search

MingSun-Tse / Collaborative-Distillation

[CVPR'20] Collaborative Distillation for Ultra-Resolution Universal Style Transfer (PyTorch)
MIT License
185 stars 23 forks source link

'NoneType' object is not callable #10

Closed zsjhhh closed 3 years ago

zsjhhh commented 4 years ago

Hi, I got an error like this,

Traceback (most recent call last): File "main.py", line 73, in net = TrainSE_With_WCTDecoder(args).cuda() File "/home/ubuntu/ZSJ/Collaborative-Distillation-master/model/model.py", line 8, in init self.BE = eval("model_original.Encoder%d" % args.stage)(args.BE, fixed=True) File "/home/ubuntu/ZSJ/Collaborative-Distillation-master/model/model_original.py", line 472, in init load_param(t7_model, 0, self.conv0) File "/home/ubuntu/ZSJ/Collaborative-Distillation-master/utils.py", line 66, in load_param_from_t7 model.get(in_layer_index).weight.float()) TypeError: 'NoneType' object is not callable

I have downloaded wct_models and BD models. Is that because the pytorch version? I use pytorch-1.4.0. If so, will you release a pytorch>=1.0 version? Thank you very much.

nickyvdz-art commented 4 years ago

not sure if this is specific to your problem; but it seems like its

pytorch==0.4.1

zsjhhh commented 4 years ago

When I used pytorch=1.14.0, I use torchfile.load instead of load_lua, and then the error appeared. I created a new virtual environment and installed pytorch=0.4.1， then the problem gone.

MingSun-Tse commented 3 years ago

At this point, I don't have more practical suggestions about how to run our code on environments other than that we suggested. So I am closing this issue. If anyone still has problem with it, please let me know.