--- dataset_info: - config_name: score_0 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 600515529750 num_examples: 132382614 download_size: 326318508165 dataset_size: 600515529750 - config_name: score_1 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 49424526987 num_examples: 8033860 download_size: 27531332308 dataset_size: 49424526987 - config_name: score_2 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 257652727098 num_examples: 29874316 download_size: 141766458398 dataset_size: 257652727098 - config_name: score_3 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 35728962911 num_examples: 2704901 download_size: 19906353964 dataset_size: 35728962911 - config_name: score_4 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 35763743641 num_examples: 2102170 download_size: 20092149434 dataset_size: 35763743641 - config_name: score_5 features: - name: id dtype: string - name: text dtype: string - name: safety_metadata struct: - name: final_harm_score dtype: int64 - name: embedding_harm_score dtype: int64 - name: llm_harm_score dtype: int64 - name: llm_score_assessment_by_chunk sequence: string - name: llm_harm_scores_by_chunk list: - name: custom_id dtype: string - name: llm_safety_score dtype: int64 - name: reason dtype: string - name: metadata struct: - name: dump dtype: string - name: url dtype: string - name: date dtype: timestamp[s] - name: file_path dtype: string - name: language dtype: string - name: language_score dtype: float64 - name: token_count dtype: int64 - name: score dtype: float64 - name: int_score dtype: int64 splits: - name: train num_bytes: 14872107115 num_examples: 704280 download_size: 8340674635 dataset_size: 14872107115 configs: - config_name: score_0 data_files: - split: train path: score_0/train-* - config_name: score_1 data_files: - split: train path: score_1/train-* - config_name: score_2 data_files: - split: train path: score_2/train-* - config_name: score_3 data_files: - split: train path: score_3/train-* - config_name: score_4 data_files: - split: train path: score_4/train-* - config_name: score_5 data_files: - split: train path: score_5/train-* ---