pere's picture
first submit
cd5ea82
|
raw
history blame
333 Bytes
metadata
language: 'no'
license: cc-by-4.0
tags:
  - norwegian
  - GPT2
  - casual language modeling
datasets:
  - oscar

Norwegian GPT-2 - Oscar

Description

This is a sample reference model trained only on the Oscar Corpus for a day on a TPU v3-8. Pretrained model on Norwegian language using a causal language modeling (CLM) objective.