, accessible, and community-driven. 🕹️ Play now: ⭐ GitHub:

🔥🔥🔥We’ve been listening to your feedback! Our latest world model HY-World 1.5 just got a major upgrade to make world generation more accessible than ever: 🛠️ Open Training Code: Fully customizable code for building and training your own models. ⚡ Accelerated Inference: Turbocharged speed and optimized VRAM for real-time interaction. 📉 Lite 5B Model: A new lightweight model that fits into small-VRAM GPUs. 🙌 Zero Waitlist: Our online app is now fully open to everyone—no application required. This is just the beginning. HY-World is building the future of spatial intelligence—open, accessible, and community-driven. 🕹️ Play now: ⭐ GitHub:

Tencent Hy

20,581 次观看 • 8 个月前

"Inference, if you look at it as a market, will be much, much bigger than cloud computing was pre-ChatGPT." Lightspeed’s Bucky Moore says inference is an underrated investment category in AI, and expects the market to break up into large, specialized platforms for each modality: "The GPU supply crunch that we're seeing right now is largely, as Dylan Patel has said on the show before, due to the fact that not only these consumer products, but also B2B products like Claude Code and Codex are just really taking off and creating insane demand for inference." "We're talking hundreds of billions in spend every year. And if that's true, I think there will be very, very large inference platforms built in each modality." "So there will be an inference platform for real-time video models, there will be an inference platform for open-source and custom language models, there will be an inference platform built specifically for long-running agents." "So I think we're just going to see that industry, which today looks like one industry, break up into many because of how big it is and how much room for specialization there is."

"Inference, if you look at it as a market, will be much, much bigger than cloud computing was pre-ChatGPT." Lightspeed’s Bucky Moore says inference is an underrated investment category in AI, and expects the market to break up into large, specialized platforms for each modality: "The GPU supply crunch that we're seeing right now is largely, as Dylan Patel has said on the show before, due to the fact that not only these consumer products, but also B2B products like Claude Code and Codex are just really taking off and creating insane demand for inference." "We're talking hundreds of billions in spend every year. And if that's true, I think there will be very, very large inference platforms built in each modality." "So there will be an inference platform for real-time video models, there will be an inference platform for open-source and custom language models, there will be an inference platform built specifically for long-running agents." "So I think we're just going to see that industry, which today looks like one industry, break up into many because of how big it is and how much room for specialization there is."

TBPN

33,278 次观看 • 6 个月前

Fireworks AI is now available first-party in Microsoft Foundry, as part of a multi-year partnership, starting today in private preview. Enterprise AI has a speed problem. Most infrastructure can't keep up with how fast open models are moving. New architectures, fine-tuned weights, models that aren't on anyone's approved list yet. The usual answer was "manage your own serving infrastructure." That's a real cost. Fireworks runs the inference. Azure owns the governance. Your weights stay yours. Beyond custom weights, teams can also access the latest open models on day zero. No waiting for the marketplace to catch up. Serverless inference for prototyping, and a path to production that doesn't require re-architecting anything. Fireworks handles the serving layer. Azure handles the rest. Your boss handles the congratulations. Want in →

Fireworks AI is now available first-party in Microsoft Foundry, as part of a multi-year partnership, starting today in private preview. Enterprise AI has a speed problem. Most infrastructure can't keep up with how fast open models are moving. New architectures, fine-tuned weights, models that aren't on anyone's approved list yet. The usual answer was "manage your own serving infrastructure." That's a real cost. Fireworks runs the inference. Azure owns the governance. Your weights stay yours. Beyond custom weights, teams can also access the latest open models on day zero. No waiting for the marketplace to catch up. Serverless inference for prototyping, and a path to production that doesn't require re-architecting anything. Fireworks handles the serving layer. Azure handles the rest. Your boss handles the congratulations. Want in →

Fireworks AI

51,657 次观看 • 6 个月前

New course: Efficient Inference with SGLang: Text and Image Generation, built in partnership with LMSys LMSYS Org and RadixArk RadixArk, and taught by Richard Chen Richard Chen, a Member of Technical Staff at RadixArk. Running LLMs in production is expensive, and much of that cost comes from redundant computation. This short course teaches you to eliminate that waste using SGLang, an open-source inference framework that caches computation already done and reuses it across future requests. When ten users share the same system prompt, SGLang processes it once, not ten times. The speedups compound quickly, especially when there's a lot of shared context across requests. Skills you'll gain: - Implement a KV cache from scratch to eliminate redundant computation within a single request - Scale caching across users and requests with RadixAttention, so shared context is only processed once - Accelerate image generation with diffusion models using SGLang's caching and multi-GPU parallelism Join and learn to make LLM inference faster and more cost-efficient at scale!

