open-llm-leaderboard-old/details_rizla__rizla54
收藏数据集概述
数据集来源
该数据集是在对模型 rizla/rizla54 进行评估运行期间自动创建的,评估运行在 Open LLM Leaderboard 上进行。
数据集组成
数据集由 63 个配置组成,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。
结果配置
一个额外的配置 "results" 存储了所有运行的聚合结果,用于计算并在 Open LLM Leaderboard 上显示聚合指标。
数据加载示例
以下是加载特定运行详细信息的示例代码: python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_rizla__rizla54", "harness_winogrande_5", split="train")
最新结果
以下是 2024-02-02T08:26:50.989261 运行的最新结果: python { "all": { "acc": 0.6070754715230492, "acc_stderr": 0.033251216013816025, "acc_norm": 0.6153378461064626, "acc_norm_stderr": 0.03397161363942805, "mc1": 0.37209302325581395, "mc1_stderr": 0.016921090118814038, "mc2": 0.5325609210892159, "mc2_stderr": 0.015366351468634187 }, "harness|arc:challenge|25": { "acc": 0.5392491467576792, "acc_stderr": 0.014566303676636586, "acc_norm": 0.5819112627986348, "acc_norm_stderr": 0.014413988396996081 }, "harness|hellaswag|10": { "acc": 0.5825532762397929, "acc_stderr": 0.004921300331285573, "acc_norm": 0.7873929496116312, "acc_norm_stderr": 0.004083157276012493 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.3, "acc_stderr": 0.04605661864718381, "acc_norm": 0.3, "acc_norm_stderr": 0.04605661864718381 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_02_02T08_26_50.989261
- 路径: **/details_harness|arc:challenge|25_2024-02-02T08-26-50.989261.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2024-02-02T08-26-50.989261.parquet
- 分割: 2024_02_02T08_26_50.989261
- 数据文件:
-
config_name: harness_gsm8k_5
- 数据文件:
- 分割: 2024_02_02T08_26_50.989261
- 路径: **/details_harness|gsm8k|5_2024-02-02T08-26-50.989261.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2024-02-02T08-26-50.989261.parquet
- 分割: 2024_02_02T08_26_50.989261
- 数据文件:
-
config_name: harness_hellaswag_10
- 数据文件:
- 分割: 2024_02_02T08_26_50.989261
- 路径: **/details_harness|hellaswag|10_2024-02-02T08-26-50.989261.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2024-02-02T08-26-50.989261.parquet
- 分割: 2024_02_02T08_26_50.989261
- 数据文件:
-
config_name: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_02_02T08_26_50.989261
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-02T08-26-50.989261.parquet
- ...
- 路径:
- 分割: latest
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-02T08-26-50.989261.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-02T08-26-50.989261.parquet
- ...
- 路径:
- 分割: 2024_02_02T08_26_50.989261
- 数据文件:



