The training parameters are the parameters required for training algorithms to work and perform well.

    • Certain parameter settings are only available for certain Training Algorithm. (parameter will be disabled when not applicable)
    • The recommended training parameters are set based on given data-set, but advanced users can also tweak to achieve better training performance.

    1. Training Algorithm,
      • currently ANSCENTER supports 7 types of Training Algorithm:
        • AdaDelta
        • AdaGrad
        • Adam
        • FSAdaGrad
        • MomentumSGD
        • RMSProp
        • SGD

    1. Loss Function
    2. Accuracy Type
    3. Metric Type
    4. Learning Rate
    5. L1 Decay
    6. L2 Decay
    7. Momentum