Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Named Entity Recognition task with my own tags. #1176

Closed
ostreech1997 opened this issue Apr 13, 2020 · 2 comments
Closed

Named Entity Recognition task with my own tags. #1176

ostreech1997 opened this issue Apr 13, 2020 · 2 comments
Assignees

Comments

@ostreech1997
Copy link

ostreech1997 commented Apr 13, 2020

Hi, I'm intrested in NER task. In documentation, I see predefind 18 tags from PERSON to CARDINAL. However, my task isn't solved by this tags. Is there any possibility to add new tags?
I want the following: use your DeepPavlov library, but with my own tags. And if it's possible, how many examples(sentences) I need?
P.S. I use this documentation link: http://docs.deeppavlov.ai/en/master/features/models/ner.html

@theanhle
Copy link
Contributor

Hi, all you need is to (1) prepare your dataset in the CoNLL2003 format (refer to the link you mentioned above), then (2) create a config file to define the model with your own hyperparameters. You should collect as many samples as possible. It's hard to say how many is enough. For instance, the CoNLL2003 dataset consists of about 15000, 3500, 3600 samples for training, validation, and test, respectively. Anyway, just try to create a dataset, build a model, train on it, and see how it works.

@ostreech1997
Copy link
Author

Ok, I got it!
Thanks a lot!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants