Машинное обучение и Большие данные: различия между версиями

Материал из Artem Aleksashkin's Wiki
Перейти к навигации Перейти к поиску
Строка 100: Строка 100:
git checkout r2.3
git checkout r2.3
./configure
./configure
bazel build --config=cuda //tensorflow/tools/pip_package:build_pip_package
bazel build --config=cuda --config=opt //tensorflow/tools/pip_package:build_pip_package
</pre>
</pre>


Версия от 18:58, 30 марта 2021

Ai-brain.jpg

Hardware

Software

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/11.2.2/local_installers/cuda-repo-ubuntu2004-11-2-local_11.2.2-460.32.03-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu2004-11-2-local_11.2.2-460.32.03-1_amd64.deb
sudo apt-key add /var/cuda-repo-ubuntu2004-11-2-local/7fa2af80.pub
sudo apt-get update
sudo apt-get -y install cuda nvidia-cuda-toolkit

Change Default Python

sudo update-alternatives --install /usr/bin/python python /usr/bin/python2 1
sudo update-alternatives --install /usr/bin/python python /usr/bin/python3 2
sudo update-alternatives --config python

Define Your Cuda Version

  • Nvidia GeForce GTX 760 4gb -> Nvidia Kepler
  • Nvidia Kepler -> CUDA SDK 10.0 – 10.2 support for compute capability 3.0 – 7.5 (Kepler, Maxwell, Pascal, Volta, Turing). Last version with support for compute capability 3.x (Kepler). 10.2 is the last official release for macOS, as support will not be available for macOS in newer releases.

Define Your Tensorflow Version

  • Check all possible TensorFlow and Cuda versions here: https://www.tensorflow.org/install/source#gpu
  • For me - tenorflow-2.3.0, cuda 10.2, nvidia-440.33.0, cuDNN 7.6, Bazel 3.1.0, GCC 7.3.1, TensorRT 6.0
    • PS cuda 10.1 won't install due 418 driver is not comportable with new kernel 5.4.0-70-generic

Installation

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-ubuntu1804.pin
sudo mv cuda-ubuntu1804.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda-repo-ubuntu1804-10-2-local-10.2.89-440.33.01_1.0-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu1804-10-2-local-10.2.89-440.33.01_1.0-1_amd64.deb
sudo apt-key add /var/cuda-repo-10-2-local-10.2.89-440.33.01/7fa2af80.pub
sudo apt-get update
sudo apt-get -y install cuda

sudo dpkg -i libcudnn7_7.6.5.32-1+cuda10.2_amd64.deb
sudo dpkg -i libcudnn7-dev_7.6.5.32-1+cuda10.2_amd64.deb
sudo dpkg -i libcudnn7-doc_7.6.5.32-1+cuda10.2_amd64.deb

sudo dpkg -i nv-tensorrt-repo-ubuntu1804-cuda10.2-trt6.0.1.8-ga-20191108_1-1_amd64.deb
sudo apt-get update
sudo apt-key add /var/nv-tensorrt-repo-cuda10.2-trt6.0.1.8-ga-20191108/7fa2af80.pub
sudo apt-get install tensorrt

sudo apt install apt-transport-https curl gnupg
curl -fsSL https://bazel.build/bazel-release.pub.gpg | gpg --dearmor > bazel.gpg
sudo mv bazel.gpg /etc/apt/trusted.gpg.d/
echo "deb [arch=amd64] https://storage.googleapis.com/bazel-apt stable jdk1.8" | sudo tee /etc/apt/sources.list.d/bazel.list
sudo apt update
sudo apt install bazel-3.2.0
sudo ln -s /usr/bin/bazel-3.2.0 /usr/bin/bazel
bazel --version  # 3.2.0

sudo apt install -y ccache
sudo /usr/sbin/update-ccache-symlinks
echo 'export PATH="/usr/lib/ccache:$PATH"' | tee -a ~/.bash_aliases
source ~/.bash_aliases

git clone https://github.com/tensorflow/tensorflow.git
cd tensorflow
git checkout r2.3
./configure
bazel build --config=cuda --config=opt //tensorflow/tools/pip_package:build_pip_package

Testing

>>> import tensorflow as tf
>>> tf.__version__
'2.3.0'
>>> tf.test.is_built_with_cuda()
True

$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:07:16_PDT_2019
Cuda compilation tools, release 10.1, V10.1.243

Python 3.8.5 (default, Jan 27 2021, 15:41:15) 
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>> tf.config.list_physical_devices("GPU")
2021-03-29 00:19:23.520023: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1
2021-03-29 00:19:23.575281: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-03-29 00:19:23.575801: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0 with properties: 
pciBusID: 0000:04:00.0 name: GeForce GTX 760 computeCapability: 3.0
coreClock: 1.15GHz coreCount: 6 deviceMemorySize: 3.94GiB deviceMemoryBandwidth: 179.05GiB/s
2021-03-29 00:19:23.576789: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2021-03-29 00:19:23.581937: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10
2021-03-29 00:19:23.583502: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10
2021-03-29 00:19:23.585776: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10
2021-03-29 00:19:23.591336: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10
2021-03-29 00:19:23.593271: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10
2021-03-29 00:19:23.701034: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2021-03-29 00:19:23.701468: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-03-29 00:19:23.702637: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-03-29 00:19:23.703486: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1657] Ignoring visible gpu device (device: 0, name: GeForce GTX 760, pci bus id: 0000:04:00.0, compute capability: 3.0) with Cuda compute capability 3.0. The minimum required Cuda capability is 3.5.
[]

git clone https://github.com/tensorflow/tensorflow.git
cd ./tensorflow
git checkout r2.2

sudo apt install apt-transport-https curl gnupg
curl -fsSL https://bazel.build/bazel-release.pub.gpg | gpg --dearmor > bazel.gpg
sudo mv bazel.gpg /etc/apt/trusted.gpg.d/
echo "deb [arch=amd64] https://storage.googleapis.com/bazel-apt stable jdk1.8" | sudo tee /etc/apt/sources.list.d/bazel.list
sudo apt update && sudo apt install bazel-2.0.0

Курсы

Большие данные

Методы

  • Теорема Байеса
  • Функции ошибки и регуляризация
  • Расстояние Кульбака-Лейблера и перекрестная энтропия
  • Градиентный спуск: основы
  • Граф вычислений и дифференцирование на нем
  • Перцептрон
  • Глубокие нейронные сети
  • Классификация
  • Кластеризация
  • Регрессия
  • Машинное зрение
  • Метод к-средних
  • word2vec

Библиотеки

Датасеты

Железо и драйверы

Темы

Face Recognition

Speech Recognition

Image Object Recognition

Anomaly Detection

Prediction

StereoVision

Некоторые полезные ресурсы

Источник — http://wiki.aleksashkin.net/index.php?title=Машинное_обучение_и_Большие_данные&oldid=1547