Crossing Linguistic Horizons

Finetuning and Comprehensive Evaluation of Vietnamese Large Language Models

Zero-Shot Sentiment Analysis Leaderboard

Models VLSP 2016 UiT-VSFC
AC↑ F1↑ AR↑ ECE↓ A@10↑ AC↑ F1↑ AR↑ ECE↓ A@10↑
URA-LLaMa 70B 0.63 ± 0.02 0.63 ± 0.02 0.74 ± 0.01 0.15 ± 0.01 0.87 ± 0.03 0.64 ± 0.01 0.54 ± 0.01 0.85 ± 0.01 0.14 ± 0.00 0.98 ± 0.01
URA-LLaMa 13B 0.52 ± 0.02 0.35 ± 0.01 0.60 ± 0.01 0.10 ± 0.01 0.64 ± 0.05 0.70 ± 0.01 0.40 ± 0.01 0.72 ± 0.01 0.23 ± 0.01 0.95 ± 0.01
URA-LLaMa 7B 0.35 ± 0.02 0.24 ± 0.01 0.54 ± 0.01 0.24 ± 0.01 0.31 ± 0.05 0.27 ± 0.01 0.18 ± 0.00 0.52 ± 0.01 0.37 ± 0.01 0.03 ± 0.01
LLaMa-2 13B 0.25 ± 0.01 0.25 ± 0.01 0.49 ± 0.01 0.39 ± 0.01 0.29 ± 0.05 0.29 ± 0.01 0.24 ± 0.01 0.52 ± 0.01 0.42 ± 0.01 0.30 ± 0.03
LLaMa-2 7B 0.15 ± 0.01 0.15 ± 0.01 0.58 ± 0.01 0.73 ± 0.01 0.12 ± 0.03 0.04 ± 0.00 0.06 ± 0.01 0.49 ± 0.01 0.79 ± 0.00 0.01 ± 0.01
Vietcuna 7B 0.11 ± 0.01 0.12 ± 0.01 0.49 ± 0.01 0.68 ± 0.01 0.11 ± 0.03 0.05 ± 0.00 0.06 ± 0.00 0.56 ± 0.01 0.73 ± 0.00 0.05 ± 0.01
MixSUra 8x7B 0.45 ± - 0.30 ± - 0.62 ± - 0.50 ± - 0.49 ± - 0.55 ± - 0.40 ± - 0.66 ± - 0.41 ± - 0.60 ± -
Gemini Pro 0.64 ± - 0.47 ± - - 0.31 ± - 0.53 ± - 0.76 ± - 0.49 ± - - 0.43 ± - 0.77 ± -
GPT-3.5 0.62 ± 0.02 0.56 ± 0.01 - 0.29 ± 0.02 0.62 ± 0.05 0.81 ± 0.31 0.68 ± 0.31 - 0.48 ± 0.01 0.83 ± 0.02
GPT-4 0.71 ± 0.01 0.68 ± 0.01 - 0.37 ± 0.01 0.70 ± 0.04 0.80 ± 0.01 0.67 ± 0.01 - 0.47 ± 0.01 0.85 ± 0.02