name: mixtral
model: mixtral:7x8B
version: 1

# Results Preferences
stop:
  - </s>
top_p: 0.95
temperature: 0.7
frequency_penalty: 0
presence_penalty: 0
max_tokens: 32768 # Infer from base config.json -> max_position_embeddings
stream: true # true | false

# Engine / Model Settings
ngl: 33 # Infer from base config.json -> num_attention_heads
ctx_len: 32768 # Infer from base config.json -> max_position_embeddings
engine: llama-cpp
prompt_template: "[INST] {prompt} [/INST]"