{
  "id": 508371,
  "title": "Notebook out of memory ",
  "url": "/competitions/birdclef-2024/discussion/508371",
  "author_name": "",
  "post_date": "2024-05-29T11:12:40.131689700Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>From yesterday night am facing notebook out of memory <br>\nTo test this, I even tested my old models which worked , even that gives me out of memory</p>\n<p>Has anyone faced such issues? </p>",
  "messages": [
    {
      "id": "2842970",
      "postDate": "05/29/2024 11:12:40",
      "content": "<p>From yesterday night am facing notebook out of memory <br>\nTo test this, I even tested my old models which worked , even that gives me out of memory</p>\n<p>Has anyone faced such issues? </p>",
      "rawMarkdown": "From yesterday night am facing notebook out of memory \nTo test this, I even tested my old models which worked , even that gives me out of memory\n\nHas anyone faced such issues?",
      "votes": null
    },
    {
      "id": "2844469",
      "postDate": "05/30/2024 05:18:52",
      "content": "<p>I also submitted the same model (convnext tiny v2), but sometimes it runs out of memory, and sometimes it works. </p>",
      "rawMarkdown": "I also submitted the same model (convnext tiny v2), but sometimes it runs out of memory, and sometimes it works.",
      "votes": null
    },
    {
      "id": "2857987",
      "postDate": "06/06/2024 08:11:44",
      "content": "<p>Using something like a Tensorflow PrefetchedDataset can address these, rather than loading the entire dataset into memory off the bat</p>",
      "rawMarkdown": "Using something like a Tensorflow PrefetchedDataset can address these, rather than loading the entire dataset into memory off the bat",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2844469,
      "author_name": "sentrankim",
      "author_url": "",
      "post_date": "05/30/2024 05:18:52",
      "content": "<p>I also submitted the same model (convnext tiny v2), but sometimes it runs out of memory, and sometimes it works. </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2857987,
      "author_name": "alexanderstokes",
      "author_url": "",
      "post_date": "06/06/2024 08:11:44",
      "content": "<p>Using something like a Tensorflow PrefetchedDataset can address these, rather than loading the entire dataset into memory off the bat</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2842970": "From yesterday night am facing notebook out of memory \nTo test this, I even tested my old models which worked , even that gives me out of memory\n\nHas anyone faced such issues?",
    "2844469": "I also submitted the same model (convnext tiny v2), but sometimes it runs out of memory, and sometimes it works.",
    "2857987": "Using something like a Tensorflow PrefetchedDataset can address these, rather than loading the entire dataset into memory off the bat"
  },
  "source": "meta"
}