遇见数据集

open-llm-leaderboard-old/details_RWKV__rwkv-4-14b-pile

收藏
Hugging Face2023-10-15 更新2024-06-22 收录
官方服务:

资源简介:

--- pretty_name: Evaluation run of RWKV/rwkv-4-14b-pile dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [RWKV/rwkv-4-14b-pile](https://huggingface.co/RWKV/rwkv-4-14b-pile) on the [Open\ \ LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard).\n\ \nThe dataset is composed of 64 configuration, each one coresponding to one of the\ \ evaluated task.\n\nThe dataset has been created from 2 run(s). Each run can be\ \ found as a specific split in each configuration, the split being named using the\ \ timestamp of the run.The \"train\" split is always pointing to the latest results.\n\ \nAn additional configuration \"results\" store all the aggregated results of the\ \ run (and is used to compute and display the agregated metrics on the [Open LLM\ \ Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)).\n\ \nTo load the details from a run, you can for instance do the following:\n```python\n\ from datasets import load_dataset\ndata = load_dataset(\"open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile\"\ ,\n\t\"harness_winogrande_5\",\n\tsplit=\"train\")\n```\n\n## Latest results\n\n\ These are the [latest results from run 2023-10-15T21:19:09.628553](https://huggingface.co/datasets/open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile/blob/main/results_2023-10-15T21-19-09.628553.json)(note\ \ that their might be results for other tasks in the repos if successive evals didn't\ \ cover the same tasks. You find each in the results and the \"latest\" split for\ \ each eval):\n\n```python\n{\n \"all\": {\n \"em\": 0.0017827181208053692,\n\ \ \"em_stderr\": 0.0004320097346039099,\n \"f1\": 0.045845218120805505,\n\ \ \"f1_stderr\": 0.0012042308580297368,\n \"acc\": 0.32904612508698977,\n\ \ \"acc_stderr\": 0.007529648846042276\n },\n \"harness|drop|3\": {\n\ \ \"em\": 0.0017827181208053692,\n \"em_stderr\": 0.0004320097346039099,\n\ \ \"f1\": 0.045845218120805505,\n \"f1_stderr\": 0.0012042308580297368\n\ \ },\n \"harness|gsm8k|5\": {\n \"acc\": 0.0037907505686125853,\n \ \ \"acc_stderr\": 0.0016927007401501789\n },\n \"harness|winogrande|5\"\ : {\n \"acc\": 0.654301499605367,\n \"acc_stderr\": 0.013366596951934373\n\ \ }\n}\n```" repo_url: https://huggingface.co/RWKV/rwkv-4-14b-pile leaderboard_url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard point_of_contact: clementine@hf.co configs: - config_name: harness_arc_challenge_25 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|arc:challenge|25_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|arc:challenge|25_2023-09-03T13:55:36.441206.parquet' - config_name: harness_drop_3 data_files: - split: 2023_10_15T21_19_09.628553 path: - '**/details_harness|drop|3_2023-10-15T21-19-09.628553.parquet' - split: latest path: - '**/details_harness|drop|3_2023-10-15T21-19-09.628553.parquet' - config_name: harness_gsm8k_5 data_files: - split: 2023_10_15T21_19_09.628553 path: - '**/details_harness|gsm8k|5_2023-10-15T21-19-09.628553.parquet' - split: latest path: - '**/details_harness|gsm8k|5_2023-10-15T21-19-09.628553.parquet' - config_name: harness_hellaswag_10 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hellaswag|10_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hellaswag|10_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-management|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-anatomy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-astronomy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-business_ethics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_biology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_chemistry|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_computer_science|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_medicine|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-college_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-computer_security|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-econometrics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-formal_logic|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-global_facts|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_biology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_geography|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_physics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-human_aging|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-human_sexuality|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-international_law|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-jurisprudence|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-machine_learning|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-management|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-marketing|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-medical_genetics|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-miscellaneous|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-moral_disputes|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-nutrition|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-philosophy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-prehistory|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_accounting|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_law|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_medicine|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-professional_psychology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-public_relations|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-security_studies|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-sociology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-virology|5_2023-09-03T13:55:36.441206.parquet' - '**/details_harness|hendrycksTest-world_religions|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_abstract_algebra_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_anatomy_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-anatomy|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-anatomy|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_astronomy_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-astronomy|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-astronomy|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_business_ethics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-business_ethics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_clinical_knowledge_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_biology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_biology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_biology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_chemistry_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_chemistry|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_computer_science_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_computer_science|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_mathematics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_mathematics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_medicine_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_medicine|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_college_physics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-college_physics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-college_physics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_computer_security_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-computer_security|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-computer_security|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_conceptual_physics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_econometrics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-econometrics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-econometrics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_electrical_engineering_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_elementary_mathematics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_formal_logic_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-formal_logic|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_global_facts_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-global_facts|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-global_facts|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_biology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_biology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_chemistry_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_computer_science_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_european_history_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_geography_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_geography|