Large Multilingual Model

by stevenhillis - opened Nov 17, 2022

Discussion

stevenhillis

Nov 17, 2022

Any plans to release an mdeberta-v3-large model? I'd really like to see a large multilingual variant!

Peterr

Jan 23, 2023

I'm unclear on whether mdeberta-v3-base was trained on all of CC100's languages or some subset (there are only 16 specifically mentioned languages). Could someone please clarify?

akhtet

Feb 7, 2023

@Peterr According to the paper, the model was trained on 2.5T CC100 multi-lingual dataset which is the same as XLM-R, so that would be all C100 languages. However, the model is evaluated on XNLI dataset which has 16 languages.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment