08_movielens_collab.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "08_movielens_collab.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPNAk4npyYdtlW5kS9b96+c",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/LukeWeidenwalker/fastai-notebooks/blob/master/08_movielens_collab.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0l03gIrnwA5o",
        "outputId": "23196928-0a5f-45cb-82cc-29b9a256519c"
      },
      "source": [
        "!pip install -Uqq fastai"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[K     |████████████████████████████████| 204kB 8.4MB/s \n",
            "\u001b[K     |████████████████████████████████| 61kB 8.7MB/s \n",
            "\u001b[?25h"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7DjVLBsQwZmp"
      },
      "source": [
        "from fastai.collab import *\n",
        "from fastai.tabular.all import *\n",
        "import pandas as pd"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17
        },
        "id": "EL6RgQtMwlaq",
        "outputId": "902d7673-5734-45cf-dac4-367135cb753e"
      },
      "source": [
        "path = untar_data(URLs.ML_100k)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              ""
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "KHoQl5pNxC3D",
        "outputId": "810dc803-77a6-4881-b290-da4eedb413e9"
      },
      "source": [
        "raw_df = pd.read_csv(path/\"u.data\", delimiter=\"\\t\", names=[\"user_id\", \"item_id\", \"rating\", \"timestamp\"])\n",
        "raw_df.head()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>item_id</th>\n",
              "      <th>rating</th>\n",
              "      <th>timestamp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>196</td>\n",
              "      <td>242</td>\n",
              "      <td>3</td>\n",
              "      <td>881250949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>186</td>\n",
              "      <td>302</td>\n",
              "      <td>3</td>\n",
              "      <td>891717742</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>22</td>\n",
              "      <td>377</td>\n",
              "      <td>1</td>\n",
              "      <td>878887116</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>244</td>\n",
              "      <td>51</td>\n",
              "      <td>2</td>\n",
              "      <td>880606923</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>166</td>\n",
              "      <td>346</td>\n",
              "      <td>1</td>\n",
              "      <td>886397596</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   user_id  item_id  rating  timestamp\n",
              "0      196      242       3  881250949\n",
              "1      186      302       3  891717742\n",
              "2       22      377       1  878887116\n",
              "3      244       51       2  880606923\n",
              "4      166      346       1  886397596"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "pIQLYUVlyQo8",
        "outputId": "72909f55-917d-4b37-f8f7-24672ba8eff3"
      },
      "source": [
        "movie_df = pd.read_csv(path/\"u.item\", delimiter=\"|\", encoding=\"latin-1\", usecols=(0,1), names=('movie','title'), header=None)\n",
        "movie_df.head()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>movie</th>\n",
              "      <th>title</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>Toy Story (1995)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>GoldenEye (1995)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>Four Rooms (1995)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>Get Shorty (1995)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>Copycat (1995)</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   movie              title\n",
              "0      1   Toy Story (1995)\n",
              "1      2   GoldenEye (1995)\n",
              "2      3  Four Rooms (1995)\n",
              "3      4  Get Shorty (1995)\n",
              "4      5     Copycat (1995)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "KVXp7VOpy7bO",
        "outputId": "85b9f9d1-2a55-4e00-9d9e-e0185c66b701"
      },
      "source": [
        "ratings = raw_df.merge(movie_df, left_on=\"item_id\", right_on=\"movie\")\n",
        "ratings.drop([\"movie\"], axis=1, inplace=True)\n",
        "ratings.head()"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>item_id</th>\n",
              "      <th>rating</th>\n",
              "      <th>timestamp</th>\n",
              "      <th>title</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>196</td>\n",
              "      <td>242</td>\n",
              "      <td>3</td>\n",
              "      <td>881250949</td>\n",
              "      <td>Kolya (1996)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>63</td>\n",
              "      <td>242</td>\n",
              "      <td>3</td>\n",
              "      <td>875747190</td>\n",
              "      <td>Kolya (1996)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>226</td>\n",
              "      <td>242</td>\n",
              "      <td>5</td>\n",
              "      <td>883888671</td>\n",
              "      <td>Kolya (1996)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>154</td>\n",
              "      <td>242</td>\n",
              "      <td>3</td>\n",
              "      <td>879138235</td>\n",
              "      <td>Kolya (1996)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>306</td>\n",
              "      <td>242</td>\n",
              "      <td>5</td>\n",
              "      <td>876503793</td>\n",
              "      <td>Kolya (1996)</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   user_id  item_id  rating  timestamp         title\n",
              "0      196      242       3  881250949  Kolya (1996)\n",
              "1       63      242       3  875747190  Kolya (1996)\n",
              "2      226      242       5  883888671  Kolya (1996)\n",
              "3      154      242       3  879138235  Kolya (1996)\n",
              "4      306      242       5  876503793  Kolya (1996)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sV4DAf920Fda"
      },
      "source": [
        "dls = CollabDataLoaders.from_df(ratings, item_name=\"title\", rating_name=\"rating\", bs=64)"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_2Vghoaf6W-k",
        "outputId": "64c46ff4-043b-4f96-ea0a-7e6feefd32a1"
      },
      "source": [
        "print(f\"Number of Users: {len(dls.classes['user_id'])}\")\n",
        "print(f\"Number of Movies: {len(dls.classes['title'])}\")"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of Users: 944\n",
            "Number of Movies: 1665\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D-85NQib2Htt",
        "outputId": "b1d4ffbb-c2ff-4b93-81a3-941d3b361e9f"
      },
      "source": [
        "# (tensor(user_ids, movie_ids), tensor(ratings))\n",
        "# shape=(64, 2)                 shape=(64, 1) \n",
        "dls.one_batch()[0].shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "torch.Size([64, 2])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 56
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7uayLZYq83T3",
        "outputId": "9f83b8eb-e740-43b4-eb6f-fdfa49761a86"
      },
      "source": [
        "batch = dls.one_batch()\n",
        "print(batch[0][:, 0])\n",
        "hot = one_hot(batch[0][:, 0], len(dls.classes['user_id'])).float()"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([405, 208, 417, 474, 311, 180, 943, 907, 764, 671,  23, 642, 561,  95,\n",
            "        109,  90, 293, 301, 160, 194, 239, 929, 780, 142, 838, 601,   6, 798,\n",
            "        618, 775,  60, 900, 403, 128, 347, 516, 354,  95, 299, 268, 585, 416,\n",
            "        207, 181, 561, 450, 892, 885, 537, 433, 405, 569, 215, 793, 417, 727,\n",
            "        252, 524, 752, 693, 829, 506, 454, 891])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v3fK9B4_6GYa"
      },
      "source": [
        "user_factors = torch.randn(len(dls.classes['user_id']), 50)"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pfKtcrPX67-3",
        "outputId": "dbb50162-b0aa-4257-e422-0747e3f9bdd1"
      },
      "source": [
        "user_factors.shape, user_factors.t().shape, hot.shape"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(torch.Size([944, 50]), torch.Size([50, 944]), torch.Size([944]))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ropZ4EDa88vD"
      },
      "source": [
        "emb = Embedding(5, 4)"
      ],
      "execution_count": 58,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pB3w4iAz9Bgu",
        "outputId": "5efd87af-7143-400c-de1b-0d49bace1fb7"
      },
      "source": [
        "for p in emb.parameters():\n",
        "  print(p)"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Parameter containing:\n",
            "tensor([[ 2.1307e-03,  2.1051e-03,  9.9638e-04,  9.9662e-03],\n",
            "        [-1.0222e-02,  3.7448e-03, -3.4844e-03,  1.5378e-03],\n",
            "        [-4.0412e-03, -1.3447e-03, -6.7373e-03,  5.2135e-03],\n",
            "        [ 4.8824e-04,  6.4750e-03, -2.1057e-03, -2.7050e-03],\n",
            "        [ 1.8325e-02, -1.9606e-02,  2.4984e-03,  4.2124e-05]],\n",
            "       requires_grad=True)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dxEwEkOh9TXY",
        "outputId": "8ce45e3b-a8e9-4534-dd61-691e5affa85c"
      },
      "source": [
        "emb(tensor([3,4]))"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor([[ 4.8824e-04,  6.4750e-03, -2.1057e-03, -2.7050e-03],\n",
              "        [ 1.8325e-02, -1.9606e-02,  2.4984e-03,  4.2124e-05]],\n",
              "       grad_fn=<EmbeddingBackward>)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mbfujjDU2L7J"
      },
      "source": [
        "class PMFModel(Module):\n",
        "  def __init__(self, n_users, n_movies, n_user_factors, n_movie_factors):\n",
        "    self.user_factors = Embedding(n_users, n_user_factors)\n",
        "    self.user_bias = Embedding(n_users, 1)\n",
        "    self.movie_factors = Embedding(n_movies, n_movie_factors)\n",
        "    self.movie_bias = Embedding(n_movies, 1)\n",
        "  \n",
        "  def forward(self, x):\n",
        "    # x -> [[(user_id, movie_id), rating], ... ]\n",
        "    res = (self.user_factors(x[:, 0]) * self.movie_factors(x[:, 1])).sum(dim=1, keepdims=True)\n",
        "    res += self.user_bias(x[:, 0]) + self.movie_bias(x[:, 1])\n",
        "    return res"
      ],
      "execution_count": 130,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "81N4_by0_Qxw"
      },
      "source": [
        "n_users = len(dls.classes['user_id'])\n",
        "n_movies = len(dls.classes['title'])\n",
        "n_user_factors = 50\n",
        "n_movie_factors = 50\n",
        "\n",
        "model = PMFModel(n_users, n_movies, n_user_factors, n_movie_factors)"
      ],
      "execution_count": 131,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cw_cLJuz2pja"
      },
      "source": [
        "learn = Learner(dls, model, loss_func=MSELossFlat())"
      ],
      "execution_count": 132,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "id": "92WSGE6EBb2D",
        "outputId": "1b2e496d-b2d3-4eeb-d8d1-109435c5f7d6"
      },
      "source": [
        "learn.fit_one_cycle(5, 5e-3)"
      ],
      "execution_count": 134,
      "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>time</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>0</td>\n",
              "      <td>1.049654</td>\n",
              "      <td>1.110791</td>\n",
              "      <td>00:07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>1.022256</td>\n",
              "      <td>1.038660</td>\n",
              "      <td>00:07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>0.891747</td>\n",
              "      <td>0.947849</td>\n",
              "      <td>00:07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>0.777282</td>\n",
              "      <td>0.869711</td>\n",
              "      <td>00:07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>0.713082</td>\n",
              "      <td>0.855310</td>\n",
              "      <td>00:07</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "TiXmHvWNBeEG",
        "outputId": "4326a313-f31d-482e-fcc4-6e8f09fec2c6"
      },
      "source": [
        "learn.recorder.plot_loss()"
      ],
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FCQu6BeoCKiK"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}