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。

🚀 Introducing Hy-Nomas: An Open Source Free-Form Interactive Video World Model Key Features: * Free-form interaction with text prompts and camera trajectories * Long-horizon generation with world memory * Lightweight LoRA adaptation for specialized domains such as robotics * Optimized for Inference on Consumer GPUs Test it and drop... show more

Tencent Hy

54,033 subscribers

48,071 次观看 • 1 天前 •via X (Twitter)

艺术教育科学技术

Anya Rossi• Live Now

Private livecam show

20 条评论

Vishal Singh 的头像

Vishal Singh1 天前

Would have been amazing if it worked on low VRAM but requires 24GB VRAM out of the box. And yes 24GB is better than previous models that require supercomputers to run, but it's still not enough. These models should support lower VRAM like 16GB or ideally even lower at launch.

Dima Nurm 的头像

Dima Nurm1 天前

@anabology take a look

xabz 的头像

xabz1 天前

world memory in a video model is the part that actually matters

PASSED K1NG 的头像

PASSED K1NG1 天前

Grateful for every open-source model, even if it hasn't quite reached the cutting edge yet

Prompt Heat 的头像

Prompt Heat1 天前

we need the best cheap strongest model and maybe free in your code desktop app :D LOL

Apollo 的头像

Apollo1 天前

Tencent’s AI ecosystem Keep evolving and expanding. From llms to world models, they keep reasong ai solutions evwn if not frontier

Sahil Sangani 🇮🇳/acc 的头像

Sahil Sangani 🇮🇳/acc1 天前

can i get access?

Zoraya Ai 的头像

Zoraya Ai1 天前

Open-source world models with persistent memory and consumer GPU support could make interactive video generation far more accessible

Dima Nurm 的头像

Dima Nurm1 天前

Awesome—today is a day for releases from the East; really cool. 👍

👁️M1nd 3xp4nd3r👁️ 的头像

👁️M1nd 3xp4nd3r👁️1 天前

Ohhhhhhh you all just changed visuals and something I been wanting to experiment for years this is amazing

Ton🇹🇭 的头像

Ton🇹🇭1 天前

😉

Alyar AI 的头像

Alyar AI1 天前

Optimized for consumer GPUs is a big win. Powerful AI should be more accessible! 💯

Clara AI | Tools & Updates 的头像

Clara AI | Tools & Updates1 天前

Free form text and camera control on consumer hardware is solid

Kyrzel ⚔️ 的头像

Kyrzel ⚔️1 天前

The future of @makooor

NeuralLoot 的头像

NeuralLoot1 天前

How it fits on one RTX 4090: DiT weights and the KV cache are stored in INT8, the decoder runs in FP8, and only 18 past latent frames stay on the GPU while the rest of the world's history waits on CPU. Same 8-step schedule, still 720x1280 at 24 FPS.

SynnJJ 的头像

SynnJJ1 天前

Just amazing, will try it right away💛💛💛

Squirrel 的头像

Squirrel1 天前

What is this

J A Z I I 的头像

J A Z I I1 天前

is there API so I can use???

Camila Rivera 的头像

Camila Rivera1 天前

Open-source world models are taking AI video to the next level

lklz 的头像

lklz1 天前

流量

相关视频

🔥🔥🔥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 • 9 个月前

We’re open-sourcing HY-World 2.0, a multimodal world model that generates, reconstructs, and simulates interactive *3D worlds* from text, images, and videos. Outputs can be integrated into game engines and embodied simulation pipelines. Key highlights: 🔹 One-click world generation Turn text or image into interactive 3D worlds automatically. 🔹 Pipeline-ready 3D outputs Editable 3D worlds for Unity and Unreal Engine, with standard 3D exports including mesh, 3DGS, and point clouds. 🔹 Unified world model system One model family for world generation and reconstruction across synthetic and real-world scenes. 🔹 Interactive character mode Explore generated 3D worlds in real time with physics-aware movement and collision support. ✨ Apply for access: 🔗 GitHub: 🤗 Hugging Face: 📄 Technical Report:

We’re open-sourcing HY-World 2.0, a multimodal world model that generates, reconstructs, and simulates interactive 3D worlds from text, images, and videos. Outputs can be integrated into game engines and embodied simulation pipelines. Key highlights: 🔹 One-click world generation Turn text or image into interactive 3D worlds automatically. 🔹 Pipeline-ready 3D outputs Editable 3D worlds for Unity and Unreal Engine, with standard 3D exports including mesh, 3DGS, and point clouds. 🔹 Unified world model system One model family for world generation and reconstruction across synthetic and real-world scenes. 🔹 Interactive character mode Explore generated 3D worlds in real time with physics-aware movement and collision support. ✨ Apply for access: 🔗 GitHub: 🤗 Hugging Face: 📄 Technical Report:

Tencent Hy

373,065 次观看 • 6 个月前

Meet Tencent HY-MT1.5 — our latest open-source translation models, featuring 1.8B and 7B models designed for seamless on-device and cloud deployment. 🌍 The 1.8B version is already trending at #1 on Hugging Face! 🥇It’s optimized for consumer hardware, delivering 0.18s latency (50 tokens) with just a 1GB memory footprint. Meanwhile, the 7B model achieves exceptional performance, surpassing many mid-sized models and rivaling the 90th percentile of Gemini 3.0 Pro. 🚀🧠 👉🏻 Try it now: 🔗 GitHub: 🤗 Hugging Face:

Meet Tencent HY-MT1.5 — our latest open-source translation models, featuring 1.8B and 7B models designed for seamless on-device and cloud deployment. 🌍 The 1.8B version is already trending at #1 on Hugging Face! 🥇It’s optimized for consumer hardware, delivering 0.18s latency (50 tokens) with just a 1GB memory footprint. Meanwhile, the 7B model achieves exceptional performance, surpassing many mid-sized models and rivaling the 90th percentile of Gemini 3.0 Pro. 🚀🧠 👉🏻 Try it now: 🔗 GitHub: 🤗 Hugging Face:

Tencent HY

17,105 次观看 • 9 个月前

📈 Scale open models with Hugging Face Text Generation Inference! 🚀 Customize with LoRA adapters ⚙️ Optimize with PEFT and TRL ☁️ Deploy on Google Cloud Platform with ease 🤖 Watch the demo from #GemmaDeveloperDay Tokyo →

📈 Scale open models with Hugging Face Text Generation Inference! 🚀 Customize with LoRA adapters ⚙️ Optimize with PEFT and TRL ☁️ Deploy on Google Cloud Platform with ease 🤖 Watch the demo from #GemmaDeveloperDay Tokyo →

Google for Developers

23,080 次观看 • 1 年前

🚀🚀🚀Introducing HY World 1.5 (WorldPlay)! We have now open-sourced the most systemized, comprehensive real-time world model framework in the industry. In HY World 1.5, we develop WorldPlay, a streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency, resolving the trade-off between speed and memory that limits current methods. You can generate and explore 3D worlds simply by inputting text or images. Walk, look around, and interact like you're playing a game. Highlights: 🔹Real-Time: Generates long-horizon streaming video at 24 FPS with superior consistency. 🔹Geometric Consistency: Achieved using a Reconstituted Context Memory mechanism to dynamically rebuild context from past frames to alleviate memory attenuation 🔹Robust Control: Uses a Dual Action Representation for robust response to user keyboard and mouse inputs. 🔹Versatile Applications: Supports both first-person and third-person perspectives, enabling applications like promptable events and infinite world extension. 👉🏻Try it now: 🌐Project Page: 🔗Github: 🤗Hugging Face: 📄Technical Report:

