polpetta/desktop-dotfiles
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(llama-swap): use --flash-attn on (flag requires explicit value)

Browse Source
This commit is contained in:
Davide Polonio 2026-04-10 15:50:25 +02:00
parent 6ea3e870bd
commit df3b927985
1 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
ollama/llama-swap.yaml
View File

@ -11,7 +11,7 @@ models:
--alias qwen3.5:9b --alias qwen3.5:9b
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 8192 --ctx-size 8192
--flash-attn --flash-attn on
--jinja --jinja
--temp 0.7 --top-p 0.9 --temp 0.7 --top-p 0.9
@ -24,7 +24,7 @@ models:
--alias qwen3.5:9b-32k --alias qwen3.5:9b-32k
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 32768 --ctx-size 32768
--flash-attn --flash-attn on
--jinja --jinja
--cache-type-k q8_0 --cache-type-k q8_0
--cache-type-v q8_0 --cache-type-v q8_0
@ -39,7 +39,7 @@ models:
--alias "hf.co/HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive:q4_k_m" --alias "hf.co/HauhauCS/Qwen3.5-9B-Uncensored-HauhauCS-Aggressive:q4_k_m"
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 32768 --ctx-size 32768
--flash-attn --flash-attn on
--jinja --jinja
--cache-type-k q8_0 --cache-type-k q8_0
--cache-type-v q8_0 --cache-type-v q8_0
@ -54,7 +54,7 @@ models:
--alias "hf.co/Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2-GGUF:q4_k_m" --alias "hf.co/Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2-GGUF:q4_k_m"
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 32768 --ctx-size 32768
--flash-attn --flash-attn on
--jinja --jinja
--cache-type-k q8_0 --cache-type-k q8_0
--cache-type-v q8_0 --cache-type-v q8_0
@ -69,7 +69,7 @@ models:
--alias "hf.co/HauhauCS/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive:q8_k_p" --alias "hf.co/HauhauCS/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive:q8_k_p"
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 8192 --ctx-size 8192
--flash-attn --flash-attn on
--jinja --jinja
"gemma4:26b-a4b": "gemma4:26b-a4b":
@ -81,5 +81,5 @@ models:
--alias "hf.co/unsloth/gemma-4-26B-A4B-it:ud-iq4_nl" --alias "hf.co/unsloth/gemma-4-26B-A4B-it:ud-iq4_nl"
--n-gpu-layers 999 --n-gpu-layers 999
--ctx-size 8192 --ctx-size 8192
--flash-attn --flash-attn on
--jinja --jinja