Fine-tuning BEREL_2.0 for Samaritan Hebrew (Torah) and Samaritan Aramaic
#2
by
johnlockejrr
- opened
Sorry for a dumb question, I'm new to BERT and I never trained a model, only OCR (kraken) for manuscript transcription. I would like to train this model for Samaritan Hebrew and Samaritan Aramaic. Should I prepare my raw texts in any way so I can train with this model? I'm very interested in same functionality, at least fill-mask etc. Any help is appreciated.