{
  "id": 372873,
  "title": "Is it possible to compete in this competition with Kaggle resources?",
  "url": "/competitions/nfl-player-contact-detection/discussion/372873",
  "author_name": "",
  "post_date": "2022-12-18T12:17:12.576286200Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi all,<br>\nI was wondering is it possible to build a descent performing model <strong>within the kaggle env</strong>. (using videos and other features as inputs instead of derived features). </p>\n<p>And can anyone share me the kind of models which can be trained entirely on Kaggle as I don't have a home setup. It will help me to skip bunch of big/high RAM consuming models.</p>",
  "messages": [
    {
      "id": "2068916",
      "postDate": "12/18/2022 12:17:12",
      "content": "<p>Hi all,<br>\nI was wondering is it possible to build a descent performing model <strong>within the kaggle env</strong>. (using videos and other features as inputs instead of derived features). </p>\n<p>And can anyone share me the kind of models which can be trained entirely on Kaggle as I don't have a home setup. It will help me to skip bunch of big/high RAM consuming models.</p>",
      "rawMarkdown": "Hi all,\nI was wondering is it possible to build a descent performing model **within the kaggle env**. (using videos and other features as inputs instead of derived features). \n\nAnd can anyone share me the kind of models which can be trained entirely on Kaggle as I don't have a home setup. It will help me to skip bunch of big/high RAM consuming models.",
      "votes": null
    },
    {
      "id": "2069008",
      "postDate": "12/18/2022 13:55:09",
      "content": "<p>Yes, its totally possible. As a matter of fact, people have already shared amazing models fully trained in Kaggle kernels that will get you started.</p>",
      "rawMarkdown": "Yes, its totally possible. As a matter of fact, people have already shared amazing models fully trained in Kaggle kernels that will get you started.",
      "votes": null
    },
    {
      "id": "2069434",
      "postDate": "12/19/2022 01:30:52",
      "content": "<p>Great! Thank you :) will try training.</p>",
      "rawMarkdown": "Great! Thank you :) will try training.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2069008,
      "author_name": "carloshuertas",
      "author_url": "",
      "post_date": "12/18/2022 13:55:09",
      "content": "<p>Yes, its totally possible. As a matter of fact, people have already shared amazing models fully trained in Kaggle kernels that will get you started.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2069434,
          "author_name": "jarupula",
          "author_url": "",
          "post_date": "12/19/2022 01:30:52",
          "content": "<p>Great! Thank you :) will try training.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2068916": "Hi all,\nI was wondering is it possible to build a descent performing model **within the kaggle env**. (using videos and other features as inputs instead of derived features). \n\nAnd can anyone share me the kind of models which can be trained entirely on Kaggle as I don't have a home setup. It will help me to skip bunch of big/high RAM consuming models.",
    "2069008": "Yes, its totally possible. As a matter of fact, people have already shared amazing models fully trained in Kaggle kernels that will get you started.",
    "2069434": "Great! Thank you :) will try training."
  },
  "source": "meta"
}