nemo_automodel.components.datasets.multimodal.loader
nemo_automodel.components.datasets.multimodal.loader
Typed construction for the packed BAGEL multimodal dataloader.
Module Contents
Classes
Data
API
Dataclass
Materialized BAGEL dataloader and its packed dataset.
dataloader
dataset
Dataclass
Construction-time configuration for the complete BAGEL input pipeline.
dataset_config
num_workers
pin_memory
prefetch_factor
Build the packed BAGEL dataset and its stateful dataloader.
Parameters:
tokenizer
Runtime tokenizer used by the packed dataset.
special_tokens
Runtime BAGEL special-token mapping.
rank
Global distributed rank used for source-data sharding and worker seeding.
world_size
Global distributed world size.
batch_size
Runtime local batch size; BAGEL packed samples require one row per step.
global_seed
Global training seed used to initialize dataloader worker RNG state.
Returns: BagelDataloaderBuild
Named result containing the dataloader and its packed dataset.