Diff
1diff --git a/dot_config/llama-server/models.ini b/dot_config/llama-server/models.ini
2index f8b5b4668f85f178791e35d5cb2be8fe2f05d219..115b59e1fedf6db8c7667cee1ddc9b1c8cc7bd27 100644
3--- a/dot_config/llama-server/models.ini
4+++ b/dot_config/llama-server/models.ini
5@@ -13,9 +13,9 @@ embedding = true
6
7 [unsloth/Qwen3.8-27B-GGUF:UD-IQ3_S]
8 hf = unsloth/Qwen3.8-27B-GGUF:UD-IQ3_S
9-alias = qwen3.8-27b-q3,reviewer
10-n-gpu-layers = auto
11-n-gpu-layers-draft = 99
12+alias = qwen3.8-27b-q3
13+n-gpu-layers = 999
14+n-gpu-layers-draft = 999
15 fit = off
16 flash-attn = on
17 jinja = true
18@@ -34,7 +34,7 @@ chat-template-kwargs = {"reasoning_effort":"medium"}
19
20 [ukisai/Swift-1.5-Qwen3.8-27B-GGUF:Q3_K_S]
21 hf = ukisai/Swift-1.5-Qwen3.8-27B-GGUF:Q3_K_S
22-alias = swift-1.5-27b-q3
23+alias = swift-1.5-27b-q3,reviewer
24 n-gpu-layers = auto
25 n-gpu-layers-draft = 99
26 fit = off
27@@ -53,6 +53,27 @@ presence-penalty = 1.5
28 reasoning = on
29 chat-template-kwargs = {"reasoning_effort":"medium"}
30
31+[ornith-ai/Ornith-1.5-9B-GGUF:Q6_K]
32+hf = ornith-ai/Ornith-1.5-9B-GGUF:Q6_K
33+alias = ornith-1.5-9B-q6,classifier
34+n-gpu-layers = 99
35+n-gpu-layers-draft = 99
36+fit = off
37+flash-attn = on
38+jinja = true
39+ctx-size = 8192
40+temp = 0.2
41+top-p = 0.95
42+top-k = 20
43+min-p = 0.0
44+parallel = 1
45+spec-type = draft-mtp
46+spec-draft-n-max = 2
47+repeat-penalty = 1.0
48+presence-penalty = 1.5
49+reasoning = on
50+chat-template-kwargs = {"reasoning_effort":"low"}
51+
52 [unsloth/gemma-4-26B-A4B-it-qat-GGUF:Q4_K_XL]
53 hf = unsloth/gemma-4-26B-A4B-it-qat-GGUF:UD-Q4_K_XL
54 alias = gemma4-26b-a4b