“This is the 31st day of my participation in the First Challenge 2022. For details: First Challenge 2022”

  • 🎉 statement: as one of the bloggers in AI field, ❤️ lives up to your time and your ❤️

    • 🍊 deep learning: # Environment building, article reading

    • 🍊 TensorFlow GPU version installation test

    • 📆 Last updated: February 22, 2022

    • 🍊 thumbs up 👍 collect ⭐ comments 📝 are all the bloggers insist on writing, updating high-quality blog post the biggest motivation!

    • 🍊 👋 Follow me 👋, join us to Get more interesting AI, charge 🚀 👋

    To build a GPU version of TensorFlow, the prerequisite is that an NVIDIA graphics card capable of supporting CUDA needs to install its basic support platform CUDA and its machine learning library cuDNN, and then build the corresponding TensorFlow GPU version on this basis

    📔 TensorFlow1.2 to 2.1 CUDA and cuDNN of GPU versions are as follows:

    • 🍊 this table may be relatively old

    📔 Conda source acceleration with Pip

    If you want to do a good job, you must first sharpen your tools. Here are two blog posts worth checking out

    • 🍊 # Linux and Windows setup PIP image source – the most practical machine learning library download acceleration setup

    • 🍊# Anaconda conda switch to domestic source, Windows and Linux configuration method, add tsinghua source —

    • 🍊 You are advised to use conda to create an independent environment, and conda is preferred for software library installation and management

    The advantage of using Conda installation is that key libraries such as CUDatoolkit-11.3.1 and CUDNn-8.2.1.32 will automatically adapt well when executing Conda install tensorflow-GPU ==2.6.0

    📕 The Linux server environment is as follows

    # # server

cat /etc/issue
 
Ubuntu 16.047. LTS \n \l


# # Cuda version
nvcc -V

Cuda compilation tools, release 10.1, V101.243.


# # graphics

NVIDIA GeForce GTX 108011 g Ti single cardCopy the code

    📗 tensorflow-gpu==1.15.0 Installation example

    
conda create -n tf15 python=3.69.

conda activate tf15

Conda is recommended for tensorflow installation
conda install tensorflow-gpu==1.15. 0

pip install opencv-python

pip install pillow
Copy the code

    📘 Checks whether TensorFlow is available

    • 🍊 Shell or CMD window to enter the Python interactive environment
    python

Python 3.69. |Anaconda, Inc.| (default, Jul 30 2019.19:07:31) 
[GCC 7.3. 0] on linux
Type "help"."copyright"."credits" or "license" for more information.
>>> import tensorflow as tf
>>> print(tf.__version__)
1.15. 0
>>> 

Copy the code

    📗 Query the tensorflow version retrieved by Conda

    Conda search TensorFlow Queries the tensorflow version retrieved by Conda

    • The versions of TensorFlow that Conda can retrieve are roughly as follows
    
conda search tensorflow

# Most items are deleted here
 
Loading channels: done
# Name Version Build Channel
tensorflow                     0.71.          py27_0  anaconda/cloud/conda-forge

tensorflow                     0.8. 0          py34_0  anaconda/cloud/conda-forge
tensorflow                     0.9. 0          py27_0  anaconda/cloud/conda-forge

tensorflow                    0.10. 0          py27_0  anaconda/cloud/conda-forge
tensorflow                    0.10. 0          py34_0  anaconda/cloud/conda-forge

tensorflow                     1.0. 0          py27_0  anaconda/cloud/conda-forge

tensorflow                     1.01.     np112py27_0  anaconda/pkgs/free  
tensorflow                     1.1. 0     np111py27_0  anaconda/pkgs/free  
tensorflow                     1.1. 0     np111py35_0  anaconda/pkgs/free  
tensorflow                     1.2. 0          py27_0  anaconda/cloud/conda-forge
tensorflow                     1.21.          py36_0  anaconda/cloud/conda-forge
tensorflow                     1.21.          py36_0  anaconda/pkgs/free  
tensorflow                     1.3. 0               0  anaconda/pkgs/free  
tensorflow                     1.3. 0          py36_0  anaconda/cloud/conda-forge
tensorflow                     1.4. 0          py27_0  anaconda/cloud/conda-forge
tensorflow                     1.41.               0  anaconda/pkgs/main  
tensorflow                     1.41.               0  pkgs/main           
tensorflow                     1.5. 0               0  anaconda/pkgs/main  

tensorflow                     1.5. 0          py36_0  anaconda/cloud/conda-forge
tensorflow                     1.51.          py27_0  anaconda/cloud/conda-forge

tensorflow                     1.6. 0               0  anaconda/pkgs/main  

tensorflow                     1.7. 0               0  anaconda/pkgs/main  
       
tensorflow                     1.8. 0               0  anaconda/pkgs/main  

tensorflow                     1.9. 0 eigen_py27hf386fcc_1  anaconda/pkgs/main  
     
tensorflow                     1.9. 0 gpu_py36h02c5d5e_1  anaconda/pkgs/main  

tensorflow                    1.10. 0 eigen_py27ha0ab958_0  anaconda/pkgs/main  


tensorflow                    1.10. 0          py36_0  anaconda/cloud/conda-forge
tensorflow                    1.11. 0 eigen_py27h06aee4b_0  anaconda/pkgs/main  
tensorflow                    1.11. 0 eigen_py27h06aee4b_0  pkgs/main           
  
