foxai.cli package

Submodules

foxai.cli.config_model module

File contains definition of CLI program config data model.

class foxai.cli.config_model.ConfigDataModel(method_config_list: List[MethodDataModel])[source]

Bases: object

Config data model for CLI program.

method_config_list: List[MethodDataModel]
class foxai.cli.config_model.MethodDataModel(explainer_with_params: ExplainerWithParams, artifact_name: Optional[str] = None)[source]

Bases: object

Method data model for CLI program.

artifact_name: Optional[str] = None
explainer_with_params: ExplainerWithParams

foxai.cli.experiment_update module

File contains CLI application for updating W&B experiment artifacts.

foxai.cli.experiment_update.download_upload_metadata(api_run: Run) → Dict[str, Any][source]

Fetch uploaded data samples metadata from logs history.

Parameters:

run_api – WandB Api object.

Returns:

Metadata of upload data sample entry.

foxai.cli.experiment_update.fetch_model_checkpoints(api: Api, model_artifact_root: str, project: str) → List[str][source]

Fetch all model checkpoints and return local paths to them.

Parameters:

  • api – WandB Api object.

  • model_artifact_root – Path to local directory where model checkpoints will be saved.

  • project – Name of a project.

Returns:

List of paths to saved downloaded model checkpoints.

foxai.cli.experiment_update.get_labels(upload_data_metadata: Dict[str, Any]) → List[int][source]

Extract and convert ground truth labels from image captions.

Parameters:

upload_data_metadata – Metadata of upload data sample entry.

Returns:

List of labels.

foxai.cli.experiment_update.load_config(config: DictConfig) → ConfigDataModel[source]

Load and validate config with explainers to apply.

Parameters:

config – DictConfig object loaded from yaml config file.

Returns:

ConfigDataModel object.

foxai.cli.experiment_update.load_images_to_tensors(filenames: List[str], artifacts_path: str) → List[Tensor][source]

Load images from local directory.

Parameters:

  • filenames – List of filepaths to images from metadata.

  • artifacts_path – Root directory of downloaded images.

Returns:

List of tensors representing images.

foxai.cli.experiment_update.main(cfg: DictConfig) → None[source]

Entry point for CLI application.

Module contents