dataquality.loggers.logger_config package

Subpackages

• dataquality.loggers.logger_config.seq2seq package
  • Submodules
  • dataquality.loggers.logger_config.seq2seq.chat module
    • Seq2SeqChatLoggerConfig
  • dataquality.loggers.logger_config.seq2seq.completion module
    • Seq2SeqCompletionLoggerConfig
  • dataquality.loggers.logger_config.seq2seq.seq2seq_base module
    • Seq2SeqLoggerConfig
      • Seq2SeqLoggerConfig.generation_config
      • Seq2SeqLoggerConfig.generation_splits
      • Seq2SeqLoggerConfig.id_to_formatted_prompt_length
      • Seq2SeqLoggerConfig.id_to_tokens
      • Seq2SeqLoggerConfig.max_input_tokens
      • Seq2SeqLoggerConfig.max_target_tokens
      • Seq2SeqLoggerConfig.model
      • Seq2SeqLoggerConfig.model_type
      • Seq2SeqLoggerConfig.response_template
      • Seq2SeqLoggerConfig.sample_length
      • Seq2SeqLoggerConfig.tokenizer
  • Module contents

Submodules

dataquality.loggers.logger_config.base_logger_config module

pydantic model BaseLoggerConfig

Bases: BaseModel

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• conditions (List[dataquality.schemas.condition.Condition])

• cur_epoch (int | None)

• cur_inference_name (str | None)

• cur_split (dataquality.schemas.split.Split | None)

• dataloader_random_sampling (bool)

• exception (str)

• existing_run (bool)

• feature_names (List[str])

• finish (Callable)

• helper_data (Dict[str, Any])

• idx_to_id_map (DefaultDict[str, List])

• inference_logged (bool)

• input_data_logged (DefaultDict[str, int])

• int_labels (bool)

• labels (Any)

• last_epoch (int)

• logged_input_ids (DefaultDict[str, Set])

• metadata_documents (Set)

• ner_labels (List[str])

• observed_labels (Any)

• observed_num_labels (Any)

• remove_embs (bool)

• report_emails (List[str])

• tagging_schema (dataquality.schemas.ner.TaggingSchema | None)

• tasks (Any)

• test_logged (bool)

• training_logged (bool)

• validation_logged (bool)

field conditions: List[Condition] = []

field cur_epoch: Optional[int] = None

field cur_inference_name: Optional[str] = None

field cur_split: Optional[Split] = None

Validated by:

• inference_sets_inference_name

field dataloader_random_sampling: bool = False

field exception: str = ''

field existing_run: bool = False

field feature_names: List[str] = []

field finish: Callable = <function BaseLoggerConfig.<lambda>>

field helper_data: Dict[str, Any] = {}

field idx_to_id_map: DefaultDict[str, List] = {}

field inference_logged: bool = False

field input_data_logged: DefaultDict[str, int] = {}

field int_labels: bool = False

field labels: Any = None

field last_epoch: int = 0

field logged_input_ids: DefaultDict[str, Set] = {}

field metadata_documents: Set = {}

field ner_labels: List[str] = []

field observed_labels: Any = None

field observed_num_labels: Any = None

field remove_embs: bool = False

field report_emails: List[str] = []

field tagging_schema: Optional[TaggingSchema] = None

field tasks: Any = None

field test_logged: bool = False

field training_logged: bool = False

field validation_logged: bool = False

reset(factory=False)

Reset all class vars

Return type:

None

dataquality.loggers.logger_config.image_classification module

pydantic model ImageClassificationLoggerConfig

Bases: TextClassificationLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• all_ids (Dict[str, set])

• observed_ids (Dict[str, set])

field all_ids: Dict[str, set] = {}

field observed_ids: Dict[str, set] = {}

dataquality.loggers.logger_config.object_detection module

class BoxFormat(value)

Bases: str, Enum

Format for bounding box representation

xyxy = 'xyxy'

tlxywh = 'tlxywh'

cxywh = 'cxywh'

pydantic model ObjectDetectionLoggerConfig

Bases: BaseLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• box_format (str)

• image_cloud_path (str)

field box_format: str = BoxFormat.xyxy

field image_cloud_path: str = ''

dataquality.loggers.logger_config.semantic_segmentation module

pydantic model SemanticSegmentationLoggerConfig

Bases: BaseLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

dataquality.loggers.logger_config.tabular_classification module

pydantic model TabularClassificationLoggerConfig

Bases: TextClassificationLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• feature_importances (Dict[str, float])

field feature_importances: Dict[str, float] = {}

dataquality.loggers.logger_config.text_classification module

pydantic model TextClassificationLoggerConfig

Bases: BaseLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• labels (List[str] | None)

• observed_labels (Set[str])

• observed_num_labels (int)

field labels: Optional[List[str]] = None

Validated by:

• clean_labels

field observed_labels: Set[str] = {}

field observed_num_labels: int = 0

dataquality.loggers.logger_config.text_multi_label module

pydantic model TextMultiLabelLoggerConfig

Bases: BaseLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• binary (bool)

• labels (List[List[str]] | None)

• observed_labels (DefaultDict[int, Set])

• observed_num_labels (List[int] | None)

• observed_num_tasks (int)

• tasks (List[str] | None)

field binary: bool = True

field labels: Optional[List[List[str]]] = None

Validated by:

• clean_labels

field observed_labels: DefaultDict[int, Set] = {}

field observed_num_labels: Optional[List[int]] = None

field observed_num_tasks: int = 0

field tasks: Optional[List[str]] = None

dataquality.loggers.logger_config.text_ner module

pydantic model TextNERLoggerConfig

Bases: BaseLoggerConfig

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Fields:

• gold_spans (Dict[str, List[Tuple[int, int, str]]])

• sample_length (Dict[str, int])

field gold_spans: Dict[str, List[Tuple[int, int, str]]] = {}

field sample_length: Dict[str, int] = {}

get_sample_key(split, sample_id)

Return type:

str

reset(factory=False)

Don’t clear the “ner” or “ner_config” variables if they are set

Return type:

None

Module contents