Mrpatekful/dialogue_generation.ipynb

## dialogue_generation.ipynb
{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "dialogue_generation.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/Mrpatekful/94aa58038cdd221cfa83a7e7334f3835/dialogue_generation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "03gBOSATlJUX",
        "colab_type": "text"
      },
      "source": [
        "# Dialogue generation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9Jy7KkYazNzB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!git clone https://github.com/bme-chatbots/dialogue-generation.git\n",
        "!python -m pip install --upgrade pip\n",
        "\n",
        "# installing apex is optional and is only useful if Colab's Tesla P100 or T4 is used\n",
        "\n",
        "# building the cython code and installing the required packages\n",
        "!cd dialogue-generation; pip install -r requirements.txt; python setup.py build_ext --inplace"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RqP1wZkkxKiT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# make sure you are using Tesla P-100 or Tesla T4\n",
        "\n",
        "!nvidia-smi"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IgAgCunuesja",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!git clone https://github.com/NVIDIA/apex\n",
        "!cd apex; pip install -v --no-cache-dir --global-option=\"--cpp_ext\" --global-option=\"--cuda_ext\" ."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SPWk27Vm-WgB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0E9Geb0Y-qU2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# the logs and models will be saved to your Google Drive\n",
        "# note that this may require large disk space 5-6 GB or more\n",
        "\n",
        "MODEL_DIR = '/drive/My Drive/chatbot'\n",
        "MODEL = 'gpt2'\n",
        "DATA = 'dailydialog'\n",
        "NAME = MODEL + DATA"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EA4Bass0yjR2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os\n",
        "\n",
        "os.makedirs(os.path.join(MODEL_DIR, MODEL, NAME), exist_ok=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m1ro1pTtmvk3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "0fe6d7e8-243a-4dcc-c1bb-62dc3f8e72b9"
      },
      "source": [
        "%load_ext tensorboard"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "The tensorboard extension is already loaded. To reload it, use:\n",
            "  %reload_ext tensorboard\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "biwtFZJznFPN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "08b6d836-2d0f-4989-c26d-5805bf051a36"
      },
      "source": [
        "%tensorboard --logdir \"{MODEL_DIR}/{MODEL}/{NAME}\""
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "Reusing TensorBoard on port 6006 (pid 1926), started 0:27:49 ago. (Use '!kill 1926' to kill it.)"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "    <div id=\"root\"></div>\n",
              "    <script>\n",
              "      (function() {\n",
              "        window.TENSORBOARD_ENV = window.TENSORBOARD_ENV || {};\n",
              "        window.TENSORBOARD_ENV[\"IN_COLAB\"] = true;\n",
              "        document.querySelector(\"base\").href = \"https://localhost:6006\";\n",
              "        function fixUpTensorboard(root) {\n",
              "          const tftb = root.querySelector(\"tf-tensorboard\");\n",
              "          // Disable the fragment manipulation behavior in Colab. Not\n",
              "          // only is the behavior not useful (as the iframe's location\n",
              "          // is not visible to the user), it causes TensorBoard's usage\n",
              "          // of `window.replace` to navigate away from the page and to\n",
              "          // the `localhost:<port>` URL specified by the base URI, which\n",
              "          // in turn causes the frame to (likely) crash.\n",
              "          tftb.removeAttribute(\"use-hash\");\n",
              "        }\n",
              "        function executeAllScripts(root) {\n",
              "          // When `script` elements are inserted into the DOM by\n",
              "          // assigning to an element's `innerHTML`, the scripts are not\n",
              "          // executed. Thus, we manually re-insert these scripts so that\n",
              "          // TensorBoard can initialize itself.\n",
              "          for (const script of root.querySelectorAll(\"script\")) {\n",
              "            const newScript = document.createElement(\"script\");\n",
              "            newScript.type = script.type;\n",
              "            newScript.textContent = script.textContent;\n",
              "            root.appendChild(newScript);\n",
              "            script.remove();\n",
              "          }\n",
              "        }\n",
              "        function setHeight(root, height) {\n",
              "          // We set the height dynamically after the TensorBoard UI has\n",
              "          // been initialized. This avoids an intermediate state in\n",
              "          // which the container plus the UI become taller than the\n",
              "          // final width and cause the Colab output frame to be\n",
              "          // permanently resized, eventually leading to an empty\n",
              "          // vertical gap below the TensorBoard UI. It's not clear\n",
              "          // exactly what causes this problematic intermediate state,\n",
              "          // but setting the height late seems to fix it.\n",
              "          root.style.height = `${height}px`;\n",
              "        }\n",
              "        const root = document.getElementById(\"root\");\n",
              "        fetch(\".\")\n",
              "          .then((x) => x.text())\n",
              "          .then((html) => void (root.innerHTML = html))\n",
              "          .then(() => fixUpTensorboard(root))\n",
              "          .then(() => executeAllScripts(root))\n",
              "          .then(() => setHeight(root, 800));\n",
              "      })();\n",
              "    </script>\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QTGtnFdBAWni",
        "colab_type": "code",
        "outputId": "b362a79e-ae22-4df8-f22b-73ddfffd3a79",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "!cd dialogue-generation; python -m src.train --model \"{MODEL}\" --fp16 --max_hist 3 --lr 1e-4 --data \"{DATA}\" --model_dir \"{MODEL_DIR}\" --batch_size 32 --name \"{NAME}\""
      ],
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Selected optimization level O2:  FP16 training with FP32 batchnorm and FP32 master weights.\n",
            "\n",
            "Defaults for this optimization level are:\n",
            "enabled                : True\n",
            "opt_level              : O2\n",
            "cast_model_type        : torch.float16\n",
            "patch_torch_functions  : False\n",
            "keep_batchnorm_fp32    : True\n",
            "master_weights         : True\n",
            "loss_scale             : dynamic\n",
            "Processing user overrides (additional kwargs that are not None)...\n",
            "After processing overrides, optimization options are:\n",
            "enabled                : True\n",
            "opt_level              : O2\n",
            "cast_model_type        : torch.float16\n",
            "patch_torch_functions  : False\n",
            "keep_batchnorm_fp32    : True\n",
            "master_weights         : True\n",
            "loss_scale             : dynamic\n",
            "\n",
            " config           |\n",
            " max_epochs       | 25\n",
            " no_cuda          | False\n",
            " fp16             | True\n",
            " lr               | 0.0001\n",
            " batch_size       | 32\n",
            " patience         | 5\n",
            " schedule         | noam\n",
            " warmup_steps     | 16000\n",
            " total_steps      | 1000000\n",
            " grad_accum_steps | 2\n",
            " local_rank       | -1\n",
            " notebook         | False\n",
            " clip_grad        |\n",
            " seed             |\n",
            " data             | dailydialog\n",
            " data_dir         | /content/dialogue-generation/src/../data\n",
            " download_dir     | /content/dialogue-generation/src/../data\n",
            " file_size        | 100000\n",
            " max_hist         | 3\n",
            " force_rebuild    | False\n",
            " max_len          | 50\n",
            " model            | gpt2\n",
            " grad_ckpt        | False\n",
            " name             | gpt2dailydialog\n",
            " model_dir        | /drive/My Drive/chatbot\n",
            " cuda             | True\n",
            " distributed      | False\n",
            "\n",
            "epoch    train_loss    train_acc    train_ppl    valid_loss    valid_acc    valid_ppl\n",
            "-------  ------------  -----------  -----------  ------------  -----------  -----------\n",
            "train 0:   0% 0/6103 [00:00<?, ?it/s]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 32768.0\n",
            "train 0:   0% 2/6103 [00:01<1:28:37,  1.15it/s, loss=48.2, acc=0.00151, ppl=inf, skip=2]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 8192.0\n",
            "train 0:   0% 4/6103 [00:02<1:02:25,  1.63it/s, loss=49.3, acc=0.00816, ppl=inf, skip=2]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 2048.0\n",
            "train 0:   0% 6/6103 [00:02<49:27,  2.05it/s, loss=50.4, acc=0.00222, ppl=inf, skip=2]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 512.0\n",
            "train 0:   2% 134/6103 [00:38<12:37,  7.88it/s, loss=17.3, acc=0.0149, ppl=1.1e+15, skip=2]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 256.0\n",
            "      0         2.453        0.375          inf         2.689        0.430       15.203\n",
            "train 1:  44% 2687/6103 [10:33<19:46,  2.88it/s, loss=1.22, acc=0.457, ppl=11.6, skip=2]Gradient overflow.  Skipping step, loss scaler 0 reducing loss scale to 2048.0\n",
            "      1         1.270        0.452       13.181         2.616        0.440       14.246\n",
            "train 2:   2% 149/6103 [00:39<10:47,  9.19it/s, loss=0.991, acc=0.555, ppl=7.26, skip=2]Exception ignored in: <bound method tqdm.__del__ of train 2:   2% 149/6103 [00:39<10:47,  9.19it/s, loss=0.991, acc=0.555, ppl=7.26, skip=2]>\n",
            "Traceback (most recent call last):\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tqdm/_tqdm.py\", line 931, in __del__\n",
            "    self.close()\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tqdm/_tqdm.py\", line 1133, in close\n",
            "    self._decr_instances(self)\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tqdm/_tqdm.py\", line 496, in _decr_instances\n",
            "    cls.monitor.exit()\n",
            "  File \"/usr/local/lib/python3.6/dist-packages/tqdm/_monitor.py\", line 52, in exit\n",
            "    self.join()\n",
            "  File \"/usr/lib/python3.6/threading.py\", line 1053, in join\n",
            "    raise RuntimeError(\"cannot join current thread\")\n",
            "RuntimeError: cannot join current thread\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3WzLZKWB_-WA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!cd dialogue-generation; python -m src.interact --model \"{MODEL}\" --max_hist 5 --model_dir \"{MODEL_DIR}\" --name \"{NAME}\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A78k_IPkfAdj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import json\n",
        "\n",
        "sns.set()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dpVEwtaUgHmp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# loading history to visualize the results which will be saved\n",
        "# as a seaborn plot to the session storage\n",
        "\n",
        "with open(os.path.join(MODEL_DIR, MODEL, NAME, 'history.json')) as fh:\n",
        "    history = pd.DataFrame(json.load(fh))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NBS1I9pvhega",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 288
        },
        "outputId": "a7a4cc03-bf82-413a-db45-476d26003b8c"
      },
      "source": [
        "figure, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(15, 4), sharex=True)\n",
        "\n",
        "sns.lineplot(data=history[['train_loss', 'valid_loss']], ax=ax1)\n",
        "sns.lineplot(data=history[['train_ppl', 'valid_ppl']], ax=ax2)\n",
        "sns.lineplot(data=history[['train_acc', 'valid_acc']], ax=ax3)\n",
        "\n",
        "figure.savefig('history.png')"
      ],
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA24AAAD7CAYAAAASEscKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVi4Sa8WcRQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1080x288 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UHRwQwvGheVc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}