Sentiment Analysis on VLSP Vietnamese 2016
95.2Accuracyfine-tuned TabularisAI
Evaluation results
| Method | Links | ||
|---|---|---|---|
| fine-tuned TabularisAISpeed (s/s)=205, Mode=fine-tuned2026.03 | 95.2 | 94.8 | |
| TabularisAI (Multilingual Base)Speed (s/s)=210, Variant=Multilingual Base2026.03 | 93.8 | 93.3 | |
| PhoBERT-baseSpeed (s/s)=1802026.03 | 93.4 | 92.8 | |
| viBERT4newsSpeed (s/s)=2002026.03 | 92.7 | 91.9 | |
| XLM-RoBERTa-baseSpeed (s/s)=2202026.03 | 91.8 | 91.2 | |
| BERT multilingual (mBERT)Speed (s/s)=2302026.03 | 90.3 | 89.8 | |
| RoBERTa-baseSpeed (s/s)=2502026.03 | 88.2 | 87.4 | |
| DistilBERT-baseSpeed (s/s)=3202026.03 | 87.5 | 86.9 | |
| GPT-2Speed (s/s)=300, Mode=Zero-shot2026.03 | 85.9 | 84.5 |