pretraining-poisoning/declarative-genre12-dot-100M

--- dataset_info: features: - name: type dtype: string - name: genre dtype: string - name: text dtype: string - name: token_count dtype: int64 - name: subtopic dtype: string - name: template_idx dtype: int64 splits: - name: train num_bytes: 451481588 num_examples: 689136 download_size: 144006437 dataset_size: 451481588 configs: - config_name: default data_files: - split: train path: data/train-* ---