open-llm-leaderboard/details_openBuddy__openbuddy-llama2-34b-v11.1-bf16
收藏Hugging Face2023-10-24 更新2024-03-04 收录
下载链接：
https://hf-mirror.com/datasets/open-llm-leaderboard/details_openBuddy__openbuddy-llama2-34b-v11.1-bf16
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 openBuddy/openbuddy-llama2-34b-v11.1-bf16 在 Open LLM Leaderboard 上的评估运行期间自动创建的。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 4 次运行中创建的，每次运行在每个配置中表示为特定的分割，分割名称使用运行的时间戳。train 分割始终指向最新的结果。一个名为 results 的额外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 还提供了如何使用 `datasets` 库中的 `load_dataset` 函数加载运行中的详细信息的示例。README 中还包含了特定运行的最新结果，显示了各种任务的 EM、F1 和准确率等指标。
This dataset was automatically generated during the evaluation run of the model openBuddy/openbuddy-llama2-34b-v11.1-bf16 on the Open LLM Leaderboard. The dataset comprises 64 configurations, each corresponding to a single evaluation task. The dataset is compiled from 4 independent runs, where each run is represented as a dedicated split under every configuration, with split names using the timestamp of the corresponding run. The 'train' split always points to the most up-to-date results. An additional configuration named 'results' stores the aggregated results across all runs, which are used to calculate and display the aggregate metrics on the Open LLM Leaderboard. The README also provides examples of how to use the `load_dataset` function from the `datasets` library to load detailed information from individual runs. Additionally, the README includes the most recent results from specific runs, showcasing metrics such as EM, F1, and accuracy across various tasks.
提供机构：
open-llm-leaderboard
原始信息汇总
数据集概述
数据集来源
该数据集是在评估模型 openBuddy/openbuddy-llama2-34b-v11.1-bf16 在 Open LLM Leaderboard 上的自动创建的。
数据集结构
- 配置数量：64个配置，每个配置对应一个评估任务。
- 运行次数：数据集由4次运行创建。每次运行在每个配置中都有一个特定的分割，分割名称使用运行的时间戳。
- 最新结果："train" 分割总是指向最新的结果。
- 结果汇总：一个额外的配置 "results" 存储所有运行的汇总结果，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_openBuddy__openbuddy-llama2-34b-v11.1-bf16", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-24T15:31:04.396852 运行的最新结果： python { "all": { "em": 0.360633389261745, "em_stderr": 0.004917536525106699, "f1": 0.4180935402684579, "f1_stderr": 0.004778710905980245, "acc": 0.5268440191410464, "acc_stderr": 0.012939810741097795 }, "harness|drop|3": { "em": 0.360633389261745, "em_stderr": 0.004917536525106699, "f1": 0.4180935402684579, "f1_stderr": 0.004778710905980245 }, "harness|gsm8k|5": { "acc": 0.3457164518574678, "acc_stderr": 0.013100422990441578 }, "harness|winogrande|5": { "acc": 0.7079715864246251, "acc_stderr": 0.012779198491754013 } }
配置详情
-
harness_arc_challenge_25
- 分割：2023_09_13T11_53_35.640501
- 路径：
**/details_harness|arc:challenge|25_2023-09-13T11-53-35.640501.parquet
- 路径：
- 分割：2023_09_13T12_14_53.531149
- 路径：
**/details_harness|arc:challenge|25_2023-09-13T12-14-53.531149.parquet
- 路径：
- 分割：latest
- 路径：
**/details_harness|arc:challenge|25_2023-09-13T12-14-53.531149.parquet
- 路径：
- 分割：2023_09_13T11_53_35.640501
-
harness_drop_3
- 分割：2023_10_24T13_56_54.496754
- 路径：
**/details_harness|drop|3_2023-10-24T13-56-54.496754.parquet
- 路径：
- 分割：2023_10_24T15_31_04.396852
- 路径：
**/details_harness|drop|3_2023-10-24T15-31-04.396852.parquet
- 路径：
- 分割：latest
- 路径：
**/details_harness|drop|3_2023-10-24T15-31-04.396852.parquet
- 路径：
- 分割：2023_10_24T13_56_54.496754
-
harness_gsm8k_5
- 分割：2023_10_24T13_56_54.496754
- 路径：
**/details_harness|gsm8k|5_2023-10-24T13-56-54.496754.parquet
- 路径：
- 分割：2023_10_24T15_31_04.396852
- 路径：
**/details_harness|gsm8k|5_2023-10-24T15-31-04.396852.parquet
- 路径：
- 分割：latest
- 路径：
**/details_harness|gsm8k|5_2023-10-24T15-31-04.396852.parquet
- 路径：
- 分割：2023_10_24T13_56_54.496754
-
harness_hellaswag_10
- 分割：2023_09_13T11_53_35.640501
- 路径：
**/details_harness|hellaswag|10_2023-09-13T11-53-35.640501.parquet
- 路径：
- 分割：2023_09_13T12_14_53.531149
- 路径：
**/details_harness|hellaswag|10_2023-09-13T12-14-53.531149.parquet
- 路径：
- 分割：latest
- 路径：
**/details_harness|hellaswag|10_2023-09-13T12-14-53.531149.parquet
- 路径：
- 分割：2023_09_13T11_53_35.640501
-
harness_hendrycksTest_5
- 分割：2023_09_13T11_53_35.640501
- 路径：
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-anatomy|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-astronomy|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-business_ethics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_biology|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_medicine|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-college_physics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-computer_security|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-econometrics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-formal_logic|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-global_facts|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_biology|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_geography|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_physics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-human_aging|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-human_sexuality|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-international_law|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-jurisprudence|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-machine_learning|5_2023-09-13T11-53-35.640501.parquet - 路径：
**/details_harness|hendrycksTest-management|5_2023-09-13T11-53-35.640501.parquet - 路径：`**/details_harness|hendrycksTest-marketing|5_2023-09-13T11-53-35.64050
- 路径：
- 分割：2023_09_13T11_53_35.640501
