pytorch_workflow.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PyTorch Image Classification Example\n",
    "\n",
    "Now we'll be looking at how covalent can be used to run actual workflows on different hardwares without making a single line change in the core code.\n",
    "\n",
    "This example is inspired from PyTorch's CIFAR10 Image Classfication tutorial which can be found [here](https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html). Some small changes have been made, things like organizing certain parts into functions and paramterizing them, etc, to covalentify the tutorial."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import covalent as ct"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Executors\n",
    "\n",
    "Creating the executor objects we'll be using in this example:\n",
    "\n",
    "1. DaskExecutor - Only requires the scheduler address of an already running Dask Cluster and every electron which has been assigned this executor will be run on that cluster.\n",
    "\n",
    "2. SlurmExecutor - Requires minimal details so that it connects to the running SLURM cluster through `ssh` and executes the electron on the node given within the `options` parameter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "dask_executor = ct.executor.DaskExecutor(scheduler_address='tcp://127.0.0.1:55201')\n",
    "\n",
    "slurm_executor = ct.executor.SlurmExecutor(\n",
    "    username=\"sankalp\",\n",
    "    address=\"beehive.agnostiq.ai\",\n",
    "    remote_workdir=\"/federation/sankalp/workdir\",\n",
    "    ssh_key_file=\"/Users/sankalpsanand/.ssh/beehive\",\n",
    "    poll_freq=10,\n",
    "    options={\n",
    "        \"partition\": \"debug\", \"ntasks\": 1, \"cpus-per-task\": 1,\n",
    "        \"chdir\": \"/federation/sankalp/workdir\", \"nodelist\": \"beehive-debug-st-t2medium-1\"},\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Covalentify\n",
    "\n",
    "The following transformation, data downloading, and dataloader objects have been enclosed in a parameterized function so they are more accessible throughout this example and can be used as we see fit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You'll notice we've added the `electron` decorator to these functions and unless they are called within a `lattice` and that `lattice` is dispatched, these will continue to act as normal functions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "@ct.electron\n",
    "def get_transform(mean_tuple, std_tuple):\n",
    "    \"\"\"Return a normalized Tensor transform\"\"\"\n",
    "\n",
    "    return transforms.Compose(\n",
    "    [transforms.ToTensor(),\n",
    "     transforms.Normalize(mean=mean_tuple, std=std_tuple)])\n",
    "\n",
    "\n",
    "@ct.electron\n",
    "def get_dataset(train, transform, root):\n",
    "    \"\"\"Return a transformed CIFAR10 dataset\"\"\"\n",
    "\n",
    "    return torchvision.datasets.CIFAR10(root=root, train=train,\n",
    "                                        download=True, transform=transform)\n",
    "\n",
    "@ct.electron\n",
    "def get_dataloader(dataset, shuffle, batch_size):\n",
    "    \"\"\"Return the dataloader for given dataset\"\"\"\n",
    "\n",
    "    return torch.utils.data.DataLoader(dataset, batch_size=batch_size,\n",
    "                                          shuffle=shuffle)\n",
    "\n",
    "\n",
    "@ct.electron\n",
    "def get_test_accuracy(net, testloader):\n",
    "    \"\"\"Return accuracy of the trained network using the test dataset\"\"\"\n",
    "\n",
    "    correct = 0\n",
    "    total = 0\n",
    "\n",
    "    # since we're not training, we don't need to calculate the gradients for our outputs\n",
    "    with torch.no_grad():\n",
    "        for data in testloader:\n",
    "            images, labels = data\n",
    "            # calculate outputs by running images through the network\n",
    "            outputs = net(images)\n",
    "            # the class with the highest energy is what we choose as prediction\n",
    "            _, predicted = torch.max(outputs.data, 1)\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "\n",
    "    return 100 * correct // total\n",
    "\n",
    "# Defining the classes present in the dataset\n",
    "classes = ('plane', 'car', 'bird', 'cat',\n",
    "            'deer', 'dog', 'frog', 'horse', 'ship', 'truck')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see how the dataset looks like:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bird  cat   deer  dog  \n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "mean_tuple = (0.5, 0.5, 0.5)\n",
    "std_tuple = (0.5, 0.5, 0.5)\n",
    "\n",
    "# functions to show an image\n",
    "\n",
    "def imshow(img):\n",
    "    img = img / 2 + 0.5     # unnormalize\n",
