- Fixes arXiv link (d3f604e61eb5af68c9f1e0cdb1f48d7115479ee1)
- fixed other arXiv link (23a440efa12e68faf799c948b34692e3c9ba3b11)
Co-authored-by: Hilmar <hlapp@users.noreply.huggingface.co>
Adds a copy of the upstream model used by Trained_model_SM.pth.
This model was downloaded from
http://data.lip6.fr/cadene/pretrainedmodels/se_resnext50_32x4d-a260b3a4.pth
but that website certificate is no longer maintained.
Saving a copy here to ensure long term availability of the model.
See https://github.com/Cadene/pretrained-models.pytorch for more
details on the se_resnext50_32x4d-a260b3a4.pth model.
Also includes two text documents containing details about
training and validation.
The model was trained using the code in the
https://github.com/hdr-bgnn/BGNN-trait-segmentation repo
by M. Maruf.
