open-llm-leaderboard-old/details_CHIH-HUNG__llama-2-13b-FINETUNE3_3.3w-r16-gate_up_down
收藏数据集概述
数据集来源
该数据集是在对模型 CHIH-HUNG/llama-2-13b-FINETUNE3_3.3w-r16-gate_up_down 进行评估运行时自动创建的。
数据集组成
数据集由64个配置组成,每个配置对应一个评估任务。
数据集创建
数据集从2次运行中创建。每个运行可以在每个配置中作为一个特定的分片找到,分片名称使用运行的时间戳。"train" 分片始终指向最新的结果。
结果汇总
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_CHIH-HUNG__llama-2-13b-FINETUNE3_3.3w-r16-gate_up_down", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-25T22:03:29.852983 运行的最新结果: python { "all": { "em": 0.18603187919463088, "em_stderr": 0.0039850799232311815, "f1": 0.22621224832214754, "f1_stderr": 0.0039981456214130545, "acc": 0.4515621662149879, "acc_stderr": 0.010508403406261944 }, "harness|drop|3": { "em": 0.18603187919463088, "em_stderr": 0.0039850799232311815, "f1": 0.22621224832214754, "f1_stderr": 0.0039981456214130545 }, "harness|gsm8k|5": { "acc": 0.12964366944655042, "acc_stderr": 0.009252657757825565 }, "harness|winogrande|5": { "acc": 0.7734806629834254, "acc_stderr": 0.011764149054698324 } }
数据集配置
以下是数据集的配置列表及其对应的数据文件路径:
-
config_name: harness_arc_challenge_25
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|arc:challenge|25_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_drop_3
- split: 2023_10_25T22_03_29.852983
- path: **/details_harness|drop|3_2023-10-25T22-03-29.852983.parquet
- split: latest
- path: **/details_harness|drop|3_2023-10-25T22-03-29.852983.parquet
- split: 2023_10_25T22_03_29.852983
-
config_name: harness_gsm8k_5
- split: 2023_10_25T22_03_29.852983
- path: **/details_harness|gsm8k|5_2023-10-25T22-03-29.852983.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2023-10-25T22-03-29.852983.parquet
- split: 2023_10_25T22_03_29.852983
-
config_name: harness_hellaswag_10
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hellaswag|10_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_5
- split: 2023_10_01T14_43_15.199560
- path: [多个路径]
- split: latest
- path: [多个路径]
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_abstract_algebra_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_anatomy_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-anatomy|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-anatomy|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_astronomy_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-astronomy|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-astronomy|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_business_ethics_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-business_ethics|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-business_ethics|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_clinical_knowledge_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_biology_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_biology|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_biology|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_chemistry_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_computer_science_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_mathematics_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_medicine_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_medicine|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_medicine|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_college_physics_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-college_physics|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-college_physics|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_computer_security_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-computer_security|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-computer_security|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_conceptual_physics_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_econometrics_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-econometrics|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-econometrics|5_2023-10-01T14-43-15.199560.parquet
- split: 2023_10_01T14_43_15.199560
-
config_name: harness_hendrycksTest_electrical_engineering_5
- split: 2023_10_01T14_43_15.199560
- path: **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-01T14-43-15.199560.parquet
- split: latest
- path: **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-01T14-43-
- split: 2023_10_01T14_43_15.199560



