{
  "id": 567325,
  "title": "Are we allowed to load pretrained model?",
  "url": "/competitions/byu-locating-bacterial-flagellar-motors-2025/discussion/567325",
  "author_name": "",
  "post_date": "2025-03-09T17:36:55.024184500Z",
  "votes": 6,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Let's say that i pretrain my model and then load it's weights for inference. Are we allowed to do that? Or submission notebook has to include training as well?</p>",
  "messages": [
    {
      "id": "3145319",
      "postDate": "03/09/2025 17:36:55",
      "content": "<p>Let's say that i pretrain my model and then load it's weights for inference. Are we allowed to do that? Or submission notebook has to include training as well?</p>",
      "rawMarkdown": "Let's say that i pretrain my model and then load it's weights for inference. Are we allowed to do that? Or submission notebook has to include training as well?",
      "votes": null
    },
    {
      "id": "3145333",
      "postDate": "03/09/2025 17:48:26",
      "content": "<p>Yes that is a great idea! As long as your notebook has no internet access, you can import and make use of whatever you like.</p>",
      "rawMarkdown": "Yes that is a great idea! As long as your notebook has no internet access, you can import and make use of whatever you like.",
      "votes": null
    },
    {
      "id": "3145343",
      "postDate": "03/09/2025 17:54:10",
      "content": "<p>Thanks! Guess i can not worry about complexity of training now!</p>",
      "rawMarkdown": "Thanks! Guess i can not worry about complexity of training now!",
      "votes": null
    },
    {
      "id": "3145643",
      "postDate": "03/10/2025 04:39:49",
      "content": "<p>Submission notebook can include a pretrained model. In fact this is a good way to handle code submissions <a href=\"https://www.kaggle.com/eikyou\" target=\"_blank\">@eikyou</a> </p>\n<p>Did you not use this in the CIBMTR competition?</p>",
      "rawMarkdown": "Submission notebook can include a pretrained model. In fact this is a good way to handle code submissions @eikyou \n\nDid you not use this in the CIBMTR competition?",
      "votes": null
    },
    {
      "id": "3145663",
      "postDate": "03/10/2025 05:21:49",
      "content": "<p><a href=\"https://www.kaggle.com/ravi20076\" target=\"_blank\">@ravi20076</a> Em, no 😅 I always submitted the notebook with training and just stopped the commit process to save quota. I'll organize my work better in this competition </p>",
      "rawMarkdown": "ravi20076 Em, no 😅 I always submitted the notebook with training and just stopped the commit process to save quota. I'll organize my work better in this competition",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3145333,
      "author_name": "andrewjdarley",
      "author_url": "",
      "post_date": "03/09/2025 17:48:26",
      "content": "<p>Yes that is a great idea! As long as your notebook has no internet access, you can import and make use of whatever you like.</p>",
      "votes": null,
      "replies": [
        {
          "id": 3145343,
          "author_name": "eikyou",
          "author_url": "",
          "post_date": "03/09/2025 17:54:10",
          "content": "<p>Thanks! Guess i can not worry about complexity of training now!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 3145643,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "03/10/2025 04:39:49",
      "content": "<p>Submission notebook can include a pretrained model. In fact this is a good way to handle code submissions <a href=\"https://www.kaggle.com/eikyou\" target=\"_blank\">@eikyou</a> </p>\n<p>Did you not use this in the CIBMTR competition?</p>",
      "votes": null,
      "replies": [
        {
          "id": 3145663,
          "author_name": "eikyou",
          "author_url": "",
          "post_date": "03/10/2025 05:21:49",
          "content": "<p><a href=\"https://www.kaggle.com/ravi20076\" target=\"_blank\">@ravi20076</a> Em, no 😅 I always submitted the notebook with training and just stopped the commit process to save quota. I'll organize my work better in this competition </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3145319": "Let's say that i pretrain my model and then load it's weights for inference. Are we allowed to do that? Or submission notebook has to include training as well?",
    "3145333": "Yes that is a great idea! As long as your notebook has no internet access, you can import and make use of whatever you like.",
    "3145343": "Thanks! Guess i can not worry about complexity of training now!",
    "3145643": "Submission notebook can include a pretrained model. In fact this is a good way to handle code submissions @eikyou \n\nDid you not use this in the CIBMTR competition?",
    "3145663": "ravi20076 Em, no 😅 I always submitted the notebook with training and just stopped the commit process to save quota. I'll organize my work better in this competition"
  },
  "source": "meta"
}