Ewe (ee) subword embeddings
| Vocab size | vocab | model | 25 dim | 50 dim | 100 dim | 200 dim | 300 dim |
|---|---|---|---|---|---|---|---|
| 1000 | vocab | model |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
| 3000 | vocab | model |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
| 5000 | vocab | model |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
txt |
bin
bokeh | umap | matrix |
Training corpus sample, encoded with different BPE vocabulary sizes
| Vocab size | eewiki sample |
|---|---|
| original | * m. senn, '' die wickiana. johann jakob wicks nachrichtensammlung aus dem 00. jahrhundert'', zürich 0000. * franz matthias mauelshagen, ''wunderkamme mia denyigba lɔ̃lɔ̃ la enu wonye woa ƒoa sia? mia tɔgbuiwo tsɔ woƒe agbe gble ɖe eta xɔe nami eɖo nye kpli woha dzi be mia wɔ miatɔ sinu nuvevi nyenye tso 0000 la, tata amu ɖoe kplikpa be ameyibɔwudodowo koe anye yeƒe nudodo. eɖoe be yemaga do yevuwu titri siwo na wodzea fifia la le ameyibɔ nyigba dz |
| 1000 | ▁* ▁m . ▁s en n , ▁'' ▁di e ▁wick ia na . ▁j oh an n ▁ja ko b ▁wick s ▁na ch rich te n sa m m l un g ▁a us ▁de m ▁00 . ▁ja h r h under t '', ▁zürich ▁0000. ▁* ▁fran z ▁ma t th ia s ▁ma ue l s ha g en , ▁'' w under ka m me ▁mia ▁de nyi gba ▁lɔ ̃ lɔ ̃ ▁la ▁enu ▁wonye ▁woa ▁ƒo a ▁sia ? ▁mia ▁tɔ gb ui wo ▁tsɔ ▁woƒe ▁agbe ▁gb le ▁ɖe ▁e ta ▁xɔe ▁na mi ▁e ɖo ▁nye ▁kp li ▁wo ha ▁dzi ▁be ▁mia ▁wɔ ▁mia tɔ ▁si nu ▁nu ve vi ▁nye nye ▁tso ▁0000 ▁la , ▁ta ta ▁amu ▁ɖo e ▁kp li kpa ▁be ▁ame yi bɔ wu dodo wo ▁ko e ▁a nye ▁ye ƒe ▁nu dodo . ▁e ɖo e ▁be ▁ye ma ga ▁do ▁yevu wu ▁ti tri ▁siwo ▁na ▁wo dze a ▁fifia ▁la ▁le ▁ame yi bɔ ▁nyi gba ▁dz |
| 3000 | ▁* ▁m . ▁sen n , ▁'' ▁die ▁wickiana . ▁johann ▁jakob ▁wick s ▁na ch rich ten sa mmlung ▁a us ▁dem ▁00. ▁ja hr h under t '', ▁zürich ▁0000. ▁* ▁fran z ▁ma t th ia s ▁ma ue ls ha gen , ▁'' w under kam me ▁mia ▁denyigba ▁lɔ ̃ lɔ ̃ ▁la ▁enu ▁wonye ▁woa ▁ƒo a ▁sia ? ▁mia ▁tɔ gbui wo ▁tsɔ ▁woƒe ▁agbe ▁gb le ▁ɖe ▁eta ▁xɔe ▁nami ▁eɖo ▁nye ▁kpli ▁wo ha ▁dzi ▁be ▁mia ▁wɔ ▁mia tɔ ▁si nu ▁nu ve vi ▁nye nye ▁tso ▁0000 ▁la , ▁tata ▁amu ▁ɖoe ▁kplikpa ▁be ▁ameyibɔ wu dodowo ▁ko e ▁a nye ▁ye ƒe ▁nu dodo . ▁eɖo e ▁be ▁ye ma ga ▁do ▁yevu wu ▁ti tri ▁siwo ▁na ▁wodze a ▁fifia ▁la ▁le ▁ameyibɔ ▁nyigba ▁dz |
| 5000 | ▁* ▁m . ▁sen n , ▁'' ▁die ▁wickiana . ▁johann ▁jakob ▁wicks ▁nach rich tensa mmlung ▁a us ▁dem ▁00. ▁jahr hunder t '', ▁zürich ▁0000. ▁* ▁fran z ▁mat th ia s ▁ma uelsha gen , ▁'' wunder kamme ▁mia ▁denyigba ▁lɔ ̃ lɔ ̃ ▁la ▁enu ▁wonye ▁woa ▁ƒo a ▁sia ? ▁mia ▁tɔgbui wo ▁tsɔ ▁woƒe ▁agbe ▁gble ▁ɖe ▁eta ▁xɔe ▁nami ▁eɖo ▁nye ▁kpli ▁wo ha ▁dzi ▁be ▁mia ▁wɔ ▁miatɔ ▁si nu ▁nu ve vi ▁nye nye ▁tso ▁0000 ▁la , ▁tata ▁amu ▁ɖoe ▁kplikpa ▁be ▁ameyibɔ wu dodowo ▁koe ▁anye ▁yeƒe ▁nu dodo . ▁eɖoe ▁be ▁ye ma ga ▁do ▁yevuwu ▁titri ▁siwo ▁na ▁wodzea ▁fifia ▁la ▁le ▁ameyibɔ ▁nyigba ▁dz |