open-llm-leaderboard-old/details_BarryFutureman__WildMarcoroni-Variant1-7B
收藏数据集概述
该数据集是在对模型 BarryFutureman/WildMarcoroni-Variant1-7B 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 63 个配置,每个配置对应一个评估任务。
- 数据集从 1 次运行中创建,每次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。
- 额外的 "results" 配置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_BarryFutureman__WildMarcoroni-Variant1-7B", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要:
python { "all": { "acc": 0.6544725411330661, "acc_stderr": 0.03209534606277316, "acc_norm": 0.6537231161998335, "acc_norm_stderr": 0.032767252968853494, "mc1": 0.5642594859241126, "mc1_stderr": 0.01735834539886313, "mc2": 0.6975837745369705, "mc2_stderr": 0.015108261944159049 }, "harness|arc:challenge|25": { "acc": 0.7098976109215017, "acc_stderr": 0.013261573677520769, "acc_norm": 0.7397610921501706, "acc_norm_stderr": 0.012821930225112571 }, "harness|hellaswag|10": { "acc": 0.7233618801035651, "acc_stderr": 0.004464217420693355, "acc_norm": 0.8860784704242183, "acc_norm_stderr": 0.0031706661225176552 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.32, "acc_stderr": 0.046882617226215034, "acc_norm": 0.32, "acc_norm_stderr": 0.046882617226215034 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_01_23T23_37_24.789593
- 路径:
**/details_harness|arc:challenge|25_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|arc:challenge|25_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: 2024_01_23T23_37_24.789593
- 数据文件:
-
config_name: harness_gsm8k_5
- 数据文件:
- 分割: 2024_01_23T23_37_24.789593
- 路径:
**/details_harness|gsm8k|5_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|gsm8k|5_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: 2024_01_23T23_37_24.789593
- 数据文件:
-
config_name: harness_hellaswag_10
- 数据文件:
- 分割: 2024_01_23T23_37_24.789593
- 路径:
**/details_harness|hellaswag|10_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|hellaswag|10_2024-01-23T23-37-24.789593.parquet
- 路径:
- 分割: 2024_01_23T23_37_24.789593
- 数据文件:
-
config_name: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_01_23T23_37_24.789593
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-anatomy|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-astronomy|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_biology|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_physics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-computer_security|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-econometrics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-global_facts|5_2024-01-23T23-37-24.789593.parquet- ...
- 路径:
- 分割: latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-anatomy|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-astronomy|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_biology|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-college_physics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-computer_security|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-econometrics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-01-23T23-37-24.789593.parquet**/details_harness|hendrycksTest-global_facts|5_2024-01-23T23-37-24.789593.parquet- ...
- 路径:
- 分割: 2024_01_23T23_37_24.789593
- 数据文件:



