5.ImageData-UsingAutoGluon.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AutoML with Image data - Using AutoGluon \n",
    "\n",
    "\n",
    "Here we demonstrate how to train models for classifying images with AutoGluon. AutoGluon can be used [for object detection](https://autogluon.mxnet.io/tutorials/object_detection/beginner.html) in a similar manner.  Our tutorial will not dive into the technical details of [convolutional neural networks](http://d2l.ai/chapter_convolutional-neural-networks/index.html) and modern [computer vision](http://d2l.ai/chapter_computer-vision/index.html); for  background on this, please study ([Zhang, 2020](http://d2l.ai/)) chapters 6, 7, and 13. \n",
    "\n",
    "\n",
    "As for tabular data, AutoGluon helps you easily train models for these vision tasks with just a single call to `fit()`.  We begin by specifying `ImageClassification` as our task of interest: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import autogluon as ag\n",
    "from autogluon import ImageClassification as task\n",
    "\n",
    "import os, mxnet\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Format Image Dataset\n",
    "\n",
    "Our image classification task is based on a subset of the [Shopee-IET dataset](https://www.kaggle.com/c/shopee-iet-machine-learning-competition/data). Each image in this data depicts a clothing item and the corresponding label specifies its clothing category.\n",
    "Our subset of the data contains the following possible labels: `BabyPants`, `BabyShirt`, `womencasualshoes`, `womenchiffontop`.  This dataset is quite small to ensure the tutorial can be run reasonably fast, so note that reported accuracy numbers may not be particularly reliable.  We can download the data subset and unzip it via the following commands:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training dataset stored in folder: data/train/\n",
      "Test dataset stored in folder: data/test/ \n",
      "\n",
      "Example files in training dataset:\n",
      "\n",
      "train/\n",
      "├── BabyPants/\n",
      "    ├── BabyPants_939.jpg\n",
      "    ├── BabyPants_463.jpg\n",
      "    ├── ...\n",
      "├── womenchiffontop/\n",
      "    ├── womenchiffontop_2633.jpg\n",
      "    ├── womenchiffontop_2454.jpg\n",
      "    ├── ...\n",
      "├── womencasualshoes/\n",
      "    ├── womencasualshoes_1060.jpg\n",
      "    ├── womencasualshoes_2775.jpg\n",
      "    ├── ...\n",
      "├── BabyShirt/\n",
      "    ├── BabyShirt_376.jpg\n",
      "    ├── BabyShirt_168.jpg\n",
      "    ├── ...\n",
      "Number of images per class: {'BabyPants': 200, 'womenchiffontop': 200, 'womencasualshoes': 200, 'BabyShirt': 200}\n",
      "\n",
      " Some (image, label) training examples:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "filename = ag.download('https://autogluon.s3.amazonaws.com/datasets/shopee-iet.zip')\n",
    "data_folder = ag.unzip(filename)\n",
    "\n",
    "train_folder = data_folder+os.sep+'train'+os.sep\n",
    "test_folder = data_folder+os.sep+'test'+os.sep\n",
    "print(\"Training dataset stored in folder:\", train_folder)\n",
    "print(\"Test dataset stored in folder:\", test_folder, \"\\n\")\n",
    "\n",
    "print(\"Example files in training dataset:\\n\\ntrain/\")\n",
    "class_names = os.listdir(train_folder)\n",
    "num_class = len(class_names)\n",
    "imgfile_perclass = []\n",
    "num_img_perclass = []\n",
    "for label in class_names:\n",
    "    print(\"├──\", label+os.sep)\n",
    "    file_names = os.listdir(train_folder+label)\n",
    "    imgfile_perclass.append(train_folder+label+os.sep+file_names[0])\n",
    "    num_img_perclass.append(len(file_names))\n",
    "    for filename in file_names[:2]:\n",
    "        print(\"    ├──\", filename)\n",
    "    print(\"    ├── ...\")\n",
    "\n",
    "    \n",
    "print(\"Number of images per class:\", {class_names[i]:num_img_perclass[i] for i in range(num_class)})\n",
    "print(\"\\n Some (image, label) training examples:\")\n",
    "fig, axs = plt.subplots(1, num_class, gridspec_kw={'wspace': 1.0}, figsize=(15,10))\n",
    "for i in range(num_class):\n",
    "    img = mxnet.image.imread(filename=imgfile_perclass[i])\n",
    "    axs[i].imshow(img.asnumpy())\n",
    "    axs[i].set_title(class_names[i])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "AutoGluon can work with image-datasets stored in either  [RecordIO format](https://mxnet.apache.org/versions/1.6/api/faq/recordio.html) or the image-folder format used for this Shopee-IET dataset, where each subdirectory corresponds to a class-label and the images from this class are stored in separate files within the subdirectory. Let's load the image data into AutoGluon:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AutoGluonObject\n"
     ]
    }
   ],
   "source": [
    "train_data = task.Dataset('data/train')\n",
    "print(train_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Unlike for tabular data, AutoGluon avoids loading the full image dataset (which can be quite large) into memory.\n",
    "We can work with the `Dataset` object like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ordered class labels: ['BabyPants', 'BabyShirt', 'womencasualshoes', 'womenchiffontop']\n"
     ]
    }
   ],
   "source": [
    "train_loaded = train_data.init() # must initialize before interacting with this object \n",
    "print(\"Ordered class labels:\", train_loaded.classes) # predictions will be class-indices corresponding to this list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fit Models with AutoGluon\n",
    "\n",
    "We now train an image classification model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "scheduler_options: Key 'training_history_callback_delta_secs': Imputing default value 60\n",
      "scheduler_options: Key 'delay_get_config': Imputing default value True\n",
      "\n",
      "Starting Experiments\n",
      "Num of Finished Tasks is 0\n",
      "Num of Pending Tasks is 2\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scheduler: FIFOScheduler(\n",
      "DistributedResourceManager{\n",
      "(Remote: Remote REMOTE_ID: 0, \n",
      "\t<Remote: 'inproc://172.31.68.247/12139/1' processes=1 threads=4, memory=64.28 GB>, Resource: NodeResourceManager(4 CPUs, 1 GPUs))\n",
      "})\n",
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a1f1d27e494244ec97ee09565cff95cb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=2.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "{'dataset': <autogluon.task.image_classification.dataset.NativeImageFolderDataset object at 0x7f5ac5fb3310>, 'net': 'ResNet50_v1b', 'optimizer': <autogluon.core.optimizer.NAG object at 0x7f5ae0d4e290>, 'loss': SoftmaxCrossEntropyLoss(batch_axis=0, w=None), 'metric': 'accuracy', 'num_gpus': 1, 'split_ratio': 0.8, 'batch_size': 64, 'input_size': 224, 'epochs': 3, 'final_fit_epochs': 3, 'verbose': True, 'num_workers': 4, 'hybridize': True, 'final_fit': False, 'tricks': Dict{'last_gamma': False, 'use_pretrained': True, 'use_se': False, 'mixup': False, 'mixup_alpha': 0.2, 'mixup_off_epoch': 0, 'label_smoothing': False, 'no_wd': False, 'teacher_name': None, 'temperature': 20.0, 'hard_weight': 0.5, 'batch_norm': False, 'use_gn': False}, 'lr_config': Dict{'lr_mode': 'cosine', 'lr_decay': 0.1, 'lr_decay_period': 0, 'lr_decay_epoch': '40,80', 'warmup_lr': 0.0, 'warmup_epochs': 0}, 'task_id': 0}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "96fc31c379954f0c8e9da8d97ee36768",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(HBox(children=(IntProgress(value=0, max=3), HTML(value=''))), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "{'dataset': <autogluon.task.image_classification.dataset.NativeImageFolderDataset object at 0x7f5ac5fb36d0>, 'net': 'ResNet50_v1b', 'optimizer': <autogluon.core.optimizer.NAG object at 0x7f5ae32da390>, 'loss': SoftmaxCrossEntropyLoss(batch_axis=0, w=None), 'metric': 'accuracy', 'num_gpus': 1, 'split_ratio': 0.8, 'batch_size': 64, 'input_size': 224, 'epochs': 3, 'final_fit_epochs': 3, 'verbose': True, 'num_workers': 4, 'hybridize': True, 'final_fit': False, 'tricks': Dict{'last_gamma': False, 'use_pretrained': True, 'use_se': False, 'mixup': False, 'mixup_alpha': 0.2, 'mixup_off_epoch': 0, 'label_smoothing': False, 'no_wd': False, 'teacher_name': None, 'temperature': 20.0, 'hard_weight': 0.5, 'batch_norm': False, 'use_gn': False}, 'lr_config': Dict{'lr_mode': 'cosine', 'lr_decay': 0.1, 'lr_decay_period': 0, 'lr_decay_epoch': '40,80', 'warmup_lr': 0.0, 'warmup_epochs': 0}, 'task_id': 1}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "65fed80a5dd84fbeab2f3f43441bafc5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(HBox(children=(IntProgress(value=0, max=3), HTML(value=''))), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "{'dataset': <autogluon.task.image_classification.dataset.NativeImageFolderDataset object at 0x7f5ae0a8ec50>, 'net': 'ResNet50_v1b', 'optimizer': <autogluon.core.optimizer.NAG object at 0x7f5ac0186710>, 'loss': SoftmaxCrossEntropyLoss(batch_axis=0, w=None), 'metric': 'accuracy', 'num_gpus': 1, 'split_ratio': 0.8, 'batch_size': 64, 'input_size': 224, 'epochs': 3, 'final_fit_epochs': 3, 'verbose': True, 'num_workers': 4, 'hybridize': True, 'final_fit': True, 'tricks': Dict{'last_gamma': False, 'use_pretrained': True, 'use_se': False, 'mixup': False, 'mixup_alpha': 0.2, 'mixup_off_epoch': 0, 'label_smoothing': False, 'no_wd': False, 'teacher_name': None, 'temperature': 20.0, 'hard_weight': 0.5, 'batch_norm': False, 'use_gn': False}, 'lr_config': Dict{'lr_mode': 'cosine', 'lr_decay': 0.1, 'lr_decay_period': 0, 'lr_decay_epoch': '40,80', 'warmup_lr': 0.0, 'warmup_epochs': 0}, 'task_id': 2}\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d565d2765b124364a7cc1c41a987e97e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(HBox(children=(IntProgress(value=0, max=3), HTML(value=''))), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Saving Training Curve in shopeeietClassifier/plot_training_curves.png\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "num_train_epochs = 3 # How many epochs to train each network. Small value used just for quick demo, you should increase it.\n",
    "num_hpo_trials = 2 # How many networks to train with different hyperparameters. Small value used just for quick demo, you should increase it.\n",
    "ngpus_per_trial = 1 # How many GPUs to use per training run. Set =0 if you don't have GPU but beware of slow run-times (otherwise make sure GPU-version of MXNet is installed).\n",
    "model_folder = 'shopeeietClassifier'\n",
    "\n",
    "classifier = task.fit(train_data, output_directory=model_folder, verbose=True,\n",
    "                      epochs=num_train_epochs, num_trials=num_hpo_trials, ngpus_per_trial=ngpus_per_trial)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Within `fit()`, the dataset is automatically split into random disjoint training and validation sets.\n",
    "Multiple networks are trained under different hyperparameter-values (two here since we set `num_hpo_trials = 2`), and finally the network with the best hyperparameter configuration is selected based on its performance on validation set.\n",
    "Below we print the validation accuracy achieved this network. In the final step of `fit()`, this best network is  retrained on our entire dataset (merging training+validation) using the same hyperparameter values, to produce the model that will be used for inference. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy on Validation Data: 0.463\n"
     ]
    }
   ],
   "source": [
    "print('Accuracy on Validation Data: %.3f' % classifier.results['best_reward'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Classifying New Images\n",
    "\n",
    "Given a new image, we can easily use our trained classifier to predict its label:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The input image is classified as [BabyShirt], with confidence = 0.49.\n",
      "Predicted probability of each class:\n",
      " {'BabyPants': 0.084875375, 'BabyShirt': 0.4944493, 'womencasualshoes': 0.13188109, 'womenchiffontop': 0.28879425}\n"
     ]
    }
   ],
   "source": [
    "image = 'data/test/BabyShirt/BabyShirt_323.jpg'\n",
    "pred_class, pred_confidence, pred_prob = classifier.predict(image, plot=True)\n",
    "\n",
    "print(\"The input image is classified as [%s], with confidence = %.2f.\" %\n",
    "      (train_loaded.classes[pred_class.asscalar()], pred_confidence.asscalar()))\n",
    "\n",
    "print(\"Predicted probability of each class:\\n\", {train_loaded.classes[i]: pred_prob[0][i].asscalar() for i in range(len(train_loaded.classes))})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Let's load a small test dataset of new images (containing 20 images from each class stored in the same image-folder format).  We can also predict-with or evaluate our classifier over all images in the test data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "27db616c89c442ecab4c4294d85ab3b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(HBox(children=(IntProgress(value=0, max=1), HTML(value=''))), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Accuracy on test data: 0.578\n"
     ]
    }
   ],
   "source": [
    "test_data = task.Dataset('data/test', train=False)\n",
    "\n",
    "test_acc = classifier.evaluate(test_data)\n",
    "print('Accuracy on test data: %.3f' % test_acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## What happened during fit()\n",
    "\n",
    "Let's first print a summary of the fitting process. From the summary below, we can see: which configuration of the hyperparameter values produced the best validation performance (`'best_config'`), as well as which hyperparameter-values were adopted for training in each trial and the resulting performance (`classification_reward` = validation accuracy in this case) after every training-epoch. Plots summarizing these results are saved in the listed files."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Summary of Fit Process:  \n",
      "{'num_classes': 4, 'best_reward': 0.4625, 'best_config': {'net▁choice': 0, 'optimizer▁learning_rate': 0.004350975063803458, 'optimizer▁wd': 0.0007124564398589832}, 'total_time': 185.57865047454834, 'reward_attr': 'classification_reward', 'classification_reward': 0.4625, 'search_space': OrderedDict([('net▁choice', Categorical['ResNet50_v1b', 'ResNet18_v1b']), ('optimizer▁learning_rate', Real: lower=0.001, upper=0.01), ('optimizer▁wd', Real: lower=0.0001, upper=0.001)]), 'search_strategy': 'random'}\n",
      "{'stop_criterion': {'time_limits': None, 'max_reward': None}, 'resources_per_trial': {'num_cpus': 4, 'num_gpus': 1}}\n",
      "Information about each trial:  \n",
      "Trial ID: 0\n",
      "{'config': {'net▁choice': 0, 'optimizer▁learning_rate': 0.0031622777, 'optimizer▁wd': 0.0003162278}, 'history': [{'epoch': 1, 'classification_reward': 0.325, 'time_this_iter': 21.69662857055664, 'time_since_start': 21.818780660629272}, {'epoch': 2, 'classification_reward': 0.38125, 'time_this_iter': 19.043954610824585, 'time_since_start': 40.8620228767395}, {'epoch': 3, 'time_this_iter': 18.81824827194214, 'time_since_start': 59.6802031993866}], 'metadata': {'epoch': 3, 'time_this_iter': 18.81824827194214, 'time_since_start': 59.6802031993866}, 'classification_reward': 0.4}\n",
      "Information about each trial:  \n",
      "Trial ID: 1\n",
      "{'config': {'net▁choice': 0, 'optimizer▁learning_rate': 0.004350975063803458, 'optimizer▁wd': 0.0007124564398589832}, 'history': [{'epoch': 1, 'classification_reward': 0.425, 'time_this_iter': 22.27428650856018, 'time_since_start': 82.18471884727478}, {'epoch': 2, 'classification_reward': 0.55625, 'time_this_iter': 18.780302047729492, 'time_since_start': 100.96479535102844}, {'epoch': 3, 'time_this_iter': 19.162839889526367, 'time_since_start': 120.12707567214966}], 'metadata': {'epoch': 3, 'time_this_iter': 19.162839889526367, 'time_since_start': 120.12707567214966}, 'classification_reward': 0.4625}\n",
      "Plot summary of models saved to file: shopeeietClassifier/SummaryOfModels.html\n",
      "Plot summary of models saved to file: SummaryOfModels.html\n",
      "Plot of HPO performance saved to file: shopeeietClassifier/PerformanceVsTrials.png\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "summary = classifier.fit_summary(output_directory=model_folder, verbosity=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's an overview of the steps automatically performed by AutoGluon during the execution of `fit()`: \n",
    "\n",
    "**1)** Preprocess raw data and split into training/validation sets.\n",
    "\n",
    "Preprocessing includes [resizing/cropping](https://mxnet.apache.org/versions/1.5.0/tutorials/gluon/transforms.html) images to ensure they are the right size for our models, as well as [normalizing](https://mxnet.apache.org/versions/1.5.0/tutorials/gluon/transforms.html) the values at each pixel based on their statistics across the dataset.\n",
    "\n",
    "\n",
    "**2)** Optionally apply [data augmentation](http://d2l.ai/chapter_computer-vision/image-augmentation.html) to increase the size of the dataset available for training. \n",
    "\n",
    "Augmenting the dataset with synthetically-generated examples may boost the resulting accuracy, but can also require longer training times. The simple-yet-effective [Mixup augmentation](https://arxiv.org/abs/1812.01187) strategy can be used by specifying the fit argument: `tricks = {'mixup': True}`. In practice, augmentation will often be applied during training when one mini-batch of training examples is loaded at a time.  \n",
    "\n",
    "<img src=\"files/images/mixup.png\" width=\"500\" height=\"400\">\n",
    "\n",
    "\n",
    "**3)** Load [pretrained convolutional neural network](https://gluon-cv.mxnet.io/model_zoo/classification.html) from\n",
    "[GluonCV](https://gluon-cv.mxnet.io/) that has previously been trained on a large image dataset.\n",
    "\n",
    "By starting from an already trained network and [fine-tuning](http://d2l.ai/chapter_computer-vision/fine-tuning.html) this network's parameters on our dataset, our resulting model can transfer its previously-learned knowledge of useful image-features to our dataset. Transfer learning enables the model to achieve good accuracy with less data than we would otherwise need to provide if training the network from scratch. Nonetheless you can still disable transfer-learning and train from scratch if you prefer, by specifying the [fit argument](https://autogluon.mxnet.io/api/autogluon.task.html#autogluon.task.ImageClassification.fit): `use_pretrained=False`. \n",
    "\n",
    "<img src=\"files/images/modelzoo.png\" width=\"350\" height=\"300\">\n",
    "\n",
    "**4)** Setup a suitable network architecture for this particular classification task.\n",
    "\n",
    "This is generally accomplished by modifying the output layer of the pretrained network. For example: if your training dataset has 10 classes but the pretrained network was previously learned with a different number of classes, then\n",
    "AutoGluon will replace the previous output layer with a new one (with randomly initialized parameters) that outputs a 10-dimensional vector (of predicted class probailities).\n",
    "\n",
    "<img src=\"files/images/finetune.svg\" width=\"400\" height=\"400\">\n",
    "\n",
    "\n",
    "**5)** Repeatedly train neural networks under different hyperparameter configurations to find the best hyperparameters. \n",
    "\n",
    "Each training-run of a network may leverage multiple GPUs, and two training-runs of networks with different hyperparameters may be run in parallel on different (sets of) GPUs as well. For example, if your machine has 4 GPUs, and you specify `ngpus_pertrial=2`, then AutoGluon will at any given time be running two training jobs in parallel, each using 2 GPUs. Hyperparameter-tuning will be further described in the next tutorial.\n",
    "\n",
    "<img src=\"files/images/autogluon_system.png\" width=\"400\" height=\"400\">\n",
    "\n",
    "\n",
    "**6)** Retrain our final model on full training+validation dataset. \n",
    "\n",
    "Here we train our final network on the entire dataset (merging training+validation) using the same hyperparameter values (including number of training epochs) as the network that previously performed best on the validation data. This is the model that will be used for inference on new data, and its accuracy benefits from being retrained on a larger dataset.  Note there is no validation data available to gauge the performance of this network, which is why it is trained under the same hyperparameter-configuration previously observed to work best.\n",
    "\n",
    "\n",
    "**7)** Optionally construct more-accurate model ensemble of multiple networks, by setting the [fit argument](https://autogluon.mxnet.io/api/autogluon.task.html#autogluon.task.ImageClassification.fit): `ensemble` to the desired number of networks in your ensemble (more networks usually lead to improved accuracy but greater inference-latency/memory-footprint).\n",
    "\n",
    "Running the code in this tutorial, we see that image-classification models tend to be much larger/slower than our models for tabular data. Thus, ensembling these models via multi-layer stacking with bagging would be quite cumbersome. We instead adopt a popular strategy known as [Deep Ensembles (Lakshminarayanan, 2017)](https://papers.nips.cc/paper/7219-simple-and-scalable-predictive-uncertainty-estimation-using-deep-ensembles.pdf): simply retrain $K$ different networks (on the same training data) all using the same hyperparameter values that previously exhibited the best validation performance. These networks only differ in their random initializations and randomness in the training process (e.g. the mini-batches of data used for stochastic gradient updates of their parameters). Since neural network optimization is highly non-convex, this simple technique suffices to create a diverse ensemble (typically values of $K$ between 3-10 are used).\n",
    "\n",
    "<img src=\"files/images/deepensemble.png\" width=\"300\" height=\"300\">\n",
    "\n",
    "Beyond these steps, AutoGluon's `fit()` offers additional `tricks` that can be used to further boost image-classification accuracy ([He et al, 2019](https://arxiv.org/abs/1812.01187)), which are listed in the [fit documentation](https://autogluon.mxnet.io/api/autogluon.task.html#autogluon.task.ImageClassification.fit)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## References\n",
    "\n",
    "[**AutoGluon Documentation** (autogluon.mxnet.io)](https://autogluon.mxnet.io/api/autogluon.task.html)\n",
    "\n",
    "Sun et al. [**Accurate image classification in 3 lines of code with AutoGluon**](https://medium.com/@zhanghang0704/image-classification-on-kaggle-using-autogluon-fc896e74d7e8). *Medium*, 2020.\n",
    "\n",
    "Zhang et al. [**Dive into Deep Learning**](http://d2l.ai/). *http:d2l.ai*, 2020.\n",
    "\n",
    "Guo et al. [**GluonCV and GluonNLP: Deep Learning in Computer Vision and Natural Language Processing**](http://jmlr.org/papers/v21/19-429.html). *Journal of Machine Learning Research*, 2020.\n",
    "\n",
    "Lakshminarayanan et al. [**Simple and Scalable Predictive Uncertainty\n",
    "Estimation using Deep Ensembles**](https://papers.nips.cc/paper/7219-simple-and-scalable-predictive-uncertainty-estimation-using-deep-ensembles.pdf). In: *NIPS*, 2017.\n",
    "\n",
    "He et al. [**Bag of Tricks for Image Classification with Convolutional Neural Networks**](https://arxiv.org/abs/1812.01187). In: *CVPR*, 2019."
   ]
  }
 ],
 "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"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "01d9cfcbfa284d33b8ecdb3ab563be4f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.719: 100%",
       "layout": "IPY_MODEL_3422f43ce62b4596b471db348dcfb05a",
       "max": 10,
       "style": "IPY_MODEL_f579431b9142422cb602de64f9c41971",
       "value": 10
      }
     },
     "0359f302ab3e4382b47d9a364c7ae4e2": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "07adeeb6d28e4fc9bc3d87c3a340943c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_01d9cfcbfa284d33b8ecdb3ab563be4f",
        "IPY_MODEL_6b8990dfac4f4ebeb0e028f0b802c2c9"
       ],
       "layout": "IPY_MODEL_0b2578a429ed4a4bb921d142f6f65d5d"
      }
     },
     "080ccde7e0914b9994899d60db0f736f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "097e2928515e47f7b0968a85606b463c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_7549d428e22044c9b1a5b009e794dd2f",
        "IPY_MODEL_9a7be6a8c79e4449923944591006a5d7"
       ],
       "layout": "IPY_MODEL_b125499a1a5049eb88ab0fe859e92824"
      }
     },
     "0a94001be2704d3383d20a0853f4f080": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_1521394fc3254c38994aee9345d26e72",
       "style": "IPY_MODEL_583780ec7fdd41728f2adf2a6850c3c3"
      }
     },
     "0b2578a429ed4a4bb921d142f6f65d5d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "0b456dd0e7db4484bc1f034b061f889b": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "0bd3f4aa76eb4929a3d8fd47164c4ffb": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "0c18c8611c2841aabfde113acaab4b07": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "0da368d400be4eb687072d2cf4051a57": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "0e4e5cf5ea5a43d99a81f4e3fe11d17e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_da48d9536d914f628c5d373d36470ecc",
       "max": 10,
       "style": "IPY_MODEL_0bd3f4aa76eb4929a3d8fd47164c4ffb",
       "value": 10
      }
     },
     "0ed3ebe63f1e42d4b57a1dbfa141bdd4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "1002b4713f864cf3b512e00728fae252": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "11aa407a7bca4002880a0927ee50f602": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "11c4a7ce156646f78f92838e72a1c470": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "14241462acd646d08c12fd60c84ffead": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_c3958eb5280d4e3b80d95ee5af4e131c",
       "style": "IPY_MODEL_4a9f70c9b6a747a99b61623933ac455e"
      }
     },
     "14c5a86cacc84723b50e45aac7e16056": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "14f32380990840b6a9b4673cdf481d3d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_4b210829d2f04cd889f7c68dc87b187a",
       "max": 10,
       "style": "IPY_MODEL_4283de7052df4c258a612b325d94147b",
       "value": 10
      }
     },
     "1521394fc3254c38994aee9345d26e72": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "16bd9b69d96646f380f831521c0fb137": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "17faab11d9c3494180be106697044df7": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "184dd4e8a26f440eb0212240df6d7f97": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "19ba0eb2c837469e89aa46eb443fe703": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "1e2e5148396643d7baada81e73c8ea52": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "1e66a41b0412430ab7ddc5d2ca50ea87": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_3176ef6be7ce4a429436129a8a3bb268",
       "style": "IPY_MODEL_ee5b2e43a0c04c7f9d94be2cb36e44ad",
       "value": " 10/10 [00:48&lt;00:00,  4.80s/it]"
      }
     },
     "1edc99628d1f4d508a88cf73ca0b8656": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "1ff104c6ea2f4c03919d466cffdd22d3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "2043115274d44ad1888f926e94ccbd93": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "20eeaace6f6542189d7aeccdcba04326": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_41e295c2db1944bb8101a117f49d54bc",
       "style": "IPY_MODEL_0da368d400be4eb687072d2cf4051a57"
      }
     },
     "223db28c4dca47158d7010644b48493f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "223e866855f84ec08e23f9c090c5bbe3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_653775d376e64a11a7a319308a84fe67",
        "IPY_MODEL_5c75a8d4493b46e7ad7880313e97a9d3"
       ],
       "layout": "IPY_MODEL_bffe466df630419683bd42b9c80faf13"
      }
     },
     "23279dd34a3148309701c67a07ad473e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_6e6e94729eca4fc698082168ccb07da0",
        "IPY_MODEL_55f31af19a76470098101853aeff3bcd"
       ],
       "layout": "IPY_MODEL_c1b4f2af95e54e9f967016f8204e7039"
      }
     },
     "25a464ad8cea41648ba31ec7b4c6a364": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "27ac5fbd17c4415488d81edd23486142": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "28ea254d2f714a6b96d8e261dd325c35": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "2c149ef4318c42749e188e17cb08808f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "2e2e2e96339a4d9897fa02dacf19ad7b": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_bbda126ca20e4347ad591eb3901311e1",
        "IPY_MODEL_a2f3bceb1ac44b3086355637971c689d"
       ],
       "layout": "IPY_MODEL_d89981a971734601a496de44111fe02f"
      }
     },
     "2ef1a0d867cf4eb99c53f2c6cf1fe348": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "30a08515e11a483bb86ee630b184db26": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_af19cc529ecd4c048ad1ad1e665d7384",
        "IPY_MODEL_88a95a0de3244951aebf45694496c304"
       ],
       "layout": "IPY_MODEL_9a2c995bbd73449ebdfeef9bb352c995"
      }
     },
     "30c7a00f781449829a3e74e84aa950d6": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "3176ef6be7ce4a429436129a8a3bb268": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "323d4b0f80854ffb967a477b2a52c1ce": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_bbfd4470204341bdb40b8805a7bb21a7",
        "IPY_MODEL_0a94001be2704d3383d20a0853f4f080"
       ],
       "layout": "IPY_MODEL_4bb15e8832f4453e80c9c0d805afbc05"
      }
     },
     "32b4685f504a4d36a8932c9aae86adde": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "32c65ef25cfc4250bb4ebdd4d006d701": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_842ab4855b064ad8ab3d86b06f73ded4",
       "style": "IPY_MODEL_e11d6dcee8c047959b8e41796cdc4479"
      }
     },
     "32f88a8a121d4473a41903577b29b23d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "3351aafe7137456c95d149ddf29866d3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "3422f43ce62b4596b471db348dcfb05a": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "350b931058e94bfb8458b4405b01c277": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_4604de51cb124e29addb52184aa95b10",
       "style": "IPY_MODEL_c482f117c8ab48f7b9c0a892d3bebd3f"
      }
     },
     "3a8bb8ff50fc441a88c702ae5144e214": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "3c4ef24d7f2542d5a551874ca7cd4ca3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_45fcea55141040d69450754b06e6c197",
       "style": "IPY_MODEL_14c5a86cacc84723b50e45aac7e16056",
       "value": " 10/10 [00:49&lt;00:00,  4.93s/it]"
      }
     },
     "3cff6f7850ac4bc8a8dfceb3d102514f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "3d8269b2eb6c4f90a876703adfc28429": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "3eca586fcc70482db6335c4a40412e97": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "40b7273c8f0c4e75b3129a06be5bc868": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_e179c10d58e74db6b1a3454c58846263",
        "IPY_MODEL_7ad65609775c49eb84566edce8e6461e"
       ],
       "layout": "IPY_MODEL_af126115b01b42f0b3bb19df29c2850f"
      }
     },
     "415cedc22060403483f5356d93286e5e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "41e295c2db1944bb8101a117f49d54bc": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "4283de7052df4c258a612b325d94147b": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "43db2286b77b4c5ba69630e5339faaf6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_e3c63d70c8984bc0a34e3fe824bcb671",
       "style": "IPY_MODEL_a6a234da915048dd9c9c20d25e14cbd5"
      }
     },
     "45fcea55141040d69450754b06e6c197": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "4604de51cb124e29addb52184aa95b10": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "474398b8f23a474aa5b78f013dc722f4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "4844a7597149419599387b83346b15cc": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_dd7d02aac9b1441d878c7239c9b7a664",
       "style": "IPY_MODEL_934ac6d7b701447491349ccefaa74422",
       "value": " 1/1 [00:03&lt;00:00,  3.32s/it]"
      }
     },
     "4889cbd6ac324f7da9b9b6b9ed9ee412": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "4a276c62e3ae495fa4b61ea32d9881f7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.575: 100%",
       "layout": "IPY_MODEL_415cedc22060403483f5356d93286e5e",
       "max": 10,
       "style": "IPY_MODEL_28ea254d2f714a6b96d8e261dd325c35",
       "value": 10
      }
     },
     "4a9f70c9b6a747a99b61623933ac455e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "4b210829d2f04cd889f7c68dc87b187a": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "4bb15e8832f4453e80c9c0d805afbc05": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "4caf89a2ede343d39be85f444e3aa774": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "4cc3a7e19b70453f809bb4f50573dfab": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_79ae0a1ea9724c249cbb4d180919f127",
       "style": "IPY_MODEL_82c4342e869b4fc483009371c204d164",
       "value": " 1/1 [00:05&lt;00:00,  5.94s/it]"
      }
     },
     "50aea63a95ec4d6394db3c6d32dc4ca3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "532eba32202943eda6b164ef6f09845d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "536345f27dbf46c1a85ff2a130c289a2": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_f061ac6f3c6d4304bfecd60ca1300228",
        "IPY_MODEL_5d989d7971e749d5b91eb79f6727f0b6"
       ],
       "layout": "IPY_MODEL_cceed64d997d40daa3a9450e11c6ff8d"
      }
     },
     "55f31af19a76470098101853aeff3bcd": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_e5f576c210de4196a426e151a431b136",
       "style": "IPY_MODEL_5aad68ade756443c90303239ba47673f",
       "value": " 10/10 [01:07&lt;00:00,  6.74s/it]"
      }
     },
     "56287f9c87c84d66966262e75ef0264d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "583780ec7fdd41728f2adf2a6850c3c3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "5a7c421263934033bd16638773c27577": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_752f4b0d41dc475d8936dea35087ffdb",
       "style": "IPY_MODEL_532eba32202943eda6b164ef6f09845d"
      }
     },
     "5aad68ade756443c90303239ba47673f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "5aed99ec8a66402cbc4e76cecac791cd": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "5c75a8d4493b46e7ad7880313e97a9d3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_8dd0b95eedd544a0ac5350b49357a9d9",
       "style": "IPY_MODEL_b84efe6c57d54f268dccd6d89d2820e0",
       "value": " 2/2 [02:38&lt;00:00, 79.04s/it]"
      }
     },
     "5d989d7971e749d5b91eb79f6727f0b6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_2043115274d44ad1888f926e94ccbd93",
       "style": "IPY_MODEL_7371fd393dc84b5e808622e428ca9070",
       "value": " 10/10 [00:47&lt;00:00,  4.80s/it]"
      }
     },
     "5eab55afbc5c42509da7ea008f655f1f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_4caf89a2ede343d39be85f444e3aa774",
       "style": "IPY_MODEL_e5fe8ba25ff84c449a0d2327b59ab1b1",
       "value": " 10/10 [01:09&lt;00:00,  6.91s/it]"
      }
     },
     "60f3be2e238746fd8336c2a75bbf303e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "62d84c6900134dd1a14d2d602448de27": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_40b7273c8f0c4e75b3129a06be5bc868",
        "IPY_MODEL_43db2286b77b4c5ba69630e5339faaf6"
       ],
       "layout": "IPY_MODEL_f14281a939f142df9c7ab9970cf6de18"
      }
     },
     "638ff7d2e5be4d31b404dbfde0ef1b2c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_23279dd34a3148309701c67a07ad473e",
        "IPY_MODEL_5a7c421263934033bd16638773c27577"
       ],
       "layout": "IPY_MODEL_c1dcd385e6bf495bb1468fddee25c8ac"
      }
     },
     "63a8e321911a4af7897583ffb286b092": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "6518397640c64b27b29413d43c664300": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "653775d376e64a11a7a319308a84fe67": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_9d721df99eff428f95806884a760ceba",
       "max": 2,
       "style": "IPY_MODEL_8b5841bcb7974afbb2edbf07bbbf4a7a",
       "value": 2
      }
     },
     "667bc2f4c288461892d6230965e01d30": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_0359f302ab3e4382b47d9a364c7ae4e2",
       "style": "IPY_MODEL_3a8bb8ff50fc441a88c702ae5144e214",
       "value": " 10/10 [01:03&lt;00:00,  6.38s/it]"
      }
     },
     "688e84420ff84e45bc707b3af528ce32": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_e032d6044823418f8d5086495bf93089",
       "style": "IPY_MODEL_6aca285a7f6949d0b34b2287db0e5907",
       "value": " 10/10 [01:02&lt;00:00,  6.29s/it]"
      }
     },
     "695fe3f0f7d040c689b93f90d17060ec": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "6aca285a7f6949d0b34b2287db0e5907": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "6b35757838b64911bae05412d8c6bc72": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_7fb1216f95634195ac300f1167c0c0c8",
        "IPY_MODEL_688e84420ff84e45bc707b3af528ce32"
       ],
       "layout": "IPY_MODEL_11c4a7ce156646f78f92838e72a1c470"
      }
     },
     "6b8990dfac4f4ebeb0e028f0b802c2c9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_32f88a8a121d4473a41903577b29b23d",
       "style": "IPY_MODEL_474398b8f23a474aa5b78f013dc722f4",
       "value": " 10/10 [00:46&lt;00:00,  4.65s/it]"
      }
     },
     "6d51e65254cb4141b1615aeb48529e7a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_223db28c4dca47158d7010644b48493f",
       "style": "IPY_MODEL_3eca586fcc70482db6335c4a40412e97"
      }
     },
     "6e6e94729eca4fc698082168ccb07da0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_801fdfc8106645588ff4a38ade119541",
       "max": 10,
       "style": "IPY_MODEL_e8900689f672439f8374fd453062a3a9",
       "value": 10
      }
     },
     "6eb684a2906c40d39ef8ea13aa4da710": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "6ebb1276a8fa46d08878e41991be7e9a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_9f312d8d15794c5892aff01d2bfdaa02",
        "IPY_MODEL_350b931058e94bfb8458b4405b01c277"
       ],
       "layout": "IPY_MODEL_2c149ef4318c42749e188e17cb08808f"
      }
     },
     "6f81326a4ef6465692df5a5961c10c24": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_714ff3333b604400a5fefc106cd02b78",
        "IPY_MODEL_b481efeae0c34bfe8027cc072cc5aef5"
       ],
       "layout": "IPY_MODEL_9e76e2ccc5224940b95ff9721c04b21b"
      }
     },
     "714ff3333b604400a5fefc106cd02b78": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_fde52457d9da460186f749cd6c3022a8",
        "IPY_MODEL_4cc3a7e19b70453f809bb4f50573dfab"
       ],
       "layout": "IPY_MODEL_7cd247e365354b23947dd3ed7afa55ee"
      }
     },
     "71a6a3ed39cb49048217582d33c0ad97": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7371fd393dc84b5e808622e428ca9070": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "747be47723d24215a5d8ad46cc078d0b": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_0e4e5cf5ea5a43d99a81f4e3fe11d17e",
        "IPY_MODEL_3c4ef24d7f2542d5a551874ca7cd4ca3"
       ],
       "layout": "IPY_MODEL_1e2e5148396643d7baada81e73c8ea52"
      }
     },
     "752f4b0d41dc475d8936dea35087ffdb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7549d428e22044c9b1a5b009e794dd2f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_75f3c0398ed742cdb313c4d8d9d9dbb4",
        "IPY_MODEL_d376dfb01a77437b94fa7eadf24d978a"
       ],
       "layout": "IPY_MODEL_16bd9b69d96646f380f831521c0fb137"
      }
     },
     "75f3c0398ed742cdb313c4d8d9d9dbb4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_9ca037db17204b518143a0d933930caf",
       "max": 40895,
       "style": "IPY_MODEL_d3fa816ee0b34f038ffe229fed320044",
       "value": 40895
      }
     },
     "762ca5246b464081a55b3c0c2a46992e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7930b3ae473145e2849a4939b0999fea": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "79ae0a1ea9724c249cbb4d180919f127": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "79c2dabbcef64214a22f1a303e8f4436": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "79f424c3238f4e0fb5fe93fa8142d9fc": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7ad65609775c49eb84566edce8e6461e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_63a8e321911a4af7897583ffb286b092",
       "style": "IPY_MODEL_50aea63a95ec4d6394db3c6d32dc4ca3",
       "value": " 1/1 [00:06&lt;00:00,  6.06s/it]"
      }
     },
     "7b52fe54a31d44849295f0d32e05dfde": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "7c26fee0e0ce42108b1a98ac6b775034": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_2e2e2e96339a4d9897fa02dacf19ad7b",
        "IPY_MODEL_edec649418354e139ff51e370727c1d9"
       ],
       "layout": "IPY_MODEL_71a6a3ed39cb49048217582d33c0ad97"
      }
     },
     "7cd247e365354b23947dd3ed7afa55ee": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "7fb1216f95634195ac300f1167c0c0c8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.725: 100%",
       "layout": "IPY_MODEL_30c7a00f781449829a3e74e84aa950d6",
       "max": 10,
       "style": "IPY_MODEL_7930b3ae473145e2849a4939b0999fea",
       "value": 10
      }
     },
     "801fdfc8106645588ff4a38ade119541": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "805fcf5d01034fe6a2c68104aefef421": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_4a276c62e3ae495fa4b61ea32d9881f7",
        "IPY_MODEL_1e66a41b0412430ab7ddc5d2ca50ea87"
       ],
       "layout": "IPY_MODEL_6518397640c64b27b29413d43c664300"
      }
     },
     "82c4342e869b4fc483009371c204d164": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "842ab4855b064ad8ab3d86b06f73ded4": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "85f48c991e36417fafe41c4f5ec76bff": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "88458be8561440b7b9cd84bcb29a5caa": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "88a95a0de3244951aebf45694496c304": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_e1723ca6fc9446f0a8aebae14106682d",
       "style": "IPY_MODEL_60f3be2e238746fd8336c2a75bbf303e",
       "value": " 2/2 [03:15&lt;00:00, 97.64s/it]"
      }
     },
     "8a585d2751464ad4a6e3a1431b0821c0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_d61ef3b9f427472fa54b3dedda57dfe2",
       "max": 40895,
       "style": "IPY_MODEL_c36d4d350bb446c7a364c315e3601a4c",
       "value": 40895
      }
     },
     "8b5841bcb7974afbb2edbf07bbbf4a7a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "8d699d515a0e4d1982074e8acffeba9d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_b0f4e1b7de13402faadcc3fba8118f05",
        "IPY_MODEL_32c65ef25cfc4250bb4ebdd4d006d701"
       ],
       "layout": "IPY_MODEL_0c18c8611c2841aabfde113acaab4b07"
      }
     },
     "8dd0b95eedd544a0ac5350b49357a9d9": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "8f3231ea2d7f483a844c9d5dde0234b7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_f7a00ee41aaf40188ae6268744367afb",
       "style": "IPY_MODEL_fae4f7513c6745eca92df0197b0b8c07"
      }
     },
     "913762445a3d4fada78ca24b2a397d1e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.662: 100%",
       "layout": "IPY_MODEL_a4d7def7985241968adb9270a2f325d3",
       "max": 10,
       "style": "IPY_MODEL_0ed3ebe63f1e42d4b57a1dbfa141bdd4",
       "value": 10
      }
     },
     "934ac6d7b701447491349ccefaa74422": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "93c50e67c55c40779b625949df168c79": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_f09d3bfeb9a24b49b3105b2f166a70ee",
        "IPY_MODEL_e77cee995ab14b788ebb76074f805354"
       ],
       "layout": "IPY_MODEL_1edc99628d1f4d508a88cf73ca0b8656"
      }
     },
     "9a2c995bbd73449ebdfeef9bb352c995": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "9a7be6a8c79e4449923944591006a5d7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_2ef1a0d867cf4eb99c53f2c6cf1fe348",
       "style": "IPY_MODEL_4889cbd6ac324f7da9b9b6b9ed9ee412"
      }
     },
     "9ca037db17204b518143a0d933930caf": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "flex": "2"
      }
     },
     "9d721df99eff428f95806884a760ceba": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "9e76e2ccc5224940b95ff9721c04b21b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "9f312d8d15794c5892aff01d2bfdaa02": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_913762445a3d4fada78ca24b2a397d1e",
        "IPY_MODEL_667bc2f4c288461892d6230965e01d30"
       ],
       "layout": "IPY_MODEL_85f48c991e36417fafe41c4f5ec76bff"
      }
     },
     "a2f3bceb1ac44b3086355637971c689d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_b8bfc8f5d68d4837bb04806f04f53b42",
       "style": "IPY_MODEL_c5fc4679c9724eaa8c79126527b4f218",
       "value": " 10/10 [00:46&lt;00:00,  4.69s/it]"
      }
     },
     "a4d7def7985241968adb9270a2f325d3": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "a6a234da915048dd9c9c20d25e14cbd5": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "ad9cbf843d9f4d4f89ab133c9f7e2e5d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_d35dbf13e7254effad04761352f143d2",
       "style": "IPY_MODEL_0b456dd0e7db4484bc1f034b061f889b"
      }
     },
     "af126115b01b42f0b3bb19df29c2850f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "af19cc529ecd4c048ad1ad1e665d7384": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_f6158ae51d894ce5bd7c8da704a596e6",
       "max": 2,
       "style": "IPY_MODEL_7b52fe54a31d44849295f0d32e05dfde",
       "value": 2
      }
     },
     "b0f4e1b7de13402faadcc3fba8118f05": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_f9d01b5ef9bd47cb8df833979372e032",
        "IPY_MODEL_4844a7597149419599387b83346b15cc"
       ],
       "layout": "IPY_MODEL_11aa407a7bca4002880a0927ee50f602"
      }
     },
     "b125499a1a5049eb88ab0fe859e92824": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "display": "inline-flex",
       "flex_flow": "row wrap",
       "width": "100%"
      }
     },
     "b3b48d7bfafc427ebbb1c6a2ec633241": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_805fcf5d01034fe6a2c68104aefef421",
        "IPY_MODEL_ca8378bfbc3249b192e1fa377c303ac9"
       ],
       "layout": "IPY_MODEL_6eb684a2906c40d39ef8ea13aa4da710"
      }
     },
     "b481efeae0c34bfe8027cc072cc5aef5": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_cf6517649c564070a1cc4895854bb5af",
       "style": "IPY_MODEL_3d8269b2eb6c4f90a876703adfc28429"
      }
     },
     "b4c578bf705641f584d9f13581fff92d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "b6afcefc627d4fb3b4edf32d56db1f87": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "b84efe6c57d54f268dccd6d89d2820e0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "b86a793be19941589e007912248b01a3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_bd87276554ca4584a3e261dd362b9119",
        "IPY_MODEL_8f3231ea2d7f483a844c9d5dde0234b7"
       ],
       "layout": "IPY_MODEL_c33f23e3f336441cad84f52e8c186fda"
      }
     },
     "b8bfc8f5d68d4837bb04806f04f53b42": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "bbda126ca20e4347ad591eb3901311e1": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.694: 100%",
       "layout": "IPY_MODEL_56287f9c87c84d66966262e75ef0264d",
       "max": 10,
       "style": "IPY_MODEL_184dd4e8a26f440eb0212240df6d7f97",
       "value": 10
      }
     },
     "bbfd4470204341bdb40b8805a7bb21a7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_14f32380990840b6a9b4673cdf481d3d",
        "IPY_MODEL_5eab55afbc5c42509da7ea008f655f1f"
       ],
       "layout": "IPY_MODEL_17faab11d9c3494180be106697044df7"
      }
     },
     "bd87276554ca4584a3e261dd362b9119": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_8a585d2751464ad4a6e3a1431b0821c0",
        "IPY_MODEL_c0a56e73b7e44e659571996ab7662678"
       ],
       "layout": "IPY_MODEL_de81fd5afb0f47098d920909bb9420fc"
      }
     },
     "bffe466df630419683bd42b9c80faf13": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "c0a56e73b7e44e659571996ab7662678": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_32b4685f504a4d36a8932c9aae86adde",
       "style": "IPY_MODEL_3cff6f7850ac4bc8a8dfceb3d102514f",
       "value": " 40895/40895 [00:02&lt;00:00, 16861.19KB/s]"
      }
     },
     "c1b4f2af95e54e9f967016f8204e7039": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "c1dcd385e6bf495bb1468fddee25c8ac": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "c2b8722965e44893a910dd409ee19b18": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "c33f23e3f336441cad84f52e8c186fda": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "display": "inline-flex",
       "flex_flow": "row wrap",
       "width": "100%"
      }
     },
     "c36d4d350bb446c7a364c315e3601a4c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "c3958eb5280d4e3b80d95ee5af4e131c": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "c482f117c8ab48f7b9c0a892d3bebd3f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "c5fc4679c9724eaa8c79126527b4f218": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "c6acf75533704a2fa955cea2c9b95a2f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_536345f27dbf46c1a85ff2a130c289a2",
        "IPY_MODEL_ad9cbf843d9f4d4f89ab133c9f7e2e5d"
       ],
       "layout": "IPY_MODEL_b6afcefc627d4fb3b4edf32d56db1f87"
      }
     },
     "c7ec84601cd542efa619dd1aec58d0ee": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "c8c3f13ac01b40afa89a911cbf3f2ec2": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_07adeeb6d28e4fc9bc3d87c3a340943c",
        "IPY_MODEL_6d51e65254cb4141b1615aeb48529e7a"
       ],
       "layout": "IPY_MODEL_d3b11e95c0484ddba3dfa7bf3510ecef"
      }
     },
     "ca8378bfbc3249b192e1fa377c303ac9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_79f424c3238f4e0fb5fe93fa8142d9fc",
       "style": "IPY_MODEL_3351aafe7137456c95d149ddf29866d3"
      }
     },
     "cceed64d997d40daa3a9450e11c6ff8d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "cd3247af1e93404888165b76d6b4859c": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "ce4a60ef24b942349e13921789ccdf57": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "cf6517649c564070a1cc4895854bb5af": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "d35dbf13e7254effad04761352f143d2": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "d376dfb01a77437b94fa7eadf24d978a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_695fe3f0f7d040c689b93f90d17060ec",
       "style": "IPY_MODEL_1002b4713f864cf3b512e00728fae252",
       "value": " 40895/40895 [00:01&lt;00:00, 28117.66KB/s]"
      }
     },
     "d3b11e95c0484ddba3dfa7bf3510ecef": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "d3fa816ee0b34f038ffe229fed320044": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "d61ef3b9f427472fa54b3dedda57dfe2": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "flex": "2"
      }
     },
     "d89981a971734601a496de44111fe02f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "da48d9536d914f628c5d373d36470ecc": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "dd7d02aac9b1441d878c7239c9b7a664": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "de81fd5afb0f47098d920909bb9420fc": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "dffc8332bda8448f8c233a24d232580a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_6b35757838b64911bae05412d8c6bc72",
        "IPY_MODEL_14241462acd646d08c12fd60c84ffead"
       ],
       "layout": "IPY_MODEL_762ca5246b464081a55b3c0c2a46992e"
      }
     },
     "e032d6044823418f8d5086495bf93089": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e04283da29ee433a90ceb57817a3cf6c": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e11d6dcee8c047959b8e41796cdc4479": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "e1723ca6fc9446f0a8aebae14106682d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e179c10d58e74db6b1a3454c58846263": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "accuracy: 0.875: 100%",
       "layout": "IPY_MODEL_c2b8722965e44893a910dd409ee19b18",
       "max": 1,
       "style": "IPY_MODEL_88458be8561440b7b9cd84bcb29a5caa",
       "value": 1
      }
     },
     "e3c63d70c8984bc0a34e3fe824bcb671": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e5f576c210de4196a426e151a431b136": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e5fe8ba25ff84c449a0d2327b59ab1b1": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "e6ac4fcf81e548088bca97c461164160": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "e77cee995ab14b788ebb76074f805354": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_e04283da29ee433a90ceb57817a3cf6c",
       "style": "IPY_MODEL_27ac5fbd17c4415488d81edd23486142",
       "value": " 2/2 [00:16&lt;00:00,  8.23s/it]"
      }
     },
     "e8900689f672439f8374fd453062a3a9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "edec649418354e139ff51e370727c1d9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "layout": "IPY_MODEL_080ccde7e0914b9994899d60db0f736f",
       "style": "IPY_MODEL_79c2dabbcef64214a22f1a303e8f4436"
      }
     },
     "ee5b2e43a0c04c7f9d94be2cb36e44ad": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "f061ac6f3c6d4304bfecd60ca1300228": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "[Epoch 10] Validation: 0.706: 100%",
       "layout": "IPY_MODEL_b4c578bf705641f584d9f13581fff92d",
       "max": 10,
       "style": "IPY_MODEL_5aed99ec8a66402cbc4e76cecac791cd",
       "value": 10
      }
     },
     "f09d3bfeb9a24b49b3105b2f166a70ee": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "100%",
       "layout": "IPY_MODEL_ce4a60ef24b942349e13921789ccdf57",
       "max": 2,
       "style": "IPY_MODEL_19ba0eb2c837469e89aa46eb443fe703",
       "value": 2
      }
     },
     "f14281a939f142df9c7ab9970cf6de18": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "f26d731ade734979b3b2103a81d7ab57": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "children": [
        "IPY_MODEL_747be47723d24215a5d8ad46cc078d0b",
        "IPY_MODEL_20eeaace6f6542189d7aeccdcba04326"
       ],
       "layout": "IPY_MODEL_cd3247af1e93404888165b76d6b4859c"
      }
     },
     "f579431b9142422cb602de64f9c41971": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "description_width": "initial"
      }
     },
     "f6158ae51d894ce5bd7c8da704a596e6": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "f7a00ee41aaf40188ae6268744367afb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {}
     },
     "f9d01b5ef9bd47cb8df833979372e032": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "accuracy: 0.90625: 100%",
       "layout": "IPY_MODEL_e6ac4fcf81e548088bca97c461164160",
       "max": 1,
       "style": "IPY_MODEL_1ff104c6ea2f4c03919d466cffdd22d3",
       "value": 1
      }
     },
     "fae4f7513c6745eca92df0197b0b8c07": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "description_width": ""
      }
     },
     "fde52457d9da460186f749cd6c3022a8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "IntProgressModel",
      "state": {
       "bar_style": "success",
       "description": "accuracy: 0.75: 100%",
       "layout": "IPY_MODEL_25a464ad8cea41648ba31ec7b4c6a364",
       "max": 1,
       "style": "IPY_MODEL_c7ec84601cd542efa619dd1aec58d0ee",
       "value": 1
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}