    "    npimg = img.numpy()\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "# get some random training images\n",
    "dataiter = iter(get_dataloader(get_dataset(True, get_transform(mean_tuple, std_tuple), \"./data\"), True, 4))\n",
    "images, labels = dataiter.next()\n",
    "\n",
    "# show images\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "# print labels\n",
    "print(' '.join(f'{classes[labels[j]]:5s}' for j in range(4)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again, here we've defined our network and put it in a function so that we can make any changes depending on parameters that might only be known during the time of execution. Once the new network has been created, we return the `Net` object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "@ct.electron\n",
    "def get_network(layers=1):\n",
    "    \"\"\"\n",
    "    Return a network object with given specifications\n",
    "    \"\"\"\n",
    "\n",
    "    class Net(nn.Module):\n",
    "        def __init__(self):\n",
    "            super().__init__()\n",
    "            self.conv1 = nn.Conv2d(3, 6, 5)\n",
    "            self.pool = nn.MaxPool2d(2, 2)\n",
    "            self.conv2 = nn.Conv2d(6, 16, 5)\n",
    "            self.fc1 = nn.Linear(16 * 5 * 5, 120)\n",
    "            self.fc2 = nn.Linear(120, 84)\n",
    "            self.fc3 = nn.Linear(84, 10)\n",
    "        \n",
    "\n",
    "        def make_nn(self):\n",
    "            return nn.Linear(84, 84)\n",
    "\n",
    "        def forward(self, x):\n",
    "            x = self.pool(F.relu(self.conv1(x)))\n",
    "            x = self.pool(F.relu(self.conv2(x)))\n",
    "            x = torch.flatten(x, 1) # flatten all dimensions except batch\n",
    "            x = F.relu(self.fc1(x))\n",
    "            x = F.relu(self.fc2(x))\n",
    "            for _ in range(layers):\n",
    "                x=F.relu(self.make_nn()(x))\n",
    "            x = self.fc3(x)\n",
    "            return x\n",
    "    \n",
    "    return Net()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training\n",
    "\n",
    "Now that we've defined everything, it's time we put them together and begin with the training process.\n",
    "\n",
    "You'll see that for the `training` function we are using the slurm executor so that we can leverage the already running high compute GPU SLURM cluster to train the classification model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.optim as optim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "@ct.electron(executor=slurm_executor)\n",
    "def training(trainloader, net, learning_rate, momentum, n_epochs):\n",
    "    \"\"\"\n",
    "    Return a trained network with cross entropy loss and SGD optimizer with given parameters\n",
    "    \"\"\"\n",
    "\n",
    "    criterion = nn.CrossEntropyLoss()\n",
    "    optimizer = optim.SGD(net.parameters(), lr=learning_rate, momentum=momentum)\n",
    "    \n",
    "    for epoch in range(n_epochs):  # loop over the dataset multiple times\n",
    "\n",
    "        running_loss = 0.0\n",
    "        for i, data in enumerate(trainloader, 0):\n",
    "            # get the inputs; data is a list of [inputs, labels]\n",
    "            inputs, labels = data\n",
    "\n",
    "            # zero the parameter gradients\n",
    "            optimizer.zero_grad()\n",
    "\n",
    "            # forward + backward + optimize\n",
    "            outputs = net(inputs)\n",
    "            loss = criterion(outputs, labels)\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "\n",
    "            # print statistics\n",
    "            running_loss += loss.item()\n",
    "            if i % 2000 == 1999:    # print every 2000 mini-batches\n",
    "                print(f'[{epoch + 1}, {i + 1:5d}] loss: {running_loss / 2000:.3f}')\n",
    "                running_loss = 0.0\n",
    "    \n",
    "    return net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "\n",
    "data_path = str(Path(\"./data\").absolute())\n",
    "\n",
    "@ct.lattice(executor=dask_executor)\n",
    "def workflow(batch_size, learning_rate, momentum, n_epochs,layers):\n",
    "    \"\"\"\n",
    "    Stitching everything together in a workflow lattice\n",
    "    \"\"\"\n",
    "    \n",
    "    transform = get_transform(mean_tuple=(0.5, 0.5, 0.5), std_tuple=(0.5, 0.5, 0.5))\n",
    "\n",
    "    trainset = get_dataset(train=True, transform=transform, root=data_path)\n",
    "    trainloader = get_dataloader(dataset=trainset, shuffle=True, batch_size=batch_size)\n",
    "\n",
    "    testset = get_dataset(train=False, transform=transform, root=data_path)\n",
    "    testloader = get_dataloader(dataset=testset, shuffle=False, batch_size=batch_size)\n",
    "\n",
    "    test_accuracies = []\n",
    "    for layer in layers:\n",
    "        net = get_network(layers=layer)\n",
    "\n",
    "        training_result = training(\n",
    "            trainloader=trainloader,\n",
    "            net=net,\n",
    "            learning_rate=learning_rate,\n",
    "            momentum=momentum,\n",
    "            n_epochs=n_epochs,\n",
    "            )\n",
    "        \n",
    "        accuracy = get_test_accuracy(net=training_result, testloader=testloader)\n",
    "        \n",
    "        test_accuracies.append(accuracy)\n",
    "\n",
    "    return test_accuracies\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's dispatch the workflow."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "17bf6e22-51eb-4365-8561-aca8b3615333\n"
     ]
    }
   ],
   "source": [
    "dispatch_id = ct.dispatch(workflow)(batch_size=4, learning_rate=0.001, momentum=0.9, n_epochs=2, layers=range(5))\n",
    "print(dispatch_id)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting the results\n",
    "\n",
    "We can get the complete `Result` object which contains all the information about the workflow."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Lattice Result\n",
      "==============\n",
      "status: COMPLETED\n",
      "result: [53, 11, 10, 10, 9]\n",
      "inputs: {'args': [], 'kwargs': {'batch_size': 4, 'learning_rate': 0.001, 'momentum': 0.9, 'n_epochs': 2, 'layers': range(0, 5)}}\n",
      "error: None\n",
      "\n",
      "start_time: 2022-06-08 16:34:42.561549+00:00\n",
      "end_time: 2022-06-08 16:41:16.092947+00:00\n",
      "\n",
      "results_dir: /Users/sankalpsanand/dev/covalent/covalent/results\n",
      "dispatch_id: 35d41a8b-4ca8-40d4-9e35-7973347b0235\n",
      "\n",
      "Node Outputs\n",
      "------------\n",
      "get_transform(0): Compose(\n",
      "    ToTensor()\n",
      "    Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5))\n",
      ")\n",
      ":parameter:(0.5, 0.5, 0.5)(1): (0.5, 0.5, 0.5)\n",
      ":parameter:(0.5, 0.5, 0.5)(2): (0.5, 0.5, 0.5)\n",
      "get_dataset(3): Dataset CIFAR10\n",
      "    Number of datapoints: 50000\n",
      "    Root location: /Users/sankalpsanand/dev/covalent/covalent/data\n",
      "    Split: Train\n",
      "    StandardTransform\n",
      "Transform: Compose(\n",
      "               ToTensor()\n",
      "               Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5))\n",
      "           )\n",
      ":parameter:True(4): True\n",
      ":parameter:/Users/sankalpsanand/dev/covalent/covalent/data(5): /Users/sankalpsanand/dev/covalent/covalent/data\n",
      "get_dataloader(6): <torch.utils.data.dataloader.DataLoader object at 0x7f8f4905dc40>\n",
      ":parameter:True(7): True\n",
      ":parameter:4(8): 4\n",
      "get_dataset(9): Dataset CIFAR10\n",
      "    Number of datapoints: 10000\n",
      "    Root location: /Users/sankalpsanand/dev/covalent/covalent/data\n",
      "    Split: Test\n",
      "    StandardTransform\n",
      "Transform: Compose(\n",
      "               ToTensor()\n",
      "               Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5))\n",
      "           )\n",
      ":parameter:False(10): False\n",
      ":parameter:/Users/sankalpsanand/dev/covalent/covalent/data(11): /Users/sankalpsanand/dev/covalent/covalent/data\n",
      "get_dataloader(12): <torch.utils.data.dataloader.DataLoader object at 0x7f8f197e6820>\n",
      ":parameter:False(13): False\n",
      ":parameter:4(14): 4\n",
      "get_network(15): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0(16): 0\n",
      "training(17): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0.001(18): 0.001\n",
      ":parameter:0.9(19): 0.9\n",
      ":parameter:2(20): 2\n",
      "get_test_accuracy(21): 53\n",
      "get_network(22): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:1(23): 1\n",
      "training(24): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0.001(25): 0.001\n",
      ":parameter:0.9(26): 0.9\n",
      ":parameter:2(27): 2\n",
      "get_test_accuracy(28): 11\n",
      "get_network(29): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:2(30): 2\n",
      "training(31): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0.001(32): 0.001\n",
      ":parameter:0.9(33): 0.9\n",
      ":parameter:2(34): 2\n",
      "get_test_accuracy(35): 10\n",
      "get_network(36): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:3(37): 3\n",
      "training(38): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0.001(39): 0.001\n",
      ":parameter:0.9(40): 0.9\n",
      ":parameter:2(41): 2\n",
      "get_test_accuracy(42): 10\n",
      "get_network(43): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:4(44): 4\n",
      "training(45): Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n",
      ":parameter:0.001(46): 0.001\n",
      ":parameter:0.9(47): 0.9\n",
      ":parameter:2(48): 2\n",
      "get_test_accuracy(49): 9\n",
      "\n"
     ]
    }
   ],
   "source": [
    "workflow_result = ct.get_result(dispatch_id)\n",
    "print(workflow_result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's see what the accuracies were in each case of adding layers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy for network with 0 added layers: 53%\n",
      "Test Accuracy for network with 1 added layers: 11%\n",
      "Test Accuracy for network with 2 added layers: 10%\n",
      "Test Accuracy for network with 3 added layers: 10%\n",
      "Test Accuracy for network with 4 added layers: 9%\n"
     ]
    }
   ],
   "source": [
    "for i in range(5):\n",
    "    print(f\"Test Accuracy for network with {i} added layers: {workflow_result.result[i]}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see what the gound truth looks like for some of the images in the test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GroundTruth:  cat   ship  ship  plane\n"
     ]
    }
   ],
   "source": [
    "testset = get_dataset(False, get_transform((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)), data_path)\n",
    "testloader = get_dataloader(testset, False, 4)\n",
    "\n",
    "dataiter = iter(testloader)\n",
    "images, labels = dataiter.next()\n",
    "\n",
    "# print images\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "print('GroundTruth: ', ' '.join(f'{classes[labels[j]]:5s}' for j in range(4)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Accessing output of a particular trained network\n",
    "\n",
    "Let's say we want the network which gave us the best accuracy, we can access it (and any other network as well) easily through the `workflow_result` object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Net(\n",
      "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
      "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
      "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
      "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "net = workflow_result._get_node_output(node_id=17)\n",
    "print(net)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "outputs = net(images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted:  frog  ship  truck plane\n"
     ]
    }
   ],
   "source": [
    "_, predicted = torch.max(outputs, 1)\n",
    "\n",
    "print('Predicted: ', ' '.join(f'{classes[predicted[j]]:5s}'\n",
    "                              for j in range(4)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see, it was about 50% right. Now let's see how did the network perform on a per-class basis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy for class: plane is 72.0 %\n",
      "Accuracy for class: car   is 55.7 %\n",
      "Accuracy for class: bird  is 53.4 %\n",
      "Accuracy for class: cat   is 36.6 %\n",
      "Accuracy for class: deer  is 44.8 %\n",
      "Accuracy for class: dog   is 46.0 %\n",
      "Accuracy for class: frog  is 49.0 %\n",
      "Accuracy for class: horse is 58.3 %\n",
      "Accuracy for class: ship  is 61.8 %\n",
      "Accuracy for class: truck is 58.8 %\n"
     ]
    }
   ],
   "source": [
    "# prepare to count predictions for each class\n",
    "correct_pred = {classname: 0 for classname in classes}\n",
    "total_pred = {classname: 0 for classname in classes}\n",
    "\n",
    "# again no gradients needed\n",
    "with torch.no_grad():\n",
    "    for data in testloader:\n",
    "        images, labels = data\n",
    "        outputs = net(images)\n",
    "        _, predictions = torch.max(outputs, 1)\n",
    "        # collect the correct predictions for each class\n",
    "        for label, prediction in zip(labels, predictions):\n",
    "            if label == prediction:\n",
    "                correct_pred[classes[label]] += 1\n",
    "            total_pred[classes[label]] += 1\n",
    "\n",
    "\n",
    "# print accuracy for each class\n",
    "for classname, correct_count in correct_pred.items():\n",
    "    accuracy = 100 * float(correct_count) / total_pred[classname]\n",
    "    print(f'Accuracy for class: {classname:5s} is {accuracy:.1f} %')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'8255f267-4d32-493b-ba7e-f72a4a4a943c'"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import covalent as ct\n",
    "\n",
    "@ct.lattice\n",
    "@ct.electron\n",
    "def task():\n",
    "    return 5 / 0\n",
    "\n",
    "ct.dispatch(task)()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "interpreter": {
   "hash": "197454157c080c19c876ac9bd28ff3f683250ebc7e59376e497fe430017b1105"
  },
  "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": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}