open-llm-leaderboard-old/details_openagi-project__OpenAGI-7B-v0.1
收藏数据集概述
数据集摘要
该数据集是在对模型 openagi-project/OpenAGI-7B-v0.1 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。数据集包含 63 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,每个运行可以在每个配置中找到特定的分割,分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。
数据集加载
要加载特定运行的详细信息,可以使用以下代码: python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_openagi-project__OpenAGI-7B-v0.1", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的示例: python { "all": { "acc": 0.6376909717416875, "acc_stderr": 0.032524006961840754, "acc_norm": 0.6396432481543075, "acc_norm_stderr": 0.033178412950361426, "mc1": 0.5189718482252142, "mc1_stderr": 0.017490896405762353, "mc2": 0.6955285302922068, "mc2_stderr": 0.015057872155924216 }, "harness|arc:challenge|25": { "acc": 0.6390784982935154, "acc_stderr": 0.014034761386175452, "acc_norm": 0.6672354948805461, "acc_norm_stderr": 0.013769863046192305 }, "harness|hellaswag|10": { "acc": 0.6885082652857997, "acc_stderr": 0.0046215681251020446, "acc_norm": 0.8612826130252937, "acc_norm_stderr": 0.003449449618650549 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695236, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695236 }, # 其他任务的结果... }
配置详情
数据集包含多个配置,每个配置对应不同的任务和分割。以下是部分配置的示例:
-
配置名称:
harness_arc_challenge_25- 数据文件:
- 分割:
2024_01_21T20_52_03.715367- 路径:
**/details_harness|arc:challenge|25_2024-01-21T20-52-03.715367.parquet
- 路径:
- 分割:
2024_01_26T17_39_46.193538- 路径:
**/details_harness|arc:challenge|25_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|arc:challenge|25_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_gsm8k_5- 数据文件:
- 分割:
2024_01_21T20_52_03.715367- 路径:
**/details_harness|gsm8k|5_2024-01-21T20-52-03.715367.parquet
- 路径:
- 分割:
2024_01_26T17_39_46.193538- 路径:
**/details_harness|gsm8k|5_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|gsm8k|5_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hellaswag_10- 数据文件:
- 分割:
2024_01_21T20_52_03.715367- 路径:
**/details_harness|hellaswag|10_2024-01-21T20-52-03.715367.parquet
- 路径:
- 分割:
2024_01_26T17_39_46.193538- 路径:
**/details_harness|hellaswag|10_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|hellaswag|10_2024-01-26T17-39-46.193538.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hendrycksTest_5- 数据文件:
- 分割:
2024_01_21T20_52_03.715367- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-anatomy|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-astronomy|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_biology|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-college_physics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-computer_security|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-econometrics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-global_facts|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-human_aging|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-international_law|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-management|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-marketing|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-01-21T20-52-03.715367.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-01-21T20-52-03.715367.parquet- `**/details_harness|hendrycksTest-nutrition|5_
- 路径:
- 分割:
- 数据文件:



