open-llm-leaderboard-old/details_DreadPoor__Nynph-7B-Model_Stock
收藏Hugging Face2024-04-15 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_DreadPoor__Nynph-7B-Model_Stock
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在评估模型DreadPoor/Nynph-7B-Model_Stock时自动生成的,包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为一个特定的分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,results配置存储了所有运行的聚合结果,用于计算和展示在Open LLM Leaderboard上的聚合指标。
该数据集是在评估模型DreadPoor/Nynph-7B-Model_Stock时自动生成的,包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为一个特定的分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,results配置存储了所有运行的聚合结果,用于计算和展示在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
数据集创建
- 创建背景:该数据集是在评估模型 DreadPoor/Nynph-7B-Model_Stock 的过程中自动创建的。
- 评估平台:数据集用于 Open LLM Leaderboard。
数据集结构
- 配置数量:数据集包含 63 个配置,每个配置对应一个评估任务。
- 数据来源:数据集从 1 次运行中创建,每次运行结果作为特定分片存储,分片名称使用运行的时间戳。
- 最新结果:"train" 分片始终指向最新结果。
- 结果汇总:额外配置 "results" 存储所有运行的汇总结果,用于计算和显示聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_DreadPoor__Nynph-7B-Model_Stock", "harness_winogrande_5", split="train")
最新结果
- 最新运行:2024-04-15T23:02:44.044199
- 结果示例: python { "all": { "acc": 0.6498025190522265, "acc_stderr": 0.031948103240068866, "acc_norm": 0.6521434250593419, "acc_norm_stderr": 0.03258320255611985, "mc1": 0.41370869033047736, "mc1_stderr": 0.0172408618120998, "mc2": 0.5810558211297511, "mc2_stderr": 0.015421310209791135 }, "harness|arc:challenge|25": { "acc": 0.6373720136518771, "acc_stderr": 0.01404910656495501, "acc_norm": 0.6689419795221843, "acc_norm_stderr": 0.013752062419817834 }, "harness|hellaswag|10": { "acc": 0.6744672376020713, "acc_stderr": 0.004676159299105417, "acc_norm": 0.857697669786895, "acc_norm_stderr": 0.0034864596026062573 }, ... }
配置详情
-
配置名称:harness_arc_challenge_25
- 数据文件:
- 分片:2024_04_15T23_02_44.044199
- 路径:
**/details_harness|arc:challenge|25_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|arc:challenge|25_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:2024_04_15T23_02_44.044199
- 数据文件:
-
配置名称:harness_gsm8k_5
- 数据文件:
- 分片:2024_04_15T23_02_44.044199
- 路径:
**/details_harness|gsm8k|5_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|gsm8k|5_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:2024_04_15T23_02_44.044199
- 数据文件:
-
配置名称:harness_hellaswag_10
- 数据文件:
- 分片:2024_04_15T23_02_44.044199
- 路径:
**/details_harness|hellaswag|10_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hellaswag|10_2024-04-15T23-02-44.044199.parquet
- 路径:
- 分片:2024_04_15T23_02_44.044199
- 数据文件:
-
配置名称:harness_hendrycksTest_5
- 数据文件:
- 分片:2024_04_15T23_02_44.044199
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-anatomy|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-astronomy|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_biology|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-college_physics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-computer_security|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-econometrics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-global_facts|5_2024-04-15T23-02-44.044199.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-04-15T23-02-44.044199.parquet...
- 路径:
- 分片:2024_04_15T23_02_44.044199
- 数据文件:



