12_intro_to_transfer_learning_resnet_workset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 25611,
     "status": "ok",
     "timestamp": 1646772985491,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "IsnbTwG_hwlE"
   },
   "outputs": [],
   "source": [
    "%%capture\n",
    "!pip install -Uqq fastai gradio\n",
    "!apt-get install tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "executionInfo": {
     "elapsed": 7859,
     "status": "ok",
     "timestamp": 1646772993345,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "_m-HZXy7h3WM"
   },
   "outputs": [],
   "source": [
    "from fastai.vision.all import *\n",
    "from fastai import *\n",
    "from torch import nn, optim\n",
    "import torch\n",
    "\n",
    "from IPython.display import display, YouTubeVideo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1646772993346,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "rcZ2oTV6Oxaf"
   },
   "outputs": [],
   "source": [
    "SIZE = {'width':1000, 'height': 600}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 37
    },
    "executionInfo": {
     "elapsed": 33554,
     "status": "ok",
     "timestamp": 1646773026896,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "_EP_49JmiA38",
    "outputId": "872a32ed-629e-4b62-f837-fa1577329a8a"
   },
   "outputs": [],
   "source": [
    "path = untar_data(URLs.PETS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 14,
     "status": "ok",
     "timestamp": 1646773026897,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "xA_M5Hz5iMZz",
    "outputId": "1e52a172-c5ed-4deb-99a9-fe360c131a16"
   },
   "outputs": [],
   "source": [
    "list(path.ls())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1646773026897,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "YWa54ndv5t4a",
    "outputId": "7119438c-58aa-4fb9-99c7-2a42da600d51"
   },
   "outputs": [],
   "source": [
    "!tree {str(path/'images')} | grep .jpg | shuf -n 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1646773026897,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "zzB5-veFii98"
   },
   "outputs": [],
   "source": [
    "def label_func(filename):\n",
    "    return '_'.join(filename.split('_')[:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "executionInfo": {
     "elapsed": 8874,
     "status": "ok",
     "timestamp": 1646773035766,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "5soie3vziYRa"
   },
   "outputs": [],
   "source": [
    "dls = ImageDataLoaders.from_name_func(\n",
    "    path=path, \n",
    "    fnames=get_image_files(path/'images'), \n",
    "    label_func=label_func, \n",
    "    valid_pct=0.2,\n",
    "    item_tfms=Resize(224),\n",
    "    batch_tfms=Normalize.from_stats(*imagenet_stats),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1646773035766,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "xe_08kADBMVT",
    "outputId": "13904dee-b57e-4558-c767-ef1f76e7a4f8"
   },
   "outputs": [],
   "source": [
    "imagenet_stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 536
    },
    "executionInfo": {
     "elapsed": 2308,
     "status": "ok",
     "timestamp": 1646773038071,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "tAwKJZh7i1AU",
    "outputId": "54132633-75e0-4174-d545-c7718ddc63c3"
   },
   "outputs": [],
   "source": [
    "dls.show_batch()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 621
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1646773038071,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "QhOZ6UrJOoY3",
    "outputId": "7fccce65-1778-4cd8-8ae3-9aeecafb9d06"
   },
   "outputs": [],
   "source": [
    "vid = YouTubeVideo('RYth6EbBUqM', **SIZE)\n",
    "display(vid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "executionInfo": {
     "elapsed": 365,
     "status": "ok",
     "timestamp": 1646773038430,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "AlfjYD-mb2Yd"
   },
   "outputs": [],
   "source": [
    "cbs = [\n",
    "    EarlyStoppingCallback(patience=3)\n",
    "]\n",
    "\n",
    "learn = cnn_learner( # Instantiate a learner object\n",
    "    dls, # Pass the dataloaders\n",
    "    models.resnet34, # Specify the architechture you want to use\n",
    "    pretrained=False, # Tell it NOT to fetch the ImageNet weights\n",
    "    cbs=cbs, # Pass the callbacks\n",
    "    metrics=[accuracy, error_rate] # Pass the metrics you want to see\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1646773038431,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "BnbaPiprNXf8",
    "outputId": "8ca87d3c-bf77-4ca3-a22e-3f26debc984e"
   },
   "outputs": [],
   "source": [
    "print(f\"\"\"\n",
    "There are {learn.dls.c} categories.\n",
    "If we guessed randomly, we would have aboiut a {100. * 1/learn.dls.c:.02f}% chance of guessing correctly.\n",
    "\"\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 244527,
     "status": "ok",
     "timestamp": 1646773282952,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "7FhMki-Ab9rs",
    "outputId": "d9c9c17a-1c06-4405-aada-00585452742d"
   },
   "outputs": [],
   "source": [
    "learn.fit_one_cycle(5, 1e-3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 67,
     "referenced_widgets": [
      "db7dcc47c89f4eaf8f12b2aad0992609",
      "1c8236784aa044dd9c2c607369336357",
      "ce6c8ed5b0b14476b6ef128ee6059d83",
      "2adafd64097e4da4a230994aa21d475f",
      "116e4a5553bf41d290e13a7e3bf160b9",
      "b48073cc65f443a69bf363fe91a895f4",
      "4ba146cd2da247c49af8cadf39496875",
      "798b951131644d80b8bd6d05b81a37f4",
      "2894eaced6184bd7b605eca51851b0e7",
      "287c5eabf4234990bf82fff0a17b4e0a",
      "28da2bd1945648ae9c97a0bf41b76951"
     ]
    },
    "executionInfo": {
     "elapsed": 2481,
     "status": "ok",
     "timestamp": 1646773285430,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "iANdnvUyjG4n",
    "outputId": "a09d5715-b38f-4c3c-a937-af6c61571704"
   },
   "outputs": [],
   "source": [
    "cbs = [\n",
    "    EarlyStoppingCallback(patience=3)\n",
    "]\n",
    "\n",
    "learn = cnn_learner( # Instantiate a learner object\n",
    "    dls, # Pass the dataloaders\n",
    "    models.resnet34, # Specify the architechture you want to use\n",
    "    pretrained=True, # Tell it to fetch the ImageNet weights\n",
    "    cbs=cbs, # Pass the callbacks\n",
    "    metrics=[accuracy, error_rate] # Pass the metrics you want to see\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1646773285430,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "CcdKalRicTgt",
    "outputId": "be6f7c90-1b2a-4434-c9f0-6042e3549317"
   },
   "outputs": [],
   "source": [
    "learn.model[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1646773285431,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "w3bm5jgol9Lm"
   },
   "outputs": [],
   "source": [
    "LR = 1e-3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 80
    },
    "executionInfo": {
     "elapsed": 47420,
     "status": "ok",
     "timestamp": 1646773332846,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "SkLhle4eZb3A",
    "outputId": "361bb3ef-2fdc-4834-aa70-57e5569256c3"
   },
   "outputs": [],
   "source": [
    "# Train the classifier head\n",
    "learn.fit_one_cycle(1, LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "executionInfo": {
     "elapsed": 278,
     "status": "ok",
     "timestamp": 1646773333122,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "HulxNy3TZj5r"
   },
   "outputs": [],
   "source": [
    "# Unfreeze the backbone\n",
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 245290,
     "status": "ok",
     "timestamp": 1646773578407,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "b0_FtS9BZl6b",
    "outputId": "04a944c7-ab25-402f-8b52-04be519c561c"
   },
   "outputs": [],
   "source": [
    "# Trainin everything together\n",
    "learn.fit_one_cycle(5, lr_max=slice(LR/10, LR))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "executionInfo": {
     "elapsed": 424,
     "status": "ok",
     "timestamp": 1646773578829,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "mJd2XiUgbmRn"
   },
   "outputs": [],
   "source": [
    "cbs = [\n",
    "    EarlyStoppingCallback(patience=3)\n",
    "]\n",
    "\n",
    "learn = cnn_learner( # Instantiate a learner object\n",
    "    dls, # Pass the dataloaders\n",
    "    resnet34, # Specify the architechture you want to use\n",
    "    pretrained=True, # Tell it to fetch the ImageNet weights\n",
    "    # loss_func=nn.CrossEntropyLoss(), # Pass the loss function\n",
    "    cbs=cbs, # Pass the callbacks\n",
    "    metrics=[accuracy, error_rate] # Pass the metrics you want to see\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 270
    },
    "executionInfo": {
     "elapsed": 291244,
     "status": "ok",
     "timestamp": 1646773870068,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "vTzn6ZrZmIog",
    "outputId": "db5f2a5e-465c-49ec-8ce7-69442be5e5e1"
   },
   "outputs": [],
   "source": [
    "learn.fine_tune(5, base_lr=LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "executionInfo": {
     "elapsed": 2755,
     "status": "ok",
     "timestamp": 1646773872820,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "qgs-iiToWbez"
   },
   "outputs": [],
   "source": [
    "import gradio as gr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 368
    },
    "executionInfo": {
     "elapsed": 12,
     "status": "ok",
     "timestamp": 1646773872824,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "k4x912f1nh-L",
    "outputId": "ad72d4c5-db79-45a6-f9bd-de2d4ba528de"
   },
   "outputs": [],
   "source": [
    "# grab one image from our training dataset\n",
    "img, cat = dls.train_ds[0]\n",
    "print(cat)\n",
    "img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 70
    },
    "executionInfo": {
     "elapsed": 385,
     "status": "ok",
     "timestamp": 1646773873202,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "qHQXZGT3rJ-h",
    "outputId": "9fb2c97f-e036-4d6b-fcad-74d8c8a5e3c3"
   },
   "outputs": [],
   "source": [
    "breed, cat_id, logits = learn.predict(img)\n",
    "print(breed, cat_id, sep='\\n')\n",
    "print(logits.sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 88
    },
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1646773873202,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "ASX1sQTMWM0p",
    "outputId": "75585f1f-d277-4f8e-a27a-7faa37219258"
   },
   "outputs": [],
   "source": [
    "def recognize_pet(img):\n",
    "    breed, cat_id, logits = learn.predict(img)\n",
    "    proba = float(logits.softmax(dim=0).max())\n",
    "    return f\"\"\"\n",
    "    Breed: {breed},\n",
    "    Probability: {proba:.04f}\n",
    "    \"\"\"\n",
    "\n",
    "print(recognize_pet(img))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sGD1-ZqavcfC"
   },
   "source": [
    "# Exercise 12.1\n",
    "\n",
    "Run the block of code below.\n",
    "Find a few images of pets to upload and observe the inferences made by the model.\n",
    "Do they work decently well?\n",
    "\n",
    "<!-- startquestion -->"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 646
    },
    "executionInfo": {
     "elapsed": 5093,
     "status": "ok",
     "timestamp": 1646773878291,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "Kr39vH55WeDr",
    "outputId": "054b47ed-32a0-4276-b611-da41abaa70f6"
   },
   "outputs": [],
   "source": [
    "gr.Interface(recognize_pet, inputs=gr.inputs.Image(source='upload'), outputs=gr.outputs.Label()).launch(inline=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1646773878291,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "e5zjV24_3Dh3"
   },
   "outputs": [],
   "source": [
    "def conv_block(in_channels, out_channels, stride=1):\n",
    "    return nn.Sequential(\n",
    "        nn.Conv2d(in_channels, out_channels, kernel_size=3, padding=1, stride=stride),\n",
    "        nn.PReLU(),\n",
    "        nn.BatchNorm2d(out_channels),\n",
    "        nn.Dropout(0.2)\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1646773878291,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "k8pvlWqMWKY1"
   },
   "outputs": [],
   "source": [
    "class ResidualConvBlock(nn.Module):\n",
    "    def __init__(self, in_channels, out_channels, stride=1):\n",
    "        super().__init__()\n",
    "        # If the channels or output shape are different, \n",
    "        # we need a projection so the channels match\n",
    "        self.projection = True if in_channels != out_channels or stride > 1 else False\n",
    "        # The projection is just a 1x1 convolutional layer that changes\n",
    "        # the number of channels to match out_channels.\n",
    "        if self.projection:\n",
    "            self.proj = nn.Conv2d(\n",
    "                in_channels, \n",
    "                out_channels, \n",
    "                kernel_size=1, \n",
    "                padding=0,\n",
    "                stride=stride # If the shape changes, this is one way to downsample spatially.\n",
    "            )\n",
    "\n",
    "        # Now just add the easy part of the network\n",
    "        self.conv1 = conv_block(in_channels, out_channels, stride)\n",
    "        self.conv2 = conv_block(out_channels, out_channels)\n",
    "        self.final_act = nn.PReLU()\n",
    "\n",
    "    def forward(self, x):\n",
    "        # Declare the identity\n",
    "        identity = x\n",
    "        # but if we need to do the projection, project x \n",
    "        # into a different number of channels.\n",
    "        if self.projection:\n",
    "            identity = self.proj(x)\n",
    "        # Pass the original x through the conv network\n",
    "        acts = self.conv2(self.conv1(x))\n",
    "        # Return the identity + the activations of the conv network\n",
    "        return self.final_act(identity + acts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1646773878292,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "ualYKVuU4yrZ"
   },
   "outputs": [],
   "source": [
    "fake_batch = torch.randn(2, 32, 28, 28)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1646773878292,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "HDoWCgt2xrSJ",
    "outputId": "faf9e860-5550-4ffd-8cc9-685d94f817d9"
   },
   "outputs": [],
   "source": [
    "res = ResidualConvBlock(32, 32)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 18,
     "status": "ok",
     "timestamp": 1646773878293,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "iiMSFHGV9wYA",
    "outputId": "48eba722-5b09-4a44-b61e-654d36ed1942"
   },
   "outputs": [],
   "source": [
    "res(fake_batch).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1646773878293,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "J8h_ejeP-UfS",
    "outputId": "30308514-d907-461e-b32c-667392e46ed8"
   },
   "outputs": [],
   "source": [
    "res = ResidualConvBlock(32, 64)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 13,
     "status": "ok",
     "timestamp": 1646773878293,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "E35p8IaJePpq",
    "outputId": "9945e70e-03f0-4e05-bf15-315a59a7b407"
   },
   "outputs": [],
   "source": [
    "res(fake_batch).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 13,
     "status": "ok",
     "timestamp": 1646773878294,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "t2BlTbMAeGCy",
    "outputId": "7757d6bb-d605-40dd-a8d2-0ac29adab631"
   },
   "outputs": [],
   "source": [
    "res = ResidualConvBlock(32, 64, 2)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1646773878294,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "ARDqU4dneLpt",
    "outputId": "33ed8770-f4d2-4ca2-b6f7-28e2e564e37c"
   },
   "outputs": [],
   "source": [
    "res(fake_batch).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1646773878294,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "Ds5s_mAZez_t"
   },
   "outputs": [],
   "source": [
    "def build_model(*args, **kwargs):\n",
    "    model = nn.Sequential(\n",
    "        nn.BatchNorm2d(3),\n",
    "        ResidualConvBlock(3, 16),\n",
    "        ResidualConvBlock(16,16),\n",
    "        ResidualConvBlock(16,32,2),\n",
    "        ResidualConvBlock(32, 64),\n",
    "        nn.AdaptiveMaxPool2d(1),\n",
    "        nn.Flatten(),\n",
    "        nn.Linear(64, 256),\n",
    "        nn.BatchNorm1d(256),\n",
    "        nn.ReLU(),\n",
    "        nn.Dropout(0.2),\n",
    "        nn.Linear(256, 37)\n",
    "    )\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1646773878295,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "9jP4mw6tfzql"
   },
   "outputs": [],
   "source": [
    "learn = cnn_learner(dls, build_model, pretrained=False, metrics=[accuracy, error_rate])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 676
    },
    "executionInfo": {
     "elapsed": 1035886,
     "status": "ok",
     "timestamp": 1646774914171,
     "user": {
      "displayName": "Mike Frantz",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GiGYCwZXGfGg-CE4_33PQcH9SW8fN6MyRkenN8-IA=s64",
      "userId": "08482506601269125384"
     },
     "user_tz": 480
    },
    "id": "rg9siiWcgHoX",
    "outputId": "0110b128-0f21-4f40-bf01-a8121aa7515e"
   },
   "outputs": [],
   "source": [
    "learn.fit_one_cycle(20, lr_max=1e-3)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "authorship_tag": "ABX9TyM1kBnFvy6Fyd23tWg0qDhV",
   "collapsed_sections": [],
   "name": "12_intro_to_transfer_learning_resnet.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "name": "python3"
  },
  "language_info": {
   "name": "python"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "116e4a5553bf41d290e13a7e3bf160b9": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_28da2bd1945648ae9c97a0bf41b76951",
      "placeholder": "​",
      "style": "IPY_MODEL_287c5eabf4234990bf82fff0a17b4e0a",
      "value": " 83.3M/83.3M [00:01&lt;00:00, 53.8MB/s]"
     }
    },
    "1c8236784aa044dd9c2c607369336357": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "287c5eabf4234990bf82fff0a17b4e0a": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "2894eaced6184bd7b605eca51851b0e7": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "28da2bd1945648ae9c97a0bf41b76951": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "2adafd64097e4da4a230994aa21d475f": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_2894eaced6184bd7b605eca51851b0e7",
      "max": 87319819,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_798b951131644d80b8bd6d05b81a37f4",
      "value": 87319819
     }
    },
    "4ba146cd2da247c49af8cadf39496875": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "798b951131644d80b8bd6d05b81a37f4": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "b48073cc65f443a69bf363fe91a895f4": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "ce6c8ed5b0b14476b6ef128ee6059d83": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_4ba146cd2da247c49af8cadf39496875",
      "placeholder": "​",
      "style": "IPY_MODEL_b48073cc65f443a69bf363fe91a895f4",
      "value": "100%"
     }
    },
    "db7dcc47c89f4eaf8f12b2aad0992609": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_ce6c8ed5b0b14476b6ef128ee6059d83",
       "IPY_MODEL_2adafd64097e4da4a230994aa21d475f",
       "IPY_MODEL_116e4a5553bf41d290e13a7e3bf160b9"
      ],
      "layout": "IPY_MODEL_1c8236784aa044dd9c2c607369336357"
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}