jnz
/

electra-ka

Inference Endpoints

Model card Files Files and versions Community

electra-ka / README.md

jnz's picture

jnz

Update README.md

020e4e2 about 4 years ago

|

history blame contribute delete

196 Bytes

electra-ka is first of its kind, Transformer based, open source Georgian language model.

The model is trained on 33GB of Georgian text collected from 4854621 pages in commoncrowl archive.