版本 pytorch0.4/pytorch0.41
小型CNN教程
https://www.jianshu.com/p/e4c7b3eb8f3d
中文文檔
https://pytorch-cn.readthedocs.io/zh/latest/package_references/torch-nn/
《深度學習框架PyTorch：入門與實踐》的對應代碼
來自 <https://github.com/chenyuntc/pytorch-book>
入門
https://zhuanlan.zhihu.com/p/29024978
博客
https://blog.csdn.net/u012436149/article/details/54627597
sort()與sorted()的不同在於，sort是在原位重新排列列表，而sorted()是產生一個新的列表。
來自 <https://www.cnblogs.com/65702708/archive/2010/09/14/1826362.html>
PyTorch 深度學習: 60分鐘快速入門
來自 <https://www.jianshu.com/p/889dbc684622>
pytorch中文教程
https://ptorch.com/docs/1/
跑貓狗大戰例子 ===================================================================
Torch: not enough memory: you tried to allocate 0GB. Buy new RAM!
來自 <https://blog.csdn.net/github_39611196/article/details/79844017>
RuntimeError: DataLoader worker (pid 16393) is killed by signal: Killed
改batch_size=1
https://blog.csdn.net/u011394059/article/details/78233731
改num_workers=0
https://github.com/udacity/P1_Facial_Keypoints/issues/6
UserWarning: volatile was removed and now has no effect
將pytorch更新到0.4.0最新版後對0.3.1版本代碼會有如下警告
來自 <https://blog.csdn.net/weixin_41797117/article/details/80237179>
confusion_matrix.add(score.data.squeeze(), label.type(t.LongTensor))
直接註釋了 /home/cwx/tnt-master/build/lib.linux-x86_64-2.7/torchnet/meter/confusionmeter.py 45行 也沒有用
git上傳 ===========================================================================
origin替換爲[email protected]:c00443756/cats_vs_dogs_pytorch_on_cpu_only.git
https://yuhongjun.github.io/tech/2017/04/28/git%E4%BB%93%E5%BA%93%E5%88%A0%E9%99%A4%E6%89%80%E6%9C%89%E6%8F%90%E4%BA%A4%E5%8E%86%E5%8F%B2%E8%AE%B0%E5%BD%95-%E6%88%90%E4%B8%BA%E4%B8%80%E4%B8%AA%E5%B9%B2%E5%87%80%E7%9A%84%E6%96%B0%E4%BB%93%E5%BA%93.html