open-llm-leaderboard-old/details_rwitz__go-bruins
收藏Hugging Face2023-12-09 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_rwitz__go-bruins
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在模型rwitz/go-bruins在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从2次运行中创建的,每次运行都可以在每个配置的特定分割中找到,分割名称使用运行的时间戳命名。train分割始终指向最新的结果。此外,还有一个名为results的配置存储了所有运行的聚合结果,并用于计算和显示Open LLM Leaderboard上的聚合指标。
该数据集是在模型rwitz/go-bruins在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从2次运行中创建的,每次运行都可以在每个配置的特定分割中找到,分割名称使用运行的时间戳命名。train分割始终指向最新的结果。此外,还有一个名为results的配置存储了所有运行的聚合结果,并用于计算和显示Open LLM Leaderboard上的聚合指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总
数据集概述
数据集组成
- 该数据集包含63个配置,每个配置对应一个评估任务。
- 数据集由2次运行结果组成,每次运行的结果以特定的时间戳命名的分片存储。
- 每个配置中有一个名为"train"的分片,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,用于计算和显示在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_rwitz__go-bruins", "harness_winogrande_5", split="train")
最新结果
- 最新结果来自2023-12-09T17:56:51.445836的运行,包含多个任务的评估结果。
- 示例结果如下:
python
{
"all": {
"acc": 0.6537762475221197,
"acc_stderr": 0.03208085743053689,
"acc_norm": 0.6538246694322897,
"acc_norm_stderr": 0.032742779319017035,
"mc1": 0.4320685434516524,
"mc1_stderr": 0.017341202394988257,
"mc2": 0.5871006945090181,
"mc2_stderr": 0.015474717474561337
},
"harness|arc:challenge|25": {
"acc": 0.6638225255972696,
"acc_stderr": 0.013804855026205761,
"acc_norm": 0.6911262798634812,
"acc_norm_stderr": 0.013501770929344003
},
"harness|hellaswag|10": {
"acc": 0.6857199761003784,
"acc_stderr": 0.004632797375289765,
"acc_norm": 0.867257518422625,
"acc_norm_stderr": 0.0033860277997584177
},
"harness|hendrycksTest-abstract_algebra|5": {
"acc": 0.32,
"acc_stderr": 0.04688261722621504,
"acc_norm": 0.32,
"acc_norm_stderr": 0.04688261722621504
},
其他任务的结果...
}
配置详情
- 每个配置包含多个数据文件,每个数据文件对应一个特定的时间戳或"latest"分片。
- 示例配置:
-
config_name: harness_arc_challenge_25split: 2023_12_09T17_47_26.960183path: **/details_harness|arc:challenge|25_2023-12-09T17-47-26.960183.parquet
split: 2023_12_09T17_56_51.445836path: **/details_harness|arc:challenge|25_2023-12-09T17-56-51.445836.parquet
split: latestpath: **/details_harness|arc:challenge|25_2023-12-09T17-56-51.445836.parquet
-
config_name: harness_gsm8k_5split: 2023_12_09T17_47_26.960183path: **/details_harness|gsm8k|5_2023-12-09T17-47-26.960183.parquet
split: 2023_12_09T17_56_51.445836path: **/details_harness|gsm8k|5_2023-12-09T17-56-51.445836.parquet
split: latestpath: **/details_harness|gsm8k|5_2023-12-09T17-56-51.445836.parquet
-
config_name: harness_hellaswag_10split: 2023_12_09T17_47_26.960183path: **/details_harness|hellaswag|10_2023-12-09T17-47-26.960183.parquet
split: 2023_12_09T17_56_51.445836path: **/details_harness|hellaswag|10_2023-12-09T17-56-51.445836.parquet
split: latestpath: **/details_harness|hellaswag|10_2023-12-09T17-56-51.445836.parquet
-
config_name: harness_hendrycksTest_5split: 2023_12_09T17_47_26.960183path: **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-anatomy|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-astronomy|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-business_ethics|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_biology|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_chemistry|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_computer_science|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_mathematics|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_medicine|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-college_physics|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-computer_security|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-econometrics|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-09T17-47-26.960183.parquetpath: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-09T17-47-26.960183.parquet}
-



