https://wiki.nixos.org/w/index.php?action=history&feed=atom&title=Talk%3ALlama-cpp 2026-05-22T21:37:26Z Revision history for this page on the wiki MediaWiki 1.45.0 https://wiki.nixos.org/w/index.php?title=Talk:Llama-cpp&diff=31925&oldid=prev 2026-05-19T12:36:04Z

<p><span class="autocomment">Review the AMD ROCm section: </span> new section</p> <p><b>New page</b></p><div>== Review the AMD ROCm section ==<br /> <br /> I wrote from memory, and I&#039;m not entirely sure my knowledge is accurate. When changing the env variable, I&#039;ve noticed performance changes in llama-cpp. But again, I&#039;m not sure. Could someone verify what I wrote? <br /> <br /> I also wonder if one should compile llama-cpp on their own machine with rocmGpuTargets ?? <br /> <br /> Thanks [[User:Woile|Woile]] ([[User talk:Woile|talk]]) 12:36, 19 May 2026 (UTC)</div>

Woile