five

open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-Open_Platypus_and_ccp_2.6w

收藏
Hugging Face2023-10-12 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-Open_Platypus_and_ccp_2.6w
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在模型CHIH-HUNG/llama-2-13b-Open_Platypus_and_ccp_2.6w在Open LLM Leaderboard上的评估过程中自动生成的。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2次运行的数据组成,每次运行在每个配置中表示为特定的分割,分割名称由运行的时间戳命名。train分割始终指向最新的结果。此外,还有一个results配置存储了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。README文件还提供了加载数据集详细信息的Python代码片段,并包含了特定运行的最新结果。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

该数据集是在对模型 CHIH-HUNG/llama-2-13b-Open_Platypus_and_ccp_2.6w 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。
  • 一个额外的配置 "results" 存储所有运行结果的聚合,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_CHIH-HUNG__llama-2-13b-Open_Platypus_and_ccp_2.6w", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-12T14:49:01.591870 运行的最新结果

python { "all": { "em": 0.003145973154362416, "em_stderr": 0.0005734993648436388, "f1": 0.06228817114093964, "f1_stderr": 0.0014101371508567083, "acc": 0.4173053896873633, "acc_stderr": 0.009418776710625477 }, "harness|drop|3": { "em": 0.003145973154362416, "em_stderr": 0.0005734993648436388, "f1": 0.06228817114093964, "f1_stderr": 0.0014101371508567083 }, "harness|gsm8k|5": { "acc": 0.06823351023502654, "acc_stderr": 0.006945358944067431 }, "harness|winogrande|5": { "acc": 0.7663772691397001, "acc_stderr": 0.011892194477183524 } }

配置详情

  • harness_arc_challenge_25

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|arc:challenge|25_2023-09-05T10:13:11.603787.parquet
  • harness_drop_3

    • 分片:2023_10_12T14_49_01.591870, latest
    • 路径:**/details_harness|drop|3_2023-10-12T14-49-01.591870.parquet
  • harness_gsm8k_5

    • 分片:2023_10_12T14_49_01.591870, latest
    • 路径:**/details_harness|gsm8k|5_2023-10-12T14-49-01.591870.parquet
  • harness_hellaswag_10

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hellaswag|10_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:多个文件路径,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_anatomy_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_astronomy_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_biology_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_college_physics_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_computer_security_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_econometrics_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-09-05T10:13:11.603787.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分片:2023_09_05T10_13_11.603787, latest
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-05T10:13:11.603787.parquet
  • **harness_hend

二维码
社区交流群
二维码
科研交流群
商业服务