woah-ceianlp/initial_dataset_pos_oversample

--- dataset_info: features: - name: original dtype: string - name: prompt_strategy dtype: string - name: system dtype: string - name: user dtype: string - name: content dtype: string - name: completion_tokens dtype: int64 - name: prompt_tokens dtype: int64 - name: is_toxic dtype: int64 - name: sample_size dtype: string - name: generation_model dtype: string splits: - name: train num_bytes: 38966001 num_examples: 8000 download_size: 1670603 dataset_size: 38966001 configs: - config_name: default data_files: - split: train path: data/train-* ---