{
  "id": 147084,
  "title": "Finetune or train from scratch?",
  "url": "/competitions/prostate-cancer-grade-assessment/discussion/147084",
  "author_name": "A.Demyanchuk",
  "post_date": "2020-04-29T12:33:07.591000",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hey guys! Happy Kaggling))</p>\n\n<p>Hope it could be more like a micro-poll than a question!\nHow do you start?\n  1. From finetuning (ImageNet pretrained) network\n  2. Train (X-archtechture Net) from scracth</p>\n\n<p>Thanks for sharing!</p>",
  "messages": [
    {
      "id": 826066,
      "postDate": "2020-04-29T12:33:07.593Z",
      "content": "<p>Hey guys! Happy Kaggling))</p>\n\n<p>Hope it could be more like a micro-poll than a question!\nHow do you start?\n  1. From finetuning (ImageNet pretrained) network\n  2. Train (X-archtechture Net) from scracth</p>\n\n<p>Thanks for sharing!</p>",
      "rawMarkdown": "Hey guys! Happy Kaggling))\n\nHope it could be more like a micro-poll than a question!\nHow do you start?\n  1. From finetuning (ImageNet pretrained) network\n  2. Train (X-archtechture Net) from scracth\n\nThanks for sharing!",
      "votes": 1
    },
    {
      "id": 826069,
      "postDate": "2020-04-29T12:37:51.713Z",
      "content": "<p>Fine tuning in 98% cases is always a good start :)</p>",
      "rawMarkdown": "Fine tuning in 98% cases is always a good start :)",
      "replies": [
        {
          "id": 826084,
          "postDate": "2020-04-29T12:54:20.327Z",
          "content": "<p>Thx )</p>",
          "rawMarkdown": "Thx )",
          "votes": 1
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 826069,
      "author_name": "DrHB",
      "author_url": "",
      "post_date": "2020-04-29T12:37:51.713000",
      "content": "<p>Fine tuning in 98% cases is always a good start :)</p>",
      "votes": 0,
      "replies": [
        {
          "id": 826084,
          "author_name": "A.Demyanchuk",
          "author_url": "",
          "post_date": "2020-04-29T12:54:20.327000",
          "content": "<p>Thx )</p>",
          "votes": 1,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "826066": "Hey guys! Happy Kaggling))\n\nHope it could be more like a micro-poll than a question!\nHow do you start?\n  1. From finetuning (ImageNet pretrained) network\n  2. Train (X-archtechture Net) from scracth\n\nThanks for sharing!",
    "826069": "Fine tuning in 98% cases is always a good start :)"
  }
}