{
  "id": 191816,
  "title": "Can I only include inference but not training in the notebook submitted?",
  "url": "/competitions/rsna-str-pulmonary-embolism-detection/discussion/191816",
  "author_name": "Swikwislkdjc",
  "post_date": "2020-10-18T19:22:20.922000",
  "votes": 0,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I heard it takes long time and takes GPU in kaggle kernel. I have GPU limits around 48 hours. I'd like to train locally and upload the trained model in the dataset. Can I only include inference but not training in the notebook submitted? Thanks</p>",
  "messages": [
    {
      "id": 1053271,
      "postDate": "2020-10-18T19:35:14.147Z",
      "content": "<p>yes. in fact, they don't make the train data available in the submitted notebook. So just include inference</p>",
      "rawMarkdown": "yes. in fact, they don't make the train data available in the submitted notebook. So just include inference",
      "votes": 3,
      "replies": [
        {
          "id": 1053306,
          "postDate": "2020-10-18T20:50:22.660Z",
          "content": "<p>Thank you !</p>",
          "rawMarkdown": "Thank you !",
          "votes": 1
        }
      ]
    },
    {
      "id": 1053264,
      "postDate": "2020-10-18T19:22:20.923Z",
      "content": "<p>I heard it takes long time and takes GPU in kaggle kernel. I have GPU limits around 48 hours. I'd like to train locally and upload the trained model in the dataset. Can I only include inference but not training in the notebook submitted? Thanks</p>",
      "rawMarkdown": "I heard it takes long time and takes GPU in kaggle kernel. I have GPU limits around 48 hours. I'd like to train locally and upload the trained model in the dataset. Can I only include inference but not training in the notebook submitted? Thanks"
    }
  ],
  "comments": [
    {
      "id": 1053271,
      "author_name": "quadcore/Richard Epstein",
      "author_url": "",
      "post_date": "2020-10-18T19:35:14.147000",
      "content": "<p>yes. in fact, they don't make the train data available in the submitted notebook. So just include inference</p>",
      "votes": 3,
      "replies": [
        {
          "id": 1053306,
          "author_name": "Swikwislkdjc",
          "author_url": "",
          "post_date": "2020-10-18T20:50:22.660000",
          "content": "<p>Thank you !</p>",
          "votes": 1,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1053271": "yes. in fact, they don't make the train data available in the submitted notebook. So just include inference",
    "1053264": "I heard it takes long time and takes GPU in kaggle kernel. I have GPU limits around 48 hours. I'd like to train locally and upload the trained model in the dataset. Can I only include inference but not training in the notebook submitted? Thanks"
  }
}