LGV pretrained models

Martin GubriFigshare 2022 · DOI · OpenAlex

Abstract

LGV models used as surrogate in the original paper. Those resnet50 models were collected along the SGD trajectory with a high learning rate. The zip file contains three random seeds in respective subfolders. Each one contains a subfolder with the original pretrained model from which the model collection started. These pretrained models were trained by <sub>Ashukha, A., et al. Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learning (2020)</sub>. <br>
BibTeX
@inproceedings{gubri2022lgv,
  author = {Martin Gubri},
  title = {{LGV pretrained models}},
  booktitle = {Figshare},
  year = {2022},
  doi = {10.6084/m9.figshare.20497821},
}

Loading related content...

Figure