1.nn.Conv2D()輸入?yún)?shù)數(shù)據(jù)格式不對
報錯：TypeError: new() received an invalid combination of arguments - got (float, int, int, int), but expected one of:
完整報錯：
File "G:/python/project/model/A2net.py", line 36, in <module>
model = A2Block(64)
File "G:/python/project/model/A2net.py", line 15, in __init__
self.dimension_reduction = nn.Conv2d(in_channels=inplanes, out_channels=inplanes/2, kernel_size=1, stride=1)
File "C:\Users\MSY\Anaconda3\lib\site-packages\torch\nn\modules\conv.py", line 297, in __init__
False, _pair(0), groups, bias)
File "C:\Users\MSY\Anaconda3\lib\site-packages\torch\nn\modules\conv.py", line 33, in __init__
out_channels, in_channels // groups, *kernel_size))
TypeError: new() received an invalid combination of arguments - got (float, int, int, int), but expected one of:
* (torch.device device)
* (torch.Storage storage)
* (Tensor other)
* (tuple of ints size, torch.device device)
* (object data, torch.device device)
問題定位：定位到報錯行為：
self.dimension_reduction = nn.Conv2d(in_channels=inplanes, out_channels=inplanes/2, kernel_size=1, stride=1)
問題分析: 根據(jù)報錯信息丸卷，是說本行代碼包含有float的數(shù)據(jù)類型向抢，通過分析可以看到糊余，只有inplanes/2可能是float類型典徘，由此想到在python3中n/2是帶有小數(shù)點(diǎn)的润樱，應(yīng)該為n//2為整數(shù)贵扰。(由于一個粗心瞭恰，報錯一個如此尷尬的bug)
問題解決：將輸出通道數(shù)inplanes/2改為inplanes//2完美解決腮郊。
2.make.sh 編譯NMS遇到問題
報錯:OSError: The CUDA lib64 path could not be located in /usr/lib64
完整報錯：
Traceback (most recent call last):
File "build.py", line 59, in <module>
CUDA = locate_cuda()
File "build.py", line 54, in locate_cuda
raise EnvironmentError('The CUDA %s path could not be located in %s' % (k, v))
OSError: The CUDA lib64 path could not be located in /usr/lib64
問題定位：打開build.py（某些項(xiàng)目為setup.py）找到
cudaconfig = {'home': home, 'nvcc': nvcc,
'include': pjoin(home, 'include'),
'lib64': pjoin(home, 'lib64')}
問題分析：lib引用的問題
問題解決：將home, 'lib64'中的lib64改為lib完美解決
3.one of the variables needed for gradient computation has been modified by an inplace operation
報錯：one of the variables needed for gradient computation has been modified by an inplace operation
完整報錯：
Traceback (most recent call last):
File "train_test.py", line 454, in <module>
train()
File "train_test.py", line 327, in train
loss.backward()
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/tensor.py", line 93, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/autograd/__init__.py", line 90, in backward
allow_unreachable=True) # allow_unreachable flag
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation
問題定位：這個bug報錯并沒有報明顯的錯誤位置是最難過的，wwwwww~~~~~
問題分析：此問題是在測試運(yùn)行網(wǎng)上關(guān)于一版Pytorch版本的SSD代碼時渺鹦，出現(xiàn)的扰法，根據(jù)網(wǎng)上的錯誤解釋，應(yīng)該時Pytorch0.4版本和0.3版本的某些不一致造成的毅厚。該問題的常用解決方法時：
1：如果使用的是pytorch0.4.0版本塞颁，回退到pytorch0.3.0版本
2：如果有inreplace參數(shù)，設(shè)為False
3：報錯原因是pytorch0.4.0版本后tensor不支持inplace操作了，所以把所有inplace錯作去掉祠锣。
后在博客modified by an inplace operation中似乎找到了合適的答案.簡單來說：x += 1 這種改成 x = x+1 原因：x+=1的值會直接在原值上面做更新酷窥，是inplace=True的情況，而后一種是先讓x+1然后賦值給x,屬于inplace=False
但是由于自己的代碼較多伴网，開始很難具體定位到哪個錯誤的位置蓬推，后來使用Beyond Compare(一款很棒的軟件，強(qiáng)推~~~)與網(wǎng)上一版正確的代碼比較澡腾，發(fā)現(xiàn)了錯誤沸伏。
x /= norm #（原本的錯誤代碼）
In-place的具體解釋可以參考。pytorch 學(xué)習(xí)筆記（二十二）：關(guān)于 inplace operation
問題解決: 將x /= norm #改為x = x / norm
后記：后來偶然發(fā)現(xiàn)动分，Pycharm原來有全局搜索的功能毅糟，上述也說大致的問題由于 /= 操作符產(chǎn)生，但是代碼過多刺啦，無法有效的找到 /=代碼所在emmmmm留特，可以使用全局搜索Pycharm中按快捷鍵Ctrl + Shift + F或從從菜單Edit-》Find-》Find in Path進(jìn)入全局查找界面，輸入 /= 即可找到大致所在玛瘸，VS code也可蜕青，自行查找即可。（白白浪費(fèi)了那么多自己查找的時間糊渊，哇的一聲~~~右核，對Pycharm還有待探索）
補(bǔ)充：后來運(yùn)行另外一個代碼的時候，發(fā)現(xiàn)報了相同的錯誤渺绒，后來找到的錯誤為：
x.unsqueeze_(1) 改為：x = x.unsqueeze(1)
4.Fan in and fan out can not be computed for tensor with less than 2 dimensions
報錯： Fan in and fan out can not be computed for tensor with less than 2 dimensions
完整報錯：
File "train_test_RFB.py", line 143, in <module>
net.extras.apply(weights_init)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 240, in apply
module.apply(fn)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 240, in apply
module.apply(fn)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 241, in apply
fn(self)
File "train_test_RFB.py", line 134, in weights_init
init.kaiming_normal_(m.state_dict()[key], mode='fan_out')
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/init.py", line 323, in kaiming_normal_
fan = _calculate_correct_fan(tensor, mode)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/init.py", line 257, in _calculate_correct_fan
fan_in, fan_out = _calculate_fan_in_and_fan_out(tensor)
File "/home/miao/anaconda3/lib/python3.6/site-packages/torch/nn/init.py", line 181, in _calculate_fan_in_and_fan_out
raise ValueError("Fan in and fan out can not be computed for tensor with less than 2 dimensions")
ValueError: Fan in and fan out can not be computed for tensor with less than 2 dimensions
問題定位:
init.kaiming_normal_(m.state_dict()[key], mode='fan_out')
問題分析:根據(jù)報錯信息贺喝，可以知道，再使用init.kaiming_normal_()進(jìn)行初始化的時候宗兼，只能初始化不小于2的維度的tensor躏鱼，經(jīng)過分析得到，在常見的使用
if 'conv' in key:
init.kaiming_normal_(m.state_dict()[key], mode='fan_out')
進(jìn)行初始化的過程中殷绍，問題出在定義的conv層染苛，回想自己的網(wǎng)絡(luò)，在conv中的定義
self.conv = nn.Sequential(nn.Conv2d(in_channels, inter_channels, 3, padding=1, bias=False),
nn.BatchNorm2d(inter_channels),
nn.ReLU())
其中包含了BatchNorm層主到，而在 BatchNorm layer維度1 , 小于2茶行。'Fan in and fan out can not be computed for tensor with less than 2 dimensions'
問題解決: 將復(fù)合的conv拆開寫，或者改寫初始化
5.libpng error: Read Error
報錯：opencv2 報錯 libpng error: Read Error
完整報錯：
libpng error: Read Error
Traceback (most recent call last):
File "main.py", line 100, in <module>
main(config)
File "main.py", line 43, in main
train.train()
File "/home/msy/project/PoolNet-master/solver.py", line 84, in train
for i, data_batch in enumerate(self.train_loader):
File "/home/msy/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 637, in __next__
return self._process_next_batch(batch)
File "/home/msy/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 658, in _process_next_batch
raise batch.exc_type(batch.exc_msg)
AttributeError: Traceback (most recent call last):
File "/home/msy/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in _worker_loop
samples = collate_fn([dataset[i] for i in batch_indices])
File "/home/msy/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in <listcomp>
samples = collate_fn([dataset[i] for i in batch_indices])
File "/home/msy/project/PoolNet-master/dataset/dataset.py", line 27, in __getitem__
sal_image = load_image(os.path.join(self.sal_root, im_name))
File "/home/msy/project/PoolNet-master/dataset/dataset.py", line 77, in load_image
if len(im.shape) != 3 or im.shape[2] != 3:
AttributeError: 'NoneType' object has no attribute 'shape'
問題定位：
im = cv2.imread(name)
len(im.shape) != 3 or im.shape[2] != 3:
問題分析：圖片格式的問題登钥，有些圖片比如說原本是jpg的格式畔师，你后綴寫成了.png或者其他類似的操作，就可能會出現(xiàn)這個問題（還是不完全理解牧牢，歡迎補(bǔ)充解答）看锉。
問題解決：
import cv2
import numpy as np
from PIL import Image
from PIL import ImageFile
import imghdr
ImageFile.LOAD_TRUNCATED_IMAGES = True
if imghdr.what(name) == "png":
Image.open(name).convert("RGB").save(name)
img = np.array(Image.open(name))
參考：
https://blog.csdn.net/andylei777/article/details/78095411
http://www.itdaan.com/blog/2016/11/22/d480f443ca62e56ddc47a7bed7cc85fd.html
6.TypeError: cannot assign 'torch.cuda.FloatTensor' as parameter 'edges' (torch.nn.Parameter or None expected)
報錯：
TypeError: cannot assign 'torch.cuda.FloatTensor' as parameter 'edges' (torch.nn.Parameter or None expected)
問題定位：
class Conv(nn.Module):
def __init__(self, in_features, out_features, bias=False):
super(Conv, self).__init__()
self.edges = Parameter(torch.Tensor(20, 20))
self.sigmoid = nn.Sigmoid()
def forward(self, input, adj):
self.edges= self.sigmoid((self.edges))
報錯分析：根據(jù)報錯的原因是不能將torch.cuda.FloatTensor作為torch.nn.Parameter的結(jié)果類型姿锭，所以只能將torch.cuda.FloatTensor的結(jié)果賦值給torch.nn.Parameter變量的data屬性。
問題解決：
self.edges.data= self.sigmoid((self.edges))
