PulseAugur
中
实时 16:53:16
Italiano(IT) 👀 FreeToken, il nuovo motore open source di UC Berkeley per far girare modelli MoE in locale su GPU consumer. FreeToken tratta il computer come un unico insieme

FreeToken使大型AI模型能在消费级GPU上运行 · 已追踪4个来源

FreeToken是加州大学伯克利分校开发的一款开源引擎,它允许像拥有7530亿参数的GLM-5.2这样的大型专家混合(MoE)模型在消费级硬件上运行。它通过将个人计算机视为一个统一的弹性推理平台,动态地将计算映射到GPU、CPU和内存上,从而实现这一点。这项创新为在边缘部署前沿规模的AI模型开辟了新的可能性。 AI

影响 使大型AI模型能够在消费级硬件上部署,有可能使前沿AI能力的获取大众化。

排序理由 发布了一款用于在消费级硬件上运行大型AI模型的新型开源引擎。

在 Mastodon — fosstodon.org 阅读 →

AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →

FreeToken使大型AI模型能在消费级GPU上运行 · 已追踪4个来源

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
发布了一款用于在消费级硬件上运行大型AI模型的新型开源引擎。
Source corroboration
4 independent sources
Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics
infra, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
46 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [4]

  1. Mastodon — fosstodon.org TIER_1 Italiano(IT) · [email protected] ·

    👀 FreeToken,加州大学伯克利分校新开源引擎,可在消费级GPU上本地运行MoE模型。FreeToken将计算机视为一个单一的集合

    👀 FreeToken, il nuovo motore open source di UC Berkeley per far girare modelli MoE in locale su GPU consumer. FreeToken tratta il computer come un unico insieme elastico di risorse 👇 https:// gomoot.com/freetoken-il-motore -di-uc-berkeley-per-far-girare-modelli-moe-in-locale-su-g…

  2. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    📋 FreeToken 是一款边缘原生 MoE 服务引擎,可在个人消费级硬件上运行前沿规模的开源模型 #opensource #LLM #AI 🧵👇

    📋 # FreeToken is an edge-native # MoE serving engine for running frontier-scale open-weight models on personal, consumer hardware # opensource # LLM # AI 🧵👇

  3. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    FreeToken 使得在单个 GPU 上运行像 GLM-5.2 这样拥有 753B 参数的模型成为可能,为边缘 AI 部署开启了新的可能性。https://www.marktechpo

    FreeToken makes it possible to run 753B parameter models like GLM-5.2 on a single GPU, opening new possibilities for edge AI deployment. https://www. marktechpost.com/2026/08/23/me et-freetoken-an-edge-native-moe-serving-engine-that-runs-753b-glm-5-2-on-a-single-workstation-gpu/ …

  4. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    FreeToken 使一个拥有 7530 亿参数的模型能在单个工作站 GPU 上运行,将个人机器视为统一的弹性推理平台。s

    FreeToken enables a 753 billion parameter model to run on a single workstation GPU by treating a personal machine as a unified elastic inference platform. The system dynamically maps computation across GPU, CPU and memory. https://www. marktechpost.com/2026/08/23/me et-freetoken-…