aakosm
Results
1
comments of
aakosm
Information on the the language composition of the pretraining dataset would also be welcome, as there are no mention on multilingual capabilities of the model in the linked blog post.