{
  "id": 21416,
  "title": "Xgb in python with customized evaluation metric",
  "url": "/competitions/expedia-hotel-recommendations/discussion/21416",
  "author_name": "FengLi",
  "post_date": "2016-06-04T01:47:55.197000",
  "votes": 0,
  "comment_count": 0,
  "views": 393,
  "content": "<p>I used my school's machine 32 core with 190G RAM.  I trained the the whole train data in which  30% of them is as validation set. I just have 100 features but it took me 5 minutes to do each round. Is it normal? Thank you. </p>",
  "messages": [
    {
      "id": 122434,
      "postDate": "2016-06-04T01:47:55.197Z",
      "content": "<p>I used my school's machine 32 core with 190G RAM.  I trained the the whole train data in which  30% of them is as validation set. I just have 100 features but it took me 5 minutes to do each round. Is it normal? Thank you. </p>",
      "rawMarkdown": "I used my school's machine 32 core with 190G RAM.  I trained the the whole train data in which  30% of them is as validation set. I just have 100 features but it took me 5 minutes to do each round. Is it normal? Thank you. "
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "122434": "I used my school's machine 32 core with 190G RAM.  I trained the the whole train data in which  30% of them is as validation set. I just have 100 features but it took me 5 minutes to do each round. Is it normal? Thank you. "
  }
}