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.

Sign up for GitHub

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

Jump to bottom

added simplified BERT support #791

Closed
wants to merge 3 commits into from
Closed

added simplified BERT support #791

wants to merge 3 commits into from

Conversation

varisd
Copy link
Member

@varisd varisd commented Feb 6, 2019 •
edited
Loading

Includes:

  • BERT like training using TransformerEncoder + SequenceLabeler
  • Data preparation script
  • tests/bert.ini with an example
  • T2T batching scheme fetching from PR T2T batching #786

@varisd varisd force-pushed the bert_simple branch 2 times, most recently from 8f7b2d6 to 1c16164 Compare February 6, 2019 19:25
jlibovicky
jlibovicky requested changes Feb 21, 2019
View reviewed changes
Copy link
Contributor

@jlibovicky jlibovicky left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tohle mi taky přijde dobrý, jenom se musíme domluvit, co tim labelerem.

neuralmonkey/decoders/sequence_labeler.py
@@ -1,12 +1,12 @@
from typing import Dict, Union
from typing import Dict
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OMG, přesně tyhlety změny mám už od října v nějaký větvi, ale čekal jsem, až Jindra dodělá tf.Dataset a mám s tím nějaké modely, co bych dál rád používal. Máš tohle už v hodně modelech?

Copy link
Member Author

@varisd varisd Feb 21, 2019 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Experimenty s BERTem mam zatim on-hold, takze no problem.

Klidne tenhle PR zavri a pouzij tu svoji vetev.

scripts/preprocess_bert.py
@@ -0,0 +1,79 @@
#!/usr/bin/env python3
# Creates training data for the BERT network training
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tohle se nepíše do komentáže, ale prostě začneš psát text do """ stringu, on se pak uloží do proměnné __doc__ a python ho považuje za dokumentaci.

@varisd
Copy link
Member Author

varisd commented Mar 14, 2019

Resolved by #798

@varisd varisd closed this Mar 14, 2019
@jlibovicky
Copy link
Contributor

Počkej, a co ten skriptíček a testíček, ty se hoděj, ne?

@varisd
Copy link
Member Author

varisd commented Mar 15, 2019

Jo, ty by se daly odsud vykuchat.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@jlibovicky jlibovicky jlibovicky requested changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@varisd @jlibovicky