open-llm-leaderboard-old/details_BFauber__opt1.3b_10e5
收藏数据集概述
数据集简介
该数据集是在对模型 BFauber/opt1.3b_10e5 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
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。
- 数据集从 1 次运行中创建,每个运行可以在每个配置中作为一个特定的分片找到,分片名称使用运行的时间戳。
- "train" 分片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_BFauber__opt1.3b_10e5", "harness_winogrande_5", split="train")
最新结果
以下是 2024-02-02T19:22:53.886717 运行的最新结果:
python { "all": { "acc": 0.25956926248319, "acc_stderr": 0.030774098043396373, "acc_norm": 0.26133458633537765, "acc_norm_stderr": 0.031590043549587034, "mc1": 0.22643818849449204, "mc1_stderr": 0.014651337324602578, "mc2": 0.38176130958688614, "mc2_stderr": 0.014298868021565533 }, "harness|arc:challenge|25": { "acc": 0.2645051194539249, "acc_stderr": 0.012889272949313368, "acc_norm": 0.295221843003413, "acc_norm_stderr": 0.013329750293382318 }, "harness|hellaswag|10": { "acc": 0.4070902210714997, "acc_stderr": 0.004902878806733043, "acc_norm": 0.5280820553674567, "acc_norm_stderr": 0.004981905293878152 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.26, "acc_stderr": 0.0440844002276808, "acc_norm": 0.26, "acc_norm_stderr": 0.0440844002276808 }, # 其他任务的结果省略... }
配置详情
-
harness_arc_challenge_25
- 分片:2024_02_02T19_22_53.886717
- 路径:
**/details_harness|arc:challenge|25_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:2024_02_02T19_22_53.886717
-
harness_gsm8k_5
- 分片:2024_02_02T19_22_53.886717
- 路径:
**/details_harness|gsm8k|5_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|gsm8k|5_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:2024_02_02T19_22_53.886717
-
harness_hellaswag_10
- 分片:2024_02_02T19_22_53.886717
- 路径:
**/details_harness|hellaswag|10_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:latest
- 路径:
**/details_harness|hellaswag|10_2024-02-02T19-22-53.886717.parquet
- 路径:
- 分片:2024_02_02T19_22_53.886717
-
harness_hendrycksTest_5
- 分片:2024_02_02T19_22_53.886717
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-anatomy|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-astronomy|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_biology|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-college_physics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-computer_security|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-econometrics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-global_facts|5_2024-02-02T19-22-53.886717.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-02-02T19-22-53.886717.parquet- `**/details_
- 路径:
- 分片:2024_02_02T19_22_53.886717



