2019 edition

goncalomcorreia · Oct 11, 2019 · 329f736 · 329f736
1 parent e384d07
commit 329f736
Show file tree

Hide file tree

Showing 8 changed files with 1,911 additions and 1,305 deletions.
diff --git a/00-intro.ipynb b/00-intro.ipynb
@@ -0,0 +1,169 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Introduction\n",
+    "\n",
+    "Material for this tutorial is here: !!LINK\n",
+    "\n",
+    "**Note:**\n",
+    "If you use PyTorch on a daily basis, you will most probably not learn a lot during this tutorial.\n",
+    "\n",
+    "**Goals:**\n",
+    "- understand PyTorch concepts (e.g. autograd, broadcasting, ...) and understand what it can and cannot do\n",
+    "- be aware of some handy tools/libraries\n",
+    "- be able to create simple neural networks\n",
+    "- learn the some tools that will help to code more complicated models in the future"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "![](img/the_real_reason.png)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# PyTorch Overview\n",
+    "\n",
+    "\n",
+    "> \"PyTorch - Tensors and Dynamic neural networks in Python\n",
+    "with strong GPU acceleration.\n",
+    "PyTorch is a deep learning framework for fast, flexible experimentation.\"\n",
+    ">\n",
+    "> -- https://pytorch.org/*\n",
+    "\n",
+    "This was the tagline prior to PyTorch 1.0.\n",
+    "Now it's:\n",
+    "\n",
+    "> \"PyTorch - From Research To Production\n",
+    "> \n",
+    "> An open source deep learning platform that provides a seamless path from research prototyping to production deployment.\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## \"Build by run\" - what is that and why do I care?"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "![](img/dynamic_graph.gif)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "A very practical reason to use PyTorch:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "from IPython.core.debugger import set_trace\n",
+    "\n",
+    "def f(x):\n",
+    "    res = x + x\n",
+    "    # set_trace()  # <-- :o\n",
+    "    return res\n",
+    "\n",
+    "x = torch.randn(1, 10)\n",
+    "f(x)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## TensorFlow and PyTorch\n",
+    "- static vs dynamic\n",
+    "- production vs prototyping "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Other neural network toolkits you might want to check out\n",
+    "- TensorFlow\n",
+    "- MXNet\n",
+    "- Keras\n",
+    "- CNTK\n",
+    "- Chainer\n",
+    "- caffe\n",
+    "- caffe2\n",
+    "- dynet\n",
+    "- many many more\n",
+    "\n",
+    "All of them are good!\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Useful Links\n",
+    "\n",
+    "- Twitter: https://twitter.com/PyTorch\n",
+    "- Forum: https://discuss.pytorch.org/\n",
+    "- Tutorials: https://pytorch.org/tutorials/\n",
+    "- Examples: https://github.com/pytorch/examples\n",
+    "- API Reference: https://pytorch.org/docs/stable/index.html\n",
+    "- Torchvision: https://pytorch.org/docs/stable/torchvision/index.html\n",
+    "- PyTorch Text: https://github.com/pytorch/text\n",
+    "- PyTorch Audio: https://github.com/pytorch/audio\n",
+    "\n",
+    "## Tutorials I based this on...\n",
+    "\n",
+    "- https://github.com/sotte/pytorch_tutorial\n",
+    "- https://github.com/erickrf/pytorch-lecture"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Some Jupyter Notebook shortcuts that might be useful:\n",
+    "\n",
+    "- `A` and `B` create a cell above and below, respectively\n",
+    "- `Shift+Enter` runs the cell and jumps to the next one/creates one below\n",
+    "- `D`, `D` deletes the cell"
+   ]
+  }
+ ],
+ "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": "3.7.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}