open-llm-leaderboard-old/details_Eurdem__megatron_1.1_MoE_2x7B
收藏数据集概述
数据集摘要
该数据集是在对模型 Eurdem/megatron_1.1_MoE_2x7B 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,每次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。
- 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_Eurdem__megatron_1.1_MoE_2x7B", "harness_winogrande_5", split="train")
最新结果
以下是 2024-01-22T14:00:29.163053 运行的最新结果:
python { "all": { "acc": 0.6536075619016702, "acc_stderr": 0.031835849880102456, "acc_norm": 0.6535987312796989, "acc_norm_stderr": 0.0324956769564807, "mc1": 0.3635250917992656, "mc1_stderr": 0.016838862883965838, "mc2": 0.5157817244014755, "mc2_stderr": 0.015241425184790871 }, "harness|arc:challenge|25": { "acc": 0.6245733788395904, "acc_stderr": 0.014150631435111726, "acc_norm": 0.6552901023890785, "acc_norm_stderr": 0.01388881628678211 }, "harness|hellaswag|10": { "acc": 0.6533559051981677, "acc_stderr": 0.004749286071559565, "acc_norm": 0.8451503684524995, "acc_norm_stderr": 0.0036102194130614605 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.29, "acc_stderr": 0.045604802157206845, "acc_norm": 0.29, "acc_norm_stderr": 0.045604802157206845 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- data_files:
- split: 2024_01_22T14_00_29.163053
- path: **/details_harness|arc:challenge|25_2024-01-22T14-00-29.163053.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2024-01-22T14-00-29.163053.parquet
- split: 2024_01_22T14_00_29.163053
- data_files:
-
config_name: harness_gsm8k_5
- data_files:
- split: 2024_01_22T14_00_29.163053
- path: **/details_harness|gsm8k|5_2024-01-22T14-00-29.163053.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2024-01-22T14-00-29.163053.parquet
- split: 2024_01_22T14_00_29.163053
- data_files:
-
config_name: harness_hellaswag_10
- data_files:
- split: 2024_01_22T14_00_29.163053
- path: **/details_harness|hellaswag|10_2024-01-22T14-00-29.163053.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2024-01-22T14-00-29.163053.parquet
- split: 2024_01_22T14_00_29.163053
- data_files:
-
config_name: harness_hendrycksTest_5
- data_files:
- split: 2024_01_22T14_00_29.163053
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-01-22T14-00-29.163053.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-01-22T14-00-29.163053.parquet ... }
- path:
- split: 2024_01_22T14_00_29.163053
- data_files:



