【2020年版】NVIDIA Jetson Nano JetPackのバージョン情報まとめ、JetPack 4.4は仕様変更の影響が大きい

【2020年版】NVIDIA Jetson Nano、Jetson Xavier NXの便利スクリプト

# Caffe doesn't currently support cuDNN 8.0 (JetPack 4.4 Product Release)
　コメントにして無効にする
# USE_CUDNN := 1
# CPU_ONLY := 1

　または USE_CUDNN = 0にする
USE_CUDNN := 0

src/caffe/layers/cudnn_conv_layer.cpp:475:11: error: ‘CUDNN_CONVOLUTION_BWD_DATA_SPECIFY_WORKSPACE_LIMIT’ was not declared in this scope
           CUDNN_CONVOLUTION_BWD_DATA_SPECIFY_WORKSPACE_LIMIT,
           ^
./include/caffe/util/cudnn.hpp:20:28: note: in definition of macro ‘CUDNN_CHECK’
     cudnnStatus_t status = condition; \
                            ^~~~~~~~~

src/caffe/layers/cudnn_conv_layer.cpp:480:54: error: ‘cudnnGetConvolutionForwardAlgorithm’ was not declared in this scope
       CUDNN_CHECK(cudnnGetConvolutionForwardAlgorithm(Caffe::cudnn_handle(0),
                   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
           fwd_bottom_descs_[i], fwd_filter_desc_, fwd_conv_descs_[i], fwd_top_descs_[i],
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
           CUDNN_CONVOLUTION_FWD_SPECIFY_WORKSPACE_LIMIT,
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
           align_down<8>(workspace_bytes / ws_groups()), &fwd_algo_[i]));
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

jetson@jetson-desktop:~ $ cat /etc/nv_tegra_release
# R32 (release), REVISION: 3.1, GCID: 18186506, BOARD: t210ref, EABI: aarch64, DATE: Tue Dec 10 06:58:34 UTC 2019

# R32 (release), REVISION: 3.1 = L4T r32.3.1

jetson@jetson-desktop:~ $ cat /sys/module/tegra_fuse/parameters/tegra_chip_id
33

24 = TX2
25 = Xavier
33 = TX1
64 = TK1

jetson@jetson-desktop:~$ uname -a
Linux jetson-desktop 4.9.140-tegra #1 SMP PREEMPT Tue Nov 5 13:43:53 PST 2019 aarch64 aarch64 aarch64 GNU/Linux

nvcc --version

/usr/local/cuda/bin/nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Mon_Mar_11_22:13:24_CDT_2019
Cuda compilation tools, release 10.0, V10.0.326

nvidia-smi

dpkg -l | grep cuda
ii  cuda-command-line-tools-10-2                  10.2.89-1                                        arm64        CUDA command-line tools
ii  cuda-compiler-10-2                            10.2.89-1                                        arm64        CUDA compiler

dpkg -l | grep CUDA
ii  cuda-command-line-tools-10-2                  10.2.89-1                                        arm64        CUDA command-line tools
ii  cuda-compiler-10-2                            10.2.89-1                                        arm64        CUDA compiler

cat /usr/local/cuda/version.txt
CUDA Version 10.0.326

JetPack 4.3まで
cat /usr/include/cudnn.h | grep CUDNN_MAJOR -A 2
#define CUDNN_MAJOR 7
#define CUDNN_MINOR 5
#define CUDNN_PATCHLEVEL 0

JetPack 4.4以降
cat /usr/include/cudnn_version.h | grep CUDNN_MAJOR -A 2
#define CUDNN_MAJOR 8
#define CUDNN_MINOR 0
#define CUDNN_PATCHLEVEL 0
--
#define CUDNN_VERSION (CUDNN_MAJOR * 1000 + CUDNN_MINOR * 100 + CUDNN_PATCHLEVEL)

cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
cat: /usr/local/cuda/include/cudnn.h: No such file or directory