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_government_and_politics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_macroeconomics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_mathematics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_microeconomics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_physics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_physics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_psychology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_statistics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_us_history_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_high_school_world_history_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_human_aging_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-human_aging|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_aging|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_human_sexuality_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-human_sexuality|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_international_law_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-international_law|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-international_law|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_jurisprudence_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-jurisprudence|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_logical_fallacies_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_machine_learning_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-machine_learning|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_management_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-management|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-management|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_marketing_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-marketing|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-marketing|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_medical_genetics_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-medical_genetics|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_miscellaneous_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-miscellaneous|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_moral_disputes_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_disputes|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_moral_scenarios_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-moral_scenarios|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_nutrition_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-nutrition|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-nutrition|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_philosophy_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-philosophy|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-philosophy|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_prehistory_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-prehistory|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-prehistory|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_professional_accounting_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_accounting|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_professional_law_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-professional_law|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_law|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_professional_medicine_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_medicine|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_professional_psychology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-professional_psychology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_public_relations_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-public_relations|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-public_relations|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_security_studies_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-security_studies|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-security_studies|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_sociology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-sociology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-sociology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_us_foreign_policy_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-us_foreign_policy|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_virology_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-virology|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-virology|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_hendrycksTest_world_religions_5 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|hendrycksTest-world_religions|5_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|hendrycksTest-world_religions|5_2023-09-03T13:55:36.441206.parquet' - config_name: harness_truthfulqa_mc_0 data_files: - split: 2023_09_03T13_55_36.441206 path: - '**/details_harness|truthfulqa:mc|0_2023-09-03T13:55:36.441206.parquet' - split: latest path: - '**/details_harness|truthfulqa:mc|0_2023-09-03T13:55:36.441206.parquet' - config_name: harness_winogrande_5 data_files: - split: 2023_10_15T21_19_09.628553 path: - '**/details_harness|winogrande|5_2023-10-15T21-19-09.628553.parquet' - split: latest path: - '**/details_harness|winogrande|5_2023-10-15T21-19-09.628553.parquet' - config_name: results data_files: - split: 2023_09_03T13_55_36.441206 path: - results_2023-09-03T13:55:36.441206.parquet - split: 2023_10_15T21_19_09.628553 path: - results_2023-10-15T21-19-09.628553.parquet - split: latest path: - results_2023-10-15T21-19-09.628553.parquet --- # Dataset Card for Evaluation run of RWKV/rwkv-4-14b-pile ## Dataset Description - **Homepage:** - **Repository:** https://huggingface.co/RWKV/rwkv-4-14b-pile - **Paper:** - **Leaderboard:** https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard - **Point of Contact:** clementine@hf.co ### Dataset Summary Dataset automatically created during the evaluation run of model [RWKV/rwkv-4-14b-pile](https://huggingface.co/RWKV/rwkv-4-14b-pile) on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard). The dataset is composed of 64 configuration, each one coresponding to one of the evaluated task. The dataset has been created from 2 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results. An additional configuration "results" store all the aggregated results of the run (and is used to compute and display the agregated metrics on the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)). To load the details from a run, you can for instance do the following: ```python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile", "harness_winogrande_5", split="train") ``` ## Latest results These are the [latest results from run 2023-10-15T21:19:09.628553](https://huggingface.co/datasets/open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile/blob/main/results_2023-10-15T21-19-09.628553.json)(note that their might be results for other tasks in the repos if successive evals didn't cover the same tasks. You find each in the results and the "latest" split for each eval): ```python { "all": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346039099, "f1": 0.045845218120805505, "f1_stderr": 0.0012042308580297368, "acc": 0.32904612508698977, "acc_stderr": 0.007529648846042276 }, "harness|drop|3": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346039099, "f1": 0.045845218120805505, "f1_stderr": 0.0012042308580297368 }, "harness|gsm8k|5": { "acc": 0.0037907505686125853, "acc_stderr": 0.0016927007401501789 }, "harness|winogrande|5": { "acc": 0.654301499605367, "acc_stderr": 0.013366596951934373 } } ``` ### Supported Tasks and Leaderboards [More Information Needed] ### Languages [More Information Needed] ## Dataset Structure ### Data Instances [More Information Needed] ### Data Fields [More Information Needed] ### Data Splits [More Information Needed] ## Dataset Creation ### Curation Rationale [More Information Needed] ### Source Data #### Initial Data Collection and Normalization [More Information Needed] #### Who are the source language producers? [More Information Needed] ### Annotations #### Annotation process [More Information Needed] #### Who are the annotators? [More Information Needed] ### Personal and Sensitive Information [More Information Needed] ## Considerations for Using the Data ### Social Impact of Dataset [More Information Needed] ### Discussion of Biases [More Information Needed] ### Other Known Limitations [More Information Needed] ## Additional Information ### Dataset Curators [More Information Needed] ### Licensing Information [More Information Needed] ### Citation Information [More Information Needed] ### Contributions [More Information Needed]

