返回全部动态
llama.cpp b10594 发布：修复 CUDA 显存浪费问题
原标题：b10594
AI 摘要
llama.cpp 发布 b10594 版本，主要修复了设备信息循环在非 TRACE 日志级别下仍会创建 GPU 上下文并导致 CUDA 后端分配 550 MB 显存的问题。该修复通过检查日志详细级别，在无输出时跳过设备循环，避免不必要的 GPU 资源占用。
以上摘要由 AI 生成，可能存在误差。事实请以原文为准。
正文节选
<details open> common : skip device_info loop if it's not going to be printed (#26692) The device_info loop iterates over the discovered devices and gets the available and total memory counts. With the CUDA backend (and possibly others too) this requires creating a GPU context, which, in case of CUDA, results in a 550 MB VRAM allocation. For this information to be used in any way, the log verbosity must be set to LOG_LEVEL_TRACE. If it's not, including in the default configuration, the contex
发布时间：2026-08-23 21:21
抓取时间：2026-08-23 22:19
来源机构：ggml-org