# JetPack 4.2.x
opencv_version
3.3.1

# JetPack 4.3 L4T r32.3.1
# JetPack 4.4 L4T r32.4.2
$ opencv_version
4.1.1

# L4T 32.4.2 = JetPack 4.4 DP
# L4T 32.4.3 = JetPack 4.4
jetson@linux:~ $ cat /etc/nv_tegra_release | grep "R32 (release), REVISION: 4."
# R32 (release), REVISION: 4.2, GCID: 20074772, BOARD: t186ref, EABI: aarch64, DATE: Thu Apr  9 01:26:40 UTC 2020
jetson@linux:~ $ echo $?
0

# L4T 32.3.1 = JetPack 4.3
jetson@linux:~ $ cat /etc/nv_tegra_release | grep "R32 (release), REVISION: 3.1"
jetson@linux:~ $ echo $?
1

jetson@linux:~$ vulkaninfo | grep Version
Vulkan Instance Version: 1.2.70
        apiVersion     = 0x402083  (1.2.131)
        driverVersion  = 134283392 (0x8010080)

jetson@linux:~$ vulkan-smoketest
terminate called after throwing an instance of 'std::runtime_error'
  what():  failed to connect to the display server
Aborted (core dumped)

SSH接続からは「画面」が無いから失敗する

sudo apt-get -y install libhdf5-serial-dev hdf5-tools libhdf5-dev zlib1g-dev zip libjpeg8-dev liblapack-dev libblas-dev gfortran
sudo apt-get -y install python3-pip
sudo pip3 install -U pip
sudo pip3 install -U pip testresources setuptools numpy==1.16.1 future==0.17.1 mock==3.0.5 h5py==2.9.0 keras_preprocessing==1.0.5 keras_applications==1.0.8 gast==0.2.2 futures protobuf pybind11

# TF-2.x
sudo pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v44 tensorflow
sudo pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v44 'tensorflow>2'

# Verifying The Installation
# TF-2.x
python3 -c "import tensorflow; print (tensorflow.__version__)"
# AttributeError: module 'tensorflow' has no attribute '__version__'

python3 -c "import tensorflow; print (tf.version.VERSION)"

# TF-1.15
sudo pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v44 'tensorflow<2'

# Verifying The Installation
python3 -c "import tensorflow; print (tensorflow.__version__)"
# 1.15.2

PyTorch v1.5.0 - JetPack 4.4 DP
Python 3.6 - torch-1.5.0-cp36-cp36m-linux_aarch64.whl
As per the PyTorch Release Notes 120, Python 2 is not longer supported

● Python 3.6 + PyTorch v1.5.0
wget https://nvidia.box.com/shared/static/3ibazbiwtkl181n95n9em3wtrca7tdzp.whl -O torch-1.5.0-cp36-cp36m-linux_aarch64.whl
sudo apt-get -y install python3-pip libopenblas-base libopenmpi-dev
pip3 install Cython
pip3 install numpy torch-1.5.0-cp36-cp36m-linux_aarch64.whl

Installing collected packages: Cython
  WARNING: The scripts cygdb, cython and cythonize are installed in '/home/jetson/.local/bin' which is not on PATH.
  Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.
Successfully installed Cython-0.29.20
jetson@jetson-desktop:~/deep_dream_tensorflow $ pip3 install numpy torch-1.5.0-cp36-cp36m-linux_aarch64.whl
Defaulting to user installation because normal site-packages is not writeable
Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (1.16.1)
Processing ./torch-1.5.0-cp36-cp36m-linux_aarch64.whl
Requirement already satisfied: future in /usr/local/lib/python3.6/dist-packages (from torch==1.5.0) (0.17.1)
Installing collected packages: torch
  WARNING: The scripts convert-caffe2-to-onnx and convert-onnx-to-caffe2 are installed in '/home/jetson/.local/bin' which is not on PATH.
  Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.
Successfully installed torch-1.5.0

PyTorch v1.4.0 - JetPack 4.4 DP
Python 2.7 - torch-1.4.0-cp27-cp27mu-linux_aarch64.whl
Python 3.6 - torch-1.4.0-cp36-cp36m-linux_aarch64.whl

● Python 3.6
wget https://nvidia.box.com/shared/static/ncgzus5o23uck9i5oth2n8n06k340l6k.whl -O torch-1.4.0-cp36-cp36m-linux_aarch64.whl
sudo apt-get -y install python3-pip libopenblas-base libopenmpi-dev
pip3 install Cython
pip3 install numpy torch-1.4.0-cp36-cp36m-linux_aarch64.whl

● Python 2.7
wget https://nvidia.box.com/shared/static/1v2cc4ro6zvsbu0p8h6qcuaqco1qcsif.whl -O torch-1.4.0-cp27-cp27mu-linux_aarch64.whl
sudo apt-get -y install libopenblas-base libopenmpi-dev
pip install future torch-1.4.0-cp27-cp27mu-linux_aarch64.whl

● torchvision v0.6.0

cd
sudo apt-get install libjpeg-dev zlib1g-dev
# see below for version of torchvision to download
git clone https://github.com/pytorch/vision torchvision --branch v0.6.0 --depth 1
cd torchvision
# sudo python setup.py install
sudo python3 setup.py install
# Searching for Pillow==7.1.2
# Best match: Pillow 7.1.2
# Searching for torch==1.5.0
# Best match: torch 1.5.0
# Searching for numpy==1.16.1
# Best match: numpy 1.16.1
# Searching for future==0.17.1
# Best match: future 0.17.1
# Using /usr/local/lib/python3.6/dist-packages
# Finished processing dependencies for torchvision==0.6.0a0+b68adcf

jetson@jetson-desktop:~/torchvision $ python3
Python 3.6.9 (default, Apr 18 2020, 01:56:04)
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torchvision
>>> print(torchvision.__version__)
0.6.0a0+b68adcf
>>> exit()

# attempting to load torchvision from build dir will result in import error
cd ../

# always needed for Python 2.7, not needed torchvision v0.5.0+ with Python 3.6
pip install 'pillow<7'

/VideoReader.cpp:6:
/home/jetson/torchvision/torchvision/csrc/cpu/decoder/defs.h:11:10: fatal error: libavcodec/avcodec.h: No such file or directory
 #include <libavcodec/avcodec.h>
          ^~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
error: command 'aarch64-linux-gnu-gcc' failed with exit status 1

sudo apt-get -y install libavcodec-dev
sudo apt-get -y install libavformat-dev

/VideoReader.cpp:6:
/home/jetson/torchvision/torchvision/csrc/cpu/decoder/defs.h:17:10: fatal error: libswscale/swscale.h: No such file or directory
 #include "libswscale/swscale.h"
          ^~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
error: command 'aarch64-linux-gnu-gcc' failed with exit status 1

sudo apt-get -y install libswscale-dev

$ python3

import torch
print(torch.__version__)
print('CUDA available: ' + str(torch.cuda.is_available()))
print('cuDNN version: ' + str(torch.backends.cudnn.version()))
a = torch.cuda.FloatTensor(2).zero_()
print('Tensor a = ' + str(a))
b = torch.randn(2).cuda()
print('Tensor b = ' + str(b))
c = a + b
print('Tensor c = ' + str(c))

import torchvision
print(torchvision.__version__)

jetson@jetson-desktop:~ $ python3

Python 3.6.9 (default, Apr 18 2020, 01:56:04)
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> print(torch.__version__)
1.5.0
>>> print('CUDA available: ' + str(torch.cuda.is_available()))
CUDA available: True
>>> print('cuDNN version: ' + str(torch.backends.cudnn.version()))
cuDNN version: 8000
>>> a = torch.cuda.FloatTensor(2).zero_()
>>> print('Tensor a = ' + str(a))
Tensor a = tensor([0., 0.], device='cuda:0')
>>> b = torch.randn(2).cuda()
>>> print('Tensor b = ' + str(b))
Tensor b = tensor([ 1.2480, -0.2380], device='cuda:0')
>>> c = a + b
>>> print('Tensor c = ' + str(c))
Tensor c = tensor([ 1.2480, -0.2380], device='cuda:0')

