Wolof (wo) 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 |
| 10000 | 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 | wowiki sample |
|---|---|
| original | turam ma ngi jóge ci latin ''october'', jóge ci ''octo'' (juroom-ñett), moo nekkoon juroom-ñetteelu weer ci arminaatu waa-rom bu jamono ju yàgg ja. jataayu jëwrini bitim reew yi nag, mooy ñaareelu campeef bi ëpp solo ci mbootaay gi, ginaaw jataayu buur yeek njiit yi. campeef bii nag dana faral di nekkoon tuubaa, ñooñu daje ci diir bu gàtt,ba nu matee ñaar fukk ak juroom ñatt,sëñ mbakke buso jiite leen,loolu lépp a ngi ame woon ci làqu gu tar,ng |
| 1000 | ▁tur am ▁ma ▁ngi ▁j ó ge ▁ci ▁l at in ▁'' o c t ob er '', ▁j ó ge ▁ci ▁'' o c t o '' ▁( j ur oom - ñ et t ), ▁moo ▁nekkoon ▁juroom - ñ et t eelu ▁weer ▁ci ▁ar m in aatu ▁waa - r om ▁bu ▁jamono ▁ju ▁yàgg ▁ja . ▁jat aayu ▁jë w ri ni ▁bi ti m ▁reew ▁yi ▁nag , ▁mooy ▁ñaar eelu ▁c am p eef ▁bi ▁ëpp ▁solo ▁ci ▁mbootaay ▁gi , ▁ginaaw ▁jat aayu ▁buur ▁y eek ▁njiit ▁yi . ▁c am p eef ▁bii ▁nag ▁dan a ▁far al ▁di ▁nekkoon ▁tuubaa , ▁ñoo ñu ▁daj e ▁ci ▁di ir ▁bu ▁g àtt , ba ▁nu ▁mat ee ▁ñaar ▁fukk ▁ak ▁juroom ▁ñatt , s ë ñ ▁mb a kk e ▁bu so ▁jiite ▁leen , l oolu ▁lépp ▁a ▁ngi ▁ame ▁woon ▁ci ▁l à qu ▁gu ▁tar , ng |
| 3000 | ▁turam ▁ma ▁ngi ▁jóge ▁ci ▁latin ▁'' o ct ob er '', ▁jóge ▁ci ▁'' o c to '' ▁( jur oom - ñ ett ), ▁moo ▁nekkoon ▁juroom - ñ ett eelu ▁weer ▁ci ▁armin aatu ▁waa - rom ▁bu ▁jamono ▁ju ▁yàgg ▁ja . ▁jataayu ▁jëwrini ▁bitim ▁reew ▁yi ▁nag , ▁mooy ▁ñaareelu ▁camp eef ▁bi ▁ëpp ▁solo ▁ci ▁mbootaay ▁gi , ▁ginaaw ▁jataayu ▁buur ▁yeek ▁njiit ▁yi . ▁camp eef ▁bii ▁nag ▁dana ▁faral ▁di ▁nekkoon ▁tuubaa , ▁ñooñu ▁daje ▁ci ▁diir ▁bu ▁gàtt , ba ▁nu ▁mat ee ▁ñaar ▁fukk ▁ak ▁juroom ▁ñatt , sëñ ▁mbakke ▁buso ▁jiite ▁leen , loolu ▁lépp ▁a ▁ngi ▁ame ▁woon ▁ci ▁làqu ▁gu ▁tar , ng |
| 5000 | ▁turam ▁ma ▁ngi ▁jóge ▁ci ▁latin ▁'' o ct ob er '', ▁jóge ▁ci ▁'' o c to '' ▁( jur oom - ñ ett ), ▁moo ▁nekkoon ▁juroom - ñ ett eelu ▁weer ▁ci ▁arminaatu ▁waa - rom ▁bu ▁jamono ▁ju ▁yàgg ▁ja . ▁jataayu ▁jëwrini ▁bitim ▁reew ▁yi ▁nag , ▁mooy ▁ñaareelu ▁campeef ▁bi ▁ëpp ▁solo ▁ci ▁mbootaay ▁gi , ▁ginaaw ▁jataayu ▁buur ▁yeek ▁njiit ▁yi . ▁campeef ▁bii ▁nag ▁dana ▁faral ▁di ▁nekkoon ▁tuubaa , ▁ñooñu ▁daje ▁ci ▁diir ▁bu ▁gàtt , ba ▁nu ▁mat ee ▁ñaar ▁fukk ▁ak ▁juroom ▁ñatt , sëñ ▁mbakke ▁buso ▁jiite ▁leen , loolu ▁lépp ▁a ▁ngi ▁ame ▁woon ▁ci ▁làqu ▁gu ▁tar , ng |
| 10000 | ▁turam ▁ma ▁ngi ▁jóge ▁ci ▁latin ▁'' o ct ob er '', ▁jóge ▁ci ▁'' o c to '' ▁( juroom - ñett ), ▁moo ▁nekkoon ▁juroom - ñett eelu ▁weer ▁ci ▁arminaatu ▁waa - rom ▁bu ▁jamono ▁ju ▁yàgg ▁ja . ▁jataayu ▁jëwrini ▁bitim ▁reew ▁yi ▁nag , ▁mooy ▁ñaareelu ▁campeef ▁bi ▁ëpp ▁solo ▁ci ▁mbootaay ▁gi , ▁ginaaw ▁jataayu ▁buur ▁yeek ▁njiit ▁yi . ▁campeef ▁bii ▁nag ▁dana ▁faral ▁di ▁nekkoon ▁tuubaa , ▁ñooñu ▁daje ▁ci ▁diir ▁bu ▁gàtt , ba ▁nu ▁matee ▁ñaar ▁fukk ▁ak ▁juroom ▁ñatt , sëñ ▁mbakke ▁buso ▁jiite ▁leen , loolu ▁lépp ▁a ▁ngi ▁ame ▁woon ▁ci ▁làqu ▁gu ▁tar , ng |