bridge.data.vlm_datasets

VLM dataset utilities.

Public API re-exports:

• Makers: functions to build conversation examples from HF datasets

• Providers: classes that build PyTorch datasets bound to HF processors

• Collate fns: model-specific batch builders

Subpackages

• bridge.data.vlm_datasets.step37_flickr8k
  • bridge.data.vlm_datasets.step37_flickr8k.flickr8k_loader
  • bridge.data.vlm_datasets.step37_flickr8k.preprocess
  • bridge.data.vlm_datasets.step37_flickr8k.samplers
  • bridge.data.vlm_datasets.step37_flickr8k.packing
  • bridge.data.vlm_datasets.step37_flickr8k.pack_transform
  • bridge.data.vlm_datasets.step37_flickr8k.template
  • bridge.data.vlm_datasets.step37_flickr8k.multimodal_utils
  • bridge.data.vlm_datasets.step37_flickr8k.provider
  • bridge.data.vlm_datasets.step37_flickr8k.packed_dataloader

Submodules

• bridge.data.vlm_datasets.hf_provider
• bridge.data.vlm_datasets.token_utils
• bridge.data.vlm_datasets.collate
• bridge.data.vlm_datasets.preloaded_provider
• bridge.data.vlm_datasets.hf_dataset_makers
• bridge.data.vlm_datasets.conversation_dataset
• bridge.data.vlm_datasets.mock_provider

Package Contents

Data

__all__

API

bridge.data.vlm_datasets.__all__

[‘make_rdr_dataset’, ‘make_cord_v2_dataset’, ‘make_medpix_dataset’, ‘make_cv17_dataset’, ‘make_raven…