KoAlpaca-RealQA-Solar-Ko-Recovery-11B-Merged

LoRA merged model from beomi/KoAlpaca-RealQA-Solar-Ko-Recovery-11B

• Developed by: beomi
• License: apache-2.0
• Finetuned from model : beomi/Solar-Ko-Recovery-11B

Eval with LogicKor

Default

Category	Single turn	Multi turn
이해(Understanding)	8.43	5.86
수학(Math)	3.86	1.00
글쓰기(Writing)	9.71	7.71
추론(Reasoning)	7.57	3.86
코딩(Coding)	7.57	7.00
문법(Grammar)	7.43	4.14

Category	Score
Single turn	7.43
Multi turn	4.93
Overall	6.18

1-Shot

Category	Single turn	Multi turn
추론(Reasoning)	7.57	4.43
코딩(Coding)	7.29	8.57
수학(Math)	2.86	1.71
글쓰기(Writing)	9.29	7.57
이해(Understanding)	7.43	5.86
문법(Grammar)	7.29	4.71

Category	Score
Single turn	6.95
Multi turn	5.48
Overall	6.21

CoT-1-Shot

Category	Single turn	Multi turn
코딩(Coding)	7.14	8.14
추론(Reasoning)	6.86	3.57
글쓰기(Writing)	9.71	7.71
수학(Math)	2.29	1.14
이해(Understanding)	7.43	7.00
문법(Grammar)	7.29	3.14

Category	Score
Single turn	6.79
Multi turn	5.12
Overall	5.95