veil.config.entity_detectors

Classes

BaseEntityDetectorConfig([priority, ...])	Marker base for entity detector configuration objects.
GlinerEntityDetectorConfig([priority, ...])	Configuration for the Gliner entity detector integration.
HostedMaskerApiEntityDetectorConfig([...])	Configuration for the hosted Masker API entity detector integration.
MaskerApiEntityDetectorConfig([priority, ...])	Marker base for masking-API entity detector configuration objects.
RegexEntityDetectorConfig([priority, ...])	Configuration wrapper for Veil’s RegexEntityDetector engine.
SpacyEntityDetectorConfig([priority, ...])	Configuration for the Spacy entity detector integration.

class veil.config.entity_detectors.BaseEntityDetectorConfig(priority=<factory>, hierarchy_position=0)

Bases: BasePolyConfig

Marker base for entity detector configuration objects.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

classmethod get_type()

hierarchy_position: int = 0

priority: Dict[str, int]

class veil.config.entity_detectors.GlinerEntityDetectorConfig(priority=<factory>, hierarchy_position=0, labels=<factory>, model='urchade/gliner_multi-v2.1', cuda_device=0, threshold=0.6, batch_size=8, max_length=384, chunk_overlap=50, nms_iou_threshold=0.8, min_span_chars=3, max_span_chars=80, top_k_per_chunk=100)

Bases: BaseEntityDetectorConfig

Configuration for the Gliner entity detector integration.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

• labels (List[str])

• model (str)

• cuda_device (int)

• threshold (float)

• batch_size (int)

• max_length (int)

• chunk_overlap (int)

• nms_iou_threshold (float)

• min_span_chars (int)

• max_span_chars (int)

• top_k_per_chunk (int)

batch_size: int = 8

chunk_overlap: int = 50

cuda_device: int = 0

classmethod get_type()

labels: List[str]

max_length: int = 384

max_span_chars: int = 80

min_span_chars: int = 3

model: str = 'urchade/gliner_multi-v2.1'

nms_iou_threshold: float = 0.8

priority: Dict[str, int]

threshold: float = 0.6

top_k_per_chunk: int = 100

class veil.config.entity_detectors.HostedMaskerApiEntityDetectorConfig(priority=<factory>, hierarchy_position=0, api_url='', headers=<factory>, model='', system_prompt='You are a helpful assistant that masks sensitive information.', max_tokens=4000, top_p=1, top_k=40, presence_penalty=0, frequency_penalty=0, temperature=0.6, timeout=30, retries=2, retry_backoff_base=0.5, retry_on_truncation=True, chunk_on_truncation=True, chunk_char_limit=4000, truncation_min_fraction=0.6)

Bases: MaskerApiEntityDetectorConfig

Configuration for the hosted Masker API entity detector integration.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

• api_url (str)

• headers (Dict[str, str])

• model (str)

• system_prompt (str)

• max_tokens (int)

• top_p (float)

• top_k (int)

• presence_penalty (float)

• frequency_penalty (float)

• temperature (float)

• timeout (float)

• retries (int)

• retry_backoff_base (float)

• retry_on_truncation (bool)

• chunk_on_truncation (bool)

• chunk_char_limit (int)

• truncation_min_fraction (float)

classmethod get_type()

class veil.config.entity_detectors.MaskerApiEntityDetectorConfig(priority=<factory>, hierarchy_position=0, api_url='', headers=<factory>, model='', system_prompt='You are a helpful assistant that masks sensitive information.', max_tokens=4000, top_p=1, top_k=40, presence_penalty=0, frequency_penalty=0, temperature=0.6, timeout=30, retries=2, retry_backoff_base=0.5, retry_on_truncation=True, chunk_on_truncation=True, chunk_char_limit=4000, truncation_min_fraction=0.6)

Bases: BaseEntityDetectorConfig

Marker base for masking-API entity detector configuration objects.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

• api_url (str)

• headers (Dict[str, str])

• model (str)

• system_prompt (str)

• max_tokens (int)

• top_p (float)

• top_k (int)

• presence_penalty (float)

• frequency_penalty (float)

• temperature (float)

• timeout (float)

• retries (int)

• retry_backoff_base (float)

• retry_on_truncation (bool)

• chunk_on_truncation (bool)

• chunk_char_limit (int)

• truncation_min_fraction (float)

api_url: str = ''

chunk_char_limit: int = 4000

chunk_on_truncation: bool = True

frequency_penalty: float = 0

classmethod get_type()

headers: Dict[str, str]

max_tokens: int = 4000

model: str = ''

presence_penalty: float = 0

priority: Dict[str, int]

retries: int = 2

retry_backoff_base: float = 0.5

retry_on_truncation: bool = True

system_prompt: str = 'You are a helpful assistant that masks sensitive information.'

temperature: float = 0.6

timeout: float = 30

top_k: int = 40

top_p: float = 1

truncation_min_fraction: float = 0.6

class veil.config.entity_detectors.RegexEntityDetectorConfig(priority=<factory>, hierarchy_position=0, enable_validation=True, min_confidence=0.0, preserve_format=True, case_sensitive=False)

Bases: BaseEntityDetectorConfig

Configuration wrapper for Veil’s RegexEntityDetector engine.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

• enable_validation (bool)

• min_confidence (float)

• preserve_format (bool)

• case_sensitive (bool)

case_sensitive: bool = False

enable_validation: bool = True

classmethod get_type()

min_confidence: float = 0.0

preserve_format: bool = True

priority: Dict[str, int]

class veil.config.entity_detectors.SpacyEntityDetectorConfig(priority=<factory>, hierarchy_position=0, model='es_core_news_sm', cuda_device=0)

Bases: BaseEntityDetectorConfig

Configuration for the Spacy entity detector integration.

Parameters:

• priority (Dict[str, int])

• hierarchy_position (int)

• model (str)

• cuda_device (int)

cuda_device: int = 0

classmethod get_type()

model: str = 'es_core_news_sm'

priority: Dict[str, int]