def __init__(
self,
save_dir: str,
corpus: Corpus,
activation_names: ActivationNames,
activations_dir: Optional[str] = None,
test_activations_dir: Optional[str] = None,
test_corpus: Optional[Corpus] = None,
model: Optional[LanguageModel] = None,
selection_func: SelectFunc = lambda sen_id, pos, example: True,
) -> None:
self.save_dir = save_dir
if not os.path.exists(save_dir):
os.mkdir(save_dir)
activations_dir, test_activations_dir = self._extract_activations(
save_dir,
corpus,
activation_names,
selection_func,
activations_dir,
test_activations_dir,
test_corpus,
model,
)
self.activation_names = activation_names
self.data_loader = DataLoader(
activations_dir,
corpus,
test_activations_dir=test_activations_dir,
test_corpus=test_corpus,
selection_func=selection_func,
)
self.classifier = LogRegCV()
def setUpClass(cls) -> None:
# Create directory if necessary
if not os.path.exists(ACTIVATIONS_DIR):
os.makedirs(ACTIVATIONS_DIR)
# Create dummy data have reader read it
create_and_dump_dummy_activations(
num_sentences=NUM_TEST_SENTENCES, activations_dim=ACTIVATIONS_DIM, max_tokens=5,
activations_dir=ACTIVATIONS_DIR, activations_name=ACTIVATIONS_NAME, num_classes=2
)
cls.data_loader = DataLoader(activations_dir=ACTIVATIONS_DIR)
cls.num_labels = cls.data_loader.data_len
def __init__(
self,
save_dir: str,
corpus: Corpus,
activation_names: ActivationNames,
activations_dir: Optional[str] = None,
test_activations_dir: Optional[str] = None,
test_corpus: Optional[Corpus] = None,
model: Optional[LanguageModel] = None,
train_selection_func: SelectionFunc = lambda sen_id, pos, example:
True,
test_selection_func: Optional[SelectionFunc] = None,
control_task: Optional[ControlTask] = None,
classifier_type: str = "logreg_torch",
save_logits: bool = False,
verbose: int = 0,
) -> None:
self.save_dir = save_dir
if not os.path.exists(save_dir):
os.makedirs(save_dir)
self.remove_callbacks = []
activations_dir, test_activations_dir = self._extract_activations(
save_dir,
corpus,
activation_names,
train_selection_func,
activations_dir,
test_activations_dir,
test_corpus,
test_selection_func,
model,
)
self.activation_names = activation_names
self.data_dict: DataDict = {}
self.data_loader = DataLoader(
activations_dir,
corpus,
test_activations_dir=test_activations_dir,
test_corpus=test_corpus,
train_selection_func=train_selection_func,
test_selection_func=test_selection_func,
control_task=control_task,
)
assert classifier_type in [
"logreg_torch",
"logreg_sklearn",
], "Classifier type not understood, should be either `logreg_toch` or `logreg_sklearn`"
self.classifier_type = classifier_type
self.save_logits = save_logits
self.verbose = verbose
def setUpClass(cls) -> None:
# Create directory if necessary
if not os.path.exists(ACTIVATIONS_DIR):
os.makedirs(ACTIVATIONS_DIR)
# Create dummy data have reader read it
cls.num_labels = create_and_dump_dummy_activations(
num_sentences=NUM_TEST_SENTENCES,
activations_dim=ACTIVATIONS_DIM,
max_sen_len=5,
activations_dir=ACTIVATIONS_DIR,
activations_name=ACTIVATIONS_NAME,
num_classes=2,
)
corpus = import_corpus(f"{ACTIVATIONS_DIR}/corpus.tsv")
cls.data_loader = DataLoader(ACTIVATIONS_DIR, corpus)
def __init__(self,
corpus: Corpus,
activations_dir: str,
activation_names: List[ActivationName],
save_dir: str,
classifier_type: str,
calc_class_weights: bool = False) -> None:
self.activation_names: List[ActivationName] = activation_names
self.save_dir = save_dir
if not os.path.exists(save_dir):
os.mkdir(save_dir)
# TODO: Allow own classifier here (should adhere to some base functions, such as .fit())
self.classifier_type = classifier_type
self.calc_class_weights = calc_class_weights
self.data_loader = DataLoader(activations_dir, corpus)
self.results: ResultsDict = defaultdict(dict)
self._reset_classifier()