search

lihaoliu-cambridge / mtmr-net

[TMI'19 & DLMIA'18] Code for "Multi-Task Deep Learning with Margin Ranking Loss for Lung Nodule Analysis".
44 stars 12 forks source link

Project requirements #1

Closed danpresil closed 6 months ago

danpresil commented 5 years ago

Can you specify the package requirements for this project?

Many thanks!

lihaoliu-cambridge commented 5 years ago

Can you specify the package requirements for this project?

Many thanks!

Hi Dan:

Please refer to the main page for requirement details.

Best Regards, Lihao.

danpresil commented 5 years ago

Hey, Thank you for the quick reply,

I'm still getting the same error with the following packages installed: numpy 1.16.3 SimpleITK 1.2.0 tensorboardX 1.6 torch 0.4.1 torchvision 0.2.2.post3

/MTMR-NET/logs/error_logs/logs: Wed, 24 Apr, 2019, 08:14:55 main.py[line:57] [INFO] Begin program Wed, 24 Apr, 2019, 08:15:17 main.py[line:62] [ERROR] Traceback (most recent call last): File "/venv_python2/local/lib/python2.7/site-packages/torch/utils/data/dataloader.py", line 42, in _worker_loop samples = collate_fn([dataset[i] for i in batch_indices]) File "/MTMR-NET/preprocessing/data_handler_2d_regression.py", line 524, in __getitem__ img_array = self.transforms(pil_image) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/transforms.py", line 60, in __call__ img = t(img) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/transforms.py", line 163, in __call__ return F.normalize(tensor, self.mean, self.std, self.inplace) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/functional.py", line 206, in normalize mean = torch.tensor(mean, dtype=torch.float32) AttributeError: 'module' object has no attribute 'float32'

Wed, 24 Apr, 2019, 08:20:38 main.py[line:57] [INFO] Begin program Wed, 24 Apr, 2019, 08:21:00 main.py[line:62] [ERROR] input and target shapes do not match: input [64 x 1], target [64] at /pytorch/aten/src/THCUNN/generic/MSECriterion.cu:12

Dan.

lihaoliu-cambridge commented 5 years ago

Hey, Thank you for the quick reply,

I'm still getting the same error with the following packages installed: numpy 1.16.3 SimpleITK 1.2.0 tensorboardX 1.6 torch 0.4.1 torchvision 0.2.2.post3

/MTMR-NET/logs/error_logs/logs: Wed, 24 Apr, 2019, 08:14:55 main.py[line:57] [INFO] Begin program Wed, 24 Apr, 2019, 08:15:17 main.py[line:62] [ERROR] Traceback (most recent call last): File "/venv_python2/local/lib/python2.7/site-packages/torch/utils/data/dataloader.py", line 42, in _worker_loop samples = collate_fn([dataset[i] for i in batch_indices]) File "/MTMR-NET/preprocessing/data_handler_2d_regression.py", line 524, in __getitem__ img_array = self.transforms(pil_image) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/transforms.py", line 60, in __call__ img = t(img) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/transforms.py", line 163, in __call__ return F.normalize(tensor, self.mean, self.std, self.inplace) File "/venv_python2/local/lib/python2.7/site-packages/torchvision/transforms/functional.py", line 206, in normalize mean = torch.tensor(mean, dtype=torch.float32) AttributeError: 'module' object has no attribute 'float32'

Wed, 24 Apr, 2019, 08:20:38 main.py[line:57] [INFO] Begin program Wed, 24 Apr, 2019, 08:21:00 main.py[line:62] [ERROR] input and target shapes do not match: input [64 x 1], target [64] at /pytorch/aten/src/THCUNN/generic/MSECriterion.cu:12

Dan.

I think you should use Pytorch 0.3.0.

I have updated the version I used.

Best Regards, Lihao.