open-llm-leaderboard-old/details_qblocks__mistral_7b_HalfEpoch_DolphinCoder
收藏数据集概述
数据集简介
该数据集是在模型qblocks/mistral_7b_HalfEpoch_DolphinCoder在Open LLM Leaderboard上的评估运行期间自动创建的。
数据集组成
- 数据集包含63个配置,每个配置对应一个评估任务。
- 数据集由1次运行创建,每个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"results"配置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_qblocks__mistral_7b_HalfEpoch_DolphinCoder", "harness_winogrande_5", split="train")
最新结果
以下是最新结果(来自2024-01-21T10:15:02.334082运行): python { "all": { "acc": 0.6142255167007181, "acc_stderr": 0.03288830350165743, "acc_norm": 0.6209744380373297, "acc_norm_stderr": 0.0335648510230416, "mc1": 0.2998776009791922, "mc1_stderr": 0.016040352966713627, "mc2": 0.4546153705308199, "mc2_stderr": 0.01483288188604126 }, "harness|arc:challenge|25": { "acc": 0.5827645051194539, "acc_stderr": 0.014409825518403077, "acc_norm": 0.6177474402730375, "acc_norm_stderr": 0.014200454049979274 }, "harness|hellaswag|10": { "acc": 0.6361282613025294, "acc_stderr": 0.004801290954387088, "acc_norm": 0.8226448914558853, "acc_norm_stderr": 0.003811883070911275 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.04512608598542129, "acc_norm": 0.28, "acc_norm_stderr": 0.04512608598542129 }, # 其他任务的结果省略... }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2024_01_21T10_15_02.334082
- path: **/details_harness|arc:challenge|25_2024-01-21T10-15-02.334082.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2024-01-21T10-15-02.334082.parquet
- split: 2024_01_21T10_15_02.334082
-
config_name: harness_gsm8k_5
- split: 2024_01_21T10_15_02.334082
- path: **/details_harness|gsm8k|5_2024-01-21T10-15-02.334082.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2024-01-21T10-15-02.334082.parquet
- split: 2024_01_21T10_15_02.334082
-
config_name: harness_hellaswag_10
- split: 2024_01_21T10_15_02.334082
- path: **/details_harness|hellaswag|10_2024-01-21T10-15-02.334082.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2024-01-21T10-15-02.334082.parquet
- split: 2024_01_21T10_15_02.334082
-
config_name: harness_hendrycksTest_5
- split: 2024_01_21T10_15_02.334082
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-01-21T10-15-02.334082.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-01-21T10-15-02.334082.parquet
- path:
- split: 2024_01_21T10_15_02.334082