>>> import torchvision
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'torchvision'
>>> print(torchvision.__version__)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
NameError: name 'torchvision' is not defined

>>> exit()

It is in a separate repository https://github.com/pytorch/vision, should be quite easy to install from source since you already compiled pytorch.

pip install torchvision
# Successfully built pillow
# Installing collected packages: pillow, torchvision
# Successfully installed pillow-7.1.2 torchvision-0.2.2.post3

jetson@jetson-desktop:~ $ python3
Python 3.6.9 (default, Apr 18 2020, 01:56:04)
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torchvision
>>> print(torchvision.__version__)
0.2.2
>>> exit()

cd
git clone https://github.com/hjptriplebee/deep_dream_tensorflow.git
cd deep_dream_tensorflow

python3 main.py --input nature_image.jpg --output nature_image_tf.jpg
python3 main.py --input paint.jpg --output paint_tf.jpg

python3 main.py --input nature_image.jpg --output nature_image_tf.jpg

  File "/usr/lib/python3.6/dist-packages/cv2/__init__.py", line 79, in bootstrap
    import cv2
ImportError: /usr/lib/aarch64-linux-gnu/libgomp.so.1: cannot allocate memory in static TLS block

cd
git clone https://github.com/CMU-Perceptual-Computing-Lab/openpose -b v1.6.0 --depth 1

cd
cd openpose
sudo bash ./scripts/ubuntu/install_deps.sh
mkdir build
cd build
cmake ..
make clean
time make -j4

-- Building with CUDA.
-- CUDA detected: 10.2
-- Found cuDNN: ver. ??? found (include: /usr/include, library: /usr/lib/aarch64-linux-gnu/libcudnn.so)
CMake Error at cmake/Cuda.cmake:263 (message):
  cuDNN version >3 is required.
Call Stack (most recent call first):
  cmake/Cuda.cmake:291 (detect_cuDNN)
  CMakeLists.txt:417 (include)

-- Configuring incomplete, errors occurred!
See also "/home/jetson/openpose/build/CMakeFiles/CMakeOutput.log".
See also "/home/jetson/openpose/build/CMakeFiles/CMakeError.log".

【2020年版】NVIDIA JetPack 4.4 DP Developer Previewで OpenPoseのビルドエラーの対策方法

【2020年版】NVIDIA Jetson Nano対応の OpenPoseをビルドする手順

USB シリアル デバイス
USB\VID_0955&PID_7020&MI_02\7&10FAEB5B&0&0002
USB\VID_0955&PID_7020&REV_0002&MI_02
USB\VID_0955&PID_7020&MI_02

CDC NCM
USB\VID_0955&PID_7020&MI_05\7&10FAEB5B&0&0005
USB\VID_0955&PID_7020&REV_0002&MI_05
USB\VID_0955&PID_7020&MI_05

New features for all Jetson products
Supports headless initial configuration
Developer kits support initial configuration from another computer connected via the same USB port used for flashing. This is an alternative to setup with display, keyboard, and mouse attached to developer kit.

Jetson Nano、Xavier NX 開発者キットを電源と USBケーブルだけの接続で最初の初期設定を行なう方法

~/.bash_profile
　が存在するので
~/.profile
　が自動的に読み込まれない。
　結果的に
~/.bashrc
　の内容がターミナル起動時に自動的に反映されない。

jetson@linux:~$ source ~/.bashrc
jetson@linux:~ $

jetson@linux:~ $ cat ~/.profile
# ~/.profile: executed by the command interpreter for login shells.
# This file is not read by bash(1), if ~/.bash_profile or ~/.bash_login
# exists.
# see /usr/share/doc/bash/examples/startup-files for examples.
# the files are located in the bash-doc package.

