checkpoint for ScanNet val seems wrong (only 6M)

[YuQiao0303]

Hi authors, Thank you for this amazing work!

I try to download the checkpoint for ScanNet val, but found in Google Drive that it's only 6M (file name is head_3dyco_scannetv2_val.pth, from link https://drive.google.com/file/d/19FYLM5SzcQSKYR4r9iChauf5cleoXC_H/view). Still, I try to use it with: python3 tools/test.py configs/scannetv2/isbnet_scannetv2.yaml dataset/pretrained/head_3dyco_scannetv2_val.pth --out out/scannetv2, and as expected, it raises the following error:

$ python3 tools/test.py configs/scannetv2/isbnet_scannetv2.yaml dataset/pretrained/head_3dyco_scannetv2_val.pth --out out/scannetv2 2023-08-23 14:38:33,330 - INFO - Load state dict from dataset/pretrained/head_3dyco_scannetv2_val.pth Traceback (most recent call last): File "tools/test.py", line 197, in main() File "tools/test.py", line 100, in main load_checkpoint(args.checkpoint, logger, model) File "/home/yuqiao/ISBNet/isbnet/util/utils.py", line 118, in load_checkpoint state_dict = torch.load(checkpoint, map_location=lambda storage, loc: storage.cuda(device)) File "/root/miniconda3/envs/jimr_pro/lib/python3.8/site-packages/torch/serialization.py", line 587, in load with _open_zipfile_reader(opened_file) as opened_zipfile: File "/root/miniconda3/envs/jimr_pro/lib/python3.8/site-packages/torch/serialization.py", line 242, in init super(_open_zipfile_reader, self).init(torch._C.PyTorchFileReader(name_or_buffer)) RuntimeError: [enforce fail at inline_container.cc:145] . PytorchStreamReader failed reading zip archive: failed finding central directory

It seems that the wrong .pth file has been uploaded to Google Drive? I wonder if it's possible for you to verify and fix it? Thank you!!

[ngoductuanlhp]

Checkpoint link is updated. Best.

[YuQiao0303]

Checkpoint link is updated. Best.

Thank you so much. I tried and it worked.