safety-aya/Nemotron-Safety-Guard-Dataset-v3-Telugu-safety

--- dataset_info: features: - name: id dtype: string - name: prompt dtype: string - name: response dtype: string - name: prompt_label dtype: string - name: response_label dtype: string - name: prompt_label_source dtype: string - name: response_label_source dtype: string - name: tag dtype: string - name: language dtype: string - name: reconstruction_id_if_redacted dtype: float64 - name: prompt_english dtype: string - name: response_english dtype: string - name: prompt_telugu dtype: string - name: response_telugu dtype: string - name: safety_label dtype: string - name: safety_categories list: string - name: safety_severity dtype: string - name: safety_confidence dtype: 'null' - name: safety_reason dtype: 'null' - name: safety_words list: 'null' - name: safety_model dtype: 'null' - name: safety_scored_at dtype: 'null' - name: safety_reasoning dtype: 'null' splits: - name: train num_bytes: 129500509 num_examples: 37645 download_size: 56616543 dataset_size: 129500509 configs: - config_name: default data_files: - split: train path: data/train-* ---