{
  "id": 54699,
  "title": "UNet Results",
  "url": "/competitions/data-science-bowl-2018/discussion/54699",
  "author_name": "",
  "post_date": "2018-04-17T00:35:01.949329100Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>First congratulations to the winners!</p>\n\n<p>I managed a 0.479 with UNet, a lot of preprocessing of the ground truth masks and augmentation. How did others do with UNet?</p>",
  "messages": [
    {
      "id": "315273",
      "postDate": "04/17/2018 00:35:01",
      "content": "<p>First congratulations to the winners!</p>\n\n<p>I managed a 0.479 with UNet, a lot of preprocessing of the ground truth masks and augmentation. How did others do with UNet?</p>",
      "rawMarkdown": "First congratulations to the winners!\n\nI managed a 0.479 with UNet, a lot of preprocessing of the ground truth masks and augmentation. How did others do with UNet?",
      "votes": null
    },
    {
      "id": "315295",
      "postDate": "04/17/2018 00:53:44",
      "content": "<p>We got 0.545 with Unet, and I bet better scores have been obtained with Unet.</p>",
      "rawMarkdown": "We got 0.545 with Unet, and I bet better scores have been obtained with Unet.",
      "votes": null
    },
    {
      "id": "315299",
      "postDate": "04/17/2018 00:56:31",
      "content": "<p>I don't doubt that. I'm just curious about what was possible. This is my first my first deep learning project and I'm just trying to learn.</p>",
      "rawMarkdown": "I don't doubt that. I'm just curious about what was possible. This is my first my first deep learning project and I'm just trying to learn.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 315295,
      "author_name": "cpmpml",
      "author_url": "",
      "post_date": "04/17/2018 00:53:44",
      "content": "<p>We got 0.545 with Unet, and I bet better scores have been obtained with Unet.</p>",
      "votes": null,
      "replies": [
        {
          "id": 315299,
          "author_name": "dredata",
          "author_url": "",
          "post_date": "04/17/2018 00:56:31",
          "content": "<p>I don't doubt that. I'm just curious about what was possible. This is my first my first deep learning project and I'm just trying to learn.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "315273": "First congratulations to the winners!\n\nI managed a 0.479 with UNet, a lot of preprocessing of the ground truth masks and augmentation. How did others do with UNet?",
    "315295": "We got 0.545 with Unet, and I bet better scores have been obtained with Unet.",
    "315299": "I don't doubt that. I'm just curious about what was possible. This is my first my first deep learning project and I'm just trying to learn."
  },
  "source": "meta"
}