--- dataset_info: - config_name: numpy__api_comp_matrix features: - name: api dtype: string - name: v_1.21 dtype: bool - name: v_1.22 dtype: bool - name: v_1.23 dtype: bool - name: v_1.24 dtype: bool - name: v_1.25 dtype: bool - name: v_1.26 dtype: bool - name: v_2.0 dtype: bool - name: v_2.1 dtype: bool - name: v_2.2 dtype: bool - name: v_2.3 dtype: bool splits: - name: train num_bytes: 79743 num_examples: 2566 download_size: 33866 dataset_size: 79743 - config_name: numpy__api_dictionary features: - name: api dtype: string - name: category dtype: string - name: doc_uri dtype: string - name: canonical_path dtype: string - name: entry_type dtype: string - name: signature_source dtype: string - name: first_version dtype: string - name: last_version dtype: string - name: is_constant dtype: bool - name: doc_versions dtype: string - name: online_doc_versions dtype: string splits: - name: train num_bytes: 51193308 num_examples: 2566 download_size: 5819733 dataset_size: 51193308 - config_name: numpy__benchmark_t1 features: - name: sample_id dtype: string - name: api_norm dtype: string - name: lib_version dtype: string - name: library dtype: string - name: group dtype: string - name: api_type dtype: string - name: prompt dtype: string - name: code_str dtype: string - name: target dtype: string - name: right_context_few_lines dtype: string - name: repo_url dtype: string - name: file_path dtype: string - name: signature_at_version dtype: string - name: docstring dtype: string splits: - name: train num_bytes: 330436307 num_examples: 72383 download_size: 55261029 dataset_size: 330436307 - config_name: numpy__benchmark_t2 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_exists dtype: bool - name: t2_balanced dtype: bool splits: - name: train num_bytes: 2446208 num_examples: 24565 download_size: 1206241 dataset_size: 2446208 - config_name: numpy__benchmark_t3 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_signature dtype: string splits: - name: train num_bytes: 3472175 num_examples: 22632 download_size: 987741 dataset_size: 3472175 - config_name: numpy__sign_comp_matrix features: - name: api dtype: string - name: signature dtype: string - name: v_1.21 dtype: bool - name: v_1.22 dtype: bool - name: v_1.23 dtype: bool - name: v_1.24 dtype: bool - name: v_1.25 dtype: bool - name: v_1.26 dtype: bool - name: v_2.0 dtype: bool - name: v_2.1 dtype: bool - name: v_2.2 dtype: bool - name: v_2.3 dtype: bool splits: - name: train num_bytes: 268130 num_examples: 2998 download_size: 86793 dataset_size: 268130 - config_name: scipy__api_comp_matrix features: - name: api dtype: string - name: v_1.7 dtype: bool - name: v_1.9 dtype: bool - name: v_1.10 dtype: bool - name: v_1.11 dtype: bool - name: v_1.12 dtype: bool - name: v_1.13 dtype: bool - name: v_1.14 dtype: bool - name: v_1.15 dtype: bool - name: v_1.16 dtype: bool splits: - name: train num_bytes: 130847 num_examples: 3642 download_size: 47964 dataset_size: 130847 - config_name: scipy__api_dictionary features: - name: api dtype: string - name: category dtype: string - name: doc_uri dtype: string - name: canonical_path dtype: string - name: entry_type dtype: string - name: signature_source dtype: string - name: first_version dtype: string - name: last_version dtype: string - name: is_constant dtype: bool - name: doc_versions dtype: string - name: online_doc_versions dtype: string splits: - name: train num_bytes: 82144033 num_examples: 3642 download_size: 10954923 dataset_size: 82144033 - config_name: scipy__benchmark_t1 features: - name: sample_id dtype: string - name: api_norm dtype: string - name: lib_version dtype: string - name: library dtype: string - name: group dtype: string - name: api_type dtype: string - name: prompt dtype: string - name: code_str dtype: string - name: target dtype: string - name: right_context_few_lines dtype: string - name: repo_url dtype: string - name: file_path dtype: string - name: signature_at_version dtype: string - name: docstring dtype: string splits: - name: train num_bytes: 43164077 num_examples: 7973 download_size: 6507377 dataset_size: 43164077 - config_name: scipy__benchmark_t2 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_exists dtype: bool - name: t2_balanced dtype: bool splits: - name: train num_bytes: 3458873 num_examples: 32501 download_size: 1615987 dataset_size: 3458873 - config_name: scipy__benchmark_t3 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_signature dtype: string splits: - name: train num_bytes: 5020231 num_examples: 28310 download_size: 1273343 dataset_size: 5020231 - config_name: scipy__sign_comp_matrix features: - name: api dtype: string - name: signature dtype: string - name: v_1.7 dtype: bool - name: v_1.9 dtype: bool - name: v_1.10 dtype: bool - name: v_1.11 dtype: bool - name: v_1.12 dtype: bool - name: v_1.13 dtype: bool - name: v_1.14 dtype: bool - name: v_1.15 dtype: bool - name: v_1.16 dtype: bool splits: - name: train num_bytes: 587437 num_examples: 4921 download_size: 146699 dataset_size: 587437 - config_name: torch__api_comp_matrix features: - name: api dtype: string - name: v_1.12 dtype: bool - name: v_1.13 dtype: bool - name: v_2.0 dtype: bool - name: v_2.1 dtype: bool - name: v_2.2 dtype: bool - name: v_2.3 dtype: bool - name: v_2.4 dtype: bool - name: v_2.5 dtype: bool - name: v_2.6 dtype: bool - name: v_2.7 dtype: bool splits: - name: train num_bytes: 215718 num_examples: 5178 download_size: 78321 dataset_size: 215718 - config_name: torch__api_dictionary features: - name: api dtype: string - name: category dtype: string - name: doc_uri dtype: string - name: canonical_path dtype: string - name: entry_type dtype: string - name: signature_source dtype: string - name: first_version dtype: string - name: last_version dtype: string - name: is_constant dtype: bool - name: doc_versions dtype: string - name: online_doc_versions dtype: string splits: - name: train num_bytes: 75716472 num_examples: 5178 download_size: 8849547 dataset_size: 75716472 - config_name: torch__benchmark_t1 features: - name: sample_id dtype: string - name: api_norm dtype: string - name: lib_version dtype: string - name: library dtype: string - name: group dtype: string - name: api_type dtype: string - name: prompt dtype: string - name: code_str dtype: string - name: target dtype: string - name: right_context_few_lines dtype: string - name: repo_url dtype: string - name: file_path dtype: string - name: signature_at_version dtype: string - name: docstring dtype: string splits: - name: train num_bytes: 648480639 num_examples: 170532 download_size: 105853265 dataset_size: 648480639 - config_name: torch__benchmark_t2 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_exists dtype: bool - name: t2_balanced dtype: bool splits: - name: train num_bytes: 5433054 num_examples: 49124 download_size: 2515118 dataset_size: 5433054 - config_name: torch__benchmark_t3 features: - name: sample_id dtype: string - name: api dtype: string - name: version dtype: string - name: library dtype: string - name: api_type dtype: string - name: gt_signature dtype: string splits: - name: train num_bytes: 8518823 num_examples: 41165 download_size: 1914070 dataset_size: 8518823 - config_name: torch__sign_comp_matrix features: - name: api dtype: string - name: signature dtype: string - name: v_1.12 dtype: bool - name: v_1.13 dtype: bool - name: v_2.0 dtype: bool - name: v_2.1 dtype: bool - name: v_2.2 dtype: bool - name: v_2.3 dtype: bool - name: v_2.4 dtype: bool - name: v_2.5 dtype: bool - name: v_2.6 dtype: bool - name: v_2.7 dtype: bool splits: - name: train num_bytes: 1272853 num_examples: 7171 download_size: 286837 dataset_size: 1272853 configs: - config_name: numpy__api_comp_matrix data_files: - split: train path: numpy__api_comp_matrix/train-* - config_name: numpy__api_dictionary data_files: - split: train path: numpy__api_dictionary/train-* - config_name: numpy__benchmark_t1 data_files: - split: train path: numpy__benchmark_t1/train-* - config_name: numpy__benchmark_t2 data_files: - split: train path: numpy__benchmark_t2/train-* - config_name: numpy__benchmark_t3 data_files: - split: train path: numpy__benchmark_t3/train-* - config_name: numpy__sign_comp_matrix data_files: - split: train path: numpy__sign_comp_matrix/train-* - config_name: scipy__api_comp_matrix data_files: - split: train path: scipy__api_comp_matrix/train-* - config_name: scipy__api_dictionary data_files: - split: train path: scipy__api_dictionary/train-* - config_name: scipy__benchmark_t1 data_files: - split: train path: scipy__benchmark_t1/train-* - config_name: scipy__benchmark_t2 data_files: - split: train path: scipy__benchmark_t2/train-* - config_name: scipy__benchmark_t3 data_files: - split: train path: scipy__benchmark_t3/train-* - config_name: scipy__sign_comp_matrix data_files: - split: train path: scipy__sign_comp_matrix/train-* - config_name: torch__api_comp_matrix data_files: - split: train path: torch__api_comp_matrix/train-* default: true - config_name: torch__api_dictionary data_files: - split: train path: torch__api_dictionary/train-* - config_name: torch__benchmark_t1 data_files: - split: train path: torch__benchmark_t1/train-* - config_name: torch__benchmark_t2 data_files: - split: train path: torch__benchmark_t2/train-* - config_name: torch__benchmark_t3 data_files: - split: train path: torch__benchmark_t3/train-* - config_name: torch__sign_comp_matrix data_files: - split: train path: torch__sign_comp_matrix/train-* --- # LibEvoBench (moved) > [!IMPORTANT] > This dataset has moved to **[JetBrains/LibEvoBench](https://huggingface.co/datasets/JetBrains/LibEvoBench)**. Please use that repository: it has the documentation, license, Croissant metadata and any future updates. This copy was the anonymous version used during review of *LibEvoBench: Probing Temporal Knowledge Stratification in Code Generation Models* (NeurIPS 2026, Evaluations & Datasets Track). It is kept unchanged so that existing links keep working. ```python from datasets import load_dataset ds = load_dataset("JetBrains/LibEvoBench", "torch__benchmark_t1", split="train") ``` Code: https://github.com/JetBrains-Research/LibEvoBench