{
  "id": 249607,
  "title": "34th place solution: Key points",
  "url": "/competitions/coleridgeinitiative-show-us-the-data/discussion/249607",
  "author_name": "Khubchandani",
  "post_date": "2021-06-29T06:37:26.498000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Thanks to the Coleridge Initiative and to Kaggle for organizing such an exciting competition. Congratulations to all the winners!</p>\n<ol>\n<li>My solution was based on the BERT MLM model published by <a href=\"https://www.kaggle.com/tungmphung\" target=\"_blank\">@tungmphung</a> early on in the competition. It is referenced on the leaderboard.</li>\n<li>It survived the leaderboard shakeup at the end (moved up 25 places).</li>\n<li>All datasets / pre-trained models used were made public before the competition deadline as stipulated.</li>\n<li>I was impacted by the 9 hour kernel training time limit. If I could circumvent this in some way, I would have gained further (more training epochs, more realistic model).</li>\n<li>I used the external datasets published by <a href=\"https://www.kaggle.com/mlconsult\" target=\"_blank\">@mlconsult</a> It improved my public score dramatically and had no adverse impact on my private score. I verified this by choosing 1 submission with this data included and 1 without, for my 2 final submissions and results were almost the same (0.372 and 0.373).</li>\n<li>I look forward to continuing to be a part of this initiative. Thanks much!! </li>\n</ol>",
  "messages": [
    {
      "id": 1369090,
      "postDate": "2021-06-29T06:37:26.497Z",
      "content": "<p>Thanks to the Coleridge Initiative and to Kaggle for organizing such an exciting competition. Congratulations to all the winners!</p>\n<ol>\n<li>My solution was based on the BERT MLM model published by <a href=\"https://www.kaggle.com/tungmphung\" target=\"_blank\">@tungmphung</a> early on in the competition. It is referenced on the leaderboard.</li>\n<li>It survived the leaderboard shakeup at the end (moved up 25 places).</li>\n<li>All datasets / pre-trained models used were made public before the competition deadline as stipulated.</li>\n<li>I was impacted by the 9 hour kernel training time limit. If I could circumvent this in some way, I would have gained further (more training epochs, more realistic model).</li>\n<li>I used the external datasets published by <a href=\"https://www.kaggle.com/mlconsult\" target=\"_blank\">@mlconsult</a> It improved my public score dramatically and had no adverse impact on my private score. I verified this by choosing 1 submission with this data included and 1 without, for my 2 final submissions and results were almost the same (0.372 and 0.373).</li>\n<li>I look forward to continuing to be a part of this initiative. Thanks much!! </li>\n</ol>",
      "rawMarkdown": "Thanks to the Coleridge Initiative and to Kaggle for organizing such an exciting competition. Congratulations to all the winners!\n1. My solution was based on the BERT MLM model published by @tungmphung early on in the competition. It is referenced on the leaderboard.\n2. It survived the leaderboard shakeup at the end (moved up 25 places).\n3. All datasets / pre-trained models used were made public before the competition deadline as stipulated.\n4. I was impacted by the 9 hour kernel training time limit. If I could circumvent this in some way, I would have gained further (more training epochs, more realistic model).\n5. I used the external datasets published by @mlconsult It improved my public score dramatically and had no adverse impact on my private score. I verified this by choosing 1 submission with this data included and 1 without, for my 2 final submissions and results were almost the same (0.372 and 0.373).\n6. I look forward to continuing to be a part of this initiative. Thanks much!! ",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1369090": "Thanks to the Coleridge Initiative and to Kaggle for organizing such an exciting competition. Congratulations to all the winners!\n1. My solution was based on the BERT MLM model published by @tungmphung early on in the competition. It is referenced on the leaderboard.\n2. It survived the leaderboard shakeup at the end (moved up 25 places).\n3. All datasets / pre-trained models used were made public before the competition deadline as stipulated.\n4. I was impacted by the 9 hour kernel training time limit. If I could circumvent this in some way, I would have gained further (more training epochs, more realistic model).\n5. I used the external datasets published by @mlconsult It improved my public score dramatically and had no adverse impact on my private score. I verified this by choosing 1 submission with this data included and 1 without, for my 2 final submissions and results were almost the same (0.372 and 0.373).\n6. I look forward to continuing to be a part of this initiative. Thanks much!! "
  }
}