open-llm-leaderboard-old/details_amu__dpo-phi2
收藏Hugging Face2024-02-09 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_amu__dpo-phi2
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对模型amu/dpo-phi2进行评估运行时自动创建的。数据集由63个配置组成,每个配置对应一个被评估的任务。数据集是从一次或多次运行中生成的,每次运行都作为每个配置中的一个特定分割存储。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了如何使用`datasets`库中的`load_dataset`函数加载数据集的示例。
该数据集是在Open LLM Leaderboard上对模型amu/dpo-phi2进行评估运行时自动创建的。数据集由63个配置组成,每个配置对应一个被评估的任务。数据集是从一次或多次运行中生成的,每次运行都作为每个配置中的一个特定分割存储。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了如何使用`datasets`库中的`load_dataset`函数加载数据集的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总
数据集概述
数据集基本信息
- 数据集名称: Evaluation run of amu/dpo-phi2
- 数据集来源: 自动创建于模型 amu/dpo-phi2 在 Open LLM Leaderboard 的评估运行中。
- 数据集结构: 包含 63 个配置,每个配置对应一个评估任务。
- 数据集创建: 从 1 次运行中创建,每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。"train" 分割始终指向最新结果。
- 额外配置: "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_amu__dpo-phi2", "harness_winogrande_5", split="train")
最新结果
- 最新结果来源: 2024-02-09T22:52:41.834873 运行
- 结果示例: python { "all": { "acc": 0.5828070162053215, "acc_stderr": 0.03369036649487999, "acc_norm": 0.5845127625459068, "acc_norm_stderr": 0.03437729917800213, "mc1": 0.30966952264381886, "mc1_stderr": 0.016185744355144912, "mc2": 0.4398875544767273, "mc2_stderr": 0.015069641700788115 }, "harness|arc:challenge|25": { "acc": 0.5836177474402731, "acc_stderr": 0.01440561827943618, "acc_norm": 0.6168941979522184, "acc_norm_stderr": 0.014206472661672874 }, "harness|hellaswag|10": { "acc": 0.5633339972117108, "acc_stderr": 0.004949589567678895, "acc_norm": 0.7513443537143996, "acc_norm_stderr": 0.004313503876346087 }, ... }
配置信息
-
配置名称: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_02_09T22_52_41.834873
- 路径: **/details_harness|arc:challenge|25_2024-02-09T22-52-41.834873.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2024-02-09T22-52-41.834873.parquet
- 分割: 2024_02_09T22_52_41.834873
- 数据文件:
-
配置名称: harness_gsm8k_5
- 数据文件:
- 分割: 2024_02_09T22_52_41.834873
- 路径: **/details_harness|gsm8k|5_2024-02-09T22-52-41.834873.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2024-02-09T22-52-41.834873.parquet
- 分割: 2024_02_09T22_52_41.834873
- 数据文件:
-
配置名称: harness_hellaswag_10
- 数据文件:
- 分割: 2024_02_09T22_52_41.834873
- 路径: **/details_harness|hellaswag|10_2024-02-09T22-52-41.834873.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2024-02-09T22-52-41.834873.parquet
- 分割: 2024_02_09T22_52_41.834873
- 数据文件:
-
配置名称: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_02_09T22_52_41.834873
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-09T22-52-41.834873.parquet
- ...
- 路径:
- 分割: latest
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-09T22-52-41.834873.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-09T22-52-41.834873.parquet
- ...
- 路径:
- 分割: 2024_02_09T22_52_41.834873
- 数据文件:



