[*] # paramètres globaux (par défaut) #ctx-size = 4096 #threads = 8 mmap = true #cache-type-k = q8_0 #cache-type-v = q8_0 cache-type-k = q4_0 cache-type-v = q4_0 n-gpu-layers = 99 mlock = true batch-size = 768 ubatch-size = 512 [Master-Of-Experts] model = /models2/Qwen3.6-35B-A3B-UD-Q4_K_XL.gguf mmproj = /models2/qwen3.6-mmproj-F16.gguf ctx-size = 262144 #ctx-size = 131072 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-Hermes] model = /models2/Carnice-9b-Q6_K.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf ctx-size = 262144 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-DeepSeek-V4-Inventif] model = /models2/Qwen3.5-9B-DeepSeek-V4-Flash-Q8_0.gguf mmproj = /models2/Qwen3.5-9B-DeepSeek-V4-Flash-mmproj.gguf ctx-size = 262144 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-DeepSeek-V4-Coder] model = /models2/Qwen3.5-9B-DeepSeek-V4-Flash-Q8_0.gguf mmproj = /models2/Qwen3.5-9B-DeepSeek-V4-Flash-mmproj.gguf ctx-size = 262144 temperature=0.7 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-General] model = /models2/Carnice-9b-Q6_K.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf ctx-size = 262144 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.6-35B-A3B-Q4-dualgpu] model = /models2/Qwen3.6-35B-A3B-UD-Q4_K_XL.gguf mmproj = /models2/qwen3.6-mmproj-F16.gguf ctx-size = 262144 #ctx-size = 131072 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-Code] model = /models2/Qwen3.5-9B-Q8_0.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf ctx-size = 131072 temperature=0.6 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=0.0 repeat-penalty=1.0 [Gemma-4-E4B-General] model = /models2/gemma-4-E4B-it-Q6_K.gguf #model = /models2/gemma-4-E4B-it-UD-Q8_K_XL.gguf mmproj = /models2/mmproj-F16-gemma-4-E4B-it.gguf ctx-size = 131072 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 reasoning=off [Qwen3.5-9B-General] model = /models2/Carnice-9b-Q6_K.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf ctx-size = 131072 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen-3.6-27B] model = /models2/Qwen3.6-27B-UD-Q3_K_XL.gguf ctx-size = 131072 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 1 [Qwen3.5-9B-General-old] model = /models2/Qwen3.5-9B-UD-Q6_K_XL.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf ctx-size = 262144 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 parallel = 2 [Qwen3.5-9B-Instruct-General] model = /models2/Qwen3.5-9B-Q8_0.gguf mmproj = /models2/mmproj-F16-Qwen3.5-9B.gguf reasoning = off ctx-size = 131072 temperature=0.7 top-p=0.8 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 [Qwen3.5-35B-A3B-General] model = /models2/Qwen3.5-35B-A3B-UD-Q4_K_L.gguf mmproj = /models2/mmproj-F16-Qwen3.5-35B-A3B.gguf ctx-size = 65536 batch-size = 2048 ubatch-size = 512 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 [Qwen3.5-35B-A3B-General-Instruct] model = /models2/Qwen3.5-35B-A3B-UD-Q4_K_L.gguf mmproj = /models2/mmproj-F16-Qwen3.5-35B-A3B.gguf reasoning = off ctx-size = 65536 batch-size = 2048 ubatch-size = 512 temperature=1.0 top-p=0.95 top-k=20 min-p=0.0 presence-penalty=1.5 repeat-penalty=1.0 [Qwen3.5-4B] model = /models2/Qwen3.5-4B-UD-Q8_K_XL.gguf mmproj = /models2/mmproj-BF16-Qwen3.5-4B.gguf ctx-size = 131072 [Nemotron-3-Nano-4B] model = /models2/NVIDIA-Nemotron-3-Nano-4B-UD-Q8_K_XL.gguf ctx-size = 131072 [Nemotron-Orchestrator] model = /models2/Nemotron-Orchestrator-8B.Q6_K.gguf ctx-size = 40960 [Gemma-4-Moe] model = /models2/gemma-4-26B-A4B-it-UD-Q4_K_XL.gguf mmproj = /models2/mmproj-google_gemma-4-26B-A4B-it-bf16.gguf ctx-size = 131072