metadata
license: cc-by-nc-4.0
datasets:
- slone/nllb-200-10M-sample
pipeline_tag: translation
language:
- ak
- am
- ar
- awa
- azj
- bm
- ban
- be
- bem
- bn
- bho
- bjn
- bug
- bg
- ca
- ceb
- cs
- cjk
- ckb
- crh
- da
- de
- dik
- dyu
- el
- en
- eo
- et
- ee
- fo
- fj
- fi
- fon
- fr
- fur
- ff
- gaz
- gd
- ga
- gl
- gn
- gu
- ht
- ha
- he
- hi
- hne
- hr
- hu
- hy
- ig
- ilo
- id
- is
- it
- jv
- ja
- kab
- kac
- kam
- kn
- ks
- ks
- ka
- kk
- kbp
- kea
- mn
- km
- ki
- rw
- ky
- kmb
- kmr
- kr
- kr
- kg
- ko
- lo
- lij
- li
- ln
- lt
- lmo
- ltg
- lb
- lua
- lg
- luo
- lus
- lv
- mag
- mai
- ml
- mr
- min
- mk
- mt
- mni
- mos
- mi
- my
- nl
- nb
- ne
- nso
- nus
- ny
- oc
- ory
- pag
- pa
- pap
- pbt
- fa
- plt
- pl
- pt
- prs
- qu
- ro
- rn
- ru
- sg
- sa
- sat
- scn
- shn
- si
- sk
- sl
- sm
- sn
- sd
- so
- st
- es
- sc
- sr
- ss
- su
- sv
- sw
- szl
- ta
- taq
- tt
- te
- tg
- tl
- ti
- tpi
- tn
- ts
- tk
- tum
- tr
- tw
- tzm
- ug
- uk
- umb
- ur
- uz
- vec
- vi
- war
- wo
- xh
- yi
- yo
- zh
- zh
- ms
- zu
It is a truncated version of NLLB-200-600M model (6 layers instead of 12, 512 hidden dimensions instead of 1024) with 175M parameters (131M of which are embeddings).
This model was fine-tuned on the slone/nllb-200-10M-sample subset of the NLLB dataset, using only the samples with BLASER score above 3.5.
It is really bad at translation, but can serve as a base model for further fine-tuning for a small number of languages. It is recommended to prune the vocabulary of this model before fine-tuning, to preserve only the tokens used with the intended languages.