Skip to content

FiftyOne dataset from LakeFS - Advanced

Category: data_generator.vision.detection.2d
Version: 1.0.0
UUID: e9fa416ae22b851219989365dd7be45360dbda4b2a15df61af785fa94de81865
Execution context: leip_af
Choice Priority: 1.0

Subcomponent Parameters

Name Synonyms Allowable Categories
Training Augmentation data.augmentation.training augmentation.vision.2d.training
Validation Augmentation data.augmentation.validation augmentation.vision.2d.validation

Value Parameters

Name Synonyms Type Values Help
Number of files per batch data.lakefs_batch_size scalar int Number of files to process in each batch when downloading from LakeFS
Download chunk size (in bytes) data.lakefs_chunk_size scalar int Size of each download chunk in bytes when fetching data from LakeFS
LakeFS commit ID that uniquely identifies the dataset version data.lakefs_commit_id scalar string Specify the commit ID in LakeFS that corresponds to the dataset version you want to use. This is the ultimate reproducibility identifier for the dataset in LakeFS.
Fiftyone Dataset Type data.fiftyone_dataset_type scalar string Specify the type of dataset in FiftyOne, e.g., 'detection', 'classification', etc.
Fiftyone Groundtruth Field Name data.fiftyone_groundtruth_field_name scalar string
Label data.fiftyone_label_map dict int Make sure the label_map is in str2idx format eg.
Maximum concurrent downloads data.lakefs_max_concurrent scalar int Maximum number of concurrent downloads when fetching data from LakeFS
LakeFS prefix data.lakefs_prefix scalar string
LakeFS repository name data.lakefs_repo scalar string
Use concurrent downloads data.lakefs_use_concurrent scalar bool Whether to use concurrent downloads when fetching data from LakeFS
Validate local data integrity data.lakefs_validate_integrity scalar bool Whether to validate local data integrity, if False will skip download if folder exists locally but wont verify validity.
Fiftyone Train View Name data.fiftyone_train_view scalar string
Fiftyone Val View Name data.fiftyone_val_view scalar string
Number of classes data.number_of_classes scalar int

Constraints

  1. Number of classes has to be positive (meta.nclasses > 0)
  2. LakeFS repository name must be provided (actual.repo != '')
  3. LakeFS commit ID must be provided (actual.commit_id != '')
  4. LakeFS prefix must be provided (actual.prefix != '')

This component fits into

Name UUID Synonyms
Composite - Class selector 06894b... data_generator.selection.underlying_dataset
Composite - MixUp Classification f19b45... data_generator.composite.mixup.dataset
Composite - Random subset feed1a... data_generator.subset.underlying_dataset
Composite - Data joiner 77071f... data_generator.composite.joiner.underlying_datasets
Composite - Data joiner b6f994... data_generator.composite.joiner.underlying_datasets
Composite - Matting 8889c0... data_generator.composite.matting.background
Composite - Matting 8889c0... data_generator.composite.matting.foreground
Composite - Class selector 10d356... data_generator.selection.underlying_dataset
Composite - Mosaic 7bebd5... data_generator.composite.mosaic.foreground
Composite - Data joiner 4e8148... data_generator.composite.joiner.underlying_datasets
Composite - Random subset 365151... data_generator.subset.underlying_dataset
Composite - Random subset 5859b6... data_generator.subset.underlying_dataset
Basic Adaptor 123217... data_generator