{
  "id": 189020,
  "title": "Runtime(Save and Run All) >> Runtime(manually). Why?!",
  "url": "/competitions/rsna-str-pulmonary-embolism-detection/discussion/189020",
  "author_name": "imnishantg",
  "post_date": "2020-10-06T10:50:00.617000",
  "votes": 0,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Hi <br>\nI'm facing this weird issue with my Kaggle notebooks (new user… so, not sure if this is known issue).<br>\nWhen I'm running a script manually, step-by-step it takes me ~30-45 mins to run it end-to-end.<br>\nWhen I run it through 'Save and Run All' option. It does not complete even in 3 hours! It seems to be running in interactive mode initially and then move to batch mode eventually. never ending.<br>\nDoes someone help why does this happen. I make sure to select GPU for the latter cases as well.</p>\n<p>My 9 hours GPU quota got wasted yesterday because of this very reason. I ran training pipeline that should've completed in 4 hours. It ended unceremoniously after 9 hours. Everything wasted.</p>\n<p>'First user' problems, I suppose. But guidance is appreciated!</p>",
  "messages": [
    {
      "id": 1040979,
      "postDate": "2020-10-07T13:33:21.347Z",
      "content": "<p>Try to reduce amount of data you process, and compare runtimes. Maybe you'll get an inight about the problem.</p>",
      "rawMarkdown": "Try to reduce amount of data you process, and compare runtimes. Maybe you'll get an inight about the problem."
    },
    {
      "id": 1039117,
      "postDate": "2020-10-06T10:50:00.617Z",
      "content": "<p>Hi <br>\nI'm facing this weird issue with my Kaggle notebooks (new user… so, not sure if this is known issue).<br>\nWhen I'm running a script manually, step-by-step it takes me ~30-45 mins to run it end-to-end.<br>\nWhen I run it through 'Save and Run All' option. It does not complete even in 3 hours! It seems to be running in interactive mode initially and then move to batch mode eventually. never ending.<br>\nDoes someone help why does this happen. I make sure to select GPU for the latter cases as well.</p>\n<p>My 9 hours GPU quota got wasted yesterday because of this very reason. I ran training pipeline that should've completed in 4 hours. It ended unceremoniously after 9 hours. Everything wasted.</p>\n<p>'First user' problems, I suppose. But guidance is appreciated!</p>",
      "rawMarkdown": "Hi \nI'm facing this weird issue with my Kaggle notebooks (new user... so, not sure if this is known issue).\nWhen I'm running a script manually, step-by-step it takes me ~30-45 mins to run it end-to-end.\nWhen I run it through 'Save and Run All' option. It does not complete even in 3 hours! It seems to be running in interactive mode initially and then move to batch mode eventually. never ending.\nDoes someone help why does this happen. I make sure to select GPU for the latter cases as well.\n\nMy 9 hours GPU quota got wasted yesterday because of this very reason. I ran training pipeline that should've completed in 4 hours. It ended unceremoniously after 9 hours. Everything wasted.\n\n'First user' problems, I suppose. But guidance is appreciated!"
    }
  ],
  "comments": [
    {
      "id": 1040979,
      "author_name": "Araik Tamazian",
      "author_url": "",
      "post_date": "2020-10-07T13:33:21.347000",
      "content": "<p>Try to reduce amount of data you process, and compare runtimes. Maybe you'll get an inight about the problem.</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1040979": "Try to reduce amount of data you process, and compare runtimes. Maybe you'll get an inight about the problem.",
    "1039117": "Hi \nI'm facing this weird issue with my Kaggle notebooks (new user... so, not sure if this is known issue).\nWhen I'm running a script manually, step-by-step it takes me ~30-45 mins to run it end-to-end.\nWhen I run it through 'Save and Run All' option. It does not complete even in 3 hours! It seems to be running in interactive mode initially and then move to batch mode eventually. never ending.\nDoes someone help why does this happen. I make sure to select GPU for the latter cases as well.\n\nMy 9 hours GPU quota got wasted yesterday because of this very reason. I ran training pipeline that should've completed in 4 hours. It ended unceremoniously after 9 hours. Everything wasted.\n\n'First user' problems, I suppose. But guidance is appreciated!"
  }
}