New course: Efficient Inference with SGLang: Text and Image Generation, built in partnership with LMSys LMSYS Org and RadixArk RadixArk, and taught by Richard Chen Richard Chen, a Member of Technical Staff at RadixArk. Running LLMs in production is expensive, and much of that cost comes from redundant computation. This short course teaches you to eliminate that waste using SGLang, an open-source inference framework that caches computation already done and reuses it across future requests. When ten users share the same system prompt, SGLang processes it once, not ten times. The speedups compound quickly, especially when there's a lot of shared context across requests. Skills you'll gain: - Implement a KV cache from scratch to eliminate redundant computation within a single request - Scale caching across users and requests with RadixAttention, so shared context is only processed once - Accelerate image generation with diffusion models using SGLang's caching and multi-GPU parallelism Join and learn to make LLM inference faster and more cost-efficient at scale!

Andrew Ng

102,747 次观看 • 5 个月前

SITUATION EXPLAINED: A new generation of AI labs are betting that world models are the path to general intelligence that can interact with the real world. We had General Intuition CEO Pim de Witte on MTS to explain what world models are, and why intuitively world models offer a clearer path towards general intelligence than LLMs. For humans, text is a technology for compressing information about the unfolding, three-dimensional world around us and communicating it to other people. When LLMs reason about the world, they do so by reconstructing the world through text. World models work by modeling the three-dimensional world at the architecture level and then allowing agents to interact with that world. Unlike an LLM, which understands space and time through text, world models perform spatial and temporal reasoning natively. General Intuition hopes to use world models to create embodied AI—machines that can reason and act in the real world. But other companies are also pushing the frontier in this space. Odyssey ML today announced a multi-agent world model, Agora-1. Their launch video shows four participants interacting with a real-time world model simulation of GoldenEye, a first-person shooter for the N64. A playable preview of the GoldenEye simulation is available on their site.

SITUATION EXPLAINED: A new generation of AI labs are betting that world models are the path to general intelligence that can interact with the real world. We had General Intuition CEO Pim de Witte on MTS to explain what world models are, and why intuitively world models offer a clearer path towards general intelligence than LLMs. For humans, text is a technology for compressing information about the unfolding, three-dimensional world around us and communicating it to other people. When LLMs reason about the world, they do so by reconstructing the world through text. World models work by modeling the three-dimensional world at the architecture level and then allowing agents to interact with that world. Unlike an LLM, which understands space and time through text, world models perform spatial and temporal reasoning natively. General Intuition hopes to use world models to create embodied AI—machines that can reason and act in the real world. But other companies are also pushing the frontier in this space. Odyssey ML today announced a multi-agent world model, Agora-1. Their launch video shows four participants interacting with a real-time world model simulation of GoldenEye, a first-person shooter for the N64. A playable preview of the GoldenEye simulation is available on their site.

MTS

11,676 次观看 • 4 个月前

Everything you love about generative models — now powered by real physics! Announcing the Genesis project — after a 24-month large-scale research collaboration involving over 20 research labs — a generative physics engine able to generate 4D dynamical worlds powered by a physics simulation platform designed for general-purpose robotics and physical AI applications. Genesis's physics engine is developed in pure Python, while being 10-80x faster than existing GPU-accelerated stacks like Isaac Gym and MJX. It delivers a simulation speed ~430,000 faster than in real-time, and takes only 26 seconds to train a robotic locomotion policy transferrable to the real world on a single RTX4090 (see tutorial: The Genesis physics engine and simulation platform is fully open source at We'll gradually roll out access to our generative framework in the near future. Genesis implements a unified simulation framework all from scratch, integrating a wide spectrum of state-of-the-art physics solvers, allowing simulation of the whole physical world in a virtual realm with the highest realism. We aim to build a universal data engine that leverages an upper-level generative framework to autonomously create physical worlds, together with various modes of data, including environments, camera motions, robotic task proposals, reward functions, robot policies, character motions, fully interactive 3D scenes, open-world articulated assets, and more, aiming towards fully automated data generation for robotics, physical AI and other applications. Open Source Code: Project webpage: Documentation: 1/n

