Ollama has released version v0.40.17-rc2, which includes an update to mlxrunner. This update adapts saved history to better manage memory pressure by retaining Metal prefix history based on observed memory headroom and execution workspace, rather than a fixed allowance. The changes aim to release unused buffers and cold history when system compression increases, while preserving reusable paths and active requests. AI
IMPACT This update improves memory management for Ollama, potentially enhancing its performance and stability for users running AI models locally.
RANK_REASON This is a release of a specific version of a software tool, not a frontier model release or significant industry event.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →