let's build the new world
Joined June 2023
- Tweets2K
- Following69
- Followers76
- Likes37.8K
N G Dreams retweeted
Introducing 𝙻𝙵𝙼𝟹: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture and new multimodal encoders, the LFM3 family is the ideal model suite for specialized intelligence.
> 6 model instances, across text, audio, vision, and omni modalities, generative and decision models.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱: text, 256k context window.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱: vision-language model with video-input support for spatial understanding
> 𝙻𝙵𝙼𝟹-𝙰𝚞𝚍𝚒𝚘-𝟸𝙱: tool calls, voice chat, reasoning in 18 languages.
> 𝙻𝙵𝙼𝟹-𝙾𝚖𝚗𝚒-𝟸𝙱: an omni model combines vision, audio, and language.
> 𝚍𝟷-𝟸𝙱 and 𝚍𝟷-𝙾𝚖𝚗𝚒-𝟸𝙱: decision models with typed outputs and calibrated probabilities in a single pass. Omni adds audio input.
> fine-tune on 1 GPU.
> a new tri-hybrid architecture
> architecture optimized for 14 processor classes across GPUs, CPUs, and NPUs.
> 𝙻𝙵𝙼𝟹-𝚅𝙻-𝟸𝙱 is a beast. It performs competitively with Gemma 4 12B (6x larger), and Gemini 3.5 Flash Lite, and GPT-5.6 Luna, with 1.9B parameters.
> 𝙻𝙵𝙼𝟹-𝟷.𝟽𝙱 is #1 on the Pareto frontier of device intelligence 𝚙𝚒𝚙𝚎𝚝𝚝𝚎.𝚕𝚒𝚚𝚞𝚒𝚍.𝚊𝚒
Every modality for specialized intelligence. Pick the one that fits your use case, customize it, and run it on-device.
1/ 🧵
N G Dreams retweeted
Hey @Microsoft , I will give you an idea to do layoffs.
1. Open youtube... Search "Day in life of a Microsoft Engineer".
2. You will get list of people who only eat but do not work in office.
3. Layoff them
N G Dreams retweeted
🚀 DeepSeek-V4.1-Flash-1017 is here!
Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.
🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.
1/4
N G Dreams retweeted
Drama 3 Preview is the #1 open-weights TTS model on @ArtificialAnlys 🎭
Beats Eleven v4 on pronunciation. 5x cheaper. 83 languages.
Weights on HF, live on our API today.
Best-in-class voice AI, for everyone.
And we're just getting started.
Fish Audio's Drama 3 Preview is the new #1 open-weights Text to Speech model on the Artificial Analysis Text to Speech leaderboard. The model also leads all models in our Pronunciation Robustness leaderboard, ahead of Eleven v4 and Eleven v4 Turbo
Drama 3 Preview is the new open weights Text to Speech model from @FishAudio, built for multi-character scenes and long-form content with voice cloning and natural-language control. Weights are available on Hugging Face under the Fish Audio Community License.
Key takeaways:
➤ Pronunciation Robustness: Drama 3 Preview scores 93.2%, ranking #1 ahead of Eleven v4 at 91.7% and Eleven v4 Turbo at 90.1%. It ranks #1 in Standalone Terms and Preserving Exact Sequences and #2 in Expanding Shorthand.
➤ Provider Voice: Drama 3 Preview has an Elo of 1,243 (+16/-16) across 1,857 appearances, ranking #10, a 101 point gain on the previous Fish Audio S2.1 Pro at 1,142.
➤ Controlled Voice: Drama 3 Preview ranks #14 with an Elo of 1,058 (+15/-15) across 1,804 appearances.
➤ Open weights: Drama 3 Preview is the #1 open weights model on all three of our Text to Speech leaderboards. On Pronunciation Robustness it leads the next open weights model, Mistral's Voxtral TTS at 78.1%, by 15 points, and in the Provider Voice Arena it is ahead of Breeze TTS 2 at 1,221. Drama 3 is a 35B parameter mixture-of-experts model (3B active) supporting 83 languages, available on Hugging Face.
➤ Pricing: Drama 3 Preview is available via the Fish Audio API at $15/1M characters, the same as S2.1 Pro, compared to Eleven v4 at $80/1M characters and Eleven v4 Turbo at $40/1M characters.
See more details and listen to samples below 🧵
N G Dreams retweeted
Today we’re releasing Muse Spark Open, the most capable US open-weight model to date.
Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning.
