Ubuntu 20.04 安裝 TenserFlow & PyTorch for RTX 30系列

將會安裝以下套件

1. TensorFlow
2. PyTorch
3. CUDA
4. cuDNN
5. Dependencies and other frameworks like Caffe and Theano(?)

系統需求

• Nvidia RTX 系列 GPUs
• Ubuntu 20.04 LTS (Focal Fossa)

安裝桌機版 Lambda Stack

LAMBDA_REPO=$(mktemp) && \
wget -O${LAMBDA_REPO} https://lambdalabs.com/static/misc/lambda-stack-repo.deb && \
sudo dpkg -i ${LAMBDA_REPO} && rm -f ${LAMBDA_REPO} && \
sudo apt-get update && sudo apt-get install -y lambda-stack-cuda --allow-downgrades

重新啟動電腦

確認是否安裝成功

$ python 
>>> import tensorflow
>>> tensorflow.__version__
'2.6.0'
>>> import torch
>>> torch.__version__
'1.9.1'

管理/升級 TensorFlow & PyTorch

sudo apt-get update && sudo apt-get dist-upgrade

確認 Nvidia Driver、CUDA、cuDNN 使否安裝成功

$ nvidia-smi
Tue Nov  2 21:04:44 2021       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.74       Driver Version: 470.74       CUDA Version: 11.4     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  On   | 00000000:01:00.0 Off |                  N/A |
|  0%   40C    P0   111W / 370W |     72MiB / 10010MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      1326      G   /usr/lib/xorg/Xorg                 56MiB |
|    0   N/A  N/A      1484      G   /usr/bin/gnome-shell               12MiB |
+-----------------------------------------------------------------------------+

$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2020 NVIDIA Corporation
Built on Mon_Oct_12_20:09:46_PDT_2020
Cuda compilation tools, release 11.1, V11.1.105
Build cuda_11.1.TC455_06.29190527_0

簡單測試 (Stanford Dog use vgg16 pretrained model by PyTorch)

1. PyTorch - StandFord Dog
2. PyTorch - MNIST