five

open-llm-leaderboard/details_facebook__opt-1.3b

收藏
Hugging Face2023-10-19 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_facebook__opt-1.3b
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在评估模型facebook/opt-1.3b时自动生成的,评估过程在Open LLM Leaderboard上进行。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次运行生成,每次运行的结果作为特定配置中的一个分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。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 of the facebook/opt-1.3b model on the Open LLM Leaderboard. It contains 64 configurations, each corresponding to one evaluation task. The dataset is derived from two evaluation runs, where the results of each run serve as a split under its respective 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 across all runs, and 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

数据集简介

该数据集是在对模型facebook/opt-1.3b进行评估运行期间自动创建的,用于Open LLM Leaderboard

数据集结构

  • 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,每个运行在每个配置中作为一个特定的分片存在,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_facebook__opt-1.3b", "harness_winogrande_5", split="train")

最新结果

以下是2023-10-19T03:17:25.770385运行的最新结果: python { "all": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346038933, "f1": 0.05017722315436251, "f1_stderr": 0.0012387308214165103, "acc": 0.2994953245415047, "acc_stderr": 0.0074273230901261535 }, "harness|drop|3": { "em": 0.0017827181208053692, "em_stderr": 0.0004320097346038933, "f1": 0.05017722315436251, "f1_stderr": 0.0012387308214165103 }, "harness|gsm8k|5": { "acc": 0.001516300227445034, "acc_stderr": 0.0010717793485492619 }, "harness|winogrande|5": { "acc": 0.5974743488555643, "acc_stderr": 0.013782866831703044 } }

配置详情

  • harness_arc_challenge_25

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|arc:challenge|25_2023-08-18T14:50:30.777525.parquet
  • harness_drop_3

    • 分片: 2023_10_19T03_17_25.770385, latest
    • 路径: **/details_harness|drop|3_2023-10-19T03-17-25.770385.parquet
  • harness_gsm8k_5

    • 分片: 2023_10_19T03_17_25.770385, latest
    • 路径: **/details_harness|gsm8k|5_2023-10-19T03-17-25.770385.parquet
  • harness_hellaswag_10

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hellaswag|10_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: 多个路径,例如**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_anatomy_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_astronomy_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_biology_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_college_physics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_computer_security_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_econometrics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-econometrics|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-18T14:50:30.777525.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分片: 2023_08_18T14_50_30.777525, latest
    • 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-18T14:50:30.777525.parquet

以上是数据集的详细概述,包括数据集的简介、结构、加载示例、最新结果以及各个配置的详细信息。

搜集汇总
数据集介绍
main_image_url
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务