Anaconda notes
Mirror site installation is recommended. https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/
Add a specific source
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/bioconda/
conda config --set show_channel_urls yes
View update source configuration
conda config --show channels
According to the necessary environment
tensorflow-gpu
install mingw libpython
keras
windows under the Path environment variable much of a problem
solution is to create a new variable, such as EPath, then part of a long variable in it. Delete Path has been placed EPath variables, then Path join% Epath% variable. In the cmd command, type Path, check whether the changes take effect or success.
NetEase mailbox, foxmail client error 554, you need time to send the message, you can send a copy to yourself.
The procedure entry point could not be located in the dynamic link library OPENSSL_sk_new_reserve, solutions, replace files of different dates, see the specific file can not locate the procedure entry point
ImportError: can not import name 'abs ' error, see the solutions to solve ImportError: can not import name 'abs '
Download the official website address Cunda
cuDNN depth neural network URL
Literature supports:
Deep learning environment to build win + python + tensorflow1.5 + CUDA9.0 + cuDNN7.0
The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.解决方案报错The TensorFlow library wasn't compiled to use SSE instructions的解决方案
import os
os.environ['TF_CPP_MIN_LOG_LEVEL']='2'
import tensorflow as tf
Depth study, the default load remote data collection, because Amazon's official website contamination, the use of home owners find ip, after the replacement, use the download tool to download is good, then the package into the folder:
C: \ the Users \ zl.keras \ datasets
modified command:
mnist.load_data (path = "mnist.zip")
using local data set it
references, The Alternative to Import from keras.datasets MNIST