Name: HadiUMN/TOFU
Creator: HadiUMN
Published: 2025-12-08 18:28:01
License: 暂无描述

下载链接：

https://hf-mirror.com/datasets/HadiUMN/TOFU

下载链接

链接失效反馈

官方服务：

资源简介：

--- dataset_info: - config_name: forget01 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 10806 num_examples: 40 download_size: 7947 dataset_size: 10806 - config_name: forget01_1 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 5403.0 num_examples: 20 download_size: 5375 dataset_size: 5403.0 - config_name: forget01_10 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 1080.6 num_examples: 4 download_size: 2922 dataset_size: 1080.6 - config_name: forget01_10_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 6758.9 num_examples: 4 download_size: 8619 dataset_size: 6758.9 - config_name: forget01_1_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 33794.5 num_examples: 20 download_size: 20159 dataset_size: 33794.5 - config_name: forget01_2 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 5403.0 num_examples: 20 download_size: 5327 dataset_size: 5403.0 - config_name: forget01_2_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 33794.5 num_examples: 20 download_size: 24273 dataset_size: 33794.5 - config_name: forget01_90 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 9725.4 num_examples: 36 download_size: 7652 dataset_size: 9725.4 - config_name: forget01_90_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 60830.1 num_examples: 36 download_size: 34538 dataset_size: 60830.1 - config_name: forget01_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 67589 num_examples: 40 download_size: 36213 dataset_size: 67589 - config_name: forget05 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 52258 num_examples: 200 download_size: 28850 dataset_size: 52258 - config_name: forget05_1 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 26129.0 num_examples: 100 download_size: 16686 dataset_size: 26129.0 - config_name: forget05_10 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 5225.8 num_examples: 20 download_size: 5216 dataset_size: 5225.8 - config_name: forget05_10_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 31900.2 num_examples: 20 download_size: 19872 dataset_size: 31900.2 - config_name: forget05_1_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 159501.0 num_examples: 100 download_size: 76095 dataset_size: 159501.0 - config_name: forget05_2 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 26129.0 num_examples: 100 download_size: 17839 dataset_size: 26129.0 - config_name: forget05_2_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 159501.0 num_examples: 100 download_size: 87118 dataset_size: 159501.0 - config_name: forget05_90 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 47032.2 num_examples: 180 download_size: 27500 dataset_size: 47032.2 - config_name: forget05_90_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 287101.8 num_examples: 180 download_size: 139138 dataset_size: 287101.8 - config_name: forget05_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 319002 num_examples: 200 download_size: 148383 dataset_size: 319002 - config_name: forget10 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 106524 num_examples: 400 download_size: 55820 dataset_size: 106524 - config_name: forget10_1 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 53262.0 num_examples: 200 download_size: 31850 dataset_size: 53262.0 - config_name: forget10_10 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 10652.4 num_examples: 40 download_size: 9630 dataset_size: 10652.4 - config_name: forget10_10_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 64438.1 num_examples: 40 download_size: 38833 dataset_size: 64438.1 - config_name: forget10_1_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 322190.5 num_examples: 200 download_size: 157241 dataset_size: 322190.5 - config_name: forget10_2 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 53262.0 num_examples: 200 download_size: 32120 dataset_size: 53262.0 - config_name: forget10_2_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 322190.5 num_examples: 200 download_size: 156288 dataset_size: 322190.5 - config_name: forget10_90 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 95871.6 num_examples: 360 download_size: 52878 dataset_size: 95871.6 - config_name: forget10_90_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 579942.9 num_examples: 360 download_size: 272037 dataset_size: 579942.9 - config_name: forget10_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 644381 num_examples: 400 download_size: 291140 dataset_size: 644381 - config_name: full features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 990798 num_examples: 4000 download_size: 499931 dataset_size: 990798 - config_name: holdout01 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 11416 num_examples: 40 download_size: 9226 dataset_size: 11416 - config_name: holdout05 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 59923 num_examples: 200 download_size: 32088 dataset_size: 59923 - config_name: holdout10 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 122394 num_examples: 400 download_size: 63227 dataset_size: 122394 - config_name: real_authors features: - name: question dtype: string - name: option1 dtype: string - name: option2 dtype: string - name: option3 dtype: string - name: option4 dtype: string - name: answer dtype: string splits: - name: train num_bytes: 15067 num_examples: 100 download_size: 13788 dataset_size: 15067 - config_name: real_authors_perturbed features: - name: question dtype: string - name: answer dtype: string - name: perturbed_answer sequence: string splits: - name: train num_bytes: 13677 num_examples: 100 download_size: 10414 dataset_size: 13677 - config_name: retain90 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 884274 num_examples: 3600 download_size: 447221 dataset_size: 884274 - config_name: retain95 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 938540 num_examples: 3800 download_size: 474083 dataset_size: 938540 - config_name: retain99 features: - name: question dtype: string - name: answer dtype: string splits: - name: train num_bytes: 979992 num_examples: 3960 download_size: 495220 dataset_size: 979992 - config_name: retain_perturbed features: - name: question dtype: string - name: answer dtype: string - name: paraphrased_answer dtype: string - name: perturbed_answer sequence: string - name: paraphrased_question dtype: string splits: - name: train num_bytes: 609892 num_examples: 400 download_size: 282368 dataset_size: 609892 - config_name: world_facts features: - name: question dtype: string - name: option1 dtype: string - name: option2 dtype: string - name: option3 dtype: string - name: option4 dtype: string - name: answer dtype: string splits: - name: train num_bytes: 15244 num_examples: 117 download_size: 14278 dataset_size: 15244 - config_name: world_facts_perturbed features: - name: question dtype: string - name: answer dtype: string - name: perturbed_answer sequence: string splits: - name: train num_bytes: 14067 num_examples: 117 download_size: 11424 dataset_size: 14067 configs: - config_name: forget01 data_files: - split: train path: forget01/train-* - config_name: forget01_1 data_files: - split: train path: forget01_1/train-* - config_name: forget01_10 data_files: - split: train path: forget01_10/train-* - config_name: forget01_10_perturbed data_files: - split: train path: forget01_10_perturbed/train-* - config_name: forget01_1_perturbed data_files: - split: train path: forget01_1_perturbed/train-* - config_name: forget01_2 data_files: - split: train path: forget01_2/train-* - config_name: forget01_2_perturbed data_files: - split: train path: forget01_2_perturbed/train-* - config_name: forget01_90 data_files: - split: train path: forget01_90/train-* - config_name: forget01_90_perturbed data_files: - split: train path: forget01_90_perturbed/train-* - config_name: forget01_perturbed data_files: - split: train path: forget01_perturbed/train-* - config_name: forget05 data_files: - split: train path: forget05/train-* - config_name: forget05_1 data_files: - split: train path: forget05_1/train-* - config_name: forget05_10 data_files: - split: train path: forget05_10/train-* - config_name: forget05_10_perturbed data_files: - split: train path: forget05_10_perturbed/train-* - config_name: forget05_1_perturbed data_files: - split: train path: forget05_1_perturbed/train-* - config_name: forget05_2 data_files: - split: train path: forget05_2/train-* - config_name: forget05_2_perturbed data_files: - split: train path: forget05_2_perturbed/train-* - config_name: forget05_90 data_files: - split: train path: forget05_90/train-* - config_name: forget05_90_perturbed data_files: - split: train path: forget05_90_perturbed/train-* - config_name: forget05_perturbed data_files: - split: train path: forget05_perturbed/train-* - config_name: forget10 data_files: - split: train path: forget10/train-* - config_name: forget10_1 data_files: - split: train path: forget10_1/train-* - config_name: forget10_10 data_files: - split: train path: forget10_10/train-* - config_name: forget10_10_perturbed data_files: - split: train path: forget10_10_perturbed/train-* - config_name: forget10_1_perturbed data_files: - split: train path: forget10_1_perturbed/train-* - config_name: forget10_2 data_files: - split: train path: forget10_2/train-* - config_name: forget10_2_perturbed data_files: - split: train path: forget10_2_perturbed/train-* - config_name: forget10_90 data_files: - split: train path: forget10_90/train-* - config_name: forget10_90_perturbed data_files: - split: train path: forget10_90_perturbed/train-* - config_name: forget10_perturbed data_files: - split: train path: forget10_perturbed/train-* - config_name: full data_files: - split: train path: full/train-* - config_name: holdout01 data_files: - split: train path: holdout01/train-* - config_name: holdout05 data_files: - split: train path: holdout05/train-* - config_name: holdout10 data_files: - split: train path: holdout10/train-* - config_name: real_authors data_files: - split: train path: real_authors/train-* - config_name: real_authors_perturbed data_files: - split: train path: real_authors_perturbed/train-* - config_name: retain90 data_files: - split: train path: retain90/train-* - config_name: retain95 data_files: - split: train path: retain95/train-* - config_name: retain99 data_files: - split: train path: retain99/train-* - config_name: retain_perturbed data_files: - split: train path: retain_perturbed/train-* - config_name: world_facts data_files: - split: train path: world_facts/train-* - config_name: world_facts_perturbed data_files: - split: train path: world_facts_perturbed/train-* ---

应用场景：