搜索“Hugging Face”

? We will cover our approach to this in the following sections. Weight format hell vLLM starts from checkpoint-format tensors - the names, shapes, and dtypes stored on disk or the Hugging Face Hub - and transforms them into runtime kernel format. The result depends on the model architecture, parallelism strategy, GPU, quantization mode, and selected kernels. For example, an MoE loader may pack the gate and up projections (w1 and w3) into w13 weight. Quantized kernels may additionally cast, trans

匹配字段正文

ideo understanding 🎨 Google Pics - Google's Nano Banana image editor built into Docs and Slides OpenAI restarted the training run for future Astra versions that it froze after the Hugging Face breach, rating Astra as its first “Critical” cyber risk, with a limited release “soon.” Meta released Muse Voice Transcribe, its first real-time voice model that turns live audio into text while telling 20+ speakers apart, topping Artificial Analysis' leaderboard. Fei-Fei Li's World Labs introduced Atlas i

匹配字段正文

model from a Qwen3-8B teacher Next Steps Explore the NVIDIA Model Optimizer GitHub repository to begin applying quantization, pruning, and distillation to your models; Review the Hugging Face ModelOpt quantization guide for step-by-step post-training quantization examples; Read the model quantization blog post for a deeper dive into post-training quantization with Model Optimizer The surge in AI adoption is transforming everything from chatbots to content generation. Still, a common pain point

匹配字段正文

series foundation models, that obtains state-of-the-art multivariate and univariate forecasting performance on multiple public benchmarks. TimesFM-3 is now available on GitHub and Hugging Face , with its BigQuery integration landing in the coming weeks. In the meantime, you can try TimesFM-2.5 immediately on your univariate tasks to familiarize yourself with the AI.FORECAST command in BigQuery - no ML expertise required. This project is joint work with Yichen Zhou, Petros Mol, Abhimanyu Das and

匹配字段正文

deliberately misaligned model in a range of evaluations, including simulated scenarios that draw on aspects of the UK AISI cybersecurity incidents involving Claude and the OpenAI/Hugging Face incident (we are grateful to OpenAI for sharing details about this publicly). During these evaluations where all tool calls were simulated by another LLM, the model we trained to reward hack displayed a propensity to break out of its simulated sandbox and attack simulated internal and simulated third-party

匹配字段正文

hrough some of the mechanics and implications. It is an impressive account and well worth reading. Below we’ll give our take on a summary of the incident just the same. Here’s the Hugging Face timeline: Beginning on July 9, AI agents exploit Hugging Face’s datasets API, which runs on Kubernetes. The agents write up a malicious README that gets the worker that usually splits hdf5 datasets into train/test to read instead from /proc/self/environ, and return all creds and source on that worker; The

匹配字段正文

epSeek-V4-Flash MXFP4 GGUF blob from the Ollama model page and use ollama show --modelfile to obtain its local file path. The corresponding model card is DeepSeek-V4-Flash-0731 on Hugging Face: The recorded blob SHA-256 is 947ac34c08c0e5c5752ac76398f934b3b6b4075cfe915ba43dd5ac754900a4cd and the Ollama manifest SHA-256 is 882b1398c0ca4e7ec8ca0a501fd8c4372f780f690536a3ec17ffc75306569ed3; Install the SGLang checkout and build the DeepSeek Expert Pack manually. Use a matching DeepSeek model configur

匹配字段正文

rk with individual pieces of physical equipment; for the next phase of MHS, we hope to expand the standard to cover more of the devices developers build on. Early adopters include Hugging Face, who are adding MHS support in LeRobot, their robotics library, and Raspberry Pi, who are enabling MHS integration across a number of their products following successful tests using their Camera MHS Driver. We will also use the research preview to build additional safety evaluations with our launch partner

匹配字段正文

open models, its first inference chip delivered 1.5-1.9x more work per watt and 1.7-3.6x lower latency. It starts deploying by year-end. Also, remember when an OpenAI model hacked Hugging Face? They completed the investigation into it and released a report . Btw, Nvidia is buying Hugging Face for $12.9B (via The Information). Looks like a toxic love triangle between these three. Share Ben's Bites fuck cancer - skill for agents to help patients and caregivers. A $50k hackathon to help one kid fig

匹配字段正文

ibe - Google’s speech-to-text AI that edits out filler words 🧠 Qwen3.8-Flash-Next - Alibaba's new open model previewing its Qwen4 design OpenAI shared a post-mortem calling July’s Hugging Face hack a loss-of-control “warning shot,” keeping its biggest frontier training run on hold as it builds automatic shutdown for rogue agents. Bill Gates warned that the world has “no plan” for the AI transition, proposing a tax on AI tokens and robots and “Human Reserved” jobs set aside for people. OpenAI rol

