From 55ac2e5568fd4115f2bdcbfebde6fbad3e43f7ca Mon Sep 17 00:00:00 2001
From: Davide Polonio <poloniodavide@gmail.com>
Date: Fri, 10 Apr 2026 15:42:19 +0200
Subject: [PATCH] feat(llama-swap): fix contexts, add flash-attn/jinja, tune
 sampling params

---
 ollama/llama-swap.yaml | 26 ++++++++++++++++++++------
 1 file changed, 20 insertions(+), 6 deletions(-)

diff --git a/ollama/llama-swap.yaml b/ollama/llama-swap.yaml
index f80b357..95f8085 100644
--- a/ollama/llama-swap.yaml
+++ b/ollama/llama-swap.yaml
@@ -11,18 +11,23 @@ models:
       --alias qwen3.5:9b
       --n-gpu-layers 999
       --ctx-size 8192
-      --temp 1 --top-k 20 --top-p 0.95 --presence-penalty 1.5
+      --flash-attn
+      --jinja
+      --temp 0.7 --top-p 0.9
 
-  "qwen3.5:9bctxSmall":
+  "qwen3.5:9b-32k":
     cmd: |
       /app/llama-server
       --host 0.0.0.0 --port ${PORT}
       --model /models/Qwen3.5-9B-Q4_K_M.gguf
       --mmproj /models/Qwen3.5-9B-mmproj-F16.gguf
-      --alias qwen3.5:9bctxSmall
+      --alias qwen3.5:9b-32k
       --n-gpu-layers 999
-      --ctx-size 131072
-      --temp 1 --top-k 20 --top-p 0.95 --presence-penalty 1.5
+      --ctx-size 32768
+      --flash-attn
+      --jinja
+      --cache-type-k q8_0 --cache-type-v q8_0
+      --temp 0.7 --top-p 0.9
 
   "qwen3.5:9b-uncensored":
     cmd: |
@@ -33,6 +38,10 @@ models:
       --alias "hf.co/HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive:q4_k_m"
       --n-gpu-layers 999
       --ctx-size 32768
+      --flash-attn
+      --jinja
+      --cache-type-k q8_0 --cache-type-v q8_0
+      --temp 0.7 --top-p 0.9
 
   "qwen3.5:9b-claude-4.6-opus-reasoning":
     cmd: |
@@ -43,13 +52,17 @@ models:
       --alias "hf.co/Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2-GGUF:q4_k_m"
       --n-gpu-layers 999
       --ctx-size 32768
-      --temp 0.6 --top-k 20 --top-p 0.95 --repeat-penalty 1
+      --flash-attn
+      --jinja
+      --cache-type-k q8_0 --cache-type-v q8_0
+      --temp 0.6 --top-p 0.95
 
   "gemma4:e4b-uncensored":
     cmd: |
       /app/llama-server
       --host 0.0.0.0 --port ${PORT}
       --jinja
+      --flash-attn
       --model /models/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive-Q8_K_P.gguf
       --mmproj /models/mmproj-Gemma-4-E4B-Uncensored-HauhauCS-Aggressive-f16.gguf
       --alias "hf.co/HauhauCS/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive:q8_k_p"
@@ -61,6 +74,7 @@ models:
       /app/llama-server
       --host 0.0.0.0 --port ${PORT}
       --jinja
+      --flash-attn
       --model /models/gemma-4-26B-A4B-it-UD-IQ4_NL.gguf
       --mmproj /models/gemma-4-26B-A4B-it-UD-IQ4_NL-mmproj-BF16.gguf
       --alias "hf.co/unsloth/gemma-4-26B-A4B-it:ud-iq4_nl"