List your company
ResearchFormerly WizWand

Machine Translation on WMT En-Fi 17 (test)

30.1BLEU (tokenized)mRASP2
19.59622.32325.0527.777May 20, 2021
Updated 8mo ago

Evaluation results

MethodLinks
mRASP2Training Strategy=unified multilingual, Model Architecture (Layers)=12 layers2021.0530.1
mRASPTraining Strategy=unified multilingual, Model Architecture (Layers)=12 layers, Fine-tuning Status=without fine-tune2021.0529.2
mBARTTraining Strategy=pre-train & fine-tuned, Fine-tuning Status=fine-tuned2021.0528.5
m-TransformerTraining Strategy=unified multilingual, Model Architecture (Layers)=12 layers2021.0528.2
mRASPTraining Strategy=pre-train & fine-tuned, Fine-tuning Status=fine-tuned2021.0528
mRASPTraining Strategy=pre-train & fine-tuned, Fine-tuning Status=fine-tuned2021.0524
mRASP2Training Strategy=unified multilingual, Model Architecture (Layers)=12 layers2021.0523.4
mBARTTraining Strategy=pre-train & fine-tuned, Fine-tuning Status=fine-tuned2021.0522.4
Multi-DistillationTraining Strategy=unified multilingual2021.0522
mRASPTraining Strategy=unified multilingual, Model Architecture (Layers)=12 layers, Fine-tuning Status=without fine-tune2021.0522
Transformer-12Training Strategy=bilingual, Model Architecture (Layers)=12 layers2021.0521.8
Multi-DistillationTraining Strategy=unified multilingual2021.0521.2
Transformer-12Training Strategy=bilingual, Model Architecture (Layers)=12 layers2021.0520.2
m-TransformerTraining Strategy=unified multilingual, Model Architecture (Layers)=12 layers2021.0520