llama-3-Korean-Bllossom-8B_ver_4.3_big_sft_2epochs Benchmark Results

llama-3-Korean-Bllossom-8B_ver_4.3_big_sft_2epochs: benchmark results

Provider: Meta. Access: Open.

Unified ELO 1404 ± 17, rank #1880 of 2656 rated models, from 10 benchmark results.

Benchmark	Score	Metric	Percentile
Open Korean LLM - KorNAT-Harmless	64.26	Normalized accuracy (%)	72
Open Korean LLM - Ko-EQ-Bench	14.83	EQ-Bench score	60.5
Open Korean LLM - KorNAT-CKA	25.15	Normalized accuracy (%)	55.4
Open Korean LLM - Ko-GSM8K (flexible extract)	16	Exact match, flexible extract (%)	46.2
Open Korean LLM - Ko-IFEval	27.89	Strict accuracy, prompt/instruction mean (%)	45.6
Open Korean LLM Leaderboard	33.52	Average Score (%)	42.4
Open Ko-LLM Leaderboard	33.52	Average (%)	42.2
Open Korean LLM - KorNAT-Helpful	42.98	Normalized accuracy (%)	20.2
Open Korean LLM - Ko-Winogrande	54.7	Accuracy (%)	17
Open Korean LLM - Ko-GPQA-Diamond	19.7	Normalized accuracy (%)	9.6