# the default umask is set in /etc/profile; for setting the umask
# for ssh logins, install and configure the libpam-umask package.
#umask 022

# if running bash
if [ -n "$BASH_VERSION" ]; then
    # include .bashrc if it exists
    if [ -f "$HOME/.bashrc" ]; then
        . "$HOME/.bashrc"
    fi
fi

# set PATH so it includes user's private bin if it exists
if [ -d "$HOME/bin" ] ; then
    PATH="$HOME/bin:$PATH"
fi

# set PATH so it includes user's private bin if it exists
if [ -d "$HOME/.local/bin" ] ; then
    PATH="$HOME/.local/bin:$PATH"
fi

jetson@linux:~ $ echo $HOME/.bashrc
/home/jetson/.bashrc

jetson@linux:~ $ cat ~/.bash_profile
export PATH="/usr/lib/llvm-7/bin:$PATH"

jetson@linux:~ $ cat  ~/.bash_login
cat: /home/jetson/.bash_login: No such file or directory

jetson@linux:~ $ echo $BASH_VERSION
4.4.20(1)-release

[libprotobuf ERROR google/protobuf/message_lite.cc:123] Can't parse message of type "caffe.NetParameter" because it is missing required fields: layer[0].clip_param.min, layer[0].clip_param.max
F0821 14:26:29.665053 22812 upgrade_proto.cpp:97] Check failed: ReadProtoFromBinaryFile(param_file, param) Failed to parse NetParameter file: models/pose/body_25/pose_iter_584000.caffemodel

【2020年版】NVIDIA Jetson Nano、Jetson Xavier NXの便利スクリプト

【2020年】Jetson Xavier NX 開発者キットが安かったので衝動買いした件、標準販売価格５万円が４万４千円！

【2020年版】NVIDIA Jetson用に最新の CMake 3.17.3をビルドしてインストールする方法

【2020年版】NVIDIA JetPack 4.4 DP Developer Previewで OpenPoseのビルドエラーの対策方法

Jetson Nano、Jetson Xavier NX 開発者キット TensorFlow on Jetson Platform

NVIDIA Jetson Nano 開発者キットを買ってみた。メモリ容量 4GB LPDDR4 RAM

【2020年版】NVIDIA Jetson Nanoの初期設定、最高速で動かす設定、空きメモリを増やす方法等

【2020年版】NVIDIA Jetson Nano対応の FFmpegをビルドする手順

【2020年版】NVIDIA Jetson Nano対応の OpenPoseをビルドする手順

NVIDIA Jetson Nanoで OpenCV 3をビルドしてインストールする方法、NVCaffe等の OpenCV 4未対応を動かす

NVIDIA Jetson Nanoで Visual Studio Code Open Sourceをビルドして実行する

【2020年版】NVIDIA Jetson Nanoで TensorFlowの StyleGANを動かして、顔画像を生成

【2020年版】NVIDIA Jetson Nanoで StyleGANの改良版の StyleGAN2で自然な画像を生成

【2020年版】Jetson Nanoで StyleGANを動かして可愛い美少女のアニメ顔を大量生産する方法

【メモリ8GB】Raspberry Pi 4 Model B 8GBを KSYで最安値で購入。ベンチマークレビュー

【Vulkan】Raspberry Pi 4 Model Bで Vulkanドライバをビルドして 3Dグラフィックのデモを動かす

PIP機能付きの 4K対応の 4入力 1出力の HDMIセレクターを買ってみた、HDMI機器が複数有る場合に便利

EDID保持機能付きの 4K対応の 4入力 2出力の マトリックス切り替え HDMIセレクター、液晶画面 2台と使用で最強