straylight: add Ornith 1.5 35B-A3B MoE to llama-server

Official ornith-ai Q4_K_M GGUF (~22 GB). Multimodal 35B MoE with ~3B active; llama-server fetches weights and mmproj on first load.
2026-09-16 05:56:27 -07:00
parent 5a734118b0
commit 85d216fce3
1 changed files with 6 additions and 0 deletions
@@ -56,6 +56,12 @@ let
    hf-repo = mradermacher/Ornith-1.5-9B-uncensored-GGUF:Q4_K_M
    dedup-cache-models = true

+    ; Ornith 1.5 35B-A3B MoE (official GGUF). ~3B active, multimodal; mmproj
+    ; comes with --hf-repo. Q4_K_M is ~22 GB.
+    [ornith-1.5-35b-a3b]
+    hf-repo = ornith-ai/Ornith-1.5-35B-A3B-GGUF:Q4_K_M
+    dedup-cache-models = true
+
    ; Qwen 3.8 27B dense, refusal-ablated. Preferred uncensored coding agent.
    ; orcarouter/Qwen3.8-27B-Uncensored-GGUF is gated; huihui Q4_K_L keeps
    ; ablation-target tensors at Q8_0. llama-server fetches on first load.