{
  "id": 313971,
  "title": "Can I first train the model parameters offline and then upload it to the notebook to run?",
  "url": "/competitions/birdclef-2022/discussion/313971",
  "author_name": "",
  "post_date": "2022-03-20T05:13:14.272602900Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Can I first train the model parameters offline and then upload it to the notebook to run?</p>",
  "messages": [
    {
      "id": "1729450",
      "postDate": "03/20/2022 05:13:14",
      "content": "<p>Can I first train the model parameters offline and then upload it to the notebook to run?</p>",
      "rawMarkdown": "Can I first train the model parameters offline and then upload it to the notebook to run?",
      "votes": null
    },
    {
      "id": "1730065",
      "postDate": "03/20/2022 22:13:21",
      "content": "<p>Yes, <a href=\"https://www.kaggle.com/yoyobar\" target=\"_blank\">@yoyobar</a>, you can train any model locally and then upload the model checkpoint or backup files into a Kaggle dataset, from there you can call the dataset, restore the model and generate predictions.</p>",
      "rawMarkdown": "Yes, @yoyobar, you can train any model locally and then upload the model checkpoint or backup files into a Kaggle dataset, from there you can call the dataset, restore the model and generate predictions.",
      "votes": null
    },
    {
      "id": "1730079",
      "postDate": "03/20/2022 22:42:01",
      "content": "<p><a href=\"https://www.kaggle.com/cv13j0\" target=\"_blank\">@cv13j0</a> Thank you, I handed in my first prediction</p>",
      "rawMarkdown": "cv13j0 Thank you, I handed in my first prediction",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1730065,
      "author_name": "cv13j0",
      "author_url": "",
      "post_date": "03/20/2022 22:13:21",
      "content": "<p>Yes, <a href=\"https://www.kaggle.com/yoyobar\" target=\"_blank\">@yoyobar</a>, you can train any model locally and then upload the model checkpoint or backup files into a Kaggle dataset, from there you can call the dataset, restore the model and generate predictions.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1730079,
      "author_name": "yoyobar",
      "author_url": "",
      "post_date": "03/20/2022 22:42:01",
      "content": "<p><a href=\"https://www.kaggle.com/cv13j0\" target=\"_blank\">@cv13j0</a> Thank you, I handed in my first prediction</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1729450": "Can I first train the model parameters offline and then upload it to the notebook to run?",
    "1730065": "Yes, @yoyobar, you can train any model locally and then upload the model checkpoint or backup files into a Kaggle dataset, from there you can call the dataset, restore the model and generate predictions.",
    "1730079": "cv13j0 Thank you, I handed in my first prediction"
  },
  "source": "meta"
}