Everything you love about generative models — now powered by real physics! Announcing the Genesis project — after a 24-month large-scale research collaboration involving over 20 research labs — a generative physics engine able to generate 4D dynamical worlds powered by a physics simulation platform designed for general-purpose robotics and physical AI applications. Genesis's physics engine is developed in pure Python, while being 10-80x faster than existing GPU-accelerated stacks like Isaac Gym and MJX. It delivers a simulation speed ~430,000 faster than in real-time, and takes only 26 seconds to train a robotic locomotion policy transferrable to the real world on a single RTX4090 (see tutorial: The Genesis physics engine and simulation platform is fully open source at We'll gradually roll out access to our generative framework in the near future. Genesis implements a unified simulation framework all from scratch, integrating a wide spectrum of state-of-the-art physics solvers, allowing simulation of the whole physical world in a virtual realm with the highest realism. We aim to build a universal data engine that leverages an upper-level generative framework to autonomously create physical worlds, together with various modes of data, including environments, camera motions, robotic task proposals, reward functions, robot policies, character motions, fully interactive 3D scenes, open-world articulated assets, and more, aiming towards fully automated data generation for robotics, physical AI and other applications. Open Source Code: Project webpage: Documentation: 1/n

Zhou Xian

3,821,512 次观看 • 1 年前

🚨 YOUR GPU IS PROBABLY WASTING MORE THAN YOU THINK. vLLM is built to squeeze far more useful work out of your GPU when serving LLMs. Running an LLM at scale isn’t just about having a powerful GPU. The real problem is how efficiently you use its memory and compute. That’s where vLLM comes in. → High-throughput LLM inference and serving → PagedAttention for smarter KV-cache memory management → Continuous batching to keep GPUs busy → Prefix caching + chunked prefill → OpenAI-compatible API out of the box → Supports a huge range of modern LLM architectures → Quantization support for running models more efficiently And the crazy part? You can start an OpenAI-compatible inference server with: `vllm serve ` So your application can talk to your own model almost like it’s talking to OpenAI. The bigger idea: Don’t just buy more GPUs. Make the GPUs you already have work harder. That’s why vLLM has become such a major project in LLM inference. 🔥 #vLLM #AI #LLM #Inference #GPU #MachineLearning #AIInfrastructure #OpenSource #AIAgents #Developers

🚨 YOUR GPU IS PROBABLY WASTING MORE THAN YOU THINK. vLLM is built to squeeze far more useful work out of your GPU when serving LLMs. Running an LLM at scale isn’t just about having a powerful GPU. The real problem is how efficiently you use its memory and compute. That’s where vLLM comes in. → High-throughput LLM inference and serving → PagedAttention for smarter KV-cache memory management → Continuous batching to keep GPUs busy → Prefix caching + chunked prefill → OpenAI-compatible API out of the box → Supports a huge range of modern LLM architectures → Quantization support for running models more efficiently And the crazy part? You can start an OpenAI-compatible inference server with: `vllm serve ` So your application can talk to your own model almost like it’s talking to OpenAI. The bigger idea: Don’t just buy more GPUs. Make the GPUs you already have work harder. That’s why vLLM has become such a major project in LLM inference. 🔥 #vLLM #AI #LLM #Inference #GPU #MachineLearning #AIInfrastructure #OpenSource #AIAgents #Developers

Vikas gupta

14,138 次观看 • 6 天前

Explore where physical AI is headed next with the people shaping it. On Feb. 10 at 9 a.m. PT, join Ming‑Yu, Vice President of Research at NVIDIA, and Chubby♨️ from Superintelligence, for a conversation on how NVIDIA Cosmos reframes AI for a world of robots, AVs, and vision-driven systems. We'll dig into: ✅ Cosmos' origins and designing models for the physical world ✅ The scientific ideas and architectures behind Cosmos and its open models ✅ How open, reasoning-capable physical AI is changing how teams prototype, deploy, and iterate ✅ What a future of AI systems that understand space, time, and causality could unlock If you’re thinking into real‑world AI, this is the conversation to catch ➡️

Explore where physical AI is headed next with the people shaping it. On Feb. 10 at 9 a.m. PT, join Ming‑Yu, Vice President of Research at NVIDIA, and Chubby♨️ from Superintelligence, for a conversation on how NVIDIA Cosmos reframes AI for a world of robots, AVs, and vision-driven systems. We'll dig into: ✅ Cosmos' origins and designing models for the physical world ✅ The scientific ideas and architectures behind Cosmos and its open models ✅ How open, reasoning-capable physical AI is changing how teams prototype, deploy, and iterate ✅ What a future of AI systems that understand space, time, and causality could unlock If you’re thinking into real‑world AI, this is the conversation to catch ➡️

NVIDIA AI Developer

26,468 次观看 • 7 个月前

Meet LingBot-World 2.0.🌍A 1.3B world model, running continuous interactive generation in real time on a single consumer GPU. 🤖 ⚡ It responds to movement, camera control, combat, and text-driven events in real time, while the scene keeps unfolding instead of ending as a fixed video clip. 🌌 Causal pretraining and MoBA help preserve textures, geometry, and scene consistency over long autoregressive rollouts. 🧠 Consistency and distribution-matching distillation compress multi-step generation and train the model on its own rollouts, reducing accumulated drift. 🛠️ The 14B Causal Pretrain backbone and Bidirectional teacher are also open, giving developers a path to post-train, distill, compress, and adapt new world models. 🔒 CC BY-NC-SA 4.0. Non-commercial use only.