原始信息汇总

数据集概述

该数据集是在对模型 RWKV/rwkv-4-14b-pile 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。

数据集组成

  • 数据集包含 64 个配置,每个配置对应一个评估任务。
  • 数据集从 2 次运行中创建,每次运行可以在每个配置中作为一个特定的分片找到,分片名称使用运行的时间戳。
  • "train" 分片始终指向最新的结果。
  • 一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示 Open LLM Leaderboard 上的聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-15T21:19:09.628553 运行的最新结果: python { "all": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346039099, "f1": 0.045845218120805505, "f1_stderr": 0.0012042308580297368, "acc": 0.32904612508698977, "acc_stderr": 0.007529648846042276 }, "harness|drop|3": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346039099, "f1": 0.045845218120805505, "f1_stderr": 0.0012042308580297368 }, "harness|gsm8k|5": { "acc": 0.0037907505686125853, "acc_stderr": 0.0016927007401501789 }, "harness|winogrande|5": { "acc": 0.654301499605367, "acc_stderr": 0.013366596951934373 } }

配置详情

  • harness_arc_challenge_25

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|arc:challenge|25_2023-09-03T13:55:36.441206.parquet
  • harness_drop_3

    • 分片:2023_10_15T21_19_09.628553, latest
    • 路径:**/details_harness|drop|3_2023-10-15T21-19-09.628553.parquet
  • harness_gsm8k_5

    • 分片:2023_10_15T21_19_09.628553, latest
    • 路径:**/details_harness|gsm8k|5_2023-10-15T21-19-09.628553.parquet
  • harness_hellaswag_10

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hellaswag|10_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:多个文件,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_anatomy_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_astronomy_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_biology_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_college_physics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_computer_security_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_econometrics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-03T13:55:36.441206.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分片:2023_09_03T13_55_36.441206, latest
    • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-03T13:55:36.441206.parquet
搜集汇总
数据集介绍
open-llm-leaderboard-old/details_RWKV__rwkv-4-14b-pile 数据集图片
构建方式
在大型语言模型评估的背景下,该数据集是专为记录和存储RWKV/rwkv-4-14b-pile模型在Open LLM Leaderboard上的评测细节而自动生成的。数据集构建过程依托于评测框架,共包含64个配置,每个配置对应一个独立的评测任务。这些任务涵盖ARC挑战、DROP、GSM8K、HellaSwag、Winogrande以及涵盖57个学科的MMLU(HendrycksTest)等经典基准。每次评测运行的结果被组织为独立的分片,以时间戳命名,而'train'分片始终指向最新一次的评测结果。此外,数据集还设有'results'配置,用于汇聚所有运行的综合指标,便于在排行榜上计算和展示聚合性能。
特点
该数据集的核心特色在于其精细化的结构化设计与版本追溯能力。它通过64个独立配置分别存储不同任务的评测细节,确保了数据的组织性与可检索性。每次评测运行均生成一个以时间戳命名的独特分片,实现了对模型性能变化轨迹的完整记录,便于研究人员进行历史对比与回归分析。'train'分片自动指向最新结果,简化了获取最新性能数据的流程。同时,'results'配置提供了全局聚合指标,使得跨任务的综合评估一目了然。数据以高效的Parquet格式存储,兼顾了存储效率与读取速度。
使用方法
使用者可通过Hugging Face的datasets库灵活加载该数据集。具体而言,调用load_dataset函数并指定数据集名称(如'open-llm-leaderboard/details_RWKV__rwkv-4-14b-pile'),同时传入目标任务的配置名称(例如'harness_winogrande_5')和所需的分片(如'train'分片以获取最新结果),即可便捷地获取特定任务的评测细节。若需访问历史运行数据,则可通过指定相应时间戳的分片名称来加载。这种设计支持对模型在不同时间点、不同任务上的表现进行精细化分析与复现,为模型迭代与评估提供了坚实的数据基础。
背景与挑战
背景概述
在大型语言模型(LLM)蓬勃发展的时代,如何系统且公正地评估模型性能成为学界与工业界共同关注的焦点。Open LLM Leaderboard 由 Hugging Face 团队于2023年发起,旨在为开源社区提供一个标准化、透明化的模型评测平台。该数据集记录了 RWKV-4-14B-Pile 模型在 leaderboard 上的完整评估结果,涵盖 ARC、HellaSwag、MMLU、Winogrande、GSM8K 及 DROP 等多项代表性任务。作为一款基于循环架构的百亿参数模型,RWKV 在 Transformer 主导的生态中独树一帜,其评测数据不仅反映了该模型在推理、常识与数学等多维度的能力边界,也为研究者对比不同架构的语言模型提供了珍贵的基准参考,推动了开源大模型评估体系的完善与普及。
当前挑战
该数据集所解决的领域问题在于,当前缺乏统一且可复现的 LLM 评测平台,不同研究团队常采用差异化的测试集与评估流程,导致模型性能难以横向对比。Open LLM Leaderboard 通过固定任务集与评测脚本,力求消除评估中的歧义与偏差。构建过程中则面临多重挑战:其一,需确保评测环境的硬件与软件配置严格一致,避免因计算资源差异引入噪声;其二,跨任务数据格式的归一化处理极为繁琐,例如 MMLU 涵盖57个子学科,每个子领域需独立适配;其三,评测结果的时间戳与版本管理要求精细,多次运行的结果需通过分片与快照机制妥善归档,以支持后续的溯源与复现分析。
常用场景
经典使用场景
在大型语言模型(LLM)评估领域,该数据集作为Open LLM Leaderboard的标准化评测记录,被广泛用于量化模型在多样化自然语言理解与推理任务上的表现。其经典使用场景涵盖从常识推理(如Winogrande)到数学求解(如GSM8K)的多维度能力评估,研究者通过加载各配置项(如harness_arc_challenge_25)下的详细结果,实现对模型性能的细粒度剖析与横向比较。
解决学术问题
该数据集有效回应了LLM领域长期存在的评估基准碎片化与不可复现问题。通过汇集ARC、HellaSwag、MMLU等主流评测任务的标准化结果,它构建了统一的性能度量体系,使学者能够系统性地探究模型在知识理解、逻辑推理与领域适应性上的优劣。其意义在于推动模型性能对比的透明化与公平性,为后续架构创新与训练策略优化提供了可靠的参照基石。
衍生相关工作
该数据集衍生了多项重要工作,包括基于其评测结果改进的RWKV架构优化研究,以及利用其细粒度数据训练性能预测器的工作。此外,它启发了多任务评估框架的标准化设计,如后续版本中引入的少样本学习与对抗性评测扩展。这些工作共同深化了人们对LLM能力边界的认知,并催生了诸如模型压缩与知识蒸馏等方向的新方法论探索。
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务