Fetching metadata from the HF Docker repository...
Feb 20 15:45 Update app.py 15:43 Update app.py Feb 5 21:44 Update app.py 21:43 Update inference.py Jan 31 23:31 Update app.py 23:07 Update app.py 22:51 Upload 2 files 22:47 Update app.py Overwriting by local 22:42 Update inference.py Overwriting by local 22:41 Update app.py Overwritting by local 22:40 Update inference.py Overwriting by local 21:23 Update inference.py Before (wrong):
https://huggingface.co/ttoosi/robust_resnet50_eps_3.00_epoch_best/blob/main/resnet50_robust.pt
After (correct):
https://huggingface.co/ttoosi/robust_resnet50_eps_3.00_epoch_best/resolve/main/resnet50_robust.pt 21:18 Update inference.py adding weights_only=False because In PyTorch 2.6, torch.load(..., weights_only=True) is the default, my checkpoints are in pickle format (old) 21:13 Update inference.py updated the robust model to the one I trained 20:54 Update inference.py testing the chekcpoint I trained 14:42 Update inference.py Back to original checkpoint 14:36 Update inference.py Trying the checkpoint, I trained with a config of normalization 14:00 Update inference.py Changing the checkpoint to what I trained (eps 3) 01:27 Upload 9 files loading stimuli 01:26 Upload 8 files uploading base files Jan 30 20:06 initial commit 
