{
  "id": 326142,
  "title": "How to choose a model from timm?",
  "url": "/competitions/birdclef-2022/discussion/326142",
  "author_name": "",
  "post_date": "2022-05-20T13:10:29.566912700Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I am not familiar with using deep neural network models. I don't know a model to start. Of course, trying a lot of models and comparing them is a part of ML engineering, but I'd like to know the virtual standard.</p>",
  "messages": [
    {
      "id": "1796101",
      "postDate": "05/20/2022 13:10:29",
      "content": "<p>I am not familiar with using deep neural network models. I don't know a model to start. Of course, trying a lot of models and comparing them is a part of ML engineering, but I'd like to know the virtual standard.</p>",
      "rawMarkdown": "I am not familiar with using deep neural network models. I don't know a model to start. Of course, trying a lot of models and comparing them is a part of ML engineering, but I'd like to know the virtual standard.",
      "votes": null
    },
    {
      "id": "1797863",
      "postDate": "05/22/2022 13:13:26",
      "content": "<p>Compare imagenet results.<br>\nTry efficientnet B0 for beggining, efnet family probably the most popular to use right now.</p>",
      "rawMarkdown": "Compare imagenet results.\nTry efficientnet B0 for beggining, efnet family probably the most popular to use right now.",
      "votes": null
    },
    {
      "id": "1814548",
      "postDate": "06/08/2022 03:35:47",
      "content": "<p><a href=\"https://www.kaggle.com/hypocrites\" target=\"_blank\">@hypocrites</a> <br>\nThanks!<br>\nI understand that Efficientnet B0 is baseline.</p>",
      "rawMarkdown": "hypocrites \nThanks!\nI understand that Efficientnet B0 is baseline.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1797863,
      "author_name": "hypocrites",
      "author_url": "",
      "post_date": "05/22/2022 13:13:26",
      "content": "<p>Compare imagenet results.<br>\nTry efficientnet B0 for beggining, efnet family probably the most popular to use right now.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1814548,
          "author_name": "hanejiyuto",
          "author_url": "",
          "post_date": "06/08/2022 03:35:47",
          "content": "<p><a href=\"https://www.kaggle.com/hypocrites\" target=\"_blank\">@hypocrites</a> <br>\nThanks!<br>\nI understand that Efficientnet B0 is baseline.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1796101": "I am not familiar with using deep neural network models. I don't know a model to start. Of course, trying a lot of models and comparing them is a part of ML engineering, but I'd like to know the virtual standard.",
    "1797863": "Compare imagenet results.\nTry efficientnet B0 for beggining, efnet family probably the most popular to use right now.",
    "1814548": "hypocrites \nThanks!\nI understand that Efficientnet B0 is baseline."
  },
  "source": "meta"
}