about model input and training

kyzhouhzau / BERT-NER

Use Google's BERT for named entity recognition （CoNLL-2003 as the dataset）.

MIT License

1.23k stars 335 forks source link

about model input and training #74

Open OKtang opened 4 years ago

OKtang commented 4 years ago

hi, i have many questions about the model input. firstly, what's the shape of 'input_ids_p'? is it an iterator? and then, how the bert do training in many epochs? lastly, which part in bert code shows the stage of Iterative training? thank you for your reply!