Overview

Utility functions for multimodal modules.

This package provides shared utility functions used across the gllm_multimodal package for media processing, file handling, and integration with external services.

Submodules

• media_utils -- Media binary retrieval and validation utilities.
• media_profiler -- Basic MIME and modality profiling for media bytes.
• image_utils -- Image processing helpers.
• video_utils -- Video processing helpers.
• video_frame_utils -- Frame decoding and resizing helpers for video analysis.
• morphology_utils -- Boolean-mask dilation and connected-region helpers.
• audio_to_text_utils -- Audio-to-text conversion utilities.
• api_utils -- API integration utilities.
• glair_vision_utils -- GLAIR Vision OCR request, response validation, and mapping helpers.
• time_utils -- Time and duration helpers.
• error_utils -- Helpers that keep errors inside the multimodal error taxonomy, including the mapping of legacy gllm-inference invoker errors.
• safe_logger -- Safe logging utilities that handle encoding issues.
• preset_utils -- Preset configuration management.
• source_utils -- Source resolution utilities.
• converter_result_utils -- Converter result formatting helpers.
• image_metadata_utils -- EXIF and image metadata extraction.
• model_utils -- Verified model-download and cache helpers.
• gdrive_utils -- Google Drive integration.
• s3_utils -- Amazon S3 integration.
• text_utils -- Text processing helpers.