Python 跑深度学习遇到的一些问题集锦

研海无涯

已于 2022-03-20 20:19:13 修改

阅读量4.4k

点赞数

分类专栏： Doing Work 文章标签： python linux tensorflow

于 2019-12-23 22:15:45 首次发布

本文链接：https://blog.csdn.net/xiaojianzhao/article/details/103673938

版权

Doing Work 专栏收录该内容

19 篇文章

订阅专栏

文章目录

TypeError: softmax() got an unexpected keyword argument ‘axis’

tensorflow gan lib

ImportError: libcublas.so.9.0: cannot open shared object file: No such file or directory

https://zhuanlan.zhihu.com/p/28997106

https://www.jianshu.com/p/a201b91b3d96

不是版本不对应的问题，是/usr/local/cuda-9.0没开通权限的问题

Sudo Chmod -R 777 /usr/local/cuda-9.0

需要安装高版本的cuda，cudnn

export PATH=/usr/local/cuda-9.0/bin:$PATH

export LD_LIBRARY_PATH=usr/local/cuda-9.0/lib64

export CUDA_HOME=$DATA_PATH/soft/cuda-9.0

LD_LIBRARY_PATH=/DATA/119/gyhu/soft/cuda-9.0/lib64

PATH==/DATA/119/gyhu/soft/cuda-9.0/bin

SyntaxError: Non-ASCII character ‘\xe9’

Pycharm查找与替换

Pycharm查找与替换l

enumerate 用法

TypeError: Expected sequence or array-like, got <class ‘NoneType’>

空的，matlab没导入数据

Conv2DTranspose

https://blog.csdn.net/nima1994/article/details/83959495

numpy的升维和降维

keras怎么特征合并到一起

https://blog.csdn.net/quiet_girl/article/details/84106625
http://blog.sina.com.cn/s/blog_a14297610102x1ze.html
http://blog.sina.com.cn/s/blog_a14297610102x1ze.html
https://keras-cn-docs.readthedocs.io/zh_CN/latest/getting_started/functional_API
https://keras-cn-docs.readthedocs.io/zh_CN/latest/getting_started/sequential_model/

Keras: Merge和merge区别、Sequencial()和Model()区别

https://blog.csdn.net/quiet_girl/article/details/84106625

deep learning 调参

https://zhuanlan.zhihu.com/p/24720954

https://www.zhihu.com/question/25097993

keras的EarlyStopping callbacks的使用与技巧

https://blog.csdn.net/silent56_th/article/details/72845912
https://blog.csdn.net/zhangxin4832/article/details/81452996

ValueError: Error when checking target: expected dense_2 to have 2 dimensions, but got array with shape (990, , 2)

https://blog.csdn.net/weixin_41735859/article/details/86288356

深度学习迭代停止条件

https://deeplearning4j.org/cn/earlystopping
https://deeplearning4j.org/cn/earlystopping

https://blog.csdn.net/hyg1985/article/details/42556847

贝叶斯深度学习，可解释性

https://www.cnblogs.com/wuliytTaotao/p/10281766.html
让DL可解释？这一份66页贝叶斯深度学习教程告诉你
http://www.zhuanzhi.ai/document/bc61534f4dfc07e4e44b205abc480319
深度学习的可解释性研究（一）：让模型「说人话」
https://www.leiphone.com/news/201805/Ti3mOdeRlWTplIlZ.html
https://www.leiphone.com/news/201806/PDfTEnrBfkUcggu2.html

如何计算参数大小

https://blog.csdn.net/hzhj2007/article/details/80164909

keras 最后结果精度不一致（结果固定）

数据清洗

https://www.cnblogs.com/jasonfreak/p/5448385.html
https://www.cnblogs.com/jasonfreak/p/5448385.html

TypeError: float() argument must be a string or a number, not ‘map’

https://blog.csdn.net/qq_30190513/article/details/78286582

Keras中利用np.random.shuffle()打乱数据集

from numpy as np
index=np.arange(2000)
np.random.shuffle(index)
print(index[0:20])
X_train=X_train[index,:,:,:] #X_train是训练集，y_train是训练标签
y_train=y_train[index]