open-llm-leaderboard/details_yeontaek__Platypus2-13B-IA3
收藏资源简介:
该数据集是在模型yeontaek/Platypus2-13B-IA3的评估运行过程中自动创建的。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从2次运行中创建的,每次运行都可以在特定配置中找到,分割以运行的时间戳命名。此外,还有一个名为“results”的配置存储了所有运行的聚合结果,并用于计算和显示在Open LLM Leaderboard上的聚合指标。
This dataset was automatically created during the evaluation runs of the model yeontaek/Platypus2-13B-IA3. It consists of 64 configurations, each corresponding to one evaluation task. The dataset is constructed from two separate runs, each of which can be found under its specific configuration, and the data splits are named after the run timestamps. Additionally, there is a configuration named "results" that stores the aggregated results of all runs, and is used to calculate and display the aggregated metrics on the Open LLM Leaderboard.
数据集概述
该数据集是在对模型 yeontaek/Platypus2-13B-IA3 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
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。
- 数据集从 2 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。
- "train" 分割总是指向最新的结果。
- 一个额外的配置 "results" 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_yeontaek__Platypus2-13B-IA3", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-16T22:47:30.949762 运行的最新结果: python { "all": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669685, "f1": 0.057964555369127514, "f1_stderr": 0.0013475913067505284, "acc": 0.43493522214636066, "acc_stderr": 0.01038750232963205 }, "harness|drop|3": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669685, "f1": 0.057964555369127514, "f1_stderr": 0.0013475913067505284 }, "harness|gsm8k|5": { "acc": 0.11296436694465505, "acc_stderr": 0.00871933902883306 }, "harness|winogrande|5": { "acc": 0.7569060773480663, "acc_stderr": 0.01205566563043104 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-18T01:02:21.186475.parquet
-
harness_drop_3
- 分割:2023_10_16T22_47_30.949762, latest
- 路径:
**/details_harness|drop|3_2023-10-16T22-47-30.949762.parquet
-
harness_gsm8k_5
- 分割:2023_10_16T22_47_30.949762, latest
- 路径:
**/details_harness|gsm8k|5_2023-10-16T22-47-30.949762.parquet
-
harness_hellaswag_10
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hellaswag|10_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:多个路径,例如
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-18T01:02:21.186475.parquet
-
harness_hendrycksTest_elementary_mathematics_5
- 分割:2023_08_18T01_02_21.186475, latest
- 路径:
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-18T01:02:21.186475.parquet
以上是数据集的详细概述,包括数据集的组成、加载示例、最新结果以及各个配置的详细信息。



