Skip to content

Config Discriminators

Source

Discriminator for NG configurations.

instantiate_algorithm_config(config)

Instantiate an algorithm configuration from a configuration dictionary.

This method uses a TypeAdapter to validate the configuration and instantiate the correct algorithm configuration. Currently only compatible with UNet-based algorithms.

Parameters:

  • config (dict[str, Any]) –

    Configuration dictionary to validate.

Returns:

  • AlgorithmConfig –

    Validated configuration as one of the UNetBasedAlgorithm configurations.

Raises:

  • ValueError –

    If the configuration is not valid.

instantiate_config(config)

Instantiate a configuration from a configuration dictionary.

This method uses a TypeAdapter to validate the configuration and instantiate the correct configuration.

Parameters:

  • config (dict[str, Any]) –

    Configuration dictionary to validate.

Returns:

  • Config –

    Validated configuration as a Config.

Raises:

  • ValueError –

    If the configuration is not valid.

instantiate_norm_config(config)

Instantiate a NormalizationConfig from a configuration dictionary.

This method uses a TypeAdapter to validate the configuration and instantiate the correct NormalizationConfig.

Parameters:

  • config (dict[str, Any]) –

    Configuration dictionary to validate.

Returns:

  • NormalizationConfig –

    Validated configuration as a NormalizationConfig.

Raises:

  • ValueError –

    If the configuration is not valid.