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

Tokenizer for ConveRT #4978

Closed
tabergma opened this issue Dec 16, 2019 · 0 comments · Fixed by #4996
Closed

Tokenizer for ConveRT #4978

tabergma opened this issue Dec 16, 2019 · 0 comments · Fixed by #4996
Assignees
@tabergma
Labels
area:rasa-oss 🎡 Anything related to the open source Rasa framework type:enhancement ✨ Additions of new features or changes to existing ones, should be doable in a single PR

Comments

@tabergma
Copy link
Contributor

Implement a tokenizer for ConveRT that allows us to use embedding for ConveRT in a sequence-fashion, for example, for the CRFEntityExtractor.

Problem: ConveRT tokenizes words into subwords and adds special characters. Thus, the token start and end does not match the entities. We need to work on an alignment so that the tokens from ConveRT match the entities.
@tabergma tabergma added type:enhancement ✨ Additions of new features or changes to existing ones, should be doable in a single PR area:rasa-oss 🎡 Anything related to the open source Rasa framework labels Dec 16, 2019
@tabergma tabergma self-assigned this Dec 16, 2019
@tabergma tabergma mentioned this issue Dec 17, 2019
Closed
4 tasks
@tabergma tabergma mentioned this issue Jan 3, 2020
Merged
4 tasks
@JulianGerhard21 JulianGerhard21 mentioned this issue Jan 30, 2020
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@tabergma tabergma

Labels
area:rasa-oss 🎡 Anything related to the open source Rasa framework type:enhancement ✨ Additions of new features or changes to existing ones, should be doable in a single PR
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

ConvertTokenizer, CLS token & features as sequence RasaHQ/rasa
Add ConveRTTokenizer RasaHQ/rasa
1 participant
@tabergma