PyTorch快速安裝-基於JupyterHub並運行K8s

PyTorch快速安裝-基於JupyterHub並運行K8s

運行PyTorch能夠直接邏輯運行、容器中運行、KubeFlow中運行以及基於JupyterHub（獨立運行或運行在K8s之上）等多種模式。這裏介紹運行在K8s上基於JupyterHub的PyTorch方法，這也是運行在雲計算環境的推薦方法。若是須要使用GPU，則須要安裝NVidia或AMD的Kubernetes下容器GPU支持，宿主機也必須同時安裝GPU驅動。

• pytorch官網資源
  • https://pytorch.org
  • https://pytorch.org/tutorials
  • https://github.com/pytorch/pytorch
• 深度學習引擎-PyTorch資源集錦
  • http://www.javashuo.com/article/p-hoyneiey-cd.html

安裝 kubernetes 的支持

• PyTorch支持Kubernetes集羣, https://my.oschina.net/u/2306127/blog/1817835
• Kubeflow 使用指南，http://www.javashuo.com/article/p-nppedcay-hp.html
• pytorch-operator 項目源碼，https://github.com/kubeflow/pytorch-operator
• pytorch-operator on Kubernetes，http://www.javashuo.com/article/p-afvqxpqf-dm.html
• 鏡像構建-iObjects Python with JupyterHub for K8s,http://www.javashuo.com/article/p-uenogwzq-cc.html
• GPU 支持的安裝
  • Kubernetes集羣升級NVidia GPU驅動版本
  • Kubernetes中調度GPU資源
  • Ubuntu快速玩起色器學習
  • Ubuntu18.04上安裝RTX 2080Ti顯卡驅動

安裝 JupyterHub/JupyterLab支持

conda install numpy
conda install scikit-image

conda install pytorch torchvision cudatoolkit=10.0 -c pytorch
# or: conda install pytorch-cpu torchvision-cpu -c pytorch

conda update --all

開始使用

獲取教程數據：

• https://download.pytorch.org/tutorial/faces.zip

使用Notebook：

# 導入支持庫
import torch 

# 確認CUDA支持及其版本
print(torch.version.cuda)

10.0.130

# 查看pytorch幫助
help(torch)

Help on package torch:

NAME
    torch

DESCRIPTION
    The torch package contains data structures for multi-dimensional
    tensors and mathematical operations over these are defined.
    Additionally, it provides many utilities for efficient serializing of
    Tensors and arbitrary types, and other useful utilities.
    
    It has a CUDA counterpart, that enables you to run your tensor computations
    on an NVIDIA GPU with compute capability >= 3.0.

PACKAGE CONTENTS
    _C
    _dl
    _jit_internal
    _nvrtc
    _ops
    _six
    _storage_docs
    _tensor_docs
    _tensor_str
    _thnn (package)
    _torch_docs
    _utils
    _utils_internal
    autograd (package)
    backends (package)
    contrib (package)
    cuda (package)
    distributed (package)
    distributions (package)
    for_onnx (package)
    functional
    hub
    jit (package)
    multiprocessing (package)
    nn (package)
    onnx (package)
    optim (package)
    random
    serialization
    sparse (package)
    storage
    tensor
    testing (package)
    utils (package)
    version
	......