mirror of https://github.com/Jittor/Jittor
139 lines
4.9 KiB
Plaintext
139 lines
4.9 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"# 计图入门教程 0 --- 介绍与安装"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"\n",
|
||
"**计图 (Jittor)** 是一个以 Python 为前端语言的深度学习框架，它 \n",
|
||
"* 效率高：可作为 NumPy，PyTorch 的替代品，可以使用 GPU 等其他加速器进行高效的数据运算。除此之外，计图还拥有多个创新点，旨在大幅提升其运算效率；\n",
|
||
"* 易使用：是一个用于实现神经网络的自动微分库，并集成了大量有关深度学习的函数库，方便您快速开展开发任务。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"\n",
|
||
"**通过本教程，您将**\n",
|
||
"* 学习并理解计图中基本类型的一般操作；\n",
|
||
"* 了解神经网络的一些基本概念，并学会如何利用计图进行神经网络的训练；\n",
|
||
"* 解决一个机器学习的经典实战问题。\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"**本教程的适用群体：** \n",
|
||
"我们的目标是，只要您会 Python 编程，即可通过本教程学习并掌握如何使用计图进行深度学习的开发。不用担心，本教程几乎对所有的关键代码都加以注释说明。只要您耐心跟着本教程一步步学习，便一定能有所斩获。"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"现在，请您开启计图快速入门之旅。"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"metadata": {},
|
||
"source": [
|
||
"\n",
|
||
"## 安装\n",
|
||
"\n",
|
||
"\n",
|
||
"Jittor框架对环境要求如下:\n",
|
||
"\n",
|
||
"\n",
|
||
"* 操作系统: **Ubuntu** >= 16.04 或 **Windows Subsystem of Linux（WSL）**\n",
|
||
"* Python：版本 >= 3.7\n",
|
||
"* C++编译器 （需要下列至少一个）\n",
|
||
" - g++ （>=5.4.0）\n",
|
||
" - clang （>=8.0）\n",
|
||
"* GPU 编译器（可选）：nvcc >=10.0\n",
|
||
"* GPU 加速库（可选）：cudnn-dev (cudnn开发版, 推荐使用tar安装方法，[参考链接](https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html#installlinux-tar))\n",
|
||
"\n",
|
||
"如果您不希望手动配置环境，我们推荐使用 Docker 进行安装。\n",
|
||
"除此之外，您还可以使用 pip 安装和手动安装。\n",
|
||
"\n",
|
||
"注意：目前Jittor通过WSL的方式在Windows操作系统上运行，WSL的安装方法请参考[微软官网](https://docs.microsoft.com/en-us/windows/wsl/install-win10)，WSL版本目前尚不支持CUDA。\n",
|
||
"\n",
|
||
"Jittor 提供了三种安装方法：docker，pip和手动安装：\n",
|
||
"\n",
|
||
"\n",
|
||
"\n",
|
||
"## Docker 安装\n",
|
||
"\n",
|
||
"我们提供了Docker安装方式，免去您配置环境，Docker安装方法如下：\n",
|
||
"\n",
|
||
"\n",
|
||
"```\n",
|
||
"# CPU only(Linux)\n",
|
||
"docker run -it --network host jittor/jittor\n",
|
||
"# CPU and CUDA(Linux)\n",
|
||
"docker run -it --network host --gpus all jittor/jittor-cuda\n",
|
||
"# CPU only(Mac and Windows)\n",
|
||
"docker run -it -p 8888:8888 jittor/jittor\n",
|
||
"# Upgrade jittor docker image\n",
|
||
"docker pull jittor/jittor\n",
|
||
"docker pull jittor/jittor-cuda\n",
|
||
"```\n",
|
||
"\n",
|
||
"关于Docker安装的详细教程，可以参考[Windows/Mac/Linux通过Docker安装计图](https://cg.cs.tsinghua.edu.cn/jittor/tutorial/2020-5-15-00-00-docker/)\n",
|
||
"\n",
|
||
"## Pip 安装\n",
|
||
"\n",
|
||
"\n",
|
||
"如果您没有准备好环境，或者使用的不是Ubuntu操作系统， 推荐使用**docker安装**， 如果您已经装好编译器和对应版本的Python,我们强烈推荐您使用这种方法\n",
|
||
"(如果无法访问github, 可以通过jittor主页下载):\n",
|
||
"\n",
|
||
"```bash\n",
|
||
"sudo apt install python3.7-dev libomp-dev\n",
|
||
"python3.7 -m pip install jittor\n",
|
||
"# or install from github(latest version)\n",
|
||
"# python3.7 -m pip install git+https://github.com/Jittor/jittor.git\n",
|
||
"python3.7 -m jittor.test.test_example\n",
|
||
"\n",
|
||
"# Upgrade jittor from pip\n",
|
||
"python3.7 -m pip install jittor -U\n",
|
||
"# Upgrade jittor from github\n",
|
||
"python3.7 -m pip install git+https://github.com/Jittor/jittor.git -U\n",
|
||
"```\n",
|
||
"\n",
|
||
"如果测试运行通过,恭喜你已经安装完成.\n",
|
||
"jittor会自动在路径中寻找合适的编译器, 如果您希望手动指定编译器, 请使用环境变量 `cc_path` 和 `nvcc_path`(可选).\n"
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 3
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython3",
|
||
"version": "3.7.5"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 4
|
||
}
|