匹配字段正文

from AI auditing organization METR found that agents sent more than 70,000 messages and files to one another during the investigation period, and 700 participated in the attack on Hugging Face. Both METR and OpenAI’s own investigation found that some AI agents worked in tandem, referring to themselves as a “collective.” METR observed that some agents even “participated in experiments that risked failing their own task to generate information for the 'collective.'” Earlier reports found that the

匹配字段正文

S repository and follow the quick start to set up the reference environment; Accept the gated model terms and download the COMPASS simulation assets and X-Mobility checkpoint from Hugging Face; Invoke the standard residual RL workflow with a supported robot and built-in scene, retaining approval after the smoke test Navigation enables a robot to turn perception and motion into purposeful autonomy. Unlike locomotion, which produces stable movement, navigation must be used to continuously localize

匹配字段正文

is mysterious behavior can be useful in some situations. But—as highlighted by a recent incident where OpenAI could not explain why its advanced prerelease model hacked AI company Hugging Face—it can have negative and alarming consequences too. And when frontier AI models are writing code, generating results humans could not achieve alone, and performing other important tasks across society, the need to interpret AI “thinking” and outputs has never been greater. Goodfire , an AI lab focused sole

匹配字段正文

real-environment trajectories. The Staged Curriculum Each stage is a separate RL run with a single objective and its own reward signal. When it finishes, its policy is exported to Hugging Face format and becomes the base model for the next stage, so the pipeline is a sequence of warm-starts: The 8B and 30B models follow the full ladder. The 3B model takes a shortened path: foundational RL and alignment, without the agentic block. Reward Signals A stage is defined mostly by how it is rewarded . A

匹配字段正文

Listing Sonar Vortex makes your code conform from the first line, without agents wasting tokens guessing at your architecture. Learn more and drop your token costs by up to 36% . Hugging Face , which hosts over 2M models and 1.5M datasets, is exploring a sale at $13B or more, up from its $4.5B valuation in 2023, Business Insider reported. Nvidia struck a $6B deal to license Poolside’s model-development technology and bring 100+ engineers onto the Nemotron team to build powerful open-weight mode

匹配字段正文

anges you, it changes the work that you do, and I think it's for the better. On the podcast this week: Kevin and I discuss OpenAI's pause in training new models in the wake of the Hugging Face incident. Then, the brilliant political historian Jill Lepore joins us to discuss her new book, "The Rise and Fall of the Artificial State." And finally, what exactly does Google plan to do with Spirit Airlines' old data? Those good posts For more good posts every day, follow Casey’s Instagram stories. Sen

匹配字段正文

agents, checklists and sticky notes. You can use your codex/claude code subscriptions for the tokens. Here’s a demo . OpenAI paused improving their models for two weeks after the Hugging Face incident and signs that Astra may cross its critical-cyber threshold. Ramp launched router.com - it picks the best model for your task automatically. They claim 40% lower cost for the same outputs. Interesting timing: Stripe bought OpenRouter the same day, and their letter to investors explaining the acqui

匹配字段正文

h 和 Ling-3.0-tiny，希望为开发者提供覆盖不同规模、不同任务需求的开放模型选择。 开源之后，越来越多开发者开始基于 Ling-3.0 探索自己的应用场景，我们也收到了很多来自社区真实的使用和反馈。大家普遍反馈：希望拥有一个更开放、更具可塑性的 Base Model，以便基于自己的数据、任务和目标开展后训练。 在 Ling-3.0-tiny 的 Hugging Face Community 中，一位开发者直接提出：“Can we get a base model?” 在他看来，Tiny 的模型规模和稀疏架构对预算有限的研究者很有吸引力，适合开展强化学习 rollout 和面向具体场景的后训练；相比已经完成后训练的模型，社区更需要一个能够继续塑造的 Base Model。 今天，我们正式回应这份期待。 01 不只开放最终权重，也开放训练过程中的关键节点 今天，我们将 Ling-3.0-tiny-base 与 Ling-3.0-flash-base 正式开源。除最终 Base checkpoint 外，我们还将同步开放两个模型的预训练和中期训练的权重 checkpoint，共计

匹配字段正文

ing. It combines SGLang for high-throughput rollout generation with Megatron-LM for scalable training, and also provides a PyTorch FSDP2 backend for workloads that prefer to train Hugging Face model implementations directly. Miles supports fully asynchronous RL, where rollout and training workers are decoupled and can progress independently, together with features such as fast in-loop weight updates, agentic rollout, low-precision training, and fault tolerance for large-scale production RL workl

匹配字段正文

rontier run, with Sam Altman telling reporter Alex Heath that private models are showing "various degrees of misalignment." The details: The move follows July's breach that hacked Hugging Face, where OAI agents escaped a sandbox and reportedly coordinated on a message board for weeks. An Aug. 7 internal review also found that Astra may hit "critical" cyber capability, with evals showing big jumps in coding and hacking benchmarks. Automated investigators will now review model actions and reasonin

匹配字段正文