tensorflow                    1.12. 0 eigen_py36hbd5f568_0  anaconda/pkgs/main  
  
tensorflow                    1.131. eigen_py27h5e92bea_0  anaconda/pkgs/main  

tensorflow                    1.131.  py37h90a7d86_1  anaconda/cloud/conda-forge
tensorflow                    1.132.      h76b4ce7_0  anaconda/cloud/conda-forge
tensorflow                    1.14. 0 eigen_py27h99c1539_0  anaconda/pkgs/main  
     
tensorflow                    1.14. 0 gpu_py27he9627f8_0  anaconda/pkgs/main  

tensorflow                    1.14. 0 mkl_py37h45c423b_0  pkgs/main           
tensorflow                    1.15. 0 eigen_py27h7b7505e_0  anaconda/pkgs/main  
        
tensorflow                     2.0. 0 eigen_py27hec4e49e_0  anaconda/pkgs/main  
tensorflow                     2.0. 0 eigen_py27hec4e49e_0  pkgs/main           

tensorflow                     2.1. 0 eigen_py27h636cc2a_0  pkgs/main           
tensorflow                     2.1. 0 eigen_py36hbb90eaf_0  anaconda/pkgs/main  
      
tensorflow                     2.2. 0 eigen_py36h84d285f_0  anaconda/pkgs/main  

tensorflow                     2.2. 0 mkl_py38h6d3daf0_0  pkgs/main           
tensorflow                     2.3. 0 eigen_py37h189e6a2_0  anaconda/pkgs/main  
tensorflow                     2.3. 0 eigen_py37h189e6a2_0  pkgs/main           

tensorflow                     2.4. 0  py37h89c1867_0  anaconda/cloud/conda-forge
tensorflow                     2.4. 0  py38h578d9bd_0  anaconda/cloud/conda-forge
tensorflow                     2.41. eigen_py37h3da6045_0  anaconda/pkgs/main  
tensorflow                     2.41. eigen_py37h3da6045_0  pkgs/main           

tensorflow                     2.43.  py36h5fab9bb_0  anaconda/cloud/conda-forge

tensorflow                     2.5. 0 eigen_py37hff93566_0  anaconda/pkgs/main  
tensorflow                     2.5. 0 eigen_py37hff93566_0  pkgs/main           

tensorflow                     2.5. 0 mkl_py39h4a0693c_0  pkgs/main           
tensorflow                     2.6. 0 cpu_py37hc107814_2  anaconda/cloud/conda-forge
tensorflow                     2.6. 0 cpu_py38h077e6c3_2  anaconda/cloud/conda-forge
tensorflow                     2.6. 0 cpu_py39hcb7c6aa_2  anaconda/cloud/conda-forge

tensorflow                     2.6. 0 eigen_py37h34b007a_0  pkgs/main           
tensorflow                     2.6. 0 eigen_py38hcc1cb13_0  anaconda/pkgs/main  
tensorflow                     2.6. 0 eigen_py38hcc1cb13_0  pkgs/main           
tensorflow                     2.6. 0 eigen_py39h4b72145_0  anaconda/pkgs/main  
tensorflow                     2.6. 0 eigen_py39h4b72145_0  pkgs/main           
tensorflow                     2.6. 0 mkl_py37h9d15365_0  anaconda/pkgs/main  
tensorflow                     2.6. 0 mkl_py37h9d15365_0  pkgs/main           
tensorflow                     2.62. cuda112py39h9333c2f_1  anaconda/cloud/conda-forge

Copy the code

    📙 tensorflow-gpu==2.6.0 Installation Example

    conda create -n tfNew python=3.8. 5

conda activate tfNew

conda search tensorflow

conda install tensorflow

conda install tensorflow-gpu==2.6. 0
Copy the code

    📙 Tensorflow – Gpu other versions

    • X version and 2.X version are widely used at present, and these two versions have many differences in functions
    • Therefore, friends in the environment to build, to distinguish
    • Other 1.X version and 2.X version of the installation, we copy the above modificationconda install tensorflow-gpu==X.X.XVersion number is ok

    🚀🚀 Mexic AI

  • 🎉 as one of the bloggers with the most dry goods in the field of AI, ❤️ lives up to his time and qing ❤️
  • ❤️ If the article is helpful to you, like, comment encourage bloggers every minute to seriously create

    • Happy learning AI, deep learning environment building: an article to read

    • 🍊 # Ubuntu install CUDa11.2 for current users

    • 🍊 # Linux and Windows setup PIP image source – the most practical machine learning library download acceleration setup

    • 🍊# Anaconda conda switch to domestic source, Windows and Linux configuration method, add tsinghua source —

    • 🍊 # Specify the GPU to run and train Python programs, deep learning single card, multi-card training GPU Settings

    • 🍊 # Install Pytorch and Torchvision in Cuda10.0 for Linux

    • 🍊 # Read SSH password login, public key authentication login

    • Install JDK 11: configure the JAVA_HOME environment variable