open-llm-leaderboard-old/details_OpenAssistant__codellama-13b-oasst-sft-v10
收藏数据集概述
数据集名称
Evaluation run of OpenAssistant/codellama-13b-oasst-sft-v10
数据集描述
该数据集是在模型 OpenAssistant/codellama-13b-oasst-sft-v10 在 Open LLM Leaderboard 上的评估运行期间自动创建的。
数据集组成
- 数据集包含 64 个配置,每个配置对应一个评估任务。
- 数据集由 4 次运行创建,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。
- "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_OpenAssistant__codellama-13b-oasst-sft-v10", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-15T06:23:43.342371 运行的最新结果: python { "all": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669533, "f1": 0.07171875000000016, "f1_stderr": 0.0015908122454952622, "acc": 0.4049487994360847, "acc_stderr": 0.011226667727964289 }, "harness|drop|3": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669533, "f1": 0.07171875000000016, "f1_stderr": 0.0015908122454952622 }, "harness|gsm8k|5": { "acc": 0.13191811978771797, "acc_stderr": 0.009321265253857515 }, "harness|winogrande|5": { "acc": 0.6779794790844514, "acc_stderr": 0.013132070202071064 } }
配置详情
-
harness_arc_challenge_25
- 分割: 2023_08_28T09_42_44.871031
- 路径: **/details_harness|arc:challenge|25_2023-08-28T09:42:44.871031.parquet
- 分割: 2023_08_28T18_08_08.712288
- 路径: **/details_harness|arc:challenge|25_2023-08-28T18:08:08.712288.parquet
- 分割: 2023_09_18T15_15_45.768968
- 路径: **/details_harness|arc:challenge|25_2023-09-18T15-15-45.768968.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2023-09-18T15-15-45.768968.parquet
- 分割: 2023_08_28T09_42_44.871031
-
harness_drop_3
- 分割: 2023_10_15T06_23_43.342371
- 路径: **/details_harness|drop|3_2023-10-15T06-23-43.342371.parquet
- 分割: latest
- 路径: **/details_harness|drop|3_2023-10-15T06-23-43.342371.parquet
- 分割: 2023_10_15T06_23_43.342371
-
harness_gsm8k_5
- 分割: 2023_10_15T06_23_43.342371
- 路径: **/details_harness|gsm8k|5_2023-10-15T06-23-43.342371.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2023-10-15T06-23-43.342371.parquet
- 分割: 2023_10_15T06_23_43.342371
-
harness_hellaswag_10
- 分割: 2023_08_28T09_42_44.871031
- 路径: **/details_harness|hellaswag|10_2023-08-28T09:42:44.871031.parquet
- 分割: 2023_08_28T18_08_08.712288
- 路径: **/details_harness|hellaswag|10_2023-08-28T18:08:08.712288.parquet
- 分割: 2023_09_18T15_15_45.768968
- 路径: **/details_harness|hellaswag|10_2023-09-18T15-15-45.768968.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2023-09-18T15-15-45.768968.parquet
- 分割: 2023_08_28T09_42_44.871031
-
harness_hendrycksTest_5
- 分割: 2023_08_28T09_42_44.871031
- 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-anatomy|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-astronomy|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_biology|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-college_physics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-computer_security|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-econometrics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-formal_logic|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-global_facts|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_biology|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_geography|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_physics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-human_aging|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-human_sexuality|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-international_law|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-jurisprudence|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-machine_learning|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details_harness|hendrycksTest-management|5_2023-08-28T09:42:44.871031.parquet
- 路径: **/details
- 分割: 2023_08_28T09_42_44.871031



