Pretrained models
·
160 commits
to main
since this release
Overview
This release contains pretrained models which can be loaded using ktrain as follows:
import ktrain
from cached_path import cached_path
model_path = cached_path("url/to/pretrained/model", extract_archive=True)
model = ktrain.load_predictor(model_path)See the README for more details.
Pretrained models
title_abstract_model.zip[URL]: trained on titles and abstracts of PubMed articles. Classifies an article as being relevant for Biofactoid curation (1) or not (0).