notebooks/neural-magic-torch-mnist/training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.022136,
     "end_time": "2021-02-03T13:27:33.986566",
     "exception": false,
     "start_time": "2021-02-03T13:27:33.964430",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Train model with Pytorch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.018835,
     "end_time": "2021-02-03T13:27:34.025541",
     "exception": false,
     "start_time": "2021-02-03T13:27:34.006706",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "papermill": {
     "duration": 2.822829,
     "end_time": "2021-02-03T13:27:36.868108",
     "exception": false,
     "start_time": "2021-02-03T13:27:34.045279",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import boto3\n",
    "import os\n",
    "from datetime import datetime\n",
    "\n",
    "import random\n",
    "from pathlib import Path\n",
    "\n",
    "import torch\n",
    "\n",
    "from torchvision import datasets\n",
    "from torchvision.transforms import ToTensor\n",
    "from torch.utils.data import DataLoader\n",
    "import torch.nn as nn\n",
    "from torch import optim\n",
    "from torch.autograd import Variable\n",
    "from torchinfo import summary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preparing data for training with DataLoaders"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Dataset retrieves our dataset’s features and labels one sample at a time. While training a model, we typically want to pass samples in “minibatches”, reshuffle the data at every epoch to reduce model overfitting, and use Python’s multiprocessing to speed up data retrieval.\n",
    "DataLoader is an iterable that abstracts this complexity for us in an easy API."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prepare MNIST data.\n",
    "train_data = datasets.MNIST(\n",
    "    root=\"../../data/raw/pytorch-mnist-dataset\",\n",
    "    train=True,\n",
    "    transform=ToTensor(),\n",
    ")\n",
    "test_data = datasets.MNIST(\n",
    "    root=\"../../data/raw/pytorch-mnist-dataset\",\n",
    "    train=False,\n",
    "    transform=ToTensor(),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'train': <torch.utils.data.dataloader.DataLoader at 0x7f9937d147c0>,\n",
       " 'test': <torch.utils.data.dataloader.DataLoader at 0x7f9937d14df0>}"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loaders = {\n",
    "    \"train\": torch.utils.data.DataLoader(\n",
    "        train_data, batch_size=128, shuffle=True, num_workers=1\n",
    "    ),\n",
    "    \"test\": torch.utils.data.DataLoader(\n",
    "        test_data, batch_size=128, shuffle=True, num_workers=1\n",
    "    ),\n",
    "}\n",
    "loaders"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define the Convolutional Neural Network model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.021457,
     "end_time": "2021-02-03T13:27:37.078460",
     "exception": false,
     "start_time": "2021-02-03T13:27:37.057003",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Set parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "papermill": {
     "duration": 0.031268,