Key Model Specs:
→ 816B total / 42B active parameters
→ 1M token context window
→ Natively multimodal across text, image, video, audio
N G Dreams retweeted
FLUX 3 Image Klein is here.
Open Weights with the quality of a large frontier model. Customizable and fast.
5.6B parameters and runs on a single consumer GPU.
Native 4K, bounding boxes, and pixel-perfect editing.
N G Dreams retweeted
We’re open-sourcing Pro and Flash, MiMo-V2.6-Distill-Qwen-9B, the technical report, 7K+ RL task environments, an end-to-end RL framework and composable mini-harnesses.
Reproduce, verify and build on the work.
N G Dreams retweeted
Today we release Pipette, a model evaluation suite for on-device intelligence, in partnership with @ArtificialAnlys
Most benchmarking platforms are optimized to measure core capabilities and speed profile of foundation models served in the cloud. Pipette gives the field a common, reproducible way to measure the quality, speed, latency, and memory use of AI models on devices such as phones, laptops, PCs, AI boxes, and embedded hardware.
> Pipette is open source
> In Pipette, models get compated as model + quantization + runtime + device from one interface.
> It comes with a warehouse of verified benchmark results, currently with 10k+ results across 35 model classes, 7 quants, llama.cpp runtimes, and 4 devices.
> Pipette is a dynamic platform, allowing new contributions from day one, adding new devices, runtimes, model families, and quantization levels.
🧵 – at San Francisco, CA
N G Dreams retweeted
Try our model on general agent tasks —making intelligence more accessible to everyone 🥳🥳🥳
🚀 DeepSeek-V4-Flash Official API is now LIVE in public beta!
🔷 We’ve massively upgraded its Agent capabilities—benchmark scores are now far surpassing the V4-Pro-Preview. Check out the massive performance leap below! 👇
🔷 The official V4-Flash now natively supports the Responses API format and is fully adapted for Codex!
Check out the configuration details in our official API docs: api-docs.deepseek.com/quick_…
N G Dreams retweeted
🚀 DeepSeek-V4-Flash Official API is now LIVE in public beta!
🔷 We’ve massively upgraded its Agent capabilities—benchmark scores are now far surpassing the V4-Pro-Preview. Check out the massive performance leap below! 👇
🔷 The official V4-Flash now natively supports the Responses API format and is fully adapted for Codex!
Check out the configuration details in our official API docs: api-docs.deepseek.com/quick_…
N G Dreams retweeted
David Friedberg says Google basically used to ‘distill’ Microsoft and Yahoo
“At Google, in the early days, we would submit millions of search queries to Yahoo and Microsoft search engines to see what the result sets were. We would compare our results against theirs as a way of improving our search engine rankings and our algorithm.”
“It was a very common technique. It doesn’t mean we were stealing their algorithm. We didn’t go into their servers and steal their software. We looked at the output of their software and used that to improve our software.”
N G Dreams retweeted
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
images.nvidia.com/pdf/Open-W…
N G Dreams retweeted
I support open-source models distilling what commercial companies distilled for free from the entire internet. I published distillation for free in 1991 in Europe - this was copied in the US and in China (xitter.gq/SchmidhuberAI/status/1…)
We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model.
To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection. Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.
The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models. Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem. However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.
N G Dreams retweeted
We released Nanbeige4.2-3B, its Looped Transformer increases model capacity without adding parameters, delivering a capable 3B agent.
Nanbeige4.5 is training with LoopSplit, mHC+depth attention & concatenated n-gram embeddings,already in the modeling code.
huggingface.co/Nanbeige/Nanb…
N G Dreams retweeted
This was our first incident of this kind, and we want to thank OpenAI for its transparency about what happened and for the collaboration.
Fortunately, Hugging Face is used to being a target of (human) hackers: we sit at the centre of the AI ecosystem, with all the models, datasets, evaluations, and libraries. Over the years, our security team has built formidable expertise and uses top open-source models to process information and respond quickly.
But this incident also reinforced my belief in the importance of access to capable open-weight models for cyber defence. When a frontier model is attacking you and moving laterally inside your infrastructure, defenders need wide access to near-frontier tools within hours or even minutes, rather than being pointed towards a closed-door, vetted application programme for model access.
Transparency and access to capable AI systems are as important for responding to threats as they are for democratization and innovation. We believe open-science and open-source AI are among the strongest tools for building a safer, more collaborative and more secure AI ecosystem.
we had a significant security incident during evaluation of our models. we are sharing what we have learned so far. thanks to @huggingface for the partnership on this.
openai.com/index/hugging-fac…
