five

open-llm-leaderboard/details_ehartford__Samantha-1.11-CodeLlama-34b

收藏
Hugging Face2023-10-22 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_ehartford__Samantha-1.11-CodeLlama-34b
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ehartford/Samantha-1.11-CodeLlama-34b的评估运行过程中自动创建的,评估在Open LLM Leaderboard上进行。数据集由64个配置组成,每个配置对应一个评估任务。数据集由2次运行创建,每次运行的结果都作为一个特定的分割存储在配置中,分割名称使用运行的时间戳。train分割始终指向最新的结果。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Open LLM Leaderboard上的聚合指标。

This dataset was automatically generated during the evaluation runs of the model ehartford/Samantha-1.11-CodeLlama-34b, where the evaluation was performed on the Open LLM Leaderboard. The dataset comprises 64 configurations, with each configuration corresponding to one individual evaluation task. The dataset was created via two separate runs; the results of each run are stored as a specific split under its corresponding configuration, with the split name being the timestamp of that run. The train split always points to the most recent evaluation results. Additionally, there is a configuration named "results" that stores the aggregated results of all runs, which is used to calculate and display the aggregated metrics on the Open LLM Leaderboard.
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集来源

  • 数据集是在评估模型 ehartford/Samantha-1.11-CodeLlama-34b 的过程中自动创建的。

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,用于计算和显示聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_ehartford__Samantha-1.11-CodeLlama-34b", "harness_winogrande_5", split="train")

最新结果

  • 最新结果来自 2023-10-22T11:44:05.953431 运行,包含以下指标: python { "all": { "em": 0.03460570469798658, "em_stderr": 0.0018718276753995871, "f1": 0.0896822567114094, "f1_stderr": 0.002178835707458538, "acc": 0.464067454416748, "acc_stderr": 0.011642141344686711 }, "harness|drop|3": { "em": 0.03460570469798658, "em_stderr": 0.0018718276753995871, "f1": 0.0896822567114094, "f1_stderr": 0.002178835707458538 }, "harness|gsm8k|5": { "acc": 0.19332827899924185, "acc_stderr": 0.010877733223180565 }, "harness|winogrande|5": { "acc": 0.7348066298342542, "acc_stderr": 0.01240654946619286 } }

配置详情

  • harness_arc_challenge_25

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|arc:challenge|25_2023-08-26T02:57:56.123943.parquet
  • harness_drop_3

    • 分割: 2023_10_22T11_44_05.953431, latest
    • 路径: **/details_harness|drop|3_2023-10-22T11-44-05.953431.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_22T11_44_05.953431, latest
    • 路径: **/details_harness|gsm8k|5_2023-10-22T11-44-05.953431.parquet
  • harness_hellaswag_10

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hellaswag|10_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: 多个路径,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-econometrics|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分割: 2023_08_26T02_57_56.123943, latest
    • 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-26T02:57:56.123943.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分割: 2023_08_26T02_
    • 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-26T02:57:56.123943.parquet
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作