得峰的专栏 [网络收藏]

【代码】Pycharm jupyter server process exited with code 1。

添加清华源conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaco

最近在使用TF2.0。运行程序出现以下错误。Failed to get convolution algorithm. This is probably because cuDNN failed to initialize一开始怀疑是CUDA和CuDNN配置错误（要求版本匹配）。反复试验后，还是有这个错误。最后发现可能是GPU显存不足造成的。需要在程序前加以下一段代码：from tensorflow.compat.v1 import ConfigProtofrom tensorflow.

在cmd中使用命令来安装python3.7： conda create -n py37 python=3.7 anacondaconda activate py37cuda安装注意自定义去掉vs选项cudnn记得配置环境变量pathYou can list all discoverable environments with conda info --envs.conda info --envsconda activate py37pip ...

GPU - CUDA - Windows Subsystem for Linux (WSL)GPU in Windows Subsystem for Linux (WSL)CUDA on Windows Subsystem for Linux (WSL)https://developer.nvidia.com/cuda/wslCUDA on WSL User GuideThe guide for using NVIDIA CUDA on Windows Subsystem for L...

TensorFlow 2.0 includes many API changes, such as reordering arguments, renaming symbols, and changing default values for parameters. Manually performing all of these modifications would be tedious and prone to error. To streamline the changes, and to make

问题Jupyter Notebook可以说是非常好用的小工具，但是不经过配置只能够在本机访问笔者参阅了文档对jupyter notebook进行配置，实现了跨主机浏览器访问安装jupyter notebook笔者使用conda包管理conda install jupyter notebook生成默认配置文件jupyter notebook --generate-config将会在用户主目录下生成.jupyter文件夹，其中jupyter_notebook_c..

pip install jupyter_contrib_nbextensionsjupyter contrib nbextension install --userpip install --user jupyter_nbextensions_configuratorjupyter nbextensions_configurator enable --user

pypi 镜像使用帮助pypi 镜像每 5 分钟同步一次。临时使用pip install -i https://pypi.tuna.tsinghua.edu.cn/simple some-package注意，simple不能少, 是https而不是http设为默认升级 pip 到最新的版本 (>=10.0.0) 后进行配置：pip install pip -Upip config set global.index-url https://pypi.tuna....

import tensorflow as tftf.config.list_physical_devices(‘GPU’)