Meet LingBot-World 2.0.🌍A 1.3B world model, running continuous interactive generation in real time on a single consumer GPU. 🤖 ⚡ It responds to movement, camera control, combat, and text-driven events in real time, while the scene keeps unfolding instead of ending as a fixed video clip. 🌌 Causal pretraining and MoBA help preserve textures, geometry, and scene consistency over long autoregressive rollouts. 🧠 Consistency and distribution-matching distillation compress multi-step generation and train the model on its own rollouts, reducing accumulated drift. 🛠️ The 14B Causal Pretrain backbone and Bidirectional teacher are also open, giving developers a path to post-train, distill, compress, and adapt new world models. 🔒 CC BY-NC-SA 4.0. Non-commercial use only.

ModelScope

33,630 次观看 • 7 天前

In just one week, Binh and I trained a full-body Unitree G1. Here's a recap: 1. Secured a Unitree G1 humanoid through a LinkedIn post 2. Deployed TWIST2 full-body teleoperation pipelines 3. Adapted TWIST2 for Zed stereo camera & collected full-body teleoperation samples (carried by Binh ) 4. Adapted & fine-tuned NVIDIA Gr00T N1.5 VLA on the TWIST2 public datasets, which I fine-tuned on an 8xNVIDIA H100 Cluster. We picked Gr00T N1.5 as it was trained with Unitree G1 embodiment data. 5. Adapted the TWIST2 codebase to stream in the actions from Gr00T via ZMQ using a co-located NVIDIA H100 for ~200ms inference latency 6. Tested the model in sim, then deployed to the real-world Unitree G1. We streamed a training sample observation to the VLA (as we didn't want to break robot in case real observations were OOD) We were the first team in the world to deploy the full TWIST2 data collection pipeline to the unitree g1 :) Much more work ahead though, which I'll work on as a side-project over the next months: 1. Exploring the various types of 'world models': video backbones, dynamics models, v-jepa-2 models. I believe these will generalize better & train much more data-efficiently than VLM backbones 2. Speeding up inference - I believe low-latency robotics inference will be a big challenge. There are many works in video diffusion which I'd like to test (e.g. SageAttention, SparseAttention, Drifting Models). Perhaps also writing custom CUDA kernels. 3. Economics of inference scaling :) What will be the compute demands as we scale inference up to millions of humanoids? Will it run on edge or on distributed 'co-located' inference clusters? These are questions I'd like to answer. Adapted TWIST2 codebase: Adapted Gr00T-N1.5 codebase: The ETH Robotics Club are doing a cool GTC Golden ticket competition with NVIDIA , so this is my submission :) The DGX Spark compute will get me a long way with initial prototyping & especially working on inference optimization for next-gen Blackwell GPUs #NVIDIAGTC #GOLDENTICKET #ETHRC

In just one week, Binh and I trained a full-body Unitree G1. Here's a recap: 1. Secured a Unitree G1 humanoid through a LinkedIn post 2. Deployed TWIST2 full-body teleoperation pipelines 3. Adapted TWIST2 for Zed stereo camera & collected full-body teleoperation samples (carried by Binh ) 4. Adapted & fine-tuned NVIDIA Gr00T N1.5 VLA on the TWIST2 public datasets, which I fine-tuned on an 8xNVIDIA H100 Cluster. We picked Gr00T N1.5 as it was trained with Unitree G1 embodiment data. 5. Adapted the TWIST2 codebase to stream in the actions from Gr00T via ZMQ using a co-located NVIDIA H100 for ~200ms inference latency 6. Tested the model in sim, then deployed to the real-world Unitree G1. We streamed a training sample observation to the VLA (as we didn't want to break robot in case real observations were OOD) We were the first team in the world to deploy the full TWIST2 data collection pipeline to the unitree g1 :) Much more work ahead though, which I'll work on as a side-project over the next months: 1. Exploring the various types of 'world models': video backbones, dynamics models, v-jepa-2 models. I believe these will generalize better & train much more data-efficiently than VLM backbones 2. Speeding up inference - I believe low-latency robotics inference will be a big challenge. There are many works in video diffusion which I'd like to test (e.g. SageAttention, SparseAttention, Drifting Models). Perhaps also writing custom CUDA kernels. 3. Economics of inference scaling :) What will be the compute demands as we scale inference up to millions of humanoids? Will it run on edge or on distributed 'co-located' inference clusters? These are questions I'd like to answer. Adapted TWIST2 codebase: Adapted Gr00T-N1.5 codebase: The ETH Robotics Club are doing a cool GTC Golden ticket competition with NVIDIA , so this is my submission :) The DGX Spark compute will get me a long way with initial prototyping & especially working on inference optimization for next-gen Blackwell GPUs #NVIDIAGTC #GOLDENTICKET #ETHRC

