tiramisu_brulee.experiment.cli package¶

Submodules¶

Common functions for predict and train CLIs Author: Jacob Reinhold <jcreinhold@gmail.com> Created on: Jul 30, 2021

tiramisu_brulee.experiment.cli.common.check_patch_size(patch_size: List[int], use_pseudo3d: bool) → None[source]¶

tiramisu_brulee.experiment.cli.common.handle_fast_dev_run(unnecessary_args: Set[str]) → Set[str][source]¶: fast_dev_run is problematic with py36 so remove it

tiramisu_brulee.experiment.cli.common.pseudo3d_dims_setup(pseudo3d_dim: Optional[List[int]], n_models: int, stage: str) → Union[List[None], List[int]][source]¶

tiramisu_brulee.experiment.cli.common.tiramisu_brulee_info(*, short: bool = True) → tiramisu_brulee.experiment.type.TiramisuBruleeInfo[source]¶: get the git commit hash and version for tiramisu-brulee

tiramisu_brulee.experiment.cli.predict

command-line interface functions for predicting lesion segmentations with Tiramisu neural network

Author: Jacob Reinhold (jcreinhold@gmail.com) Created on: May 25, 2021

tiramisu_brulee.experiment.cli.predict.predict(args: Optional[Union[argparse.Namespace, jsonargparse.Namespace, Iterable[str]]] = None) → int[source]¶: use a Tiramisu CNN for prediction

tiramisu_brulee.experiment.cli.predict.predict_image(args: Optional[Union[argparse.Namespace, jsonargparse.Namespace, Iterable[str]]] = None) → int[source]¶: use a Tiramisu CNN for prediction for a single time-point

tiramisu_brulee.experiment.cli.train

command-line interface functions for training lesion segmentation Tiramisu neural networks

Author: Jacob Reinhold (jcreinhold@gmail.com) Created on: May 25, 2021

tiramisu_brulee.experiment.cli.train.train(args: Optional[Union[argparse.Namespace, jsonargparse.Namespace, Iterable[str]]] = None, *, return_best_model_paths: bool = False) → Union[List[pathlib.Path], int][source]¶: train a Tiramisu CNN for segmentation

Training and prediction CLI for tiramisu-brulee