接入 API · 个人 AI 解读连接自己的模型解读资讯,浏览新闻无需配置。
返回资讯列表
算力与芯片官方来源国际

ONNX Runtime 发布: ONNX Runtime v1.28.1

ONNX Runtime 发布 · 来源更新
今日摘要使用自己的 API,仅供个人查看

来源摘要

This is a patch release on top of [v1.28.0](https://github.com/microsoft/onnxruntime/releases/tag/v1.28.0), containing support for device-free WebGPU compilation, improved compatibility with sandboxed Windows processes, and targeted graph-validation fixes. ## WebGPU EP - Added support for device-free compile-only sessions, enabling offline graph transformation and optimized-model serialization without access to GPU hardware ([#29681](https://github.com/microsoft/onnxruntime/pull/29681)) ## Bug Fixes - Prevented an access violation in Windows processes under Win32k lockdown by skipping DXGI device discovery ([#29755](https://github.com/microsoft/onnxruntime/pull/29755)) - Allowed zero-input `EPContext` nodes, aligning their schema with support for compiling zero-input models ([#29799](https://github.com/microsoft/onnxruntime/pull/29799)) - Hardened FastGelu fusion to skip malformed `Mul` and `Pow` patterns ([#32016](https://github.com/microsoft/onnxruntime/pull/32016)) - Added validation for in-memory external initializer references, rejecting unregistered or mismatched data before graph transformation ([#32042](https://github.com/microsoft/onnxruntime/pull/32042)) ## Contributors Thanks to our 4 contributors for this release! [@apsonawane](https://github.com/apsonawane), [@shiyi9801](https://github.com/shiyi9801), [@adrastogi](https://github.com/adrastogi), [@mingmingtasd](https://github.com/mingmingtasd) Full Changelog: [v1.28.0...v1.28.1](https://github.com/microsoft/onnxruntime/compare/v1.28.0...v1.28.1)

阅读原始来源
来源
ONNX Runtime 发布 · 官方来源
来源发布
2026/08/19 07:19
来源更新
2026/08/19 07:29
首次采集
2026/09/19 13:21

本文为公开信息索引与摘要,详情及后续变化请以原始来源为准。

把 AI 雷达放到桌面

在支持安装的浏览器中,可以将本站作为应用打开。

安装入口取决于浏览器;应用和网站使用同一份最新内容。

查看完整安装指南