open-llm-leaderboard-old/details_yam-peleg__Experiment8-7B
收藏数据集概述
该数据集是在对模型 yam-peleg/Experiment8-7B 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。
- 数据集由 1 次运行创建,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。
- 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_yam-peleg__Experiment8-7B", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要:
python { "all": { "acc": 0.6566450072812023, "acc_stderr": 0.0320051313605246, "acc_norm": 0.6575118910656207, "acc_norm_stderr": 0.032652021481169954, "mc1": 0.5605875152998776, "mc1_stderr": 0.017374520482513704, "mc2": 0.7024529326790527, "mc2_stderr": 0.015023401127362113 }, "harness|arc:challenge|25": { "acc": 0.7013651877133106, "acc_stderr": 0.013374078615068745, "acc_norm": 0.7209897610921502, "acc_norm_stderr": 0.013106784883601334 }, "harness|hellaswag|10": { "acc": 0.7120095598486357, "acc_stderr": 0.004519011688417168, "acc_norm": 0.8812985461063533, "acc_norm_stderr": 0.0032277587155455987 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.31, "acc_stderr": 0.04648231987117316, "acc_norm": 0.31, "acc_norm_stderr": 0.04648231987117316 }, # 其他任务的结果... }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2024_02_11T22_11_17.659177
- path: **/details_harness|arc:challenge|25_2024-02-11T22-11-17.659177.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2024-02-11T22-11-17.659177.parquet
- split: 2024_02_11T22_11_17.659177
-
config_name: harness_gsm8k_5
- split: 2024_02_11T22_11_17.659177
- path: **/details_harness|gsm8k|5_2024-02-11T22-11-17.659177.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2024-02-11T22-11-17.659177.parquet
- split: 2024_02_11T22_11_17.659177
-
config_name: harness_hellaswag_10
- split: 2024_02_11T22_11_17.659177
- path: **/details_harness|hellaswag|10_2024-02-11T22-11-17.659177.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2024-02-11T22-11-17.659177.parquet
- split: 2024_02_11T22_11_17.659177
-
config_name: harness_hendrycksTest_5
- split: 2024_02_11T22_11_17.659177
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-11T22-11-17.659177.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-11T22-11-17.659177.parquet
- path:
- split: 2024_02_11T22_11_17.659177