     "end_time": "2021-02-03T13:27:37.130774",
     "exception": false,
     "start_time": "2021-02-03T13:27:37.099506",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# MNIST dataset parameters.\n",
    "num_classes = 10  # total classes (0-9 digits).\n",
    "batch_size = 128"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.022834,
     "end_time": "2021-02-03T13:27:37.477155",
     "exception": false,
     "start_time": "2021-02-03T13:27:37.454321",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Create CNN Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "papermill": {
     "duration": 0.167411,
     "end_time": "2021-02-03T13:27:37.667600",
     "exception": false,
     "start_time": "2021-02-03T13:27:37.500189",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class CNN(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(CNN, self).__init__()\n",
    "        self.conv1 = nn.Sequential(\n",
    "            nn.Conv2d(\n",
    "                in_channels=1,\n",
    "                out_channels=32,\n",
    "                kernel_size=5,\n",
    "                stride=1,\n",
    "                padding=2,\n",
    "            ),\n",
    "            nn.ReLU(),\n",
    "            nn.MaxPool2d(kernel_size=2),\n",
    "        )\n",
    "        self.conv2 = nn.Sequential(\n",
    "            nn.Conv2d(\n",
    "                in_channels=32, out_channels=64, kernel_size=5, stride=1, padding=2\n",
    "            ),\n",
    "            nn.ReLU(),\n",
    "            nn.MaxPool2d(kernel_size=2),\n",
    "        )\n",
    "        #         self.drop = nn.Dropout(0.5)\n",
    "\n",
    "        # fully connected layer, output 10 classes\n",
    "        self.out = nn.Linear(64 * 7 * 7, num_classes)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = self.conv2(x)\n",
    "        # flatten the output of conv2 to (batch_size, 64 * 7 * 7)\n",
    "        x = x.view(x.size(0), -1)\n",
    "        #         x = self.drop(x)\n",
    "        output = self.out(x)\n",
    "        return output, x  # return x for visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CNN(\n",
      "  (conv1): Sequential(\n",
      "    (0): Conv2d(1, 32, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
      "    (1): ReLU()\n",
      "    (2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (conv2): Sequential(\n",
      "    (0): Conv2d(32, 64, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
      "    (1): ReLU()\n",
      "    (2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (out): Linear(in_features=3136, out_features=10, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "cnn = CNN()\n",
    "print(cnn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "==========================================================================================\n",
       "Layer (type:depth-idx)                   Output Shape              Param #\n",
       "==========================================================================================\n",
       "CNN                                      --                        --\n",
       "├─Sequential: 1-1                        [128, 32, 14, 14]         --\n",
       "│    └─Conv2d: 2-1                       [128, 32, 28, 28]         832\n",
       "│    └─ReLU: 2-2                         [128, 32, 28, 28]         --\n",
       "│    └─MaxPool2d: 2-3                    [128, 32, 14, 14]         --\n",
       "├─Sequential: 1-2                        [128, 64, 7, 7]           --\n",
       "│    └─Conv2d: 2-4                       [128, 64, 14, 14]         51,264\n",
       "│    └─ReLU: 2-5                         [128, 64, 14, 14]         --\n",
       "│    └─MaxPool2d: 2-6                    [128, 64, 7, 7]           --\n",
       "├─Linear: 1-3                            [128, 10]                 31,370\n",
       "==========================================================================================\n",
       "Total params: 83,466\n",
       "Trainable params: 83,466\n",
       "Non-trainable params: 0\n",
       "Total mult-adds (G): 1.37\n",
       "==========================================================================================\n",
       "Input size (MB): 0.40\n",
       "Forward/backward pass size (MB): 38.55\n",
       "Params size (MB): 0.33\n",
       "Estimated Total Size (MB): 39.28\n",
       "=========================================================================================="
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary(cnn, input_size=(batch_size, 1, 28, 28))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define the loss function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CrossEntropyLoss()"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss_func = nn.CrossEntropyLoss()\n",
    "loss_func"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a Optimization Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Adam (\n",
       "Parameter Group 0\n",
       "    amsgrad: False\n",
       "    betas: (0.9, 0.999)\n",
       "    eps: 1e-08\n",
       "    lr: 0.01\n",
       "    weight_decay: 0\n",
       ")"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "optimizer = optim.Adam(cnn.parameters(), lr=0.01)\n",
    "optimizer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_epochs = 1\n",
    "\n",
    "\n",
    "def train(num_epochs, cnn, loaders, loss_func, optimizer):\n",
    "    cnn.train()\n",
    "\n",
    "    # Train the model\n",
    "    total_step = len(loaders[\"train\"])\n",
    "\n",
    "    for epoch in range(num_epochs):\n",
    "        for i, (images, labels) in enumerate(loaders[\"train\"]):\n",
    "            # gives batch data, normalize x when iterate train_loader\n",
    "            b_x = Variable(images)  # batch x\n",
    "            b_y = Variable(labels)  # batch y\n",
    "            output = cnn(b_x)[0]\n",
    "            loss = loss_func(output, b_y)\n",
    "\n",
    "            # clear gradients for this training step\n",
    "            optimizer.zero_grad()\n",
    "\n",
    "            # backpropagation, compute gradients\n",
    "            loss.backward()\n",
    "            # apply gradients\n",
    "            optimizer.step()\n",
    "\n",
    "            if (i + 1) % 100 == 0:\n",
    "                print(\n",
    "                    \"Epoch [{}/{}], Step [{}/{}], Loss: {:.4f}\".format(\n",
    "                        epoch + 1, num_epochs, i + 1, total_step, loss.item()\n",
    "                    )\n",
    "                )\n",
    "                pass\n",
    "\n",
    "            pass\n",
    "\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [1/1], Step [100/469], Loss: 0.1514\n",
      "Epoch [1/1], Step [200/469], Loss: 0.0564\n",
      "Epoch [1/1], Step [300/469], Loss: 0.0237\n",
      "Epoch [1/1], Step [400/469], Loss: 0.0872\n"
     ]
    }
   ],
   "source": [
    "train(num_epochs, cnn, loaders, loss_func, optimizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate the model on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test():\n",
    "    # Test the model\n",
    "    cnn.eval()\n",
    "\n",
    "    with torch.no_grad():\n",
    "        correct = 0\n",
    "        total = 0\n",
    "        for images, labels in loaders[\"test\"]:\n",
    "            test_output, last_layer = cnn(images)\n",
    "            pred_y = torch.max(test_output, 1)[1].data.squeeze()\n",
    "            accuracy = (pred_y == labels).sum().item() / float(labels.size(0))\n",
    "            pass\n",
    "\n",
    "    print(\"Test Accuracy of the model on the 10000 test images: %.2f\" % accuracy)\n",
    "\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy of the model on the 10000 test images: 0.94\n"
     ]
    }
   ],
   "source": [
    "test()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Store trained model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "use_ceph = bool(int(os.getenv(\"USE_CEPH\", 0)))\n",
    "automation = bool(int(os.getenv(\"AUTOMATION\", 0)))\n",
    "\n",
    "time_version = f\"torch-{datetime.now():%y%m%d%H%M%S}-{random.getrandbits(64):08x}\"\n",
    "\n",
    "# Path to data\n",
    "directory_path = Path.cwd().parents[0]\n",
    "trained_model_path = directory_path.joinpath(\n",
    "    str(os.environ.get(\"TRAINED_MODEL_PATH\", \"../models\"))\n",
    ")\n",
    "\n",
    "if not os.path.exists(f\"{trained_model_path}/{time_version}\"):\n",
    "    Path(f\"{trained_model_path}/{time_version}\").mkdir(parents=True, exist_ok=True)\n",
    "\n",
    "\n",
    "torch.save(cnn.state_dict(), f\"{trained_model_path}/{time_version}/pytorch_model.pt\")\n",
    "\n",
    "if automation or use_ceph:\n",
    "    # Download files from S3\n",
    "    s3_endpoint_url = os.environ[\"OBJECT_STORAGE_ENDPOINT_URL\"]\n",
    "    s3_access_key = os.environ[\"AWS_ACCESS_KEY_ID\"]\n",
    "    s3_secret_key = os.environ[\"AWS_SECRET_ACCESS_KEY\"]\n",
    "    s3_bucket = os.environ[\"OBJECT_STORAGE_BUCKET_NAME\"]\n",
    "\n",
    "    # Create an S3 client\n",
    "    s3 = boto3.client(\n",
    "        service_name=\"s3\",\n",
    "        aws_access_key_id=s3_access_key,\n",
    "        aws_secret_access_key=s3_secret_key,\n",
    "        endpoint_url=s3_endpoint_url,\n",
    "    )\n",
    "\n",
    "    p = Path(f\"{trained_model_path}/{time_version}/pytorch_model.pt\")\n",
    "    key = f\"{project_name}/models{p}\"\n",
    "    print(key)\n",
    "    s3.upload_file(\n",
    "        Bucket=s3_bucket,\n",
    "        Key=key,\n",
    "        Filename=f\"{trained_model_path}/{time_version}/pytorch_model.pt\",\n",
    "    )"
   ]
  }
 ],
 "metadata": {
  "dependency_resolution_engine": "pipenv",
  "kernelspec": {
   "display_name": "Python (pytorch-training)",
   "language": "python",
   "name": "jupyterlab-requirements"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "papermill": {
   "default_parameters": {},
   "duration": null,
   "end_time": null,
   "environment_variables": {},
   "exception": null,
   "input_path": "/opt/app-root/src/elyra-aidevsecops-tutorial/notebooks/training.ipynb",
   "output_path": "/opt/app-root/src/elyra-aidevsecops-tutorial/notebooks/training.ipynb",
   "parameters": {},
   "start_time": "2021-02-03T13:27:32.988886",
   "version": "2.2.0"
  },
  "requirements": "{\"packages\": {\"torchvision\": \"*\", \"torch\": \"*\", \"boto3\": \"*\", \"torchinfo\": \"*\"}, \"dev-packages\": {}, \"source\": [{\"url\": \"https://pypi.org/simple\", \"verify_ssl\": true, \"name\": \"pypi\"}], \"requires\": {\"python_version\": \"3.8\"}}",
  "requirements_lock": "{\"_meta\": {\"sources\": [{\"url\": \"https://pypi.org/simple\", \"verify_ssl\": true, \"name\": \"pypi\"}], \"requires\": {\"python_version\": \"3.8\"}, \"hash\": {\"sha256\": \"705f94b8a66792b9e5be6663b6933a230108b37ae01006ccbacfd2d539b26655\"}, \"pipfile-spec\": 6}, \"default\": {\"boto3\": {\"version\": \"==1.18.44\", \"hashes\": [\"sha256:3a270f002818703d5f2eef5296c2fd8b44ef21a3f3290a716ec2202da8dd464e\", \"sha256:8bc3211a7d7767c2c72ae9b226edb5eec5bb96989c83696832b8a5c35feb356a\"], \"index\": \"pypi\"}, \"botocore\": {\"version\": \"==1.21.44\", \"hashes\": [\"sha256:2e134c9f799015e448086ed2b809fe50cc776f6600f093d1a44772288e61260f\", \"sha256:c7640cb49c0e009bea4ad767715acbe0d305b7007235f52422bf31b5d23be8f1\"], \"markers\": \"python_version >= '3.6'\"}, \"jmespath\": {\"version\": \"==0.10.0\", \"hashes\": [\"sha256:b85d0567b8666149a93172712e68920734333c0ce7e89b78b3e987f71e5ed4f9\", \"sha256:cdf6525904cc597730141d61b36f2e4b8ecc257c420fa2f4549bac2c2d0cb72f\"], \"markers\": \"python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'\"}, \"numpy\": {\"version\": \"==1.21.2\", \"hashes\": [\"sha256:09858463db6dd9f78b2a1a05c93f3b33d4f65975771e90d2cf7aadb7c2f66edf\", \"sha256:209666ce9d4a817e8a4597cd475b71b4878a85fa4b8db41d79fdb4fdee01dde2\", \"sha256:298156f4d3d46815eaf0fcf0a03f9625fc7631692bd1ad851517ab93c3168fc6\", \"sha256:30fc68307c0155d2a75ad19844224be0f2c6f06572d958db4e2053f816b859ad\", \"sha256:423216d8afc5923b15df86037c6053bf030d15cc9e3224206ef868c2d63dd6dc\", \"sha256:426a00b68b0d21f2deb2ace3c6d677e611ad5a612d2c76494e24a562a930c254\", \"sha256:466e682264b14982012887e90346d33435c984b7fead7b85e634903795c8fdb0\", \"sha256:51a7b9db0a2941434cd930dacaafe0fc9da8f3d6157f9d12f761bbde93f46218\", \"sha256:52a664323273c08f3b473548bf87c8145b7513afd63e4ebba8496ecd3853df13\", \"sha256:550564024dc5ceee9421a86fc0fb378aa9d222d4d0f858f6669eff7410c89bef\", \"sha256:5de64950137f3a50b76ce93556db392e8f1f954c2d8207f78a92d1f79aa9f737\", \"sha256:640c1ccfd56724f2955c237b6ccce2e5b8607c3bc1cc51d3933b8c48d1da3723\", \"sha256:7fdc7689daf3b845934d67cb221ba8d250fdca20ac0334fea32f7091b93f00d3\", \"sha256:805459ad8baaf815883d0d6f86e45b3b0b67d823a8f3fa39b1ed9c45eaf5edf1\", \"sha256:92a0ab128b07799dd5b9077a9af075a63467d03ebac6f8a93e6440abfea4120d\", \"sha256:9f2dc79c093f6c5113718d3d90c283f11463d77daa4e83aeeac088ec6a0bda52\", \"sha256:a5109345f5ce7ddb3840f5970de71c34a0ff7fceb133c9441283bb8250f532a3\", \"sha256:a55e4d81c4260386f71d22294795c87609164e22b28ba0d435850fbdf82fc0c5\", \"sha256:a9da45b748caad72ea4a4ed57e9cd382089f33c5ec330a804eb420a496fa760f\", \"sha256:b160b9a99ecc6559d9e6d461b95c8eec21461b332f80267ad2c10394b9503496\", \"sha256:b342064e647d099ca765f19672696ad50c953cac95b566af1492fd142283580f\", \"sha256:b5e8590b9245803c849e09bae070a8e1ff444f45e3f0bed558dd722119eea724\", \"sha256:bf75d5825ef47aa51d669b03ce635ecb84d69311e05eccea083f31c7570c9931\", \"sha256:c01b59b33c7c3ba90744f2c695be571a3bd40ab2ba7f3d169ffa6db3cfba614f\", \"sha256:d96a6a7d74af56feb11e9a443150216578ea07b7450f7c05df40eec90af7f4a7\", \"sha256:dd0e3651d210068d13e18503d75aaa45656eef51ef0b261f891788589db2cc38\", \"sha256:e167b9805de54367dcb2043519382be541117503ce99e3291cc9b41ca0a83557\", \"sha256:e42029e184008a5fd3d819323345e25e2337b0ac7f5c135b7623308530209d57\", \"sha256:f545c082eeb09ae678dd451a1b1dbf17babd8a0d7adea02897a76e639afca310\", \"sha256:fde50062d67d805bc96f1a9ecc0d37bfc2a8f02b937d2c50824d186aa91f2419\"], \"markers\": \"python_version < '3.11' and python_version >= '3.7'\"}, \"pillow\": {\"version\": \"==8.3.2\", \"hashes\": [\"sha256:0412516dcc9de9b0a1e0ae25a280015809de8270f134cc2c1e32c4eeb397cf30\", \"sha256:04835e68ef12904bc3e1fd002b33eea0779320d4346082bd5b24bec12ad9c3e9\", \"sha256:06d1adaa284696785375fa80a6a8eb309be722cf4ef8949518beb34487a3df71\", \"sha256:085a90a99404b859a4b6c3daa42afde17cb3ad3115e44a75f0d7b4a32f06a6c9\", \"sha256:0b9911ec70731711c3b6ebcde26caea620cbdd9dcb73c67b0730c8817f24711b\", \"sha256:10e00f7336780ca7d3653cf3ac26f068fa11b5a96894ea29a64d3dc4b810d630\", \"sha256:11c27e74bab423eb3c9232d97553111cc0be81b74b47165f07ebfdd29d825875\", \"sha256:11eb7f98165d56042545c9e6db3ce394ed8b45089a67124298f0473b29cb60b2\", \"sha256:13654b521fb98abdecec105ea3fb5ba863d1548c9b58831dd5105bb3873569f1\", \"sha256:15ccb81a6ffc57ea0137f9f3ac2737ffa1d11f786244d719639df17476d399a7\", \"sha256:18a07a683805d32826c09acfce44a90bf474e6a66ce482b1c7fcd3757d588df3\", \"sha256:19ec4cfe4b961edc249b0e04b5618666c23a83bc35842dea2bfd5dfa0157f81b\", \"sha256:1c3ff00110835bdda2b1e2b07f4a2548a39744bb7de5946dc8e95517c4fb2ca6\", \"sha256:27a330bf7014ee034046db43ccbb05c766aa9e70b8d6c5260bfc38d73103b0ba\", \"sha256:2b11c9d310a3522b0fd3c35667914271f570576a0e387701f370eb39d45f08a4\", \"sha256:2c661542c6f71dfd9dc82d9d29a8386287e82813b0375b3a02983feac69ef864\", \"sha256:2cde7a4d3687f21cffdf5bb171172070bb95e02af448c4c8b2f223d783214056\", \"sha256:2d5e9dc0bf1b5d9048a94c48d0813b6c96fccfa4ccf276d9c36308840f40c228\", \"sha256:2f23b2d3079522fdf3c09de6517f625f7a964f916c956527bed805ac043799b8\", \"sha256:35d27687f027ad25a8d0ef45dd5208ef044c588003cdcedf05afb00dbc5c2deb\", \"sha256:35d409030bf3bd05fa66fb5fdedc39c521b397f61ad04309c90444e893d05f7d\", \"sha256:4326ea1e2722f3dc00ed77c36d3b5354b8fb7399fb59230249ea6d59cbed90da\", \"sha256:4abc247b31a98f29e5224f2d31ef15f86a71f79c7f4d2ac345a5d551d6393073\", \"sha256:4d89a2e9219a526401015153c0e9dd48319ea6ab9fe3b066a20aa9aee23d9fd3\", \"sha256:4e59e99fd680e2b8b11bbd463f3c9450ab799305d5f2bafb74fefba6ac058616\", \"sha256:548794f99ff52a73a156771a0402f5e1c35285bd981046a502d7e4793e8facaa\", \"sha256:56fd98c8294f57636084f4b076b75f86c57b2a63a8410c0cd172bc93695ee979\", \"sha256:59697568a0455764a094585b2551fd76bfd6b959c9f92d4bdec9d0e14616303a\", \"sha256:6bff50ba9891be0a004ef48828e012babaaf7da204d81ab9be37480b9020a82b\", \"sha256:6cb3dd7f23b044b0737317f892d399f9e2f0b3a02b22b2c692851fb8120d82c6\", \"sha256:7dbfbc0020aa1d9bc1b0b8bcf255a7d73f4ad0336f8fd2533fcc54a4ccfb9441\", \"sha256:838eb85de6d9307c19c655c726f8d13b8b646f144ca6b3771fa62b711ebf7624\", \"sha256:8b68f565a4175e12e68ca900af8910e8fe48aaa48fd3ca853494f384e11c8bcd\", \"sha256:8f284dc1695caf71a74f24993b7c7473d77bc760be45f776a2c2f4e04c170550\", \"sha256:963ebdc5365d748185fdb06daf2ac758116deecb2277ec5ae98139f93844bc09\", \"sha256:a048dad5ed6ad1fad338c02c609b862dfaa921fcd065d747194a6805f91f2196\", \"sha256:a1bd983c565f92779be456ece2479840ec39d386007cd4ae83382646293d681b\", \"sha256:a66566f8a22561fc1a88dc87606c69b84fa9ce724f99522cf922c801ec68f5c1\", \"sha256:bcb04ff12e79b28be6c9988f275e7ab69f01cc2ba319fb3114f87817bb7c74b6\", \"sha256:bd24054aaf21e70a51e2a2a5ed1183560d3a69e6f9594a4bfe360a46f94eba83\", \"sha256:be25cb93442c6d2f8702c599b51184bd3ccd83adebd08886b682173e09ef0c3f\", \"sha256:c691b26283c3a31594683217d746f1dad59a7ae1d4cfc24626d7a064a11197d4\", \"sha256:cc9d0dec711c914ed500f1d0d3822868760954dce98dfb0b7382a854aee55d19\", \"sha256:ce2e5e04bb86da6187f96d7bab3f93a7877830981b37f0287dd6479e27a10341\", \"sha256:ce651ca46d0202c302a535d3047c55a0131a720cf554a578fc1b8a2aff0e7d96\", \"sha256:d0c8ebbfd439c37624db98f3877d9ed12c137cadd99dde2d2eae0dab0bbfc355\", \"sha256:d675a876b295afa114ca8bf42d7f86b5fb1298e1b6bb9a24405a3f6c8338811c\", \"sha256:dde3f3ed8d00c72631bc19cbfff8ad3b6215062a5eed402381ad365f82f0c18c\", \"sha256:e5a31c07cea5edbaeb4bdba6f2b87db7d3dc0f446f379d907e51cc70ea375629\", \"sha256:f514c2717012859ccb349c97862568fdc0479aad85b0270d6b5a6509dbc142e2\", \"sha256:fc0db32f7223b094964e71729c0361f93db43664dd1ec86d3df217853cedda87\", \"sha256:fd4fd83aa912d7b89b4b4a1580d30e2a4242f3936882a3f433586e5ab97ed0d5\", \"sha256:feb5db446e96bfecfec078b943cc07744cc759893cef045aa8b8b6d6aaa8274e\"], \"markers\": \"python_version >= '3.6'\"}, \"python-dateutil\": {\"version\": \"==2.8.2\", \"hashes\": [\"sha256:0123cacc1627ae19ddf3c27a5de5bd67ee4586fbdd6440d9748f8abb483d3e86\", \"sha256:961d03dc3453ebbc59dbdea9e4e11c5651520a876d0f4db161e8674aae935da9\"], \"markers\": \"python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'\"}, \"s3transfer\": {\"version\": \"==0.5.0\", \"hashes\": [\"sha256:50ed823e1dc5868ad40c8dc92072f757aa0e653a192845c94a3b676f4a62da4c\", \"sha256:9c1dc369814391a6bda20ebbf4b70a0f34630592c9aa520856bf384916af2803\"], \"markers\": \"python_version >= '3.6'\"}, \"six\": {\"version\": \"==1.16.0\", \"hashes\": [\"sha256:1e61c37477a1626458e36f7b1d82aa5c9b094fa4802892072e49de9c60c4c926\", \"sha256:8abb2f1d86890a2dfb989f9a77cfcfd3e47c2a354b01111771326f8aa26e0254\"], \"markers\": \"python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'\"}, \"torch\": {\"version\": \"==1.9.1\", \"hashes\": [\"sha256:06435080ba0a2c8f88b65af0550b973c5aa7771eacd9b17f69057fc7436a8ae2\", \"sha256:1fb49ca0ca8edefbb3f47f6801482144c3a746ec21a65eb3f0839a1d8fb24705\", \"sha256:335961a5c893f7b33b29aecbc19382a1a1b0106b3457a1c45148e1e14f8f5e09\", \"sha256:351dda9f483486bec66ed838234e96f077e6886c88110bb1e2f4a708ed2356ce\", \"sha256:35ec703bc535bde7e8790ab9500f02d4413d995ac981520501fde95e268781e1\", \"sha256:42ca081a2e0e759844e70cad7efd8fcfb2f81634dffa73a226564eb83d989e5b\", \"sha256:54dacb6a3f63c54334fadbf22fb6e9ee865085a4e0368962edff5babda057606\", \"sha256:936d303c5e1d60259fb71d95a33e84d84fececa25a0fae112f6a23286ff183c8\", \"sha256:a198332e2d344d25e423ae2df98d56d83060f19e9f4cf23164dffc8d403efeb8\", \"sha256:b92f934b3c95578b3fd37cc06afca208d63f02b0d01b806e979cb4e46124a7f8\", \"sha256:dd3ca91dc1a9fe3fbcddf035cb2fb8be44d57a527b845cd196ba69249adecccf\", \"sha256:e470697006a4c08e4fb6a645e8ca49b0d36c8e7ccf413deef5161335bd7399f1\"], \"index\": \"pypi\"}, \"torchinfo\": {\"version\": \"==1.5.3\", \"hashes\": [\"sha256:46ec87260016b70e548e57dcb4a9cc40878c83ab3532fe6a3d6c60437a1ce58f\", \"sha256:49911e5d499f9ff97e270455d1f0d01f205c87cbb7df2bb4c5ba175d88940307\"], \"index\": \"pypi\"}, \"torchvision\": {\"version\": \"==0.10.1\", \"hashes\": [\"sha256:1c186f42b4f8aa9a01c56c3a758693b0447aa169afb9fba0051177f8fecbd691\", \"sha256:453e935212193e89b4bbb8d51082d8138631c2f8a420390284b1946d893df6eb\", \"sha256:46a70a30ea7aeab63e67504778f2565fbb1c153fdd8e1a8c6a22193aec4dbddd\", \"sha256:4ebffeee5468a0934952030eaba1de1dbb08154132235ee1d9049e41dfb1600d\", \"sha256:6c8fe90213be4bce590ac9647b34db022d5d1ae94f309a733b9a64e65232173a\", \"sha256:99d3e01e1d67d12bcc88e826431b70cad5b8e4729a277c04601f83358a120508\", \"sha256:ac8dfbe4933013dda898b815e2476ebbc35e3a16b9352dfdd66e773c77755bec\", \"sha256:bc99a984b162ee5626787eaee885d9fec1a5f16837f9d0c8223cca3269b9e47d\", \"sha256:cd7e2b1a89d5a08f24325fc12441f5ba2822f407489377ac7841bf351a1f4d37\", \"sha256:d6420bf21b9d0bdbabe55d64c8b11c61f8eb077948a55d5707946fcb17d97cec\", \"sha256:d7c2d6c20244404fc9ca3568c88c305cb5a81d526d5912d52d22c64999bd4353\", \"sha256:e504d9d51eae60a98925aee4a3fd58655abd5669659ad7431f7791a93af166fc\"], \"index\": \"pypi\"}, \"typing-extensions\": {\"version\": \"==3.10.0.2\", \"hashes\": [\"sha256:49f75d16ff11f1cd258e1b988ccff82a3ca5570217d7ad8c5f48205dd99a677e\", \"sha256:d8226d10bc02a29bcc81df19a26e56a9647f8b0a6d4a83924139f4a8b01f17b7\", \"sha256:f1d25edafde516b146ecd0613dabcc61409817af4766fbbcfb8d1ad4ec441a34\"]}, \"urllib3\": {\"version\": \"==1.26.6\", \"hashes\": [\"sha256:39fb8672126159acb139a7718dd10806104dec1e2f0f6c88aab05d17df10c8d4\", \"sha256:f57b4c16c62fa2760b7e3d97c35b255512fb6b59a259730f36ba32ce9f8e342f\"], \"markers\": \"python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4' and python_version < '4'\"}}, \"develop\": {}}",
  "toc-autonumbering": false,
  "toc-showcode": false,
  "toc-showmarkdowntxt": false,
  "toc-showtags": false
 },
 "nbformat": 4,
 "nbformat_minor": 4
}