Granite Switch 4.1：单端点实现聊天、决策与语音

AI 摘要

IBM 在 Hugging Face 博客发布 Granite Switch 4.1（8B 与 30B），将通用决策 aLoRA、安全决策 Guardian aLoRA 以及 OpenASR 榜单上最快的 ASR 模型 Granite-Speech-5.0-Turbo-CTC 组合进单一托管端点，显存占用与基础模型相当。由于采用缓存高效的 Activated LoRA，基础模型的 KV cache 可在后续适配器决策中复用，配合 vLLM 低秩适配推理优化，模型能在聊天、决策与语音转写请求间快速切换。官方称 30B 版本在 5 步 RAG 工作流中快于 Jev-27B，并计划发布 Adapter SDK 供社区自定义组合适配器。

正文节选

Introduced last spring, the Granite Switch project enables a modular approach to model development. Now Granite Switch 4.1 models, armed with a Decision adapter and the fastest ASR model on the OpenASR leaderboard, can move efficiently between natural-language responses, general and specialized decisions, and speech-to-text tasks—all from one hosted endpoint. Specifically, Granite-Switch-4.1-8B and -30B models are configured to include: - A new Decision aLoRA for general-purpose decision-making

发布时间：—

抓取时间：2026-10-21 23:05

来源机构：Hugging Face