Constants

gllm_multimodal.constants

Constants for multimodal operations in Gen AI applications.

DEFAULT_ROUTER_EMBEDDING_MODEL = 'twelvelabs/marengo3.0' module-attribute

AudioFormatConstants

Constants for audio format related keys and MIME types.

FORMAT_ALIASES = {FORMAT_MPEG: FORMAT_MP3, FORMAT_OGA: FORMAT_OGG} class-attribute instance-attribute

FORMAT_FLAC = 'flac' class-attribute instance-attribute

FORMAT_M4A = 'm4a' class-attribute instance-attribute

FORMAT_MP3 = 'mp3' class-attribute instance-attribute

FORMAT_MPEG = 'mpeg' class-attribute instance-attribute

FORMAT_MPGA = 'mpga' class-attribute instance-attribute

FORMAT_OGA = 'oga' class-attribute instance-attribute

FORMAT_OGG = 'ogg' class-attribute instance-attribute

FORMAT_TO_MIME_TYPE = {FORMAT_MP3: MIME_MPEG, FORMAT_MPEG: MIME_MPEG, FORMAT_MPGA: MIME_MPEG, FORMAT_M4A: MIME_MP4, FORMAT_WAV: MIME_WAV, FORMAT_WAVE: MIME_WAV, FORMAT_FLAC: MIME_FLAC, FORMAT_OGG: MIME_OGG, FORMAT_OGA: MIME_OGG} class-attribute instance-attribute

FORMAT_WAV = 'wav' class-attribute instance-attribute

FORMAT_WAVE = 'wave' class-attribute instance-attribute

MIME_FLAC = 'audio/flac' class-attribute instance-attribute

MIME_MP4 = 'audio/mp4' class-attribute instance-attribute

MIME_MPEG = 'audio/mpeg' class-attribute instance-attribute

MIME_OGG = 'audio/ogg' class-attribute instance-attribute

MIME_WAV = 'audio/wav' class-attribute instance-attribute

CaptionConstants

Constants for caption operations in Gen AI applications.

CAPTION_DEFAULT_JSON_KEY = 'captions' class-attribute instance-attribute

DEFAULT_NUMBER_OF_CAPTIONS = 5 class-attribute instance-attribute

DOMAIN_KNOWLEDGE = 'domain_knowledge' class-attribute instance-attribute

IS_STRUCTURED_CAPTION = 'is_structured_caption' class-attribute instance-attribute

MEDIA_METADATA = 'media_metadata' class-attribute instance-attribute

MULTIMODAL_CONTEXT = 'multimodal_context' class-attribute instance-attribute

NOT_GIVEN = 'not_given' class-attribute instance-attribute

NUMBER_OF_CAPTIONS = 'number_of_captions' class-attribute instance-attribute

TEXT_CONTEXT = 'text_context' class-attribute instance-attribute

TEXT_ONE_LINER = 'text_one_liner' class-attribute instance-attribute

CaptionOutputFormattingStrategy

Bases: StrEnum

Defines how output should be formatted.

STRUCTURED = 'structured' class-attribute instance-attribute

TEXT = 'text' class-attribute instance-attribute

ConverterResultTag

Bases: StrEnum

Defines converter result tags.

CAPTION = 'captions' class-attribute instance-attribute

MERMAID = 'mermaid' class-attribute instance-attribute

OCR = 'ocr' class-attribute instance-attribute

STRUCTURED_CAPTION = 'structured_caption' class-attribute instance-attribute

DetectedObjectClass

Class names assigned to detected objects produced by OCR and detection converters.

HANDWRITING = 'handwriting' class-attribute instance-attribute

TABLE = 'table' class-attribute instance-attribute

TEXT = 'text' class-attribute instance-attribute

DetectedObjectMetadataKey

Metadata keys set on detected objects produced by OCR and detection converters.

COLUMN_COUNT = 'column_count' class-attribute instance-attribute

PAGE_NUMBER = 'page_number' class-attribute instance-attribute

ROW_COUNT = 'row_count' class-attribute instance-attribute

DocumentConstants

Constants for document loading used by OCR and related converters.

ALLOWED_MIME_TYPES = ['image/', 'application/pdf'] class-attribute instance-attribute

ExifConstants

Constants for EXIF tag operations in image metadata extraction.

GPS_LATITUDE = 'GPS GPSLatitude' class-attribute instance-attribute

GPS_LATITUDE_REF = 'GPS GPSLatitudeRef' class-attribute instance-attribute

GPS_LONGITUDE = 'GPS GPSLongitude' class-attribute instance-attribute

GPS_LONGITUDE_REF = 'GPS GPSLongitudeRef' class-attribute instance-attribute

GPS_SOUTH = 'S' class-attribute instance-attribute

GPS_WEST = 'W' class-attribute instance-attribute

LATITUDE = 'latitude' class-attribute instance-attribute

LONGITUDE = 'longitude' class-attribute instance-attribute

GLAIRVisionOcrConstants

Constants for GLAIR Vision OCR API requests and responses.

