6b4fc91ca7632d16bf11b02650bb91d2d6a262cc

 1diff --git a/dot_config/llama-server/models.ini b/dot_config/llama-server/models.ini
 2index 869982cea7feeda8ceee2acde014afe3cb178e9a..dc201d241d60eec841964cc43fbdc97ee9a6a7f0 100644
 3--- a/dot_config/llama-server/models.ini
 4+++ b/dot_config/llama-server/models.ini
 5@@ -31,25 +31,6 @@ presence-penalty = 1.5
 6 reasoning = on
 7 chat-template-kwargs = {"reasoning_effort":"low"}
 8 
 9-[unsloth/gemma-4-12B-it-qat-GGUF:UD-Q4_K_XL]
10-hf = unsloth/gemma-4-12B-it-qat-GGUF:UD-Q4_K_XL
11-alias = gemma4-12b,jp
12-n-gpu-layers = 99
13-n-gpu-layers-draft = 99
14-fit = off
15-flash-attn = on
16-np = 2
17-kv-unified = on
18-ctx-size = 32768
19-batch-size = 1024
20-ubatch-size = 1024
21-temp = 1.0
22-top-p = 0.95
23-top-k = 64
24-spec-type = draft-mtp
25-spec-draft-n-max = 4
26-reasoning = off
27-
28 [unsloth/gemma-4-26B-A4B-it-qat-GGUF:Q4_K_XL]
29 hf = unsloth/gemma-4-26B-A4B-it-qat-GGUF:UD-Q4_K_XL
30 alias = gemma4-26b-a4b