This page includes Deep Learning environment setup instruction and other useful projects This instruction is for linux 16.04 with Nivida GPU
- Nvidia Graphics driver installation
$ sudo add-apt-repository ppa:graphics-drivers/ppa
$ sudo apt-get update
$ sudo apt-get install nvidia-375
After this, reboot
$ sudo reboot
When u type $ nvidia-smi , you should see the following message.
- CUDA toolkit installation
please download cuda_8.0.61_375.26_linux.run from the following link
https://drive.google.com/open?id=1f_HjxCg30BCD9r8OfhtsaXyKChhAiy8s
Then run the following command for installation
$ sudo sh cuda_8.0.61_375.26_linux.run
You will see long license. Just keep entering until you see the following questions. Please answer the questions as follows
- Run folloiinwg commands for the environment settings
$ echo -e "\n## CUDA and cuDNN paths" >> ~/.bashrc
$ echo 'export PATH=/usr/local/cuda-8.0/bin:${PATH}' >> ~/.bashrc
$ echo 'export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64:${LD_LIBRARY_PATH}' >> ~/.bashrc
Check if the following lines are created in the ~./bashrc
export PATH = /usr/local/cuda-8.0/bin : $ { PATH }
export LD_LIBRARY_PATH = /usr/local/cuda-8.0/lib64 : $ { LD_LIBRARY_PATH }
-
Run following commands to check if CUDA is installed properly
$ source ~/.bashrc
$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2016 NVIDIA Corporation
Built on Tue_Jan_10_13:22:03_CST_2017
Cuda compilation tools, release 8.0, V8.0.61 -
As a final step of CUDA installation, check the CUDA path $ which nvcc
/usr/local/cuda-8.0/bin/nvcc
- Join the site in the following link and download CuDNN v5.1 Library (cudnn-8.0-linux-x64-v5.1.tgz) for Linux https://developer.nvidia.com/rdp/cudnn-download
You can also download cudnn-8.0-linux-x64-v5.1.tgz from the following link
https://drive.google.com/open?id=1f_HjxCg30BCD9r8OfhtsaXyKChhAiy8s
- Run the following commands Note: If nvcc result is not "/usr/local/cuda-8.0/bin/nvcc", use nvcc output instead of "/usr/local/cuda-8.0/bin/nvcc" after "which nvcc"
$ tar xzvf cudnn-8.0-linux-x64-v5.1.tgz
$ which nvcc
/usr/local/cuda-8.0/bin/nvcc
$ sudo cp cuda/lib64/* /usr/local/cuda-8.0/lib64/
$ sudo cp cuda/include/* /usr/local/cuda-8.0/include/
$ sudo chmod a+r /usr/local/cuda-8.0/lib64/libcudnn*
$ sudo chmod a+r /usr/local/cuda-8.0/include/cudnn.h
sudo apt-get install libcupti-dev
Install python 3.6.3 first. Keep Enterng and select yes for all question during installation process
$ bash Anaconda3-4.3.0-Linux-x86_64.sh
$ source ~/.bashrc
Run python and check version
$ python
Python 3.6.0 |Anaconda 4.3.0 (64-bit)| (default, Dec 23 2016, 12:22:00)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Install tensorflow following instruction in the following link
https://www.tensorflow.org/install/install_linux
Note: There are multiple installatio method introduced in the website. Just use pip3 based installation unless you really want isolated development environment.
Here is the best tensorflow tutorial
https://github.com/Lab930boss/DeepLearningZeroToAll.git
Though the vidoe lecture is in Korean, the example is self-contained and well coded.
You will immediately tell what each lab does when you read lab project file name
Fyi there is lecture note link. The lecture note is written in English
https://goo.gl/jPtWNt
https://github.com/jayrambhia/Install-OpenCV
all about Yolo https://github.com/AlexeyAB/darknet#how-to-train-tiny-yolo-to-detect-your-custom-objects
Best Yolo tutorial website https://blog.paperspace.com/how-to-implement-a-yolo-object-detector-in-pytorch/