schrodinger.application.transforms.scorers.ldml module

LDML inference scorer for AL-Pareto active learning.

class schrodinger.application.transforms.scorers.ldml.InferencePropertyNames(*, prediction: str, uncertainty: str)

Bases: BaseModel

Property names produced by InferenceScorer.

Parameters:

• prediction – Property name for the ML prediction value.

• uncertainty – Property name for the ML uncertainty value.

prediction: str

uncertainty: str

model_config: ClassVar[ConfigDict] = {'frozen': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

schrodinger.application.transforms.scorers.ldml.inference_property_names(basename: str) → InferencePropertyNames

Return property names for the given basename.

Parameters:

basename – Property basename (e.g. 'my_dataset').

Returns:

Property names using the basename.

class schrodinger.application.transforms.scorers.ldml.InferenceScorerConfig(*, ldml_url: HttpUrl, ld_url: pydantic.networks.HttpUrl | None = None, jsession_id: str | None = None, polling_interval: int = 600, max_retries: int = 5, dataset_display_name: str, property_basename: str | None = None)

Bases: InferenceConfig

Configuration for InferenceScorer.

Extends InferenceConfig with an optional property_basename that controls the generated property names. If not provided, the basename is derived by sanitizing dataset_display_name.

Parameters:

property_basename – Basename for inference property names.

property_basename: str | None

property property_names: InferencePropertyNames

Property names derived from property_basename.

model_config: ClassVar[ConfigDict] = {'frozen': True}

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

class schrodinger.application.transforms.scorers.ldml.InferenceScorer(**kwargs)

Bases: AsyncScorer

Score structures using LDML inference.

Submits structures to the LDML inference service, polls for results, and annotates structures with prediction and uncertainty values. Structures that fail inference are dropped.

Property names default to values derived from dataset_display_name but can be overridden via property_basename.

See InferenceScorerConfig for available configuration options.

SCORER_ID: str = 'ldml_inference'

config_class

alias of InferenceScorerConfig