logLevel: info
healthCheckTimeout: 120

models:
  DuckAi-GeneralFast:
    proxy: http://localhost:8091
    cmd: /app/Scripts/startGeneralFast.sh
    cmdStop: /app/Scripts/stopGeneralFast.sh
    checkEndpoint: /health

  DuckAi-General:
    proxy: http://localhost:8092
    cmd: /app/Scripts/startGeneral.sh
    cmdStop: /app/Scripts/stopGeneral.sh
    checkEndpoint: /health
    ttl: 600

  DuckAi-Chat:
    proxy: http://localhost:8093
    cmd: /app/Scripts/startChat.sh
    cmdStop: /app/Scripts/stopChat.sh
    checkEndpoint: /health
    ttl: 600

  DuckAi-Coder:
    proxy: http://localhost:8094
    cmd: /app/Scripts/startCoder.sh
    cmdStop: /app/Scripts/stopCoder.sh
    checkEndpoint: /health
    ttl: 600

  DuckAi-EmbeddingFast:
    proxy: http://localhost:8095
    cmd: /app/Scripts/startEmbeddingFast.sh
    cmdStop: /app/Scripts/stopEmbeddingFast.sh
    checkEndpoint: /health
    ttl: 600

  DuckAi-Embedding:
    proxy: http://localhost:8096
    cmd: /app/Scripts/startEmbedding.sh
    cmdStop: /app/Scripts/stopEmbedding.sh
    checkEndpoint: /health
    ttl: 600

groups:
  default-models:
    swap: false
    exclusive: false
    persistent: true
    members:
      - DuckAi-GeneralFast
      - DuckAi-Chat
      - DuckAi-Embedding

hooks:
  on_startup:
    preload:
      - DuckAi-GeneralFast