polyglot-ko-12.8B-Chat-QLoRA-Merge_v3 Benchmark Results

polyglot-ko-12.8B-Chat-QLoRA-Merge_v3: benchmark results

Provider: Other. Access: Open.

Unified ELO 1384 ± 18, rank #2026 of 2656 rated models, from 10 benchmark results.

Benchmark	Score	Metric	Percentile
Open Korean LLM - KorNAT-CKA	28.11	Normalized accuracy (%)	74.1
Open Korean LLM - Ko-GPQA-Diamond	23.74	Normalized accuracy (%)	55.6
Open Korean LLM - Ko-IFEval	29.98	Strict accuracy, prompt/instruction mean (%)	53.7
Open Korean LLM - Ko-EQ-Bench	0.23	EQ-Bench score	44.2
Open Korean LLM - Ko-Winogrande	58.48	Accuracy (%)	39.2
Open Korean LLM Leaderboard	32.7	Average Score (%)	35.4
Open Ko-LLM Leaderboard	32.7	Average (%)	35.2
Open Korean LLM - KorNAT-Helpful	42.43	Normalized accuracy (%)	16.2
Open Korean LLM - KorNAT-Harmless	60.32	Normalized accuracy (%)	13.7
Open Korean LLM - Ko-GSM8K (flexible extract)	0.83	Exact match, flexible extract (%)	10.8