Kikuyu (ki) 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 | kiwiki sample |
|---|---|
| original | {| border="0" cellpadding="0" cellspacing="0" align="right" style="margin-left:0em; background:#f0f0f0; border: 0px #aaa solid; border-collapse: colla o ũrĩa mahinda marathiĩ na ũtaramu ũgathie ũkĩambatagĩra noguo kambuni irĩa irenda gũkorwo ikĩoa ikinya rĩa gũthomothia ciana maũndũ matiganĩte irath ciohi ici cithondekagwo nĩ oo atomu kũruta erektroni na kũgwatanira na atomu iria ĩngĩ. ngerekano ya kinyamũ, kĩna kiohi kia mũthemba ũyũ nĩ oxigen . |
| 1000 | ▁{| ▁border ="0" ▁cellpadding ="0" ▁cellspacing ="0" ▁align =" right " ▁style =" margin - le ft : 0 em ; ▁background : # f 0 f 0 f 0; ▁border : ▁0 px ▁# aa a ▁solid ; ▁border - collapse : ▁c oll a ▁o ▁ũrĩa ▁ma hinda ▁mara thiĩ ▁na ▁ũ tara mu ▁ũ ga thie ▁ũ kĩ a mba ta gĩ ra ▁no guo ▁kambuni ▁irĩa ▁i renda ▁gũ korwo ▁i kĩ o a ▁i kin ya ▁rĩa ▁gũ thomo thia ▁ciana ▁maũndũ ▁mati g anĩte ▁ira th ▁cio hi ▁ici ▁ci thondeka gwo ▁nĩ ▁o o ▁atomu ▁kũ ruta ▁erektroni ▁na ▁kũ gwa tani ra ▁na ▁atomu ▁iria ▁ĩ ngĩ . ▁ngere kano ▁ya ▁kin ya mũ , ▁kĩ na ▁ki o hi ▁kia ▁mũ themba ▁ũyũ ▁nĩ ▁ox ig en ▁. |
| 3000 | ▁{| ▁border ="0" ▁cellpadding ="0" ▁cellspacing ="0" ▁align =" right " ▁style =" margin - left :0 em ; ▁background :# f 0 f 0 f 0; ▁border : ▁0 px ▁# aaa ▁solid ; ▁border - collapse : ▁c oll a ▁o ▁ũrĩa ▁mahinda ▁mara thiĩ ▁na ▁ũ tara mu ▁ũga thie ▁ũ kĩa mbata gĩra ▁no guo ▁kambuni ▁irĩa ▁i renda ▁gũkorwo ▁i kĩo a ▁i kinya ▁rĩa ▁gũ thomo thia ▁ciana ▁maũndũ ▁mati g anĩte ▁ira th ▁cio hi ▁ici ▁ci thondekagwo ▁nĩ ▁oo ▁atomu ▁kũruta ▁erektroni ▁na ▁kũ gwa tani ra ▁na ▁atomu ▁iria ▁ĩngĩ . ▁ngerekano ▁ya ▁kinya mũ , ▁kĩna ▁kiohi ▁kia ▁mũthemba ▁ũyũ ▁nĩ ▁oxigen ▁. |
| 5000 | ▁{| ▁border ="0" ▁cellpadding ="0" ▁cellspacing ="0" ▁align =" right " ▁style =" margin - left :0 em ; ▁background :# f 0 f 0 f 0; ▁border : ▁0 px ▁# aaa ▁solid ; ▁border - collapse : ▁c oll a ▁o ▁ũrĩa ▁mahinda ▁mara thiĩ ▁na ▁ũ tara mu ▁ũga thie ▁ũkĩa mbata gĩra ▁no guo ▁kambuni ▁irĩa ▁i renda ▁gũkorwo ▁i kĩoa ▁ikinya ▁rĩa ▁gũ thomo thia ▁ciana ▁maũndũ ▁matig anĩte ▁ira th ▁cio hi ▁ici ▁ci thondekagwo ▁nĩ ▁oo ▁atomu ▁kũruta ▁erektroni ▁na ▁kũ gwa tanira ▁na ▁atomu ▁iria ▁ĩngĩ . ▁ngerekano ▁ya ▁kinya mũ , ▁kĩna ▁kiohi ▁kia ▁mũthemba ▁ũyũ ▁nĩ ▁oxigen ▁. |