Command Palette
Search for a command to run...
Traduction automatique
La traduction automatique est une tâche importante dans le traitement du langage naturel, visant à convertir des phrases d'une langue source en expressions équivalentes dans une langue cible. Ces dernières années, les modèles de réseaux neuronaux basés sur le mécanisme d'attention encodeur-décodeur, comme BERT, ont connu des progrès significatifs, améliorant considérablement la qualité de la traduction. Les métriques d'évaluation courantes incluent BLEU, METEOR et NIST, tandis que les ensembles de données de la série WMT sont des ressources importantes largement utilisées pour les tests de référence.
WMT2014 English-German
Transformer Cycle (Rev)
WMT2014 English-French
Transformer+BT (ADMIN init)
IWSLT2014 German-English
PiNMT
WMT2016 English-Romanian
DeLighT
WMT2016 Romanian-English
MLM pretraining
ACES
HWTSC-Teacher-Sim
WMT2014 German-English
Bi-SimCut
IWSLT2015 German-English
PS-KD
WMT2016 English-German
IWSLT2015 English-Vietnamese
EnViT5 + MTet
IWSLT2015 English-German
PS-KD
WMT2016 German-English
WMT2015 English-German
ByteNet
IWSLT2014 English-German
FLoRes-200
GenTranslate-7B
WMT 2017 Latvian-English
WMT2016 English-Russian
Attentional encoder-decoder + BPE
WMT 2017 English-Chinese
DynamicConv
FRMT (Portuguese - Brazil)
FRMT (Portuguese - Portugal)
FRMT (Chinese - Taiwan)
FRMT (Chinese - Mainland)
flores95-devtest X-eng
flores95-devtest eng-X
SeamlessM4T Large
WMT2014 French-English
WMT2017 Chinese-English
StrokeNet
WMT 2018 Finnish-English
20NEWS
tensorflow/tensor2tensor
IWSLT2017 English-French
Transformer base + BPE-Dropout
WMT2014 English-Czech
Evolved Transformer Big
IWSLT2017 German-English
Adaptively Sparse Transformer (alpha-entmax)
IWSLT2017 French-English
NLLB-200
Arba Sicula
Larger
IWSLT2017 English-Arabic
Transformer base + BPE-Dropout
IWSLT2017 Arabic-English
IWSLT2015 Vietnamese-English
WMT2019 English-German
Facebook FAIR (ensemble)
WMT2017 Turkish-English
Itihasa
WMT2017 Russian-English
OmniNetP
WMT2019 German-English
WMT2019 Finnish-English
WMT2019 English-Japanese
WMT2015 English-Russian
C2-50k Segmentation
WMT2016 Czech-English
Attentional encoder-decoder + BPE
WMT2016 English-Czech
WMT2016 English-French
DeLighT
WMT2017 English-French
OmniNetP
WMT2017 Finnish-English
WMT2017 English-German
OmniNetP
WMT2016 Finnish-English
WMT 2018 English-Estonian
Multi-pass backtranslated adapted transformer
WMT2016 Russian-English
WMT2017 English-Finnish
OmniNetP
V_B (trained on T_H)
ACCURAT balanced test corpus for under resourced languages Russian-Estonian
Multilingual Transformer
ACCURAT balanced test corpus for under resourced languages Estonian-Russian
Alexa Point of View
T5
Business Scene Dialogue EN-JA
Business Scene Dialogue JA-EN
Transformer-base
IWSLT 2017
GPT-4o (HPT)
IWSLT2015 Chinese-English
BP-Transformer
IWSLT2015 Thai-English
Seq-KD + Seq-Inter + Word-KD
Multi Lingual Bug Reports
ChatGPT
slone/myv_ru_2022 myv-ru
slone/mbart-large-51-myv-mul-v1
slone/myv_ru_2022 ru-myv
Tatoeba (EL-to-EN)
Tatoeba (EN-to-EL)
PENELOPIE Transformers-based NMT (EN2EL)
V_A (trained on T_H)
M_C
WMT 2022 Russian-English
V_C (trained on T_H)
WMT 2017 English-Latvian
WMT 2018 English-Finnish
Transformer trained on highly filtered data
WMT 2018 Estonian-English
Multi-pass backtranslated adapted transformer
WMT 2022 Chinese-English
Vega-MT
WMT 2022 Czech-English
WMT 2022 English-Chinese
WMT 2022 English-Czech
WMT 2022 English-German
WMT 2022 English-Japanese
WMT 2022 English-Russian
Vega-MT
WMT 2022 German-English
WMT 2022 Japanese-English