{
  "id": 162591,
  "title": "How to use pip-installed libraries during submission?",
  "url": "/competitions/birdsong-recognition/discussion/162591",
  "author_name": "",
  "post_date": "2020-06-29T13:13:47.800936900Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>I am using pytorch-lightning in my notebook, but how can I use it in the submission kernel since that doesn't allow internet?</p>",
  "messages": [
    {
      "id": "906642",
      "postDate": "06/29/2020 13:13:47",
      "content": "<p>I am using pytorch-lightning in my notebook, but how can I use it in the submission kernel since that doesn't allow internet?</p>",
      "rawMarkdown": "I am using pytorch-lightning in my notebook, but how can I use it in the submission kernel since that doesn't allow internet?",
      "votes": null
    },
    {
      "id": "906726",
      "postDate": "06/29/2020 14:08:16",
      "content": "<p>You can create a kaggle dataset (or check if one exists already) where you put you the <code>.whl</code> file, and install it by adding external kaggle dataset (which doesn't  require internet).</p>",
      "rawMarkdown": "You can create a kaggle dataset (or check if one exists already) where you put you the `.whl` file, and install it by adding external kaggle dataset (which doesn't  require internet).",
      "votes": null
    },
    {
      "id": "907182",
      "postDate": "06/29/2020 18:26:02",
      "content": "<p>In one notebook run <code>!pip download PACKAGE_NAME</code>. That will download <code>WHL</code> files. Then put those <code>WHL</code> files into a Kaggle dataset. Then <code>!pip install PATH_TO_WHL</code>. (Install all the <code>WHL</code> files, one at a time).</p>",
      "rawMarkdown": "In one notebook run `!pip download PACKAGE_NAME`. That will download `WHL` files. Then put those `WHL` files into a Kaggle dataset. Then `!pip install PATH_TO_WHL`. (Install all the `WHL` files, one at a time).",
      "votes": null
    },
    {
      "id": "907505",
      "postDate": "06/30/2020 02:45:33",
      "content": "<p>Thanks.</p>",
      "rawMarkdown": "Thanks.",
      "votes": null
    },
    {
      "id": "907562",
      "postDate": "06/30/2020 03:14:58",
      "content": "<p>Thanks Pranav. I found one dataset <a href=\"https://www.kaggle.com/higepon/pytorchlightning-071\">https://www.kaggle.com/higepon/pytorchlightning-071</a></p>",
      "rawMarkdown": "Thanks Pranav. I found one dataset https://www.kaggle.com/higepon/pytorchlightning-071",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 906726,
      "author_name": "pranavkasela",
      "author_url": "",
      "post_date": "06/29/2020 14:08:16",
      "content": "<p>You can create a kaggle dataset (or check if one exists already) where you put you the <code>.whl</code> file, and install it by adding external kaggle dataset (which doesn't  require internet).</p>",
      "votes": null,
      "replies": [
        {
          "id": 907562,
          "author_name": "krishsg",
          "author_url": "",
          "post_date": "06/30/2020 03:14:58",
          "content": "<p>Thanks Pranav. I found one dataset <a href=\"https://www.kaggle.com/higepon/pytorchlightning-071\">https://www.kaggle.com/higepon/pytorchlightning-071</a></p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 907182,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "06/29/2020 18:26:02",
      "content": "<p>In one notebook run <code>!pip download PACKAGE_NAME</code>. That will download <code>WHL</code> files. Then put those <code>WHL</code> files into a Kaggle dataset. Then <code>!pip install PATH_TO_WHL</code>. (Install all the <code>WHL</code> files, one at a time).</p>",
      "votes": null,
      "replies": [
        {
          "id": 907505,
          "author_name": "krishsg",
          "author_url": "",
          "post_date": "06/30/2020 02:45:33",
          "content": "<p>Thanks.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "906642": "I am using pytorch-lightning in my notebook, but how can I use it in the submission kernel since that doesn't allow internet?",
    "906726": "You can create a kaggle dataset (or check if one exists already) where you put you the `.whl` file, and install it by adding external kaggle dataset (which doesn't  require internet).",
    "907182": "In one notebook run `!pip download PACKAGE_NAME`. That will download `WHL` files. Then put those `WHL` files into a Kaggle dataset. Then `!pip install PATH_TO_WHL`. (Install all the `WHL` files, one at a time).",
    "907505": "Thanks.",
    "907562": "Thanks Pranav. I found one dataset https://www.kaggle.com/higepon/pytorchlightning-071"
  },
  "source": "meta"
}