Vit_Large_ranger.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Vit_Large_ranger.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyO+wuOkP1TYQLWWXsz99tSV"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6wsfLg2oqb7S"
      },
      "source": [
        "In this notebook, I have tested the performance of a [Vision Transformer](https://arxiv.org/pdf/2010.11929.pdf) on the [ImageWoof](https://github.com/fastai/imagenette#imagewoof) dataset. \n",
        "\n",
        "\n",
        "I have used a pretrained ViT - Large model, with 24 layers and 307M parameters, and have fine - tuned as well as evaluated it on the ImageWoof dataset. The fine - tuning is done using the ranger optimizer. The ranger is based on two seperate papers : (1) [On the Varience of the Adaptive Learning rate and Beyond, RAdam](https://arxiv.org/pdf/1908.03265.pdf) (2) [Lookahead Optimizer: k steps forward, 1 step back](https://arxiv.org/pdf/1907.08610.pdf)\n",
        "\n",
        "I have also customized the ViT - Large model, that can be used in case it is required to freeze some parts of the model (please see the custom_ViT() function in the Model (4th) section).\n",
        "\n",
        "I have used [fastai](https://fastai) for this project."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oTDYNTSCstTB"
      },
      "source": [
        "# 1. **Installing and importing the dependencies**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "snMD_c0jHDkP"
      },
      "source": [
        "!pip install fastai timm nbdev -q --upgrade"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JY44yh80HNFy"
      },
      "source": [
        "from fastai.vision.all import *\n",
        "import torch\n",
        "import timm\n",
        "import math\n",
        "import warnings"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aSU5ZXw4s2_o"
      },
      "source": [
        "# 2. **Downloading the ImageWoof dataset**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_EpWrEucHODS",
        "outputId": "1a32c2a6-64dd-4a8a-ca00-5959b8f3fda3"
      },
      "source": [
        "path = untar_data(URLs.IMAGEWOOF)\n",
        "path.ls()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(#3) [Path('/root/.fastai/data/imagewoof2/val'),Path('/root/.fastai/data/imagewoof2/train'),Path('/root/.fastai/data/imagewoof2/noisy_imagewoof.csv')]"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DSOwys3Vs9Qw"
      },
      "source": [
        "# 3. **Data Preprocessing**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qZ3OXefws_x0"
      },
      "source": [
        "Each image in the dataset is first resized to 300 x 300, using the [Resize()](https://docs.fast.ai/vision.augment.html#Resize) class. Then, while training the model on each batch, each image, in the batch, is further reduced to the size of 224 x 224, alongwith normalizing them with Imagenet statistics, using [aug_transforms](https://docs.fast.ai/vision.augment.html#aug_transforms) and [Normalize](https://docs.fast.ai/data.transforms.html#Normalize), respectively.\n",
        "\n",
        "Batch size is set to 4."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OrIdNbLZHOn1"
      },
      "source": [
        "item_tfms = Resize(size = 300)\n",
        "batch_tfms = [*aug_transforms(size = 224, min_scale = 0.35), Normalize.from_stats(*imagenet_stats)]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cuEbxGNPHPHs"
      },
      "source": [
        "bs = 4"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D0SPk-RmHPsQ"
      },
      "source": [
        "# From \"https://walkwithfastai.com/ImageWoof\"\n",
        "label_to_dict = dict(\n",
        "  n02086240= 'Shih-Tzu',\n",
        "  n02087394= 'Rhodesian ridgeback',\n",
        "  n02088364= 'Beagle',\n",
        "  n02089973= 'English foxhound',\n",
        "  n02093754= 'Australian terrier',\n",
        "  n02096294= 'Border terrier',\n",
        "  n02099601= 'Golden retriever',\n",
        "  n02105641= 'Old English sheepdog',\n",
        "  n02111889= 'Samoyed',\n",
        "  n02115641= 'Dingo'\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UGrpLo9GHPzI"
      },
      "source": [
        "imagewoof = DataBlock(blocks = (ImageBlock, CategoryBlock),\n",
        "                      get_items = get_image_files,\n",
        "                      splitter = GrandparentSplitter(valid_name = 'val'),\n",
        "                      get_y = Pipeline([parent_label, label_to_dict.__getitem__]),\n",
        "                      item_tfms = item_tfms,\n",
        "                      batch_tfms = batch_tfms)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xstst75NHP4-",
        "outputId": "eea51898-3e8f-4d93-b555-3c28c4f2d3a0"
      },
      "source": [
        "dls = imagewoof.dataloaders(path, bs = bs)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/_tensor.py:1023: UserWarning: torch.solve is deprecated in favor of torch.linalg.solveand will be removed in a future PyTorch release.\n",
            "torch.linalg.solve has its arguments reversed and does not return the LU factorization.\n",
            "To get the LU factorization see torch.lu, which can be used with torch.lu_solve or torch.lu_unpack.\n",
            "X = torch.solve(B, A).solution\n",
            "should be replaced with\n",
            "X = torch.linalg.solve(A, B) (Triggered internally at  /pytorch/aten/src/ATen/native/BatchLinearAlgebra.cpp:760.)\n",
            "  ret = func(*args, **kwargs)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 373
        },
        "id": "St3snsu0HP9L",
        "outputId": "1bf9ac21-63bd-447e-ecf0-f67d9162c864"
      },
      "source": [
        "dls.show_batch(max_n = 9)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x432 with 4 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hqqTTl-BtsIq"
      },
      "source": [
        "# 4. **Model**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pj4yrR9stuxL"
      },
      "source": [
        "I have defined a module named EmbedBlock, to concatenate the class token and the input tokens, as well as to add the position embeddings."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2EDYwdYlHQBG"
      },
      "source": [
        "# Heavily inspired by \"https://github.com/rwightman/pytorch-image-models/blob/5f9aff395c224492e9e44248b15f44b5cc095d9c/timm/models/vision_transformer.py\"\n",
        "\n",
        "class EmbedBlock (Module) :\n",
        "  def __init__ (self, num_pathces, embed_dim) :\n",
        "    self.cls_tokens = nn.Parameter(torch.zeros(1, 1, embed_dim))\n",
        "    self.pos_embeds = nn.Parameter(torch.zeros(1, num_pathces + 1, embed_dim))\n",
        "\n",
        "  def forward (self, x) :\n",
        "    B = x.shape[0]\n",
        "    cls_tokens = self.cls_tokens.expand(B, -1, -1)\n",
        "    x = torch.cat((cls_tokens, x), dim = 1)\n",
        "    x = x + self.pos_embeds\n",
        "    return x "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Js0wWkR2t3Of"
      },
      "source": [
        "I have also defined a module Header that acts as the final head of our model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "exsD4DGBHcM8"
      },
      "source": [
        "# Heavily inspired by \"https://github.com/rwightman/pytorch-image-models/blob/5f9aff395c224492e9e44248b15f44b5cc095d9c/timm/models/vision_transformer.py\"\n",
        "\n",
        "class Header (Module) :\n",
        "  def __init__ (self, ni, num_classes) :\n",
        "    self.head = nn.Linear(ni, num_classes)\n",
        "  \n",
        "  def forward (self, x) :\n",
        "    x = x[:, 0]                  # Extracting the clsass token, which is used for the classification task. \n",
        "    x = self.head(x)\n",
        "    return x"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7ESOg0jUt59U"
      },
      "source": [
        "I have defined a custom_ViT() function that wraps the different modules of the model as a sequential object. This is helpful in case it is required to freeze some parts of the model (mostly the body) while, train the other parts (mostly the header)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QuPndl9gHco8"
      },
      "source": [
        "def custom_ViT (timm_model_name, num_pathces, embed_dim, ni, num_classes, pretrained = True) :\n",
        "  model = timm.create_model(timm_model_name, pretrained) \n",
        "  module_layers = list(model.children())\n",
        "  return nn.Sequential(\n",
        "      module_layers[0],\n",
        "      EmbedBlock(num_pathces, embed_dim),\n",
        "      nn.Sequential(*module_layers[1:-1]),\n",
        "      Header(ni, num_classes)\n",
        "  ) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gciDPl9WHcxJ"
      },
      "source": [
        "model_test = custom_ViT('vit_large_patch16_224', 196, 1024, 1024, dls.c, True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HSM6uFIeHdCD"
      },
      "source": [
        "# Heavily inspired by \"https://github.com/rwightman/pytorch-image-models/blob/5f9aff395c224492e9e44248b15f44b5cc095d9c/timm/models/vision_transformer.py\"\n",
        "\n",
        "def _no_grad_trunc_normal_(tensor, mean, std, a, b):\n",
        "    # Cut & paste from PyTorch official master until it's in a few official releases - RW\n",
        "    # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf\n",
        "    def norm_cdf(x):\n",
        "        # Computes standard normal cumulative distribution function\n",
        "        return (1. + math.erf(x / math.sqrt(2.))) / 2.\n",
        "\n",
        "    if (mean < a - 2 * std) or (mean > b + 2 * std):\n",
        "        warnings.warn(\"mean is more than 2 std from [a, b] in nn.init.trunc_normal_. \"\n",
        "                      \"The distribution of values may be incorrect.\",\n",
        "                      stacklevel=2)\n",
        "\n",
        "    with torch.no_grad():\n",
        "        # Values are generated by using a truncated uniform distribution and\n",
        "        # then using the inverse CDF for the normal distribution.\n",
        "        # Get upper and lower cdf values\n",
        "        l = norm_cdf((a - mean) / std)\n",
        "        u = norm_cdf((b - mean) / std)\n",
        "\n",
        "        # Uniformly fill tensor with values from [l, u], then translate to\n",
        "        # [2l-1, 2u-1].\n",
        "        tensor.uniform_(2 * l - 1, 2 * u - 1)\n",
        "\n",
        "        # Use inverse cdf transform for normal distribution to get truncated\n",
        "        # standard normal\n",
        "        tensor.erfinv_()\n",
        "\n",
        "        # Transform to proper mean, std\n",
        "        tensor.mul_(std * math.sqrt(2.))\n",
        "        tensor.add_(mean)\n",
        "\n",
        "        # Clamp to ensure it's in the proper range\n",
        "        tensor.clamp_(min=a, max=b)\n",
        "        return tensor\n",
        "\n",
        "\n",
        "def trunc_normal_(layer, param, mean=0., std=1., a=-2., b=2.):\n",
        "    # type: (Tensor, float, float, float, float) -> Tensor\n",
        "    r\"\"\"Fills the input Tensor with values drawn from a truncated\n",
        "    normal distribution. The values are effectively drawn from the\n",
        "    normal distribution :math:`\\mathcal{N}(\\text{mean}, \\text{std}^2)`\n",
        "    with values outside :math:`[a, b]` redrawn until they are within\n",
        "    the bounds. The method used for generating the random values works\n",
        "    best when :math:`a \\leq \\text{mean} \\leq b`.\n",
        "    Args:\n",
        "        tensor: an n-dimensional `torch.Tensor`\n",
        "        mean: the mean of the normal distribution\n",
        "        std: the standard deviation of the normal distribution\n",
        "        a: the minimum cutoff value\n",
        "        b: the maximum cutoff value\n",
        "    Examples:\n",
        "        >>> w = torch.empty(3, 5)\n",
        "        >>> nn.init.trunc_normal_(w)\n",
        "    \"\"\"\n",
        "    tensor = layer.get_parameter(param)\n",
        "    return _no_grad_trunc_normal_(tensor, mean, std, a, b)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QBkOBpvot_8m"
      },
      "source": [
        "The function, defined above, is used to intialise the parameters of EmbedBlock."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ECcz31_NHl1M"
      },
      "source": [
        "# apply_init(model_test[1], lambda x: trunc_normal_(x, param = 'cls_tokens', std = 0.2))\n",
        "# apply_init(model_test[1], lambda x: trunc_normal_(x, param = 'pos_embeds', std = 0.2))\n",
        "trunc_normal_(model_test[1], 'cls_tokens')\n",
        "trunc_normal_(model_test[1], 'pos_embeds')\n",
        "apply_init(model_test[3], nn.init.kaiming_normal_) "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EjqsNiFOuEV2"
      },
      "source": [
        "\n",
        "Checking if the parameters of the EmbedBlock are initialised :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zzea9p-aHdFy",
        "outputId": "b4a72c77-4bf0-4d23-d697-57a1e34d02ea"
      },
      "source": [
        "model_test[1].get_parameter('pos_embeds')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Parameter containing:\n",
              "tensor([[[ 0.8015, -0.9923,  0.2519,  ...,  0.9063, -1.1533,  0.1561],\n",
              "         [-1.6486,  0.4017, -1.1665,  ...,  0.1081, -0.7906, -0.1275],\n",
              "         [ 0.8124,  1.9816,  0.1916,  ...,  1.3498, -0.3083,  0.3760],\n",
              "         ...,\n",
              "         [-1.2626,  1.3094,  0.1387,  ..., -0.0055, -1.4397, -1.9857],\n",
              "         [-0.3869,  1.4453,  0.0590,  ..., -1.3351, -0.8866, -0.8725],\n",
              "         [-0.6007, -1.6362,  0.6578,  ..., -0.5168, -0.2878, -0.2628]]],\n",
              "       requires_grad=True)"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VCjSbj7xHdJN",
        "outputId": "04b9514a-74e9-42f9-cec9-a9b563559d9d"
      },
      "source": [
        "model_test[1].get_parameter('cls_tokens')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Parameter containing:\n",
              "tensor([[[ 0.2635, -0.8517,  1.5480,  ..., -0.0430,  0.8127, -0.8280]]],\n",
              "       requires_grad=True)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HDghiCZLuOZh"
      },
      "source": [
        "We are going to unfreeze the full model, as the ImageWoof dataset is significantly complex."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9FruVNF4HQEl"
      },
      "source": [
        "learn = Learner(dls, model_test, loss_func = LabelSmoothingCrossEntropy(), metrics = accuracy, opt_func = ranger).to_fp16()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pwlUhGjvuLYE"
      },
      "source": [
        "Sumary of the model :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "JfuD9zKFHmEM",
        "outputId": "cf4728d9-c121-4f78-9449-9b6b9932975d"
      },
      "source": [
        "learn.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "Sequential (Input shape: 4)\n",
              "============================================================================\n",
              "Layer (type)         Output Shape         Param #    Trainable \n",
              "============================================================================\n",
              "                     4 x 1024 x 14 x 14  \n",
              "Conv2d                                    787456     True      \n",
              "Identity                                                       \n",
              "EmbedBlock                                                     \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 3072      \n",
              "Linear                                    3148800    True      \n",
              "Dropout                                                        \n",
              "Linear                                    1049600    True      \n",
              "Dropout                                                        \n",
              "Identity                                                       \n",
              "LayerNorm                                 2048       True      \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 4096      \n",
              "Linear                                    4198400    True      \n",
              "GELU                                                           \n",
              "____________________________________________________________________________\n",
              "                     4 x 197 x 1024      \n",
              "Linear                                    4195328    True      \n",
              "Dropout                                                        \n",
              "LayerNorm                                 2048       True      \n",
              "Identity                                                       \n",
              "____________________________________________________________________________\n",
              "                     4 x 10              \n",
              "Linear                                    10250      True      \n",
              "____________________________________________________________________________\n",
              "\n",
              "Total params: 303,109,130\n",
              "Total trainable params: 303,109,130\n",
              "Total non-trainable params: 0\n",
              "\n",
              "Optimizer used: <function ranger at 0x7f0763640050>\n",
              "Loss function: LabelSmoothingCrossEntropy()\n",
              "\n",
              "Callbacks:\n",
              "  - TrainEvalCallback\n",
              "  - MixedPrecision\n",
              "  - Recorder\n",
              "  - ProgressCallback"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kRMatvPOubKT"
      },
      "source": [
        "I have used [lr_find()](https://docs.fast.ai/callback.schedule.html#Learner.lr_find) method for finding a suitable learning rate."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "vevxGBDVHmMI",
        "outputId": "5c7a4273-3959-4e0f-901c-386a823eaae7"
      },
      "source": [
        "learn.lr_find()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SuggestedLRs(valley=0.00013182566908653826)"
            ]
          },
          "metadata": {},
          "execution_count": 20
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uL23LJC0u1nD"
      },
      "source": [
        "# 5. **Training and evaluation**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Klh2WlJwuveS"
      },
      "source": [
        "Here, I have trained (fine - tuned) the model for 2 epochs using the [fit_flat_cos()](https://docs.fast.ai/callback.schedule.html#Learner.fit_flat_cos) method."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "tjfvZZ6SHmQd",
        "outputId": "992d5720-2858-492e-c08a-f0fc9b1782a8"
      },
      "source": [
        "learn.fit_flat_cos(2, 1e-3)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: left;\">\n",
              "      <th>epoch</th>\n",
              "      <th>train_loss</th>\n",
              "      <th>valid_loss</th>\n",
              "      <th>accuracy</th>\n",
              "      <th>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>2.175490</td>\n",
              "      <td>2.120081</td>\n",
              "      <td>0.229575</td>\n",
              "      <td>1:22:55</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>2.072094</td>\n",
              "      <td>2.037750</td>\n",
              "      <td>0.272843</td>\n",
              "      <td>1:22:58</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VxvPGI9fvHPB"
      },
      "source": [
        "\n",
        "So, the accuracy after fine - tuning for 2 epochs is 27.28%"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "r-Krzj2hH06o",
        "outputId": "dbda9669-7d23-41cf-dc99-5ee3f12fd9a3"
      },
      "source": [
        "learn.recorder.plot_loss()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "id": "27Fi87pfH1A4",
        "outputId": "1a32856a-f1cf-4d66-c216-1743b9b5c1ad"
      },
      "source": [
        "interp = ClassificationInterpretation.from_learner(learn)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9a8p_FOyvNU0"
      },
      "source": [
        "The confusion matrix :"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 743
        },
        "id": "cl1h9tRdH1Ee",
        "outputId": "7f18bd3f-d51f-4b39-aaa1-fb1dbeae9411"
      },
      "source": [
        "interp.plot_confusion_matrix(figsize = (12, 12), dpi = 60)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NVenrcgGH1H_",
        "outputId": "6aa9d897-2e6c-44c4-9858-1d932b5e820c"
      },
      "source": [
        "interp.most_confused(min_val = 20)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[('Shih-Tzu', 'Old English sheepdog', 119),\n",
              " ('Samoyed', 'Old English sheepdog', 108),\n",
              " ('Old English sheepdog', 'Samoyed', 100),\n",
              " ('Beagle', 'English foxhound', 91),\n",
              " ('Border terrier', 'Rhodesian ridgeback', 89),\n",
              " ('Golden retriever', 'Rhodesian ridgeback', 86),\n",
              " ('Shih-Tzu', 'Samoyed', 81),\n",
              " ('Border terrier', 'Beagle', 72),\n",
              " ('Dingo', 'Rhodesian ridgeback', 66),\n",
              " ('Australian terrier', 'Beagle', 59),\n",
              " ('Australian terrier', 'Old English sheepdog', 58),\n",
              " ('Dingo', 'Golden retriever', 57),\n",
              " ('Rhodesian ridgeback', 'Border terrier', 56),\n",
              " ('English foxhound', 'Beagle', 52),\n",
              " ('Shih-Tzu', 'Beagle', 52),\n",
              " ('Border terrier', 'Golden retriever', 45),\n",
              " ('Rhodesian ridgeback', 'Beagle', 45),\n",
              " ('Rhodesian ridgeback', 'Golden retriever', 45),\n",
              " ('Old English sheepdog', 'Beagle', 43),\n",
              " ('Australian terrier', 'Border terrier', 42),\n",
              " ('Golden retriever', 'Old English sheepdog', 40),\n",
              " ('Old English sheepdog', 'Shih-Tzu', 40),\n",
              " ('Shih-Tzu', 'Australian terrier', 39),\n",
              " ('Beagle', 'Old English sheepdog', 38),\n",
              " ('Golden retriever', 'Border terrier', 38),\n",
              " ('Golden retriever', 'Dingo', 38),\n",
              " ('Samoyed', 'Shih-Tzu', 36),\n",
              " ('Australian terrier', 'Dingo', 35),\n",
              " ('Dingo', 'Beagle', 35),\n",
              " ('Rhodesian ridgeback', 'Dingo', 35),\n",
              " ('Australian terrier', 'Samoyed', 34),\n",
              " ('Border terrier', 'Old English sheepdog', 34),\n",
              " ('Dingo', 'Australian terrier', 32),\n",
              " ('Beagle', 'Samoyed', 31),\n",
              " ('Old English sheepdog', 'English foxhound', 31),\n",
              " ('Australian terrier', 'Shih-Tzu', 30),\n",
              " ('Golden retriever', 'Beagle', 30),\n",
              " ('Golden retriever', 'English foxhound', 29),\n",
              " ('Samoyed', 'Beagle', 29),\n",
              " ('Dingo', 'Border terrier', 28),\n",
              " ('Golden retriever', 'Australian terrier', 28),\n",
              " ('Shih-Tzu', 'Golden retriever', 28),\n",
              " ('Australian terrier', 'Golden retriever', 27),\n",
              " ('Beagle', 'Shih-Tzu', 27),\n",
              " ('Border terrier', 'Australian terrier', 27),\n",
              " ('Dingo', 'English foxhound', 27),\n",
              " ('Beagle', 'Australian terrier', 26),\n",
              " ('Beagle', 'Dingo', 26),\n",
              " ('English foxhound', 'Old English sheepdog', 26),\n",
              " ('Beagle', 'Rhodesian ridgeback', 25),\n",
              " ('Border terrier', 'Dingo', 25),\n",
              " ('Border terrier', 'Shih-Tzu', 25),\n",
              " ('Dingo', 'Samoyed', 25),\n",
              " ('Dingo', 'Old English sheepdog', 23),\n",
              " ('Australian terrier', 'Rhodesian ridgeback', 22),\n",
              " ('Beagle', 'Golden retriever', 22),\n",
              " ('Old English sheepdog', 'Australian terrier', 22)]"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "da2CznGDvZxt"
      },
      "source": [
        "Even after fine - tuning the model for 2 epochs, the model is highly inaccurate, compared to the performance we observed with Adam optimizer. With ranger, the model should be fine - tuned for more epochs, to get an accuracy similar to what we observed when using Adam."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oamO4RrCH1KQ",
        "outputId": "94a0ef4f-17a5-499b-85c6-73c38b708931"
      },
      "source": [
        "learn.save('stage - 1')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Path('models/stage - 1.pth')"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MgitV_6LvR-z"
      },
      "source": [
        "\n",
        "Thanks a lot for taking the time to go through this notebook !"
      ]
    }
  ]
}