Optimum-Intel v2.2.0 与 OpenVINO GenAI 2026.4.0 更新

AI 摘要

Hugging Face 博客发布 Optimum-Intel v2.2.0 与 OpenVINO GenAI 2026.4.0 更新。Optimum Intel 2.2 扩展了模型导出支持，新增 Mistral 3、DeepSeek-OCR-2、Gemma 4、Qwen-Image、LTX-2、Fun-ASR 等模型，并支持 DFlash 与 MTP 投机解码。OpenVINO GenAI 2026.4 新增图像生成与语音识别流水线，扩展 VLMPipeline，为 Mamba 2 层引入 Paged Attention，并改进可观测性与 Node.js API。

正文节选

These libraries cover different steps in deploying a model: - Optimum Intel exports your model to the OpenVINO format - NNCF applies quantization on your converted model, to help lower memory and compute requirements - OpenVINO executes the optimized model efficiently on Intel CPUs, GPUs, and NPUs - OpenVINO GenAI provides APIs for generative AI tasks, including text generation, visual understanding, image generation, and speech recognition Together, they provide a complete workflow, taking a mo

发布时间：—

抓取时间：2026-09-24 02:35

来源机构：Hugging Face