open-llm-leaderboard/details_zarakiquemparte__zarablend-1.1-l2-7b
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_zarakiquemparte__zarablend-1.1-l2-7b
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是在Open LLM Leaderboard上对模型zarakiquemparte/zarablend-1.1-l2-7b进行评估时自动生成的。数据集包含64个配置，每个配置对应一个被评估的任务。数据集由2次运行生成，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，用于在Open LLM Leaderboard上计算和显示聚合指标。README还提供了如何使用Python中的datasets库加载运行细节的示例。
This dataset was automatically generated during the evaluation of the model zarakiquemparte/zarablend-1.1-l2-7b on the Open LLM Leaderboard. The dataset consists of 64 configurations, each corresponding to one evaluated task. The dataset is compiled from two separate runs, where each run is represented as a dedicated split under each configuration, with the split name being the timestamp of the corresponding run. The `train` split consistently points to the most up-to-date evaluation results. Additionally, a `results` configuration stores the aggregated results across all runs, which are utilized to compute and display the aggregate metrics on the Open LLM Leaderboard. The README also includes examples of how to load run details using the `datasets` library in Python.
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 zarakiquemparte/zarablend-1.1-l2-7b 进行评估运行期间自动创建的，用于 Open LLM Leaderboard。
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：数据集从2次运行中创建。每个运行结果作为一个特定的分片存储在每个配置中，分片名称使用运行的时间戳。
- 训练分片："train" 分片始终指向最新的结果。
- 结果配置：额外配置 "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_zarakiquemparte__zarablend-1.1-l2-7b", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-24T16:53:24.152575 运行 的最新结果： python { "all": { "em": 0.2829278523489933, "em_stderr": 0.0046127395092502785, "f1": 0.3596476510067135, "f1_stderr": 0.004549657562733716, "acc": 0.38580685542430376, "acc_stderr": 0.009136475194671255 }, "harness|drop|3": { "em": 0.2829278523489933, "em_stderr": 0.0046127395092502785, "f1": 0.3596476510067135, "f1_stderr": 0.004549657562733716 }, "harness|gsm8k|5": { "acc": 0.045489006823351025, "acc_stderr": 0.005739657656722217 }, "harness|winogrande|5": { "acc": 0.7261247040252565, "acc_stderr": 0.012533292732620292 } }
配置详情
以下是部分配置及其数据文件路径：
-
harness_arc_challenge_25
- 分片：2023_10_08T20_34_38.320909
- 路径：
**/details_harness|arc:challenge|25_2023-10-08T20-34-38.320909.parquet
- 路径：
- 分片：latest
- 路径：
**/details_harness|arc:challenge|25_2023-10-08T20-34-38.320909.parquet
- 路径：
- 分片：2023_10_08T20_34_38.320909
-
harness_drop_3
- 分片：2023_10_24T16_53_24.152575
- 路径：
**/details_harness|drop|3_2023-10-24T16-53-24.152575.parquet
- 路径：
- 分片：latest
- 路径：
**/details_harness|drop|3_2023-10-24T16-53-24.152575.parquet
- 路径：
- 分片：2023_10_24T16_53_24.152575
-
harness_gsm8k_5
- 分片：2023_10_24T16_53_24.152575
- 路径：
**/details_harness|gsm8k|5_2023-10-24T16-53-24.152575.parquet
- 路径：
- 分片：latest
- 路径：
**/details_harness|gsm8k|5_2023-10-24T16-53-24.152575.parquet
- 路径：
- 分片：2023_10_24T16_53_24.152575
-
harness_hellaswag_10
- 分片：2023_10_08T20_34_38.320909
- 路径：
**/details_harness|hellaswag|10_2023-10-08T20-34-38.320909.parquet
- 路径：
- 分片：latest
- 路径：
**/details_harness|hellaswag|10_2023-10-08T20-34-38.320909.parquet
- 路径：
- 分片：2023_10_08T20_34_38.320909
-
harness_hendrycksTest_5
- 分片：2023_10_08T20_34_38.320909
- 路径：
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-anatomy|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-astronomy|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-business_ethics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_biology|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_chemistry|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_computer_science|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_mathematics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_medicine|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-college_physics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-computer_security|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-econometrics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-formal_logic|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-global_facts|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_biology|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_geography|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_physics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-human_aging|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-human_sexuality|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-international_law|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-jurisprudence|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-machine_learning|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-management|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-marketing|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-medical_genetics|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-miscellaneous|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-moral_disputes|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-08T20-34-38.320909.parquet - 路径：
**/details_harness|hendrycksTest-nutrition|5_2023-10-08T20-34-38.320909.parquet - 路径：`**/details_harness|hendrycksTest-philosophy|5_2023-10-0
- 路径：
- 分片：2023_10_08T20_34_38.320909
