open-llm-leaderboard-old/details_jpquiroga__Mistral_7B_dare_slerp_merge_instruct_open_orca
收藏数据集概述
数据集简介
该数据集是在模型 jpquiroga/Mistral_7B_dare_slerp_merge_instruct_open_orca 在 Open LLM Leaderboard 上的评估运行期间自动创建的。
数据集结构
- 数据集包含 63 个配置,每个配置对应一个评估任务。
- 数据集从 1 次运行中创建,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。
- 额外的配置 "results" 存储所有运行的聚合结果,用于计算并在 Open LLM Leaderboard 上显示聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_jpquiroga__Mistral_7B_dare_slerp_merge_instruct_open_orca", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要: python { "all": { "acc": 0.6161554493970413, "acc_stderr": 0.032936724283737305, "acc_norm": 0.6211432753839832, "acc_norm_stderr": 0.033602576002331234, "mc1": 0.40269277845777235, "mc1_stderr": 0.01716883093518722, "mc2": 0.5634809148432152, "mc2_stderr": 0.015447882111793024 }, "harness|arc:challenge|25": { "acc": 0.5810580204778157, "acc_stderr": 0.014418106953639013, "acc_norm": 0.6262798634812287, "acc_norm_stderr": 0.014137708601759091 }, "harness|hellaswag|10": { "acc": 0.6211909978092014, "acc_stderr": 0.004840990593494689, "acc_norm": 0.8169687313284206, "acc_norm_stderr": 0.0038590186619620053 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.35, "acc_stderr": 0.0479372485441102, "acc_norm": 0.35, "acc_norm_stderr": 0.0479372485441102 }, # 其他任务的结果... }
配置详情
-
harness_arc_challenge_25
- 分割: 2024_04_16T15_14_56.072253
- 路径:
**/details_harness|arc:challenge|25_2024-04-16T15-14-56.072253.parquet - 分割: latest
- 路径:
**/details_harness|arc:challenge|25_2024-04-16T15-14-56.072253.parquet
-
harness_gsm8k_5
- 分割: 2024_04_16T15_14_56.072253
- 路径:
**/details_harness|gsm8k|5_2024-04-16T15-14-56.072253.parquet - 分割: latest
- 路径:
**/details_harness|gsm8k|5_2024-04-16T15-14-56.072253.parquet
-
harness_hellaswag_10
- 分割: 2024_04_16T15_14_56.072253
- 路径:
**/details_harness|hellaswag|10_2024-04-16T15-14-56.072253.parquet - 分割: latest
- 路径:
**/details_harness|hellaswag|10_2024-04-16T15-14-56.072253.parquet
-
harness_hendrycksTest_5
- 分割: 2024_04_16T15_14_56.072253
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-anatomy|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-astronomy|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_biology|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-college_physics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-computer_security|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-econometrics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-04-16T15-14-56.072253.parquet**/details_harness|hendrycksTest-global_facts|5_2024-04-16T15-14-56.072253.parquet
其他路径...