🚀🚀🚀Introducing HY World 1.5 (WorldPlay)! We have now open-sourced the most systemized, comprehensive real-time world model framework in the industry. In HY World 1.5, we develop WorldPlay, a streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency, resolving the trade-off between speed and memory that limits current methods. You can generate and explore 3D worlds simply by inputting text or images. Walk, look around, and interact like you're playing a game. Highlights: 🔹Real-Time: Generates long-horizon streaming video at 24 FPS with superior consistency. 🔹Geometric Consistency: Achieved using a Reconstituted Context Memory mechanism to dynamically rebuild context from past frames to alleviate memory attenuation 🔹Robust Control: Uses a Dual Action Representation for robust response to user keyboard and mouse inputs. 🔹Versatile Applications: Supports both first-person and third-person perspectives, enabling applications like promptable events and infinite world extension. 👉🏻Try it now: 🌐Project Page: 🔗Github: 🤗Hugging Face: 📄Technical Report:

Tencent Hy

190,264 次观看 • 10 个月前

🚀 Excited to release LongLive 2.0! 🎬 An end-to-end infrastructure for long video generation, with FP4 and parallelism at the core of both training and inference. ⚡45.7 FPS generation speed on 5B model⚡ ✨ LongLive 2.0 supports real-video training, few-step distillation, multi-shot training/inference, sequence-parallel acceleration, NVFP4 KV cache, and async VAE decoding deployment. 🧩 To our knowledge, this is the first open-source 4-bit long video generation infra that covers both training and inference. 🙌 Welcome to check it out, try it, and share feedback! 🔗 Code: 📰 Paper: 🎥 Demo: #LongVideoGeneration #VideoGeneration #Realtime #AIInfra #EfficientAI #FP4 #Parallel #NVIDIA

🚀 Excited to release LongLive 2.0! 🎬 An end-to-end infrastructure for long video generation, with FP4 and parallelism at the core of both training and inference. ⚡45.7 FPS generation speed on 5B model⚡ ✨ LongLive 2.0 supports real-video training, few-step distillation, multi-shot training/inference, sequence-parallel acceleration, NVFP4 KV cache, and async VAE decoding deployment. 🧩 To our knowledge, this is the first open-source 4-bit long video generation infra that covers both training and inference. 🙌 Welcome to check it out, try it, and share feedback! 🔗 Code: 📰 Paper: 🎥 Demo: #LongVideoGeneration #VideoGeneration #Realtime #AIInfra #EfficientAI #FP4 #Parallel #NVIDIA

Yukang Chen

60,053 次观看 • 5 个月前

1/ Tencent releases HY World 1.5 (WorldPlay), an open-source real-time world model with 24fps A streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency In direct comparison to Genie 3, it's technically inferior, and there's no information on how long the consistency will last. However, Tencent has released a significant world model as open source for the first time, making it available to everyone. A few more examples, a thread 🧵

1/ Tencent releases HY World 1.5 (WorldPlay), an open-source real-time world model with 24fps A streaming video diffusion model that enables real-time, interactive world modeling with long-term geometric consistency In direct comparison to Genie 3, it's technically inferior, and there's no information on how long the consistency will last. However, Tencent has released a significant world model as open source for the first time, making it available to everyone. A few more examples, a thread 🧵

Chubby♨️

37,776 次观看 • 10 个月前

🚀 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

11,943 次观看 • 1 年前

🔥 Ming-UniAudio: The 「Nano Banana」moment for speech is here! A single model for universal understanding, generation & free-form editing. First Unified Continuous Tokenizer 「MingTok-Audio」and Unified Und & Gen Speech LLM built on it. First Universal Free-form Speech Editing Model Without Timestamp Condition. First FreeForm Audio Editing Benchmark. 💻 GitHub: 🤗 Tokenizer: 🤗 Model: base: edit: 🤗 Benchmark: 🌍 blog: #AI #Speech #SpeechLLM #LLM #GenerativeAI #Audio #ASR #TTS #SpeechEditing

🔥 Ming-UniAudio: The 「Nano Banana」moment for speech is here! A single model for universal understanding, generation & free-form editing. First Unified Continuous Tokenizer 「MingTok-Audio」and Unified Und & Gen Speech LLM built on it. First Universal Free-form Speech Editing Model Without Timestamp Condition. First FreeForm Audio Editing Benchmark. 💻 GitHub: 🤗 Tokenizer: 🤗 Model: base: edit: 🤗 Benchmark: 🌍 blog: #AI #Speech #SpeechLLM #LLM #GenerativeAI #Audio #ASR #TTS #SpeechEditing

Ant Ling

21,365 次观看 • 1 年前

🚀Introducing Wan2.2-S2V — a 14B parameter model designed for film-grade, audio-driven human animation. 🎬Going beyond basic talking heads to deliver professional-level quality for film, TV, and digital content. And it’s open-source! ✨ Key features: 🔹 Long-video dynamic consistency 🔹 Cinema-quality audio-to-video generation 🔹 Advanced motion and environment control via instruction Perfect for filmmakers, content creators, and developers crafting immersive AI-powered stories. Try it now : Github: Project: Hugging Face Demo: Modelscope Demo: Hugging Face Weights: ModelScope Weights:

🚀Introducing Wan2.2-S2V — a 14B parameter model designed for film-grade, audio-driven human animation. 🎬Going beyond basic talking heads to deliver professional-level quality for film, TV, and digital content. And it’s open-source! ✨ Key features: 🔹 Long-video dynamic consistency 🔹 Cinema-quality audio-to-video generation 🔹 Advanced motion and environment control via instruction Perfect for filmmakers, content creators, and developers crafting immersive AI-powered stories. Try it now : Github: Project: Hugging Face Demo: Modelscope Demo: Hugging Face Weights: ModelScope Weights:

Wan

128,597 次观看 • 1 年前

Woow Nvidia has a 2.6B open-source world model 🔥 Turn an image, a text prompt, and a camera trajectory into a controllable world. All on a single GPU. - Code available on GitHub - Paper on arXiv - Useful for embodied AI, robotics research, and simulations Runs on GPUs like an RTX 5090 or H100, making world-model experiments possible without a multi-GPU setup.

Woow Nvidia has a 2.6B open-source world model 🔥 Turn an image, a text prompt, and a camera trajectory into a controllable world. All on a single GPU. - Code available on GitHub - Paper on arXiv - Useful for embodied AI, robotics research, and simulations Runs on GPUs like an RTX 5090 or H100, making world-model experiments possible without a multi-GPU setup.

Simplifying AI

21,151 次观看 • 10 天前

We're thrilled to release & open-source Hunyuan3D World Model 1.0! This model enables you to generate immersive, explorable, and interactive 3D worlds from just a sentence or an image. It's the industry's first open-source 3D world generation model, compatible with CG pipelines for full editability & simulation. Set to transform game development, VR, digital content creation and so on. Get started now👇🏻 Project Page： Try it now： Github： Hugging Face：

We're thrilled to release & open-source Hunyuan3D World Model 1.0! This model enables you to generate immersive, explorable, and interactive 3D worlds from just a sentence or an image. It's the industry's first open-source 3D world generation model, compatible with CG pipelines for full editability & simulation. Set to transform game development, VR, digital content creation and so on. Get started now👇🏻 Project Page： Try it now： Github： Hugging Face：

Tencent Hy

1,231,203 次观看 • 1 年前

Microsoft just dropped MineWorld on Hugging Face a Real-Time and Open-Source Interactive World Model on Minecraft

Microsoft just dropped MineWorld on Hugging Face a Real-Time and Open-Source Interactive World Model on Minecraft

AK

95,337 次观看 • 1 年前

Introducing LingBot-World: An open-source world simulator pushing the boundaries of video generation. 🚀 🌍 High-Fidelity: Realistic, scientific, & stylized. 🧠 Long-Term Memory: Minute-level consistency. ⚡ Real-Time: <1s latency at 16 FPS. 📜 Apache 2.0 Licensed. Model: Github:

Introducing LingBot-World: An open-source world simulator pushing the boundaries of video generation. 🚀 🌍 High-Fidelity: Realistic, scientific, & stylized. 🧠 Long-Term Memory: Minute-level consistency. ⚡ Real-Time: <1s latency at 16 FPS. 📜 Apache 2.0 Licensed. Model: Github:

ModelScope

28,809 次观看 • 8 个月前

1/5 🚀 Thrilled to open-source OSCAR 🤖 — an action-conditioned world model for robotics, led by the visiting student in my group Zhuoyuan Wu! It generalizes across different robot embodiments with precise action controllability. All trained on a single GH200 GPU, and outperforms existing open-sourced baselines, which have larger model capacity and need more compute. Everything is public, including training data. 📄 Paper: 🌐 Project: 💻 Code: 🤗 Robot data: 🤗 Human data: 🤗 Weights: #Robotics #WorldModels #AI #OpenSource

1/5 🚀 Thrilled to open-source OSCAR 🤖 — an action-conditioned world model for robotics, led by the visiting student in my group Zhuoyuan Wu! It generalizes across different robot embodiments with precise action controllability. All trained on a single GH200 GPU, and outperforms existing open-sourced baselines, which have larger model capacity and need more compute. Everything is public, including training data. 📄 Paper: 🌐 Project: 💻 Code: 🤗 Robot data: 🤗 Human data: 🤗 Weights: #Robotics #WorldModels #AI #OpenSource

Jun Gao

105,237 次观看 • 4 个月前

Today, we are open-sourcing Hunyuan World 1.1 (WorldMirror), a universal feed-forward 3D reconstruction model. 🚀🚀🚀 While our previously released Hunyuan World 1.0 (open-sourced, lite version deployable on consumer GPUs) focused on generating 3D worlds from text or single-view images, Hunyuan World 1.1 significantly expands the input scope by unlocking video-to-3D and multi-view-to-3D world creation. Highlights: 🔹Any Input, Maximized Flexibility and Fidelity: Flexibly integrates diverse geometric priors (camera poses, intrinsics, depth maps) to resolve structural ambiguities and ensure geometrically consistent 3D outputs. 🔹Any Output, SOTA Results：This elegant architecture simultaneously generates multiple 3D representations: dense point clouds, multi-view depth maps, camera parameters, surface normals, and 3D Gaussian Splattings. 🔹Single-GPU & Fast Inference: As an all-in-one, feed-forward model, Hunyuan World 1.1 runs on a single GPU and delivers all 3D attributes in a single forward pass, within seconds. 🌐Project Page: 🔗Github： 🤗Hugging Face： ✨Demo: 📄Technical Report:

Today, we are open-sourcing Hunyuan World 1.1 (WorldMirror), a universal feed-forward 3D reconstruction model. 🚀🚀🚀 While our previously released Hunyuan World 1.0 (open-sourced, lite version deployable on consumer GPUs) focused on generating 3D worlds from text or single-view images, Hunyuan World 1.1 significantly expands the input scope by unlocking video-to-3D and multi-view-to-3D world creation. Highlights: 🔹Any Input, Maximized Flexibility and Fidelity: Flexibly integrates diverse geometric priors (camera poses, intrinsics, depth maps) to resolve structural ambiguities and ensure geometrically consistent 3D outputs. 🔹Any Output, SOTA Results：This elegant architecture simultaneously generates multiple 3D representations: dense point clouds, multi-view depth maps, camera parameters, surface normals, and 3D Gaussian Splattings. 🔹Single-GPU & Fast Inference: As an all-in-one, feed-forward model, Hunyuan World 1.1 runs on a single GPU and delivers all 3D attributes in a single forward pass, within seconds. 🌐Project Page: 🔗Github： 🤗Hugging Face： ✨Demo: 📄Technical Report:

Tencent Hy

168,984 次观看 • 1 年前

We are excited to unveil HunyuanVideo 1.5, the strongest open-source video generation model. Built upon DiT architecture, it redefines the open-source SOTA for accessibility and performance.🚀🚀🚀 HunyuanVideo 1.5 delivers state-of-the-art visual quality and motion coherence while drastically lowering the entry barrier for developers and creators: ⚡️ Unmatched Accessibility: Ultra-light 8.3B parameters, deployable on consumer GPUs with only 14GB VRAM. 🖥️ HD Cinematic Quality: Natively generates 5–10 second 480p/720p HD videos, with super-resolution support for 1080p cinematic quality. By merging SOTA performance with high hardware efficiency, HunyuanVideo 1.5 sets the new technical baseline for the open-source community. 🌐Project Page: 🔗Github: 🤗Hugging Face： 📄Technical Report:

We are excited to unveil HunyuanVideo 1.5, the strongest open-source video generation model. Built upon DiT architecture, it redefines the open-source SOTA for accessibility and performance.🚀🚀🚀 HunyuanVideo 1.5 delivers state-of-the-art visual quality and motion coherence while drastically lowering the entry barrier for developers and creators: ⚡️ Unmatched Accessibility: Ultra-light 8.3B parameters, deployable on consumer GPUs with only 14GB VRAM. 🖥️ HD Cinematic Quality: Natively generates 5–10 second 480p/720p HD videos, with super-resolution support for 1080p cinematic quality. By merging SOTA performance with high hardware efficiency, HunyuanVideo 1.5 sets the new technical baseline for the open-source community. 🌐Project Page: 🔗Github: 🤗Hugging Face： 📄Technical Report:

Tencent Hy

128,008 次观看 • 11 个月前

The first truly open-source audio-video model. LTX-2 is a DiT-based foundation model with all core video generation capabilities in one unified model. Designed to run locally on consumer GPUs. - text-to-video - image-to-video - and video-to-video modes 100% open-source.

The first truly open-source audio-video model. LTX-2 is a DiT-based foundation model with all core video generation capabilities in one unified model. Designed to run locally on consumer GPUs. - text-to-video - image-to-video - and video-to-video modes 100% open-source.

Akshay 🚀

66,268 次观看 • 9 个月前

Introducing Chatterbox Multilingual 🌎 23 languages. One model. Open source. On GitHub , Hugging Face , and Let's dive in 🧵

Introducing Chatterbox Multilingual 🌎 23 languages. One model. Open source. On GitHub , Hugging Face , and Let's dive in 🧵

Resemble AI

24,082 次观看 • 1 年前

We're releasing HY-Embodied-0.5, a family of foundation models for real-world embodied agents. The 2B model is now open source. It strengthens spatial-temporal perception and embodied reasoning for prediction, interaction, and planning. 🤖 The suite includes: 🔹 2B for edge deployment 🔹 32B for complex reasoning Key innovations: 🔹 Mixture-of-Transformers (MoT) architecture for modality-specific computation 🔹 Latent tokens for improved perceptual representation 🔹 Self-evolving post-training 🔹 On-policy distillation from large to small models Across 22 benchmarks, the 2B model outperforms similarly sized SOTA systems on 16 tasks. The 32B model approaches frontier-level performance. 🔗 GitHub: 🤗 Hugging Face:

We're releasing HY-Embodied-0.5, a family of foundation models for real-world embodied agents. The 2B model is now open source. It strengthens spatial-temporal perception and embodied reasoning for prediction, interaction, and planning. 🤖 The suite includes: 🔹 2B for edge deployment 🔹 32B for complex reasoning Key innovations: 🔹 Mixture-of-Transformers (MoT) architecture for modality-specific computation 🔹 Latent tokens for improved perceptual representation 🔹 Self-evolving post-training 🔹 On-policy distillation from large to small models Across 22 benchmarks, the 2B model outperforms similarly sized SOTA systems on 16 tasks. The 32B model approaches frontier-level performance. 🔗 GitHub: 🤗 Hugging Face:

Tencent Hy

34,317 次观看 • 6 个月前