luyaoochen / uninformed-students-pytorch Goto Github PK

python teacher_train.py
...
Traceback (most recent call last):
File "teacher_train.py", line 61, in
model = _Teacher(patch_size).to(device)
File "/home/dit-train-3/anaconda3/envs/pytorch_test/lib/python3.8/site-packages/torch/nn/modules/module.py", line 612, in to
return self._apply(convert)
File "/home/dit-train-3/anaconda3/envs/pytorch_test/lib/python3.8/site-packages/torch/nn/modules/module.py", line 359, in _apply
module._apply(fn)
File "/home/dit-train-3/anaconda3/envs/pytorch_test/lib/python3.8/site-packages/torch/nn/modules/module.py", line 359, in _apply
module._apply(fn)
File "/home/dit-train-3/anaconda3/envs/pytorch_test/lib/python3.8/site-packages/torch/nn/modules/module.py", line 381, in _apply
param_applied = fn(param)
File "/home/dit-train-3/anaconda3/envs/pytorch_test/lib/python3.8/site-packages/torch/nn/modules/module.py", line 610, in convert
return t.to(device, dtype if t.is_floating_point() else None, non_blocking)
RuntimeError: CUDA error: invalid device ordinal

What's happened?

你好，感谢你的复现！
我在跑evaluate.py的时候：
plt.imshow(plt_img, cmap='jet') # 这里报错: TypeError: Invalid dimensions for image data，上一步plt_img的shape为[256, 256, 1]，改为[256, 256]后，还是报这个错。
后来我把plt这部分注释掉了，跑后面的程序，然后发现结果图 res.jpg 效果不是特别理想，有些确实很不错（我跑的carpet数据，patch_size[17, 33, 65], 3个student）。请问有无提升的计划，我建了一个群（QQ: 689772351），大家一起讨论，嘻嘻~

Hello, thanks for your work! Can I ask if there is any difference between the feature map extracted from student and teacher? When I visualize the feature maps, they seem to be same.

Hello, thanks for your codes! Can I ask why you fetch the first 512-d vector from the fc's output? I guess we should extract from the fc's input, that is, a 512-d feature.