This repository provides a new siamese neural models (sentence transformers) trained on biomedical data. These models can be used in several NLP tasks like Document classification, Question Answering etc.
Our models are deployed in HuggingFace via this link https://huggingface.co/menadsa/
script is the traning script of our models
Our training and validation set dataset are available via this drive link https://drive.google.com/drive/folders/0ALX47Q5owRX_Uk9PVA?hl=fr
We have created 2 datasets. The first one for the first training step with 3.79 millions pairs for the training and 18940 for the validation set. The second consist of 6.75 millions pairs for the training and 18,557 for the validation.