there are some different betwween BERT — transformers 4.3.0 documentation and BERT — transformers 4.3.0 documentation
There are labels and next_sentence_label in BertForPretraining, but nothing in TFBertForPretraining.
Does it means there are some different between BertForPretraining and TFBertForPretaining?
Or there is a wrong in TF doc.
There is indeed something wrong in the TF doc: the next_sentence_label is there but not documented. If you want to open a PR to fix this that would be awesome!
Thank you for your reply,I have create a PR at fix tf doc bug by Sniper970119 · Pull Request #10570 · huggingface/transformers · GitHub