context_size: 4096
#mirostat: 2
#mirostat_tau: 5.0
#mirostat_eta: 0.1
f16: true
#low_vram: true
threads: 11
gpu_layers: 90

name: llava
mmap: true
backend: llama-cpp
roles:
  user: "USER:"
  assistant: "ASSISTANT:"

  system: "SYSTEM:"
parameters:
  #model: openbuddy-llama2-34b-v11.1-bf16.Q3_K_S.gguf
  #model: openbuddy-mistral-7b-v13.Q6_K.gguf
  model: ggml-model-q4_k.gguf
  #model: openbuddy-llama2-13b-v11.1.Q6_K.gguf
  #model: openbuddy-llama2-34b-v11.1-bf16.Q4_K_S.gguf
  #model: llama2-22b-daydreamer-v3.ggmlv3.q6_K.bin

  temperature: 0.2

  top_k: 40
  top_p: 0.95
  #ngqa: 8
template:
  chat: chat-simple
mmproj: mmproj-model-f16.gguf