Skip to main content
Version: 1.0.0

Models

Models is a mapping for Machine Learning models that need to be imported together with the configuration to run inference sessions using ONNX and make predictions for the agents. Here you can map them to actionModel and volumeModel which will be called by the agent when a new step needs to be performed.

Format​

nametype
typearray of {
    class: string,
    model...: {
        alias: string,
        location: string
    }
}
requiredtrue

Example​

models:
  - class: "lp"
    actionModel:
      alias: "bancor_lp_action"
      location: "models/action_user.onnx"
    volumeModel:
      alias: "bancor_lp_volume"
      location: "models/amount_user.onnx"
  - class: "arb"
  - class: "trader"
    volumeModel:
      alias: "bancor_trader"
      location: "models/trader.onnx"