Crossing Linguistic Horizons

Finetuning and Comprehensive Evaluation of Vietnamese Large Language Models

Zero-Shot Knowledge Leaderboard

Models ZaloE2E ViMMRC
EM F1 AC F1 AR ECE A@10
URA-LLaMa 70B 0.28 ± 0.02 0.44 ± 0.02 0.80 ± 0.02 0.80 ± 0.02 0.85 ± 0.01 0.10 ± 0.02 0.96 ± 0.03
URA-LLaMa 13B 0.12 ± 0.01 0.22 ± 0.01 0.40 ± 0.02 0.31 ± 0.02 0.57 ± 0.02 0.48 ± 0.02 0.42 ± 0.08
URA-LLaMa 7B 0.09 ± 0.01 0.20 ± 0.02 0.30 ± 0.02 0.10 ± 0.01 0.56 ± 0.02 0.27 ± 0.02 0.56 ± 0.07
LLaMa-2 13B 0.06 ± 0.01 0.10 ± 0.01 0.52 ± 0.02 0.41 ± 0.02 0.64 ± 0.02 0.33 ± 0.02 0.73 ± 0.07
LLaMa-2 7B 0.03 ± 0.01 0.07 ± 0.01 0.37 ± 0.02 0.25 ± 0.02 0.51 ± 0.02 0.35 ± 0.02 0.29 ± 0.06
Vietcuna 7B 0.03 ± 0.01 0.06 ± 0.01 0.32 ± 0.02 0.22 ± 0.02 0.50 ± 0.00 0.07 ± 0.02 0.33 ± 0.07
GPT-3.5 0.37 ± 0.02 0.56 ± 0.02 0.90 ± 0.01 0.72 ± 0.01 - 0.65 ± 0.01 0.90 ± 0.04
GPT-4 0.38 ± 0.02 0.55 ± 0.02 0.92 ± 0.01 0.73 ± 0.06 - 0.67 ± 0.01 0.90 ± 0.04