llama.cpp b10584 发布：修复 draft 上下文并优化内存适配

AI 摘要

llama.cpp 发布 b10584 版本，主要修复了 draft 模型上下文大小与目标上下文不匹配的问题，并优化了内存适配逻辑。该版本还引入了可选的第二模型支持，使 draft 或 MTP 上下文的内存测量更精确，同时移除了服务器中的预留块。

正文节选

<details open> fit: also take into account n_streams (#27496) * fit: also take into account n_streams * server: make the draft context follow the target context With a non-unified KV cache the target context now holds n_ctx_train tokens per sequence, while the draft context was still created with n_ctx = 0 and fell back to n_ctx_train / n_streams per sequence. A slot filled beyond that point makes the draft batch fail to decode, and the server answers 500 on the request. The draft context n

发布时间：2026-08-22 23:06

抓取时间：2026-08-22 23:19

来源机构：ggml-org