how to fine tune bert for ner on custom data?
I want to fine tune bert on my custom dataset but i don't know how to tag sentences and do fine tuning if anyone knows please suggest me.
Hello, I am looking for the same thing. There is this link to a colab notebook : https://colab.research.google.com/github/tensorflow/tpu/blob/master/tools/colab/bert_finetuning_with_cloud_tpus.ipynb But it is restricted to MRPC, CoLA, MNLI tasks for which you need an appropriate formated dataset from your corpus. I would have expected to be able to fine -tune on my training corpus (unlabeled) using masked language model task. Have you found smthg else ?
@Anticsss @sofieneJ I am also looking for this. Did any of you manage to find a solution?
I was also trying to use BERT for custom NER. I found a slotion kind of. Check out https://www.kaggle.com/amrkeleg/fine-tuning-bert-models-for-bio-entity-recognition used bert for biological NER.
@Ankan1998 what is this? is this a building dataset format for bert? \