Issues loading RegionVIT pre-trained checkpoints

lucidrains / vit-pytorch

Implementation of Vision Transformer, a simple way to achieve SOTA in vision classification with only a single transformer encoder, in Pytorch

MIT License

20.4k stars 3.03k forks source link

Issues loading RegionVIT pre-trained checkpoints #235

Open PrithivirajDamodaran opened 2 years ago

PrithivirajDamodaran commented 2 years ago

Hello there - I am trying to load pre-trained checkpoints from original author's page. Are there changes in the implementation of the model here? None of the checkpoints are loading citing differences in layers.

varunponda commented 1 year ago

Hey @PrithivirajDamodaran Try this, model.load_state_dict(state_dict, strict=False).