ModelsSelected

Amazon launches Strands Decider 2B model

亚马逊推出 Strands Decider 2B 开源决策模型,支持本地部署

Why we picked it

This model enables local AI decision-making without cloud dependency, addressing privacy concerns for sensitive applications. Developers can implement it for real-time decision tasks in edge computing environments, with performance competitive to larger models while requiring minimal resources.

Amazon's Strands Agents team released Strands Decider 2B on 1 Oct 2024.

Body

The decision model is open-source on GitHub with weights available on Hugging Face.It runs on local CPU/GPU hardware and achieves rank 3 on JevBench dataset among 2B-level models.

The model shows 113ms median decision latency on common hardware and 153ms on NVIDIA GeForce RTX 3090.

Original · IT之家

亚马逊推出 Strands Decider 2B 开源决策模型,支持本地部署

IT之家 10 月 3 日消息,Amazon(亚马逊)Strands Agents 团队当地时间本月 1 日宣布推出 Strands Decider 2B 决策模型。其已在 GitHub 上开源,权重则可在 Hugging Face 上找到; 可在本地的 CPU / GPU 上运行 。 Strands Decider 2B 基于预训练的 Qwen3.5-2B“躯干”, 以具备评分功能的指针“头部”替换拥有文本生成能力的原有语言模型“头部” 。新“头部”规模较小,总参数仅略超过一百万;“躯干”则通过一个 rank-16 LoRA 适配器进行微调。 Strands Decider 2B 在 JevBench 公开数据集上 具有不错的准确率和校准度表现 。其在 2B 级别模型中排名第 3,优于所有严格不超过 2B 的竞争对手。 在市面常见硬件上本地运行该模型时,Strands Decider 2B 交出了 113ms 的决策时延中位数 。其在 NVIDIA GeForce RTX 3090 上执行小型决策任务的中位数时延是 153ms。

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Strands Decider 2B, Amazon, Qwen3.5-2B, open-source models.

Source attribution: curated from IT之家. Cross-source corroboration: 2 independent sources have covered this development (including marktechpost), confirming industry-wide signal strength.