Open ahmetidea opened 3 years ago
Dear Adam,
I really appreciate the effort of yours to make unlabelled data meaningful. This work, for sure fills the gap of a ready-to-use code of semi-supervised BERT based NER application. I work with Turkish text. Normally, I use the pre-trained models for Turkish at Hugginface such as: https://huggingface.co/dbmdz/bert-base-turkish-uncased and https://huggingface.co/dbmdz/bert-base-turkish-128k-uncased.
Is there an easy way of using these pre-trained models for Turkish within your code?
Thank you very much. A.Yildirim
Dear Adam,
I really appreciate the effort of yours to make unlabelled data meaningful. This work, for sure fills the gap of a ready-to-use code of semi-supervised BERT based NER application. I work with Turkish text. Normally, I use the pre-trained models for Turkish at Hugginface such as: https://huggingface.co/dbmdz/bert-base-turkish-uncased and https://huggingface.co/dbmdz/bert-base-turkish-128k-uncased.
Is there an easy way of using these pre-trained models for Turkish within your code?
Thank you very much. A.Yildirim