firefly-ziya-13b / README.md
leaderboard-pr-bot's picture
Adding Evaluation Results
2a1d82b
|
raw
history blame
908 Bytes

该模型基于YeungNLP/Ziya-LLaMA-13B-Pretrain-v1,使用百万中英文指令数据,进行指令微调。

更多详情请查看Firefly项目

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 47.58
ARC (25-shot) 55.38
HellaSwag (10-shot) 78.47
MMLU (5-shot) 45.18
TruthfulQA (0-shot) 49.29
Winogrande (5-shot) 74.82
GSM8K (5-shot) 6.9
DROP (3-shot) 23.01