Rainbow

the error when load the pre-train module resnet101

1

RuntimeError: While copying the parameter named rpn_cls_score_net.weight, whose dimensions in the model are torch.Size([18, 512, 1, 1]) and whose dimensions in the checkpoint are torch.Size([24, 512, 1, 1]).

the probelm is: when i use a image, to initial the Surface, error was arise: _**UFuncTypeError: Cannot cast ufunc 'add' output from dtype('float64') to dtype('uint8') with casting rule 'same_kind'**_ after...

result of slicing is an empty tensor

1

this problem happens randomly............... Traceback (most recent call last): File "./tools/trainval_net.py", line 141, in max_iters=args.max_iters) File "/home/user/PycharmProjects/MAttNet/pyutils/mask-faster-rcnn/tools/../lib/model/train_val.py", line 367, in train_net sw.train_model(max_iters) File "/home/user/PycharmProjects/MAttNet/pyutils/mask-faster-rcnn/tools/../lib/model/train_val.py", line 284, in train_model self.net.train_step(blobs, self.optimizer)...

code publication

1

hi, thank you for your brilliant work, i have got many from your work. I'm ready to do related research, but i can't find the code of NeurIPS 2022 paper...

Hello, thank you for this brilliant work. As shown in title, i want to konw that do you provide other model trained weights except the CpnRestNeXt101UNet, like CpnU22,CpnResUNet ? Looking...