Arnie Ramesh

23,236 次观看 • 7 个月前

$After taking some time off post-Rapid, I'm excited to share what I’ve been up to since: Datawizz AI! We’ve raised a $12.5M Seed led by Human Capital to make AI 10x cheaper, 2x more accurate and 15x faster by transitioning from LLMs to SLMs. AI is eating the world. But unit economics are eating AI. Looking at the fastest growing AI products, they all share two traits - growing fast, and painful inference bills. General-purpose LLMs are just too expensive to run. A big reason for that is we train LLMs to be good at everything - answer any question, be an expert on any topic. The big labs dub this "generalisation", but for real-world applications, it is unnecessary. In reality - many AI applications need models to be experts in one thing - and do that thing extremely well. Your coding model doesn’t need to memorize ancient recipes for Garum sauce. This is where Datawizz comes in - we sit between the AI applications and automatically create smaller (100x-1,000x) specialized models to handle specific aspects of your work. By focusing the model and combining industry-data in the distillation process - we end up with models that beat SOTA LLMs at a fraction of the cost. We created Datawizz to make AI specialized and scalable. We’re early in the journey, but have already been able to save companies 90%+ on their inference bill and speed up their apps by 10x. Excited to build better AI platforms? Join the Datawizz team (link in first comment)$

After taking some time off post-Rapid, I'm excited to share what I’ve been up to since: Datawizz AI! We’ve raised a $12.5M Seed led by Human Capital to make AI 10x cheaper, 2x more accurate and 15x faster by transitioning from LLMs to SLMs. AI is eating the world. But unit economics are eating AI. Looking at the fastest growing AI products, they all share two traits - growing fast, and painful inference bills. General-purpose LLMs are just too expensive to run. A big reason for that is we train LLMs to be good at everything - answer any question, be an expert on any topic. The big labs dub this "generalisation", but for real-world applications, it is unnecessary. In reality - many AI applications need models to be experts in one thing - and do that thing extremely well. Your coding model doesn’t need to memorize ancient recipes for Garum sauce. This is where Datawizz comes in - we sit between the AI applications and automatically create smaller (100x-1,000x) specialized models to handle specific aspects of your work. By focusing the model and combining industry-data in the distillation process - we end up with models that beat SOTA LLMs at a fraction of the cost. We created Datawizz to make AI specialized and scalable. We’re early in the journey, but have already been able to save companies 90%+ on their inference bill and speed up their apps by 10x. Excited to build better AI platforms? Join the Datawizz team (link in first comment)

Iddo Gino 🐙

21,928 次观看 • 11 个月前

Most AI world models can generate beautiful scenes. Keeping those scenes alive for an hour without falling apart is the real challenge. That's what caught my attention about LingBot-World 2.0 (LingBot-World-Infinity) from Robbyant Instead of chasing longer videos, it focuses on something much harder: persistent, interactive worlds that stay coherent while you explore. A few highlights: • Generates worlds from a single frame and continuously responds to live user actions through a causal world model. • Streams stable 720p at 60 FPS in real time. The team reports a continuous 60 minute stress test across 20 different scenarios with no noticeable visual degradation. • Uses a Brain-Cerebellum co-simulation framework where a VLM plans events while the video model turns them into consistent world evolution. • Pilot and Director Agents help drive character behavior and introduce new objects and events. • Open sourced with a 14B flagship model, while the paper also describes a lightweight 1.3B version for a single consumer GPU. There is also an online interactive demo. The biggest takeaway? We're moving beyond AI that generates clips. We're getting closer to AI that generates living, evolving worlds you can actually interact with. And that feels like a much bigger shift than another jump in video quality. Explore more: 💻 Github: 🤗 Weights: 🌐 Website-with videos you can use : 🎮 Try it online: #Robbyant #LingBot #WorldModel #EmbodiedAI #OpenSource #Robotics #ad

Most AI world models can generate beautiful scenes. Keeping those scenes alive for an hour without falling apart is the real challenge. That's what caught my attention about LingBot-World 2.0 (LingBot-World-Infinity) from Robbyant Instead of chasing longer videos, it focuses on something much harder: persistent, interactive worlds that stay coherent while you explore. A few highlights: • Generates worlds from a single frame and continuously responds to live user actions through a causal world model. • Streams stable 720p at 60 FPS in real time. The team reports a continuous 60 minute stress test across 20 different scenarios with no noticeable visual degradation. • Uses a Brain-Cerebellum co-simulation framework where a VLM plans events while the video model turns them into consistent world evolution. • Pilot and Director Agents help drive character behavior and introduce new objects and events. • Open sourced with a 14B flagship model, while the paper also describes a lightweight 1.3B version for a single consumer GPU. There is also an online interactive demo. The biggest takeaway? We're moving beyond AI that generates clips. We're getting closer to AI that generates living, evolving worlds you can actually interact with. And that feels like a much bigger shift than another jump in video quality. Explore more: 💻 Github: 🤗 Weights: 🌐 Website-with videos you can use : 🎮 Try it online: #Robbyant #LingBot #WorldModel #EmbodiedAI #OpenSource #Robotics #ad

Alif Khan

84,542 次观看 • 2 个月前

OPENAI JUST CUT MODEL PRICES AS THE AI PRICE WAR HEATS UP 🔥 OpenAI is cutting the price of Luna by 80% and Terra by 20%. It is also rolling out a faster mode that runs at twice the speed, but costs more. The bigger picture is clear -- cheaper Chinese models and tighter corporate AI budgets are forcing the biggest AI companies to compete on price, not just performance.

OPENAI JUST CUT MODEL PRICES AS THE AI PRICE WAR HEATS UP 🔥 OpenAI is cutting the price of Luna by 80% and Terra by 20%. It is also rolling out a faster mode that runs at twice the speed, but costs more. The bigger picture is clear -- cheaper Chinese models and tighter corporate AI budgets are forcing the biggest AI companies to compete on price, not just performance.

CryptosRus

12,382 次观看 • 1 个月前

Real-time world models represent a fundamental shift in AI. reactor is building the platform for real-time generative video infrastructure, supporting developers who need the tech for use across entertainment, physical AI, and robotics. Co-founders Alberto and Bryce Schmidtchen joined us last week on The Investment Memo, hosted by Partners Bucky Moore and Amber Yang, to talk about the era of world models. The conversation centered around the infrastructure Reactor is building, why real-time models are the edge right now, and current use cases for the product. Alberto and Bryce agreed that world models are shaping the way simulations are created, and that developers need a streamlined platform that can support their ideas. We believe Reactor is positioned to be at the frontier of research into real-time generative models. We look forward to seeing how these models apply across industries. Chapters 00:00 Introduction & Overview of Reactor 01:08 Meet the Hosts & Founders 02:18 The Origin Story: From 3D Assets to World Models 05:07 Real-Time Video Applications Across Industries 06:55 The Open Source World Model Explosion 07:23 Why Infrastructure Is the Opportunity 08:42 Parallels to Past Technology Waves 09:51 Bridging the Research-to-Production Gap 13:13 What Developers Are Building with World Models 16:41 Lessons from Luma AI 18:23 What Apple Vision Pro Taught Bryce About Real-Time Systems 20:48 Company Values & Team Culture 22:40 Series A: What the Capital Unlocks 24:13 Reactor's Five-Year Vision 26:09 Closing Remarks

Real-time world models represent a fundamental shift in AI. reactor is building the platform for real-time generative video infrastructure, supporting developers who need the tech for use across entertainment, physical AI, and robotics. Co-founders Alberto and Bryce Schmidtchen joined us last week on The Investment Memo, hosted by Partners Bucky Moore and Amber Yang, to talk about the era of world models. The conversation centered around the infrastructure Reactor is building, why real-time models are the edge right now, and current use cases for the product. Alberto and Bryce agreed that world models are shaping the way simulations are created, and that developers need a streamlined platform that can support their ideas. We believe Reactor is positioned to be at the frontier of research into real-time generative models. We look forward to seeing how these models apply across industries. Chapters 00:00 Introduction & Overview of Reactor 01:08 Meet the Hosts & Founders 02:18 The Origin Story: From 3D Assets to World Models 05:07 Real-Time Video Applications Across Industries 06:55 The Open Source World Model Explosion 07:23 Why Infrastructure Is the Opportunity 08:42 Parallels to Past Technology Waves 09:51 Bridging the Research-to-Production Gap 13:13 What Developers Are Building with World Models 16:41 Lessons from Luma AI 18:23 What Apple Vision Pro Taught Bryce About Real-Time Systems 20:48 Company Values & Team Culture 22:40 Series A: What the Capital Unlocks 24:13 Reactor's Five-Year Vision 26:09 Closing Remarks

Lightspeed

144,942 次观看 • 3 个月前

🚨 Open-source AI just got serious. Alibaba just released Qwen 3.5-Plus 🤯 397B parameters. Only 17B active per inference. Performance comparable to Gemini 3 Pro — at just $0.8 per million tokens (1/18 the cost). But here’s what makes it different 👇 This isn’t just another benchmark drop. In the demo, Qwen runs directly inside the terminal: ➙ Reviews GitHub PRs ➙ Fetches diffs automatically ➙ Runs CLI & smoke tests ➙ Generates integration tests ➙ Executes /context ➙ Shows real-time token usage breakdown That’s not a chatbot. That’s a real developer workflow. Under the hood: ⚡ 60% lower deployment memory ⚡ Inference throughput increased by 8.6× (32K context) ⚡ Inference throughput increased by up to 19× (256K ultra-long context) ⚡ Native multimodal model ⚡ Supports direct video input (up to 2 hours) And the ecosystem is massive: ✔ 201 languages ✔ 250K vocabulary ✔ 400+ open-sourced models ✔ 1B+ downloads ✔ 200K+ derivative models Available now via Qwen App, MoDa Community, Hugging Face, and Alibaba Cloud Bailian. ✅ The AI race isn’t about who’s biggest anymore. → It’s about who’s usable. 🔥 Watch the demo 👇 #Qwen #OpenSourceAI #AI #DevTools

🚨 Open-source AI just got serious. Alibaba just released Qwen 3.5-Plus 🤯 397B parameters. Only 17B active per inference. Performance comparable to Gemini 3 Pro — at just $0.8 per million tokens (1/18 the cost). But here’s what makes it different 👇 This isn’t just another benchmark drop. In the demo, Qwen runs directly inside the terminal: ➙ Reviews GitHub PRs ➙ Fetches diffs automatically ➙ Runs CLI & smoke tests ➙ Generates integration tests ➙ Executes /context ➙ Shows real-time token usage breakdown That’s not a chatbot. That’s a real developer workflow. Under the hood: ⚡ 60% lower deployment memory ⚡ Inference throughput increased by 8.6× (32K context) ⚡ Inference throughput increased by up to 19× (256K ultra-long context) ⚡ Native multimodal model ⚡ Supports direct video input (up to 2 hours) And the ecosystem is massive: ✔ 201 languages ✔ 250K vocabulary ✔ 400+ open-sourced models ✔ 1B+ downloads ✔ 200K+ derivative models Available now via Qwen App, MoDa Community, Hugging Face, and Alibaba Cloud Bailian. ✅ The AI race isn’t about who’s biggest anymore. → It’s about who’s usable. 🔥 Watch the demo 👇 #Qwen #OpenSourceAI #AI #DevTools

Toha Khan

123,375 次观看 • 7 个月前

New short course with Mistral AI ! Mistral's open-source Mixtral 8x7B model uses a "mixture of experts" (MoE) architecture. Unlike a standard transformer, an MoE model has multiple expert feed-forward networks (8 in this case), with a gating network selecting two experts at inference time. This enables MoE to match the performance of a large model but faster inference. Mixtral 8x7B has 46.7B parameters but activates only 12.9B at inference to predict the next token. In Getting Started with Mistral, you’ll learn from Mistral’s Sophia Yang, Ph.D. to: - Explore Mistral's open-source models (Mistral 7B, Mixtral 8x7B) and commercial models via API calls and Mistral AI's Le Chat website - Implement JSON mode to generate structured outputs to integrate directly into larger software systems. - Use function calling for Tool Use, such as calling custom Python code that queries tabular data - Ground your LLM's response with external knowledge sources using RAG - Build a Mistral-powered chat interface that can reference external documents This course will help deepen your prompt engineering skills. Please sign up here:

New short course with Mistral AI ! Mistral's open-source Mixtral 8x7B model uses a "mixture of experts" (MoE) architecture. Unlike a standard transformer, an MoE model has multiple expert feed-forward networks (8 in this case), with a gating network selecting two experts at inference time. This enables MoE to match the performance of a large model but faster inference. Mixtral 8x7B has 46.7B parameters but activates only 12.9B at inference to predict the next token. In Getting Started with Mistral, you’ll learn from Mistral’s Sophia Yang, Ph.D. to: - Explore Mistral's open-source models (Mistral 7B, Mixtral 8x7B) and commercial models via API calls and Mistral AI's Le Chat website - Implement JSON mode to generate structured outputs to integrate directly into larger software systems. - Use function calling for Tool Use, such as calling custom Python code that queries tabular data - Ground your LLM's response with external knowledge sources using RAG - Build a Mistral-powered chat interface that can reference external documents This course will help deepen your prompt engineering skills. Please sign up here:

Andrew Ng

387,172 次观看 • 2 年前

Excited to share that we’ve raised $300M in our Series B round, led by Radical Ventures, bringing our total funding to more than $450M, with leading technology companies joining as both customers and investors. We continue accelerating the path to AGI through our two core pillars: ultra-optimized infrastructure for AI workloads, and realtime world models built on top of it. Today, we’re also launching DOS (Decart Optimization Stack) 2.0 – our next-generation inference and training platform, delivering over 1,600 tokens per second for agentic inference and over 100 FPS for world models across major hardware platforms. Alongside DOS, we will launch new versions of our world models in the coming weeks: Lucy, for immersive realtime experiences, and Oasis, for physical AI. Grateful to our partners, customers, and backers across media & entertainment, Physical AI, chips, hyperscalers, and the broader AI ecosystem. Radical Ventures, Adobe, Alpha Partners, Amazon and Amazon Web Services, Atreides Management, LP, Benchmark, eBay, NVIDIA, Sequoia Capital, Toyota USA, @valorep, oren zeev; Andrej Karpathy, Michael Eisner, Yamauchi-No.10 Family Office, Moritz Baier-Lentz and more. We made a film for this moment with Decart CEO Dean Leitersdorf and Moritz Baier-Lentz - a look at the company, the technology, and what comes next.

Excited to share that we’ve raised $300M in our Series B round, led by Radical Ventures, bringing our total funding to more than $450M, with leading technology companies joining as both customers and investors. We continue accelerating the path to AGI through our two core pillars: ultra-optimized infrastructure for AI workloads, and realtime world models built on top of it. Today, we’re also launching DOS (Decart Optimization Stack) 2.0 – our next-generation inference and training platform, delivering over 1,600 tokens per second for agentic inference and over 100 FPS for world models across major hardware platforms. Alongside DOS, we will launch new versions of our world models in the coming weeks: Lucy, for immersive realtime experiences, and Oasis, for physical AI. Grateful to our partners, customers, and backers across media & entertainment, Physical AI, chips, hyperscalers, and the broader AI ecosystem. Radical Ventures, Adobe, Alpha Partners, Amazon and Amazon Web Services, Atreides Management, LP, Benchmark, eBay, NVIDIA, Sequoia Capital, Toyota USA, @valorep, oren zeev; Andrej Karpathy, Michael Eisner, Yamauchi-No.10 Family Office, Moritz Baier-Lentz and more. We made a film for this moment with Decart CEO Dean Leitersdorf and Moritz Baier-Lentz - a look at the company, the technology, and what comes next.

Decart

99,072 次观看 • 4 个月前

The biggest driver of AI compute now is real-world deployment. At NVIDIA GTC, IREN's Kent Draper, alongside Lenovo's Vlad Rozanovich, shared what they’re actually seeing on the ground from AI infrastructure demand, and the signal is coming from enterprise use. He pointed to agents, internal enterprise models, and real-world deployments as the workloads now driving demand, and explained that the challenge isn't finding customers, it’s how quickly they can build capacity to keep up. Their strategy? A vertically integrated model focused entirely on power-dense computing, which allows them to scale faster, control the environment, and deliver infrastructure purpose-built for AI rather than multi-tenant data centers trying to support dozens of different workloads.

The biggest driver of AI compute now is real-world deployment. At NVIDIA GTC, IREN's Kent Draper, alongside Lenovo's Vlad Rozanovich, shared what they’re actually seeing on the ground from AI infrastructure demand, and the signal is coming from enterprise use. He pointed to agents, internal enterprise models, and real-world deployments as the workloads now driving demand, and explained that the challenge isn't finding customers, it’s how quickly they can build capacity to keep up. Their strategy? A vertically integrated model focused entirely on power-dense computing, which allows them to scale faster, control the environment, and deliver infrastructure purpose-built for AI rather than multi-tenant data centers trying to support dozens of different workloads.

Six Five Media

7,422,478 次观看 • 5 个月前