Datasets:

cfahlgren1
/

class-label-string-token-repro

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train.csv
dataset_info:
  features:
    - name: text
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            '0': negative
            '1': positive

class-label-string-token-repro

Minimal repro dataset for class-label columns whose source CSV token is a label name instead of an integer id.