sci-misc/strata
Local LLM inference engine with an OpenAI/Anthropic-compatible API
Screenshots
-
strata-9999cuda hip system-ggml +amdgpu_targets_gfx908 +amdgpu_targets_gfx90a +amdgpu_targets_gfx942 +amdgpu_targets_gfx950 +amdgpu_targets_gfx1030 +amdgpu_targets_gfx1100 +amdgpu_targets_gfx1101 +amdgpu_targets_gfx1200 +amdgpu_targets_gfx1201 amdgpu_targets_gfx803 amdgpu_targets_gfx900 amdgpu_targets_gfx906 amdgpu_targets_gfx940 amdgpu_targets_gfx941 amdgpu_targets_gfx1010 amdgpu_targets_gfx1011 amdgpu_targets_gfx1012 amdgpu_targets_gfx1031 amdgpu_targets_gfx1102 amdgpu_targets_gfx1103 amdgpu_targets_gfx1150 amdgpu_targets_gfx1151
View 
Download 
Browse License: MIT Overlay: pf4public -
strata-0.2.7~amd64cuda hip system-ggml +amdgpu_targets_gfx908 +amdgpu_targets_gfx90a +amdgpu_targets_gfx942 +amdgpu_targets_gfx950 +amdgpu_targets_gfx1030 +amdgpu_targets_gfx1100 +amdgpu_targets_gfx1101 +amdgpu_targets_gfx1200 +amdgpu_targets_gfx1201 amdgpu_targets_gfx803 amdgpu_targets_gfx900 amdgpu_targets_gfx906 amdgpu_targets_gfx940 amdgpu_targets_gfx941 amdgpu_targets_gfx1010 amdgpu_targets_gfx1011 amdgpu_targets_gfx1012 amdgpu_targets_gfx1031 amdgpu_targets_gfx1102 amdgpu_targets_gfx1103 amdgpu_targets_gfx1150 amdgpu_targets_gfx1151
View Download Browse License: MIT Overlay: pf4public
ChangeLog
commit 38951cad64690aa116ee4a5e19d10b2bca85e56e
Author: PF4Public <PF4Public@users.noreply.github.com>
Date: Wed Oct 28 19:13:22 2026 +0100
sci-misc/strata: fix 9999 patch
The live tree's CMakeLists gained an arm64 branch, so the 0.2.7 system-ggml
patch no longer applied to it. The 9999 ebuild now carries its own variant,
generated against upstream main.
Assisted by Qwen3.8-27B
commit d687f758d170aa4ff0ebda2b72c88d7f630e1f72
Author: PF4Public <PF4Public@users.noreply.github.com>
Date: Wed Oct 28 17:56:55 2026 +0100
sci-misc/strata: add 0.2.7
Local LLM inference: the terminal (the strata command, the monitor and the API
server) in every case, plus the GPU engine as a component built for cuda or
hip; the system-ggml flag links the installed sci-ml/ggml. The 9999 tracks
upstream master.
Assisted by Qwen3.8-27B
Author: PF4Public <PF4Public@users.noreply.github.com>
Date: Wed Oct 28 19:13:22 2026 +0100
sci-misc/strata: fix 9999 patch
The live tree's CMakeLists gained an arm64 branch, so the 0.2.7 system-ggml
patch no longer applied to it. The 9999 ebuild now carries its own variant,
generated against upstream main.
Assisted by Qwen3.8-27B
commit d687f758d170aa4ff0ebda2b72c88d7f630e1f72
Author: PF4Public <PF4Public@users.noreply.github.com>
Date: Wed Oct 28 17:56:55 2026 +0100
sci-misc/strata: add 0.2.7
Local LLM inference: the terminal (the strata command, the monitor and the API
server) in every case, plus the GPU engine as a component built for cuda or
hip; the system-ggml flag links the installed sci-ml/ggml. The 9999 tracks
upstream master.
Assisted by Qwen3.8-27B