open-llm-leaderboard/details_IGeniusDev__llama13B-quant8-testv1-openorca-customdataset
收藏数据集概述
数据集简介
该数据集是在评估模型 IGeniusDev/llama13B-quant8-testv1-openorca-customdataset 在 Open LLM Leaderboard 上的运行过程中自动创建的。
数据集结构
- 配置数量:64个配置,每个配置对应一个评估任务。
- 创建来源:数据集从3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。
- 最新结果:"train" 分割始终指向最新的结果。
- 汇总结果:一个额外的配置 "results" 存储所有运行的汇总结果,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_IGeniusDev__llama13B-quant8-testv1-openorca-customdataset", "harness_winogrande_5", split="train")
最新结果
以下是 2023-09-22T22:58:12.026440 运行 的最新结果: python { "all": { "em": 0.001363255033557047, "em_stderr": 0.0003778609196460785, "f1": 0.061706166107382464, "f1_stderr": 0.0013539743732266353, "acc": 0.4288700212365805, "acc_stderr": 0.010174848411278824 }, "harness|drop|3": { "em": 0.001363255033557047, "em_stderr": 0.0003778609196460785, "f1": 0.061706166107382464, "f1_stderr": 0.0013539743732266353 }, "harness|gsm8k|5": { "acc": 0.10083396512509477, "acc_stderr": 0.008294031192126607 }, "harness|winogrande|5": { "acc": 0.7569060773480663, "acc_stderr": 0.012055665630431043 } }
配置详情
以下是数据集的配置详情:
配置:harness_arc_challenge_25
- 分割:2023_08_25T21_11_04.087350
- 路径:
**/details_harness|arc:challenge|25_2023-08-25T21:11:04.087350.parquet
- 路径:
- 分割:2023_08_28T19_50_33.159802
- 路径:
**/details_harness|arc:challenge|25_2023-08-28T19:50:33.159802.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-28T19:50:33.159802.parquet
- 路径:
配置:harness_drop_3
- 分割:2023_09_22T22_58_12.026440
- 路径:
**/details_harness|drop|3_2023-09-22T22-58-12.026440.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|drop|3_2023-09-22T22-58-12.026440.parquet
- 路径:
配置:harness_gsm8k_5
- 分割:2023_09_22T22_58_12.026440
- 路径:
**/details_harness|gsm8k|5_2023-09-22T22-58-12.026440.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-09-22T22-58-12.026440.parquet
- 路径:
配置:harness_hellaswag_10
- 分割:2023_08_25T21_11_04.087350
- 路径:
**/details_harness|hellaswag|10_2023-08-25T21:11:04.087350.parquet
- 路径:
- 分割:2023_08_28T19_50_33.159802
- 路径:
**/details_harness|hellaswag|10_2023-08-28T19:50:33.159802.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-08-28T19:50:33.159802.parquet
- 路径:
配置:harness_hendrycksTest_5
- 分割:2023_08_25T21_11_04.087350
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-anatomy|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-astronomy|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_biology|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-college_physics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-computer_security|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-econometrics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-global_facts|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_geography|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_physics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-human_aging|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-human_sexuality|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-international_law|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-jurisprudence|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-machine_learning|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-management|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-marketing|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-medical_genetics|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-miscellaneous|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-moral_disputes|5_2023-08-25T21:11:04.087350.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-25T21:11:04.087350.parquet- `**/details_harness|hendrycksTest-nutrition|5_2023-08-25T21:1
- 路径:



