{
  "id": 446899,
  "title": "XGBoost 2.0.0 (whl file)",
  "url": "/competitions/child-mind-institute-detect-sleep-states/discussion/446899",
  "author_name": "",
  "post_date": "2023-10-13T14:28:31.601688800Z",
  "votes": 35,
  "comment_count": 2,
  "views": 0,
  "content": "<p>In our notebooks we have access to XGBoost  version 1.7.6. </p>\n<p>As many will know last September 12th there was a major new release: <a href=\"https://github.com/dmlc/xgboost/releases/tag/v2.0.0\" target=\"_blank\">XGBoost 2.0.0</a> and in an informative post  by <a href=\"https://www.kaggle.com/siukeitin\" target=\"_blank\">@siukeitin</a> on another competition it seems that <a href=\"https://www.kaggle.com/competitions/playground-series-s3e23/discussion/445637\" target=\"_blank\">\"xgboost 2.0.0 runs significantly faster on CPU\"</a>.  The  release notes also mention \"<em>…In addition to performance, we have reduced CPU memory usage</em>\", something that will be very useful in this competition!</p>\n<p>In view of this I have taken the liberty of saving the <code>whl</code> file of XGBoost 2.0.0 to a dataset that we can use to install this major new version in our notebooks even with the \"internet off\" setting.</p>\n<p>Simply attach the <a href=\"https://www.kaggle.com/datasets/carlmcbrideellis/xgboost-2-0-0-whl\" target=\"_blank\"><code>XGBoost (2.0.0) whl</code></a> dataset to ones notebook, then install via</p>\n<pre><code>!pip install -q /kaggle//xgboost----whl/xgboost--py3-none-manylinux2014_x86_64.whl\n</code></pre>\n<p>All the best and good luck!</p>\n<p>carl</p>",
  "messages": [
    {
      "id": "2480752",
      "postDate": "10/13/2023 14:28:31",
      "content": "<p>In our notebooks we have access to XGBoost  version 1.7.6. </p>\n<p>As many will know last September 12th there was a major new release: <a href=\"https://github.com/dmlc/xgboost/releases/tag/v2.0.0\" target=\"_blank\">XGBoost 2.0.0</a> and in an informative post  by <a href=\"https://www.kaggle.com/siukeitin\" target=\"_blank\">@siukeitin</a> on another competition it seems that <a href=\"https://www.kaggle.com/competitions/playground-series-s3e23/discussion/445637\" target=\"_blank\">\"xgboost 2.0.0 runs significantly faster on CPU\"</a>.  The  release notes also mention \"<em>…In addition to performance, we have reduced CPU memory usage</em>\", something that will be very useful in this competition!</p>\n<p>In view of this I have taken the liberty of saving the <code>whl</code> file of XGBoost 2.0.0 to a dataset that we can use to install this major new version in our notebooks even with the \"internet off\" setting.</p>\n<p>Simply attach the <a href=\"https://www.kaggle.com/datasets/carlmcbrideellis/xgboost-2-0-0-whl\" target=\"_blank\"><code>XGBoost (2.0.0) whl</code></a> dataset to ones notebook, then install via</p>\n<pre><code>!pip install -q /kaggle//xgboost----whl/xgboost--py3-none-manylinux2014_x86_64.whl\n</code></pre>\n<p>All the best and good luck!</p>\n<p>carl</p>",
      "rawMarkdown": "In our notebooks we have access to XGBoost  version 1.7.6. \n\nAs many will know last September 12th there was a major new release: [XGBoost 2.0.0](https://github.com/dmlc/xgboost/releases/tag/v2.0.0) and in an informative post  by @siukeitin on another competition it seems that [\"xgboost 2.0.0 runs significantly faster on CPU\"](https://www.kaggle.com/competitions/playground-series-s3e23/discussion/445637).  The  release notes also mention \"*...In addition to performance, we have reduced CPU memory usage*\", something that will be very useful in this competition!\n\nIn view of this I have taken the liberty of saving the `whl` file of XGBoost 2.0.0 to a dataset that we can use to install this major new version in our notebooks even with the \"internet off\" setting.\n\nSimply attach the [`XGBoost (2.0.0) whl`](https://www.kaggle.com/datasets/carlmcbrideellis/xgboost-2-0-0-whl) dataset to ones notebook, then install via\n\n```python\n!pip install -q /kaggle/input/xgboost-2-0-0-whl/xgboost-2.0.0-py3-none-manylinux2014_x86_64.whl\n```\n\nAll the best and good luck!\n\ncarl",
      "votes": null
    },
    {
      "id": "2482987",
      "postDate": "10/15/2023 12:04:27",
      "content": "<p>Will try this out, thanks</p>",
      "rawMarkdown": "Will try this out, thanks",
      "votes": null
    },
    {
      "id": "2512890",
      "postDate": "11/05/2023 01:43:24",
      "content": "<p>You can take adventage of the latest <code>2.0.1</code> over here - <a href=\"https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package\" target=\"_blank\">https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package</a></p>",
      "rawMarkdown": "You can take adventage of the latest `2.0.1` over here - https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2482987,
      "author_name": "yutaoc2",
      "author_url": "",
      "post_date": "10/15/2023 12:04:27",
      "content": "<p>Will try this out, thanks</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2512890,
      "author_name": "jirkaborovec",
      "author_url": "",
      "post_date": "11/05/2023 01:43:24",
      "content": "<p>You can take adventage of the latest <code>2.0.1</code> over here - <a href=\"https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package\" target=\"_blank\">https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2480752": "In our notebooks we have access to XGBoost  version 1.7.6. \n\nAs many will know last September 12th there was a major new release: [XGBoost 2.0.0](https://github.com/dmlc/xgboost/releases/tag/v2.0.0) and in an informative post  by @siukeitin on another competition it seems that [\"xgboost 2.0.0 runs significantly faster on CPU\"](https://www.kaggle.com/competitions/playground-series-s3e23/discussion/445637).  The  release notes also mention \"*...In addition to performance, we have reduced CPU memory usage*\", something that will be very useful in this competition!\n\nIn view of this I have taken the liberty of saving the `whl` file of XGBoost 2.0.0 to a dataset that we can use to install this major new version in our notebooks even with the \"internet off\" setting.\n\nSimply attach the [`XGBoost (2.0.0) whl`](https://www.kaggle.com/datasets/carlmcbrideellis/xgboost-2-0-0-whl) dataset to ones notebook, then install via\n\n```python\n!pip install -q /kaggle/input/xgboost-2-0-0-whl/xgboost-2.0.0-py3-none-manylinux2014_x86_64.whl\n```\n\nAll the best and good luck!\n\ncarl",
    "2482987": "Will try this out, thanks",
    "2512890": "You can take adventage of the latest `2.0.1` over here - https://www.kaggle.com/datasets/jirkaborovec/xgboost-python-package"
  },
  "source": "meta"
}