Downloads · 30 days
14
1% of all-time downloads
Jeux2BERT
Massinissa/Jeux2BERT
Jeux2BERT is a feature extraction model from Massinissa. Use it when you need embeddings to search or compare text. It is set up for transformers.
Jeux2BERT is a Flaubert language model augmented by the lexico-semantic network JeuxDeMots. Thus, this model tries to capture the distributional and relational properties of words, but also tries to discriminate the d…
All-time downloads
1.3K
Public
Repo size
436 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin218 MB · 99%
At a glance
- Task
- Feature Extraction
- Library
- transformers
- Model type
- flaubert
- Access
- Public
- Created
- Jun 2, 2022
- Updated
- Jun 22, 2022
- SHA
- 2299c0c1
Model card
From the Hugging Face model README
Jeux2BERT
Jeux2BERT is a Flaubert language model augmented by the lexico-semantic network JeuxDeMots. Thus, this model tries to capture the distributional and relational properties of words, but also tries to discriminate the different relational properties between words or syntagms.
The Web application includes three Tasks : Link Prediction (Classification de triplets), Relation Prediction (Prédiction de Relation) and Triple Ranking (Classement de triplets)
Web App
[https://github.com/atmani-massinissa/Jeux2BERT_APP/tree/main]
Demo
[https://share.streamlit.io/atmani-massinissa/jeux2bert_app/main/app.py?page=Classement+de+triplets]
The task Triple Ranking (Classement de triplets) don't run smoothly on the streamlit server because of the inference's time, so it's better to run it locally instead on the demo's server.
- Task
- Feature Extraction
- Library
- transformers
- Type
- flaubert
- Created
- Jun 2, 2022
- Updated
- Jun 22, 2022