mlwiz

• mlwiz.data
  • data.dataset
    • Cora
      • Cora._load_dataset()
      • Cora._save_dataset()
      • Cora.dim_input_features
      • Cora.dim_target
      • Cora.process_dataset()
    • DatasetInterface
      • DatasetInterface._load_dataset()
      • DatasetInterface._save_dataset()
      • DatasetInterface.dataset_filename
      • DatasetInterface.dataset_filepath
      • DatasetInterface.dataset_folder
      • DatasetInterface.dataset_name
      • DatasetInterface.dim_input_features
      • DatasetInterface.dim_target
      • DatasetInterface.name
      • DatasetInterface.process_dataset()
      • DatasetInterface.raw_dataset_folder
    • IterableDatasetInterface
      • IterableDatasetInterface.dataset_filepaths
      • IterableDatasetInterface.dataset_folder
      • IterableDatasetInterface.dataset_name
      • IterableDatasetInterface.dim_input_features
      • IterableDatasetInterface.dim_target
      • IterableDatasetInterface.process_dataset()
      • IterableDatasetInterface.raw_dataset_folder
      • IterableDatasetInterface.set_eval()
      • IterableDatasetInterface.shuffle_urls()
      • IterableDatasetInterface.shuffle_urls_elements()
      • IterableDatasetInterface.splice()
      • IterableDatasetInterface.subset()
      • IterableDatasetInterface.url_indices
    • MNIST
      • MNIST.dim_input_features
      • MNIST.dim_target
      • MNIST.process_dataset()
    • MNISTTemporal
      • MNISTTemporal.dim_input_features
      • MNISTTemporal.dim_target
      • MNISTTemporal.process_dataset()
    • NCI1
      • NCI1._load_dataset()
      • NCI1._save_dataset()
      • NCI1.dim_input_features
      • NCI1.dim_target
      • NCI1.process_dataset()
    • ToyIterableDataset
      • ToyIterableDataset.dim_input_features
      • ToyIterableDataset.dim_target
      • ToyIterableDataset.process_dataset()
      • ToyIterableDataset.url_indices
    • _ReshapeMNISTTemporal
  • data.provider
    • DataProvider
      • DataProvider._get_dataset()
      • DataProvider._get_loader()
      • DataProvider._get_splitter()
      • DataProvider._require_exp_seed()
      • DataProvider._require_outer_and_inner_k()
      • DataProvider._require_outer_k()
      • DataProvider.get_dim_input_features()
      • DataProvider.get_dim_target()
      • DataProvider.get_inner_train()
      • DataProvider.get_inner_val()
      • DataProvider.get_outer_test()
      • DataProvider.get_outer_train()
      • DataProvider.get_outer_val()
      • DataProvider.set_exp_seed()
      • DataProvider.set_inner_k()
      • DataProvider.set_outer_k()
    • IterableDataProvider
      • IterableDataProvider._get_loader()
    • SingleGraphDataProvider
      • SingleGraphDataProvider._get_dataset()
      • SingleGraphDataProvider._get_loader()
      • SingleGraphDataProvider._get_splitter()
      • SingleGraphDataProvider.get_inner_train()
      • SingleGraphDataProvider.get_inner_val()
      • SingleGraphDataProvider.get_outer_test()
      • SingleGraphDataProvider.get_outer_train()
      • SingleGraphDataProvider.get_outer_val()
    • SubsetTrainEval
    • _iterable_worker_init_fn()
    • seed_worker()
  • data.sampler
    • RandomSampler
  • data.splitter
    • Fold
    • InnerFold
      • InnerFold.todict()
    • OuterFold
      • OuterFold.todict()
    • SingleGraphSplitter
      • SingleGraphSplitter.split()
    • Splitter
      • Splitter._get_splitter()
      • Splitter._splitter_args()
      • Splitter.check_splits_overlap()
      • Splitter.get_targets()
      • Splitter.load()
      • Splitter.save()
      • Splitter.split()
      • Splitter.stratify
    • _NoShuffleTrainTestSplit
      • _NoShuffleTrainTestSplit.split()
  • data.util
    • check_argument()
    • get_or_create_dir()
    • load_dataset()
    • preprocess_data()
    • single_graph_collate()
• mlwiz.evaluation
  • evaluation.config
    • Config
      • Config.get()
      • Config.items()
      • Config.keys()
  • evaluation.evaluator
    • RiskAssesser
      • RiskAssesser._create_dataset_getter()
      • RiskAssesser._request_termination()
      • RiskAssesser.compute_best_hyperparameters()
      • RiskAssesser.compute_final_runs_score_per_fold()
      • RiskAssesser.compute_risk_assessment_result()
      • RiskAssesser.model_selection()
      • RiskAssesser.process_config_results_across_inner_folds()
      • RiskAssesser.process_model_selection_runs()
      • RiskAssesser.risk_assessment()
      • RiskAssesser.run_final_model()
      • RiskAssesser.wait_configs()
    • _get_ray_num_gpus_per_task()
    • _make_termination_checker()
    • _mean_std_ci()
    • _push_progress_update()
    • _set_cuda_memory_limit_from_env()
    • extract_and_sum_elapsed_seconds()
    • run_test()
    • run_valid()
    • send_telegram_update()
  • evaluation.grid
    • Grid
      • Grid._gen_configs()
      • Grid._gen_helper()
      • Grid._list_helper()
      • Grid.exp_name
      • Grid.num_configs
  • evaluation.random_search
    • RandomSearch
      • RandomSearch._dict_helper()
      • RandomSearch._gen_helper()
      • RandomSearch._sampler_helper()
  • evaluation.util
    • _collect_metric_samples()
    • _df_to_latex_table()
    • _list_outer_fold_ids()
    • _load_final_run_metric_samples()
    • _summarize_samples()
    • choice()
    • create_dataframe()
    • create_latex_table_from_assessment_results()
    • filter_experiments()
    • get_scores_from_assessment_results()
    • get_scores_from_outer_results()
    • instantiate_data_provider_from_config()
    • instantiate_dataset_from_config()
    • instantiate_model_from_config()
    • load_checkpoint()
    • loguniform()
    • normal()
    • randint()
    • retrieve_best_configuration()
    • retrieve_experiments()
    • statistical_significance()
    • uniform()
• mlwiz.experiment
  • experiment.experiment
    • Experiment
      • Experiment._return_class_and_args()
      • Experiment.create_engine()
      • Experiment.create_model()
      • Experiment.run_test()
      • Experiment.run_valid()
• mlwiz.log
  • log.logger
    • Logger
      • Logger.log()
• mlwiz.model
  • model.interface
    • ModelInterface
      • ModelInterface.forward()
• mlwiz.training
  • training.callback
    • callback.early_stopping
      • EarlyStopper
      • PatienceEarlyStopper
    • callback.engine_callback
      • EngineCallback
      • IterableEngineCallback
    • callback.gradient_clipping
      • GradientClipper
    • callback.metric
      • AdditiveLoss
      • AllocatedGPUMemory
      • Classification
      • MeanAbsoluteError
      • MeanSquareError
      • Metric
      • MultiScore
      • MulticlassAccuracy
      • MulticlassClassification
      • Regression
      • SingleGraphMulticlassAccuracy
      • SingleGraphMulticlassClassification
      • ToyMetric
    • callback.optimizer
      • Optimizer
    • callback.plotter
      • Plotter
      • _NullSummaryWriter
    • callback.scheduler
      • EpochScheduler
      • MetricScheduler
      • Scheduler
  • training.event
    • event.dispatcher
      • EventDispatcher
    • event.handler
      • EventHandler
    • event.state
      • State
  • training.engine
    • DataStreamTrainingEngine
      • DataStreamTrainingEngine._loop()
    • TrainingEngine
      • TrainingEngine._check_termination()
      • TrainingEngine._loop()
      • TrainingEngine._loop_helper()
      • TrainingEngine._restore_checkpoint_and_best_results()
      • TrainingEngine._to_data_list()
      • TrainingEngine._to_list()
      • TrainingEngine._train()
      • TrainingEngine.cumulative_batch_unsent_time
      • TrainingEngine.cumulative_epoch_unsent_time
      • TrainingEngine.infer()
      • TrainingEngine.set_device()
      • TrainingEngine.set_eval_mode()
      • TrainingEngine.set_training_mode()
      • TrainingEngine.train()
    • fmt()
    • log()
    • reorder()
  • training.profiler
    • Profiler
      • Profiler.report()
      • Profiler.total_elapsed_time
  • training.util
    • atomic_torch_save()

mlwiz.util

General-purpose helper functions for MLWiz.

Includes dotted-path resolution, config parsing helpers, and atomic dill serialization utilities.

mlwiz.util.atomic_dill_save(data: object, filepath: str) → object

Saves a dill object to a file. :param data: the dill object to save :param filepath: the path to the dill object to save

mlwiz.util.dill_load(filepath: str) → object

Loads a dill file. :param filepath: the path to the dill file :return: an object

mlwiz.util.return_class_and_args(config: dict, key: str, return_class_name: bool = False) → Tuple[Callable[[...], object], dict]

Returns the class and arguments associated to a specific key in the configuration file.

Parameters:

• config (dict) – the configuration dictionary

• key (str) – a string representing a particular class in the configuration dictionary

• return_class_name (bool) – if True, returns the class name as a string rather than the class object

Returns:

a tuple (class, dict of arguments), or (None, None) if the key is not present in the config dictionary

mlwiz.util.s2c(class_name: str) → Callable[[...], object]

Converts a dotted path to the corresponding class

Parameters:

class_name (str) – dotted path to class name

Returns:

the class to be used