Datasets:

Modalities:
Text
Formats:
parquet
Libraries:
Datasets
Dask
BOLD-Embeddings / README.md
vshulev's picture
Upload dataset
e2d3725 verified
metadata
dataset_info:
  features:
    - name: processid
      dtype: string
    - name: embeddings
      sequence: float32
  splits:
    - name: train
      num_bytes: 14332206772
      num_examples: 4636068
  download_size: 17138874377
  dataset_size: 14332206772
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

DNA Embeddings produced by LofiAmazon/BarcodeBERT-Entire-BOLD.

Features:

  • processid The process ID from BOLD
  • embeddings The DNA embeddings produced by the model

NOTE: This is a small sample of all data available in BOLD. Generating embeddings for the entire BOLD database would require ~20GB of space.