ALL_TEXTS_KEY = 'all_texts' class-attribute instance-attribute

API_KEY_HEADER = 'x-api-key' class-attribute instance-attribute

CELLS_KEY = 'cells' class-attribute instance-attribute

COLUMN_COUNT_KEY = 'column_count' class-attribute instance-attribute

COLUMN_INDEX_KEY = 'column_index' class-attribute instance-attribute

COLUMN_SPAN_KEY = 'column_span' class-attribute instance-attribute

CONFIDENCE_KEY = 'confidence' class-attribute instance-attribute

HANDWRITING_TEXT_TYPE = 'HW' class-attribute instance-attribute

IMAGE_FIELD = 'image' class-attribute instance-attribute

IS_HEADER_KEY = 'is_header' class-attribute instance-attribute

PAGE_INDEX_KEY = 'page_index' class-attribute instance-attribute

POLYGON_COORDINATE_COUNT = 2 class-attribute instance-attribute

POLYGON_KEY = 'polygon' class-attribute instance-attribute

POLYGON_POINT_COUNT = 4 class-attribute instance-attribute

READ_KEY = 'read' class-attribute instance-attribute

REASON_KEY = 'reason' class-attribute instance-attribute

ROW_COUNT_KEY = 'row_count' class-attribute instance-attribute

ROW_INDEX_KEY = 'row_index' class-attribute instance-attribute

ROW_SPAN_KEY = 'row_span' class-attribute instance-attribute

STATUS_KEY = 'status' class-attribute instance-attribute

SUCCESS_STATUS = 'SUCCESS' class-attribute instance-attribute

TABLES_KEY = 'tables' class-attribute instance-attribute

TABLE_FIELD = 'table' class-attribute instance-attribute

TABLE_REQUEST_VALUE = 'true' class-attribute instance-attribute

TEXT_TYPE_KEY = 'text_type' class-attribute instance-attribute

UNKNOWN_ERROR_REASON = 'unknown error' class-attribute instance-attribute

VALUE_KEY = 'value' class-attribute instance-attribute

ImageToTextConstants

Constants for image-to-text operations in Gen AI applications.

FILENAME = 'filename' class-attribute instance-attribute

NO_IMAGE_FILENAME = 'no_image_filename' class-attribute instance-attribute

MediaBackend

Bases: StrEnum

Defines supported media processing backends.

FFMPEG = 'ffmpeg' class-attribute instance-attribute

GSTREAMER = 'gstreamer' class-attribute instance-attribute

Modality

Bases: StrEnum

Defines supported modalities.

AUDIO = 'audio' class-attribute instance-attribute

IMAGE = 'image' class-attribute instance-attribute

MIXED = 'mixed' class-attribute instance-attribute

TEXT = 'text' class-attribute instance-attribute

UNKNOWN = 'unknown' class-attribute instance-attribute

VIDEO = 'video' class-attribute instance-attribute

ModalityConverterApproach

Bases: StrEnum

Defines supported modality converter approach types.

ASR = 'asr' class-attribute instance-attribute

GLAIR_VISION = 'glair_vision' class-attribute instance-attribute

HYBRID = 'hybrid' class-attribute instance-attribute

LM_BASED = 'lm_based' class-attribute instance-attribute

TRANSCRIPT_FETCH = 'transcript_fetch' class-attribute instance-attribute

ModalityConverterBuildStrategy

Bases: StrEnum

Defines how a modality converter should be constructed from its configuration.

KWARGS = 'kwargs' class-attribute instance-attribute

LMRP = 'lmrp' class-attribute instance-attribute

PRESET = 'preset' class-attribute instance-attribute

ModalityConverterTask

Bases: StrEnum

Defines supported modality converter tasks.

AUTO = 'auto' class-attribute instance-attribute

CAPTIONING = 'captioning' class-attribute instance-attribute

MERMAID = 'mermaid' class-attribute instance-attribute

OCR = 'ocr' class-attribute instance-attribute

TRANSCRIPT = 'transcript' class-attribute instance-attribute

ModalityTransformerRouterPreset

Bases: StrEnum

Defines supported modality transformer router presets.

DOMAIN_SPECIFIC = 'domain_specific' class-attribute instance-attribute

MULTIMODAL = 'multimodal' class-attribute instance-attribute

ModalityTransformerType

Bases: StrEnum

Defines supported modality transformer types.

GENERIC = 'generic' class-attribute instance-attribute

STANDARD = 'standard' class-attribute instance-attribute

ModelProvider

Bases: StrEnum

Defines supported model providers for modality converters.

GOOGLE = 'google' class-attribute instance-attribute

GOOGLE_CLOUD = 'google_cloud' class-attribute instance-attribute

OPENAI = 'openai' class-attribute instance-attribute

OPENAI_CHAT_COMPLETIONS = 'openai-chat-completions' class-attribute instance-attribute

PROSA = 'prosa' class-attribute instance-attribute

QWEN = 'qwen' class-attribute instance-attribute

SNOWFLAKE = 'snowflake' class-attribute instance-attribute

YOUTUBE = 'youtube' class-attribute instance-attribute

OcrConstants

Constants for OCR operations in Gen AI applications.

BOX_2D_KEY = 'box_2d' class-attribute instance-attribute

MULTIMODAL_CONTEXT = 'multimodal_context' class-attribute instance-attribute

TEXT_KEY = 'text' class-attribute instance-attribute

TEXT_REGIONS_KEY = 'text_regions' class-attribute instance-attribute

TextConstants

Constants for text processing in Gen AI applications.

CR = '\r' class-attribute instance-attribute

CRLF = '\r\n' class-attribute instance-attribute

LINE_SEPARATOR = '\n' class-attribute instance-attribute

WhisperResponseKeys

String keys used when parsing OpenAI Whisper API responses.

AUDIO_TRANSCRIPTS = 'audio_transcripts' class-attribute instance-attribute

LANGUAGE = 'language' class-attribute instance-attribute

SEGMENTS = 'segments' class-attribute instance-attribute

TEXT = 'text' class-attribute instance-attribute

WORDS = 'words' class-attribute instance-attribute