{
  "id": 99086,
  "title": "Training image",
  "url": "/competitions/recursion-cellular-image-classification/discussion/99086",
  "author_name": "",
  "post_date": "2019-07-08T18:11:43.353227Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Can anyone please help me with \"ON which image should I train my model\"?\nLike there are 2 sites with 6 channels each, if I convert both sites' images to rgb, there are still two images for one entry, on which image should I train my model? Please elaborate anyone!</p>",
  "messages": [
    {
      "id": "570739",
      "postDate": "07/08/2019 18:11:43",
      "content": "<p>Can anyone please help me with \"ON which image should I train my model\"?\nLike there are 2 sites with 6 channels each, if I convert both sites' images to rgb, there are still two images for one entry, on which image should I train my model? Please elaborate anyone!</p>",
      "rawMarkdown": "Can anyone please help me with \"ON which image should I train my model\"?\nLike there are 2 sites with 6 channels each, if I convert both sites' images to rgb, there are still two images for one entry, on which image should I train my model? Please elaborate anyone!",
      "votes": null
    },
    {
      "id": "570815",
      "postDate": "07/08/2019 19:46:40",
      "content": "<p>You can train on both of them. Just treat them as separate examples that happen to have the same label.</p>",
      "rawMarkdown": "You can train on both of them. Just treat them as separate examples that happen to have the same label.",
      "votes": null
    },
    {
      "id": "570916",
      "postDate": "07/08/2019 23:44:11",
      "content": "<p>To be clear, there were 2 sets of 6 channel imagery done on each sample - You should be able to train on both </p>",
      "rawMarkdown": "To be clear, there were 2 sets of 6 channel imagery done on each sample - You should be able to train on both",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 570815,
      "author_name": "mrosett",
      "author_url": "",
      "post_date": "07/08/2019 19:46:40",
      "content": "<p>You can train on both of them. Just treat them as separate examples that happen to have the same label.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 570916,
      "author_name": "gidutz",
      "author_url": "",
      "post_date": "07/08/2019 23:44:11",
      "content": "<p>To be clear, there were 2 sets of 6 channel imagery done on each sample - You should be able to train on both </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "570739": "Can anyone please help me with \"ON which image should I train my model\"?\nLike there are 2 sites with 6 channels each, if I convert both sites' images to rgb, there are still two images for one entry, on which image should I train my model? Please elaborate anyone!",
    "570815": "You can train on both of them. Just treat them as separate examples that happen to have the same label.",
    "570916": "To be clear, there were 2 sets of 6 channel imagery done on each sample - You should be able to train on both"
  },
